Informix Handbook

Front cover
Draft Document for Review August 23, 2010 10:53 am
SG24-7884-00
IBM Informix
Developers Handbook
Learn application development with
supported APIs, drivers, and interfaces
Understand Informix supported
programming environments
Practical examples
Whei-Jen Chen
David Jay
Javier Sagrera
Krishna Doddi
Manoj Ghogale
ibm.com/redbooks
7884edno.fm
International Technical Support Organization

IBM Informix Developers Handbook
October 2010
SG24-7884-00
7884edno.fm
Note: Before using this information and the product it supports, read the information in
Notices on page ix.
First Edition (October 2010)

This edition applies to IBM Informix Version 11.5.
This document created or updated on August 23, 2010.
Copyright International Business Machines Corporation 2010. All rights reserved.

Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP
Schedule Contract with IBM Corp.
7884TOC.fm
Contents
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
The team who wrote this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Become a published author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Chapter 1. Introduction to IBM Informix . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Server options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Informix servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Informix tools for developers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2.1 I-Connect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.2 Client SDK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.3 4GL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.4 Ruby on Rails . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.5 Informix DataBlade Developers Kit . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.2.6 Informix Spatial DataBlade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.2.7 PHP on Informix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.3 Informix overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3.1 Architecture overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3.2 Informix developer environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.3.3 Informix capabilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Chapter 2. Setting up an Informix development environment . . . . . . . . . 23
2.1 Informix deployment and setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.1 Installation planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.2 Informix Server installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.3 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Client setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2.1 Client options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2.2 Setting up IBM Informix Client Software Development Kit . . . . . . . . 33
2.2.3 Setting up IBM Data Server drivers. . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.2.4 Setting up Informix JDBC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Chapter 3. Working with the ODBC driver . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1 ODBC and Informix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.2 Setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Copyright IBM Corp. 2010. All rights reserved.
iii
7884TOC.fm
3.2.1 IBM Informix ODBC Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

3.2.2 IBM Data Server Driver for ODBC and CLI . . . . . . . . . . . . . . . . . . . . 66
3.2.3 Verifying connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.3 Developing an ODBC application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.3.1 Connecting to the database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.3.2 Type mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.3.3 Performing database operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.3.4 Handling special data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.3.5 Error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
3.3.6 Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Chapter 4. Working with ESQL/C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.1 Informix ESQL/C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
4.2 Setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.3 Developing an ESQL/C application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4.3.1 Creating an ESQL/C application . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4.3.2 Performing database operations . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4.3.3 Data types mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.3.4 Handling special data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
4.3.5 Exception handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Chapter 5. Working with the JDBC drivers . . . . . . . . . . . . . . . . . . . . . . . . 149
5.1 JDBC drivers for an Informix database . . . . . . . . . . . . . . . . . . . . . . . . . . 150
5.1.1 JDBC drivers for Informix database . . . . . . . . . . . . . . . . . . . . . . . . 150
5.2.1 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.2.2 Verify Connectivity with the Informix JDBC driver . . . . . . . . . . . . . . 153
5.2.3 Verify Connectivity with the Data Server Driver . . . . . . . . . . . . . . . 155
5.3 JDBC type mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
5.4 Performing database operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.4.1 Connection to the database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.4.2 Manipulating data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
5.5 Informix additional features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.5.1 Batch inserts, updates, and resultset metadata . . . . . . . . . . . . . . . 169
5.5.2 BIGSERIAL data type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.5.3 Informix Smart Large Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.5.4 Secure Socket Layer (SSL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
5.5.5 Typical errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
5.5.6 Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Chapter 6. IBM Informix with Hibernate . . . . . . . . . . . . . . . . . . . . . . . . . . 183
6.1 Hibernate for Java . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
6.1.1 Hibernate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
iv
7884TOC.fm
6.1.2 Hibernate concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

6.2.1 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
6.2.2 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.3 Using Hibernate with an Informix database. . . . . . . . . . . . . . . . . . . . . . . 192
6.3.1 Components of a Hibernate application . . . . . . . . . . . . . . . . . . . . . 192
6.3.2 Working with a Hibernate object . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
6.3.3 Using annotations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Chapter 7. Working with the IBM Informix OLE DB Provider . . . . . . . . . 209
7.1 IBM Informix OLE DB Provider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
7.2.1 Installation and setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
7.2.2 Verifying connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
7.3 Developing an OLE DB application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
7.3.1 Supported interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.3.2 Connecting to database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.3.3 Type mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.3.4 Cursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.3.5 Typical database operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.4 Visual Basic, ADO.NET, and SQL Server . . . . . . . . . . . . . . . . . . . . . . . . 232
7.4.1 OLE DB with Visual Basic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
7.4.2 ADO.NET and the OLEDB bridge . . . . . . . . . . . . . . . . . . . . . . . . . . 234
7.4.3 SQL Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
7.5 Troubleshooting and tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
7.5.1 Typical errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7.5.2 Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Chapter 8. Working with the .NET Data Providers . . . . . . . . . . . . . . . . . . 247
8.1 Informix and .NET data providers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
8.2.1 IBM Informix .Net Provider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
8.2.2 IBM Data Server Provider for .NET . . . . . . . . . . . . . . . . . . . . . . . . . 249
8.3 Developing a .NET application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
8.3.2 Data type mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
8.3.4 Handling Informix specific data types . . . . . . . . . . . . . . . . . . . . . . . 269
8.4 Visual Studio Add-In for Visual Studio. . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Chapter 9. Working with PHP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
9.1 Informix and PHP extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
Contents
7884TOC.fm

9.2.1 Installation of OAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
9.2.2 PDO_INFORMIX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
9.2.3 PDO_IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
9.2.4 Verifying the PDO setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
9.3 Developing a PHP application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
9.3.3 Handling complex data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
9.3.4 Working with PHP extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
9.3.5 Exception handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Chapter 10. User defined routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
10.1 An Overview of UDRs and database extensions . . . . . . . . . . . . . . . . . 324
10.1.1 Considerations for user defined routines . . . . . . . . . . . . . . . . . . . 324
10.1.2 About user-defined routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
10.1.3 Considerations for extending data types . . . . . . . . . . . . . . . . . . . . 328
10.2 Developing user defined routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
10.2.1 User defined routines in SQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
10.2.2 User defined routines in Java . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
10.2.3 User defined routines in C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
10.3 DataBlades and bladelets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
10.3.1 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
10.3.2 IBM Informix provided DataBlades . . . . . . . . . . . . . . . . . . . . . . . . 349
10.3.3 Developing a bladelet routine . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Chapter 11. Working with Ruby on Rails. . . . . . . . . . . . . . . . . . . . . . . . . . 355
11.1 A brief overview of Ruby on Rails . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
11.1.1 Architecture of Ruby on Rails . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
11.1.2 Ruby driver and Rails adapter. . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
11.2 Setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
11.2.1 Ruby Informix driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
11.2.2 Data Server Ruby driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
11.2.3 Rails adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
11.3 Database operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
11.4 Using the Rails adapter with Ruby Informix . . . . . . . . . . . . . . . . . . . . . 376
11.5 Using the Rails adapter with IBM_DB . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Chapter 12. Informix 4GL Web Services . . . . . . . . . . . . . . . . . . . . . . . . . . 389
12.1 Basic concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
12.1.1 IBM Informix 4GL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
12.1.2 Service Oriented Architecture and Web Services . . . . . . . . . . . . . 390
vi
7884TOC.fm
12.1.3 Web Services development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

12.1.4 Informix 4GL and Web Services . . . . . . . . . . . . . . . . . . . . . . . . . . 391
12.1.5 Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
12.2 Setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
12.2.1 Prerequisites and supported platforms . . . . . . . . . . . . . . . . . . . . . 392
12.2.2 Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
12.3 I4GL Web Service tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
12.3.1 Web Service Compiler w4glc . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
12.3.2 The w4gl utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
12.3.3 Web Service Description Language Parser (wsdl_parser) . . . . . . 397
12.3.4 I4GL Web Services process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
12.4 Developing a Web Service with I4GL . . . . . . . . . . . . . . . . . . . . . . . . . . 398
12.4.1 Example I4GL function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
12.4.2 Host and application details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
12.4.3 Definition of the Web Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
12.4.4 Generate the configuration file . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
12.4.5 Deployment of the Web Service . . . . . . . . . . . . . . . . . . . . . . . . . . 409
12.4.6 Packaging of the Web Service . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
12.4.7 Starting the Axis2 application server . . . . . . . . . . . . . . . . . . . . . . . 411
12.4.8 Consuming the I4GL Web Service . . . . . . . . . . . . . . . . . . . . . . . . 411
12.5 Consuming a Web Service with I4GL . . . . . . . . . . . . . . . . . . . . . . . . . . 414
12.5.1 Web service to consume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
12.5.2 Compiling the wrapper code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
12.5.3 Using the Web Service from an I4GL application . . . . . . . . . . . . . 416
12.6 Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
12.6.1 Typical problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
12.6.2 Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Chapter 13. Application development considerations. . . . . . . . . . . . . . . 425
13.1 Concurrency and locking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
13.1.1 Types of locks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
13.1.2 Lock duration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
13.1.3 Lock granularity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
13.2 Locking issues and performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
13.2.1 Deadlocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
13.3 Isolation levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
13.4 Configuration options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
13.4.1 Server identification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
13.4.2 Storage space identifiers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
13.4.3 Limiters and limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
13.4.4 Java configuration parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
13.5 Working with your database administrator . . . . . . . . . . . . . . . . . . . . . . 445
13.5.1 Parameters for negotiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Contents
vii
7884TOC.fm
13.5.2 Monitoring isolation levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

13.5.3 Monitoring locks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
13.5.4 Monitoring user threads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Appendix A. Onconfig parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Appendix B. Accommodating distributed transactions . . . . . . . . . . . . . 455
B.1 Distributed transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
B.2 TP/XA Transaction Manager XA Interface Library . . . . . . . . . . . . . . . . . 456
B.3 XA_TOOL ESQL/C sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Other publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
How to get Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
viii
7884spec.fm
Notices
This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area.
Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM
product, program, or service may be used. Any functionally equivalent product, program, or service that
does not infringe any IBM intellectual property right may be used instead. However, it is the user's
responsibility to evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in this document.
The furnishing of this document does not give you any license to these patents. You can send license
inquiries, in writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785 U.S.A.
The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer
of express or implied warranties in certain transactions, therefore, this statement may not apply to you.
This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may
make improvements and/or changes in the product(s) and/or the program(s) described in this publication at
any time without notice.
Any references in this information to non-IBM websites are provided for convenience only and do not in any
manner serve as an endorsement of those websites. The materials at those websites are not part of the
materials for this IBM product and use of those websites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate without
incurring any obligation to you.
Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm
the accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on
the capabilities of non-IBM products should be addressed to the suppliers of those products.
This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the
sample programs are written. These examples have not been thoroughly tested under all conditions. IBM,
therefore, cannot guarantee or imply reliability, serviceability, or function of these programs.
ix
7884spec.fm
Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business
Machines Corporation in the United States, other countries, or both. These and other IBM trademarked
terms are marked on their first occurrence in this information with the appropriate symbol ( or ),
indicating US registered or common law trademarks owned by IBM at the time this information was
published. Such trademarks may also be registered or common law trademarks in other countries. A current
list of IBM trademarks is available on the Web at http://www.ibm.com/legal/copytrade.shtml
The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
AIX
DataBlade
DB2 Universal Database
DB2
developerWorks
Distributed Relational Database

Architecture
DRDA
IBM
Informix
MQSeries
Rational
Redbooks
Redpaper
Redbooks (logo)
UC2
WebSphere
The following terms are trademarks of other companies:

Java, and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other
countries, or both.
Microsoft, Windows, and the Windows logo are trademarks of Microsoft Corporation in the United States,
other countries, or both.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Linux is a trademark of Linus Torvalds in the United States, other countries, or both.
Other company, product, or service names may be trademarks or service marks of others.
7884pref.fm
Preface
IBM Informix is a low-administration, easy-to-use, and embeddable database
that is idea for application development. It supports a wide range of development
platforms, such as Java, .NET, PHP, and Web services, enabling developers to
build database applications in their language of choice. Informix is designed to
handle RDBMS data and XML out of the box, and can easily be extended to
handle new data sets.
This IBM Redbooks publication provides fundamentals of Informix application
development. It covers the Informix Clients installation and configuration for
application development environments. It discusses the skills and techniques for
building Informix applications with Java, ESQL/C, OLE DB, .NET, PHP, Ruby on
Rails, DataBlade, and Hibernate. Code examples are used to demonstrate
how to develop an Informix application with various drivers, APIs, and interfaces.
This book also provides application development trouble-shooting and
considerations for performance.
The team who wrote this book

This book was produced by a team of specialists from around the world working
at the International Technical Support Organization, Rochester Center.
Whei-Jen Chen is a Project Leader at the International Technical Support
Organization, San Jose Center. She has extensive experience in application
development, database design and modeling, and DB2 system administration.
Whei-Jen is an IBM Certified Solutions Expert in Database Administration and
Application Development, and an IBM Certified IT Specialist
David Jay is a Staff Software Engineer in the IBM North
Americas Technical Support Team, providing advanced
technical support, defect discovery, and support training for
Informix products and SolidDB. David joined the Informix
Support team in September 1996, and has served in various
software and support roles since the 1980's. He has a
Bachelor of Science from Pennsylvania State University, and
enjoys public speaking assignments whenever he gets them
through Toastmasters and IBM.
xi
7884pref.fm
Javier Sagrera is a software engineer on the Common Client

Technologies (CCT) group. He joined the Informix team in 2000
and has over 15 years experience in application development
for Informix database servers and Informix clients. Currently
based on the IBM UK Bedfont Lab in London, he has extensive
knowledge on all the Microsoft technologies and is
considered as a subject matter expert worldwide on all the
Informix development tools.
Krishna Doddi (also known as Prasad) has been with IBM
Informix since 1996. He worked in Informix Advanced Support
for Client Products. The main products he supported were
Informix Client SDK, mostly on Windows platform. After the
IBM acquisition in 2001, he moved to DB2 Advanced Support,
again in the Client Products division. After five years in DB2, he
moved to Informix QA for Informix Client Products including
JCC Common Client and PureQuery. He is currently the focal
for Informix Integration projects.
Manoj Ghogale is a Tech Lead in the Informix team at ISL
India. Manoj has 9 years of industry experience and has been
with IBM for 5 years. He has worked on numerous automation
projects and new features of Informix server. He holds a
Bachelor's degree in Engineering from National Institute of
Engineering, Mysore, India.
Acknowledgements
Thanks to the following people for their contributions to this project:
Jacques Roy
Ted Wasserman
Rakeshkumar Naik
Jonathan Leffler
Robert Uleman
Richard Snoke
Guy Bowerman
IBM Software Group
Greg Holmes
Alberto Bortolan
Adam Hattrell
IBM Bedfont Laboratory
xii
7884pref.fm
Emma Jacobs
International Technical Support Organization, Rochester Center
Become a published author

Here's an opportunity to spotlight your skills, grow your career, and become a
published author - all at the same time! Join an ITSO residency project and help
write a book in your area of expertise, while honing your experience using
leading-edge technologies. Your efforts will help to increase product acceptance
and customer satisfaction, as you expand your network of technical contacts and
relationships. Residencies run from two to six weeks in length, and you can
participate either in person or as a remote resident working from your home
base.
Find out more about the residency program, browse the residency index, and
apply online at:
ibm.com/redbooks/residencies.html
Comments welcome
Your comments are important to us!
We want our books to be as helpful as possible. Send us your comments about
this book or other IBM Redbooks publications in one of the following ways:
Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
Send your comments in an e-mail to:
redbooks@us.ibm.com
Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400
Preface
xiii
7884pref.fm
xiv
7884ch01.fm
Chapter 1.
Introduction to IBM Informix

We live in a business world that requires a variety of software components to fully
handle the scope of business needs. When considered as a group, the products
have to work together, and they must scale well to cover the demands of
business to assure continued growth. Frequently, a significant criteria list may be
needed, to specify all the facets encountered in daily, monthly, and yearly periods
of business operations. Developers must be sensitive to how the database
engine functions, what are server side capabilities that can be brought into play,
and what development tasks need to occur on the client, or application side.
This chapter provides an overview of Informix products and capabilities. The
approach is somewhat reference oriented. In addition to the overview of Informix
products, we also provide an awareness of the architecture of IBM Informix from
both the server-side and the client-side, to better acquaint the developer with
perspectives you may need to know.
Lastly, we discuss the capabilities of the products in the Informix suite.The IBM
Informix products that we describe demonstrate how to manage business
performance, handle all types of business applications, minimize administration
through various self-automated and self-monitored options, and at all times, the
setup can help minimize costs for maintenance and resource allocations.
7884ch01.fm
1.1 Server options

There are a number of options for IBM Informix servers, designed to meet the
needs of business enterprises that range in size from a small group with a few
employees, up to several thousand people. The database needs to meet a
variety of requirements of a diverse group. What follows is a brief discussion of
each of the server possibilities, and how they are generally used.
With the wealth of different server types available, unless otherwise noted, each
comes with the opportunity to use the Client Software Development Kit (Client
SDK) for development of applications. Prior to renaming the brand editions in
May 2010, the IBM Informix has had several versions. You can expect that each
version is backward and forward compatible, beginning with 7.2x, and moving
forward through 7.3x, 9.2x, 9.30, 9.40, and 10.00. The Cheetah series began
with Informix Dynamic Server 11.10, and quickly advanced into the 11.50 family.
At this writing we are now into the IBM Informix Ultimate series that begins with
11.50.xC7 and following,
It is helpful to be aware of version numbering conventions. UC or UD at the
end of a version number indicates 32-bit UNIX. A TC indicates a 32-bit
Windows version. An FC indicates a 64-bit version (UNIX or Windows). This
becomes important in application and memory addressability. A 32-bit
application works with a 64-bit server, but the application will only be able to
access 32-bit memory.
1.1.1 Informix servers

Several options are available for IBM Informix Servers, each one designed to
address specific needs in business.
No-charge editions
The following no-charge editions are available as separate offerings subject to
the IBM International License Agreement for Non-warranted Programs (ILAN):
Informix Developer Edition

The Developer Edition provides all the features of Informix at no cost, when
solely used for application development and testing. It can be used for application
development and prototyping, with no time limits. Support is available by way of
the informix user community through internet forums.
The Informix Developer Edition offers the application development these choices:
C, C++, .NET, Java, Ruby on Rails, Perl, Python, PHP, and 4GL. The operating
systems that the Developer Edition supports includes AIX, HP, Unix, Linux,
7884ch01.fm
Macintosh, Sun Solaris, and Windows. This Informix edition is limited to 1 CPU, 1
GB memory, and 8 GB storage.
Informix Developer Edition includes the following bundle:
Informix Client Software Development Kit
Informix DataBlade Developers Kit
Informix Spatial Datablade
Informix Innovator-C Edition

The Innovator-C Edition provides a robust and powerful environment designed to
support small production workloads. It features the most widely used data
processing functionality, including limited Enterprise Replication and High
Availability clustering.
This edition offers application development these choices: C, C++, .NET, Java,
Ruby on Rails, Perl, and 4GL. The operating systems the Innovator-C Edition
supports include AIX, HP, Unix, Linux, Macintosh, Sun Solaris, and Windows.
This edition is limited to 1 socket with no more than 4 cores and a total of 2 GB of
RAM operating from the same Install.
The Innovator-C no-charge Informix edition is an offering to be used for
development, test, and end-user production workloads without a license fee. This
edition can only be used by end-user organizations. It cannot be re-distributed
without signing a re-distribution contract. Support is community-based though an
optional for-charge service and support package is available.
Licensed versions
In a business enterprise that demands constant uptime and consistent delivery
requirements, it helps to have a license and professional, timely support to cover
every need. There are several licensing options available. In this section we
discuss the Informix server versions that are licensed subscriptions. Table 1-1 on
page 3 describes the terms and abbreviations for these editions. The criteria may
be subject to change over time, consult with a sales representative for the most
accurate definitions.
Table 1-1 Informix licensed versions
Licensing term
Defined
Processor Value Unit

(PVU)
(also known as processor-based
pricing)
Calculated using the number of processor cores in

the physical server multiplied by the corresponding
value units based on processor architecture. This
is an unlimited user or connection license and is
usually the optimal choice when the user or
session load cannot be controlled or counted.
Chapter 1. Introduction to IBM Informix
7884ch01.fm
Licensing term
Defined
Authorized User
Single Install
A single named user or specific individual

accessing one installation of Informix on each
physical or virtual server. That authorized user can
establish multiple connections to an Informix
instance on the server. Each connection is for the
exclusive use of that one authorized user from a
single client device.
Concurrent Session
A single logical connection from a client device to

an Informix instance on each physical or logical
server. Each connection, whether active or not,
requires a license, regardless of whether it comes
from one client device with multiple users or from a
single user establishing multiple connections. The
number of concurrent sessions is counted from the
client device, rather than at the Server level,
regardless of whether the connection is directly to
the Informix instance or indirectly; whether it is
through application servers, persistent connectivity
layers, connection multiplexers / concentrators, or
any other technology inserted between the actual
end-user and the Informix instance.
Limited Use Socket

(LU Socket)
Available only on Informix Growth Edition, this

allows for licensing on a physical socket potentially
containing multiple cores. A LU Socket license is
required for each active processor socket. This
licensing metric can only be used on a physical
server with no more than four physical sockets. You
may purchase licenses for up to four physical
sockets and use up to 16 cores.
Install
An Install is an installed copy of an informix

product on a physical server (or partition thereof) or
in a virtual machine image. For example, if a
physical server is segmented into partitions,
whether logical (also know as LPARs) or physical,
each partition containing Informix is considered a
separate IBM Informix Install for licensing
purposes and restrictions. The concept of an
Install applies to the licensing limits specified for
all Informix Editions.
Informix Choice Edition

Informix Choice Edition is ideal for small to medium size business. This edition is
available for both Microsoft Windows and Apple Macintosh operating systems.
This edition is limited to a total of eight cores over a maximum of two sockets and
7884ch01.fm
8GB of RAM operating. The Informix Choice Edition includes limited Enterprise
Replication (ER) clustering with 2-root nodes to send or receive data updates
within the cluster. This edition also provides limited High Availability (H/A) cluster
functionality; along with the Primary server, you can have one secondary node,
either an HDR secondary or RSS secondary. The H/A cluster secondary node
can be used for SQL operations. This edition does not support use of the Shared
Disk secondary (SD secondary) node type.
Informix Growth Edition

The Informix Growth Edition is available on all platforms, Linux, UNIX, Windows,
and Apple Macintosh. The Growth Edition is ideal for mid-sized companies or
departmental servers in an enterprise deployment. It can be deployed on up to
16 cores over a maximum of 4 sockets and 16GB of RAM operating from the
same Install. License options include Authorized User Single Install, Concurrent
Session, PVU, and Limited Use Socket (LU Socket) metrics. LU Socket enables
licensing by physical processor socket and is limited to physical servers with no
more than four physical processor sockets.
Informix Growth Edition gives you additional database functionality including
unlimited ER cluster nodes of any type for sending or receiving data updates
within the cluster. From a licensing perspective, since ER nodes are stand-alone,
each ER node must be fully licensed. Informix Growth edition supports up to two
(2) H/A cluster secondary nodes of any type. As long as the secondary node(s)
are only functioning as a backup secondary, they can be deployed without
charge. However, if you use a secondary node for SQL operations (read or
write), the secondary node must be fully licensed.
Informix Ultimate Edition

The IBM Informix Ultimate Edition includes nearly all the Informix features and
functionality with unlimited scalability required for the highest OLTP and
warehousing performance.
The operating systems the Informix Ultimate Edition supports include AIX, HP,
Unix, Linux, Macintosh, Sun Solaris, and Windows. This edition can be licensed
by PVU, Concurrent Session, or Authorized User Single Install metrics.
With this edition, full H/A cluster and ER functionality is available including
unlimited ER nodes and all H/A cluster secondary instance types. From a
licensing perspective, since ER nodes are stand-alone, each ER node must be
fully licensed but H/A secondary nodes can be deployed without charge if they
are only functioning as a backup secondary. If you use any secondary node for
SQL operations (read or write), the secondary node must be fully licensed.
This edition offers the following Add-on:
The Geodetic and Excalibur DataBlades.
7884ch01.fm
The Storage Optimization Feature was released with Informix version11.5

xC4 and following. The Storage Optimization Feature provides data
compression to dramatically reduce data storage and backup/recovery costs
and administration. The reduced data footprint also provides a significant
increase in performance for data retrieval.
Extended Parallel Server

The IBM Informix Extended Parallel Server (XPS) is a high-end database that is
typically used for scalable data warehousing with fast data loading and
comprehensive data management. It is designed for a broad range of enterprises
that require complex, query-intensive analytical applications.
The IBM Informix Extended Parallel Server usage examples include:
Reliable data manipulation

Ad hoc queries from data warehouses
A combined data warehouse and data mart
Rapid and concurrent data loading and query execution
IBM Informix Extended Parallel Server provides full parallel query processing,
while being able to use hardware resources to deliver mainframe-caliber
scalability, manageability, and performance with minimal OS and administrative
overhead. For more information, see Database Strategies: Using Informix XPS
and DB2 Universal Database, SG24-6437.
Informix OnLine
IBM Informix OnLine is an easy to use, embeddable relational database server
for low-to-medium workloads.
It has less features and functionality than the Informix Ultimate Server, but it
scales well while providing online transaction processing support and the
assurance of data integrity. Informix OnLine has rich multimedia data
management capabilities, supporting the storage of a wide range of media such
as documents, images, and audio by way of text and byte columns. Also it is not
designed to handle extended data types or replication.
This server supports a wide variety of application development tools, along with a
large number of other third-party tools, through support for the ODBC and JDBC
industry standards for client connectivity
There are three edition options for this server:
Informix OnLine Extended Edition 5.20
This edition option is a full-featured, easy-to-use SQL database with low
administrative overhead. It contains two popular Informix products, Informix
OnLine and Informix STAR, and provides distributed computing with a proven
7884ch01.fm
database server. It allows the Rapid Application Development tools to use

pipes to communicate with OnLine Extended Edition 5.20 servers running on
the same machine, instead of having to use a TCP/IP loopback connection
with Informix Star or Informix Net. It supports greater than 2GB chunk offsets,
to utilize the entire capacity of the disk drive without partitioning it into smaller
logical devices and employing logical volume managers.
The operating systems this edition supports include AIX, HP Unix, and Sun
Solaris.
Informix OnLine Extended Edition for Linux
This edition is same as the Informix OnLine Extended Edition 5.20 but is but
only available for Linux.
Informix OnLine Personal Edition
This is the single-user product version, available only on Linux. It provides the
same functionality as IBM Informix OnLine Extended Edition, at an
economical cost. It enables you to quickly become familiar with the
ease-of-use and multimedia capability of this proven relational database
management system.
Standard Engine
IBM Informix Standard Engine (SE) is an embeddable database server that runs
on UNIX, Linux and Windows. It provides an ideal solution for developing small to
medium-sized applications that need the power of SQL without database
administration requirements (low-maintenance, high-reliability). For limited scale
databases, it delivers excellent performance, adheres to data consistency
standards, and still provides client/server capabilities. It can be seamlessly
integrated with Informix application development tools and third-party
development tools compliant with the ODBC and JDBC standards.
The operating systems this edition supports include AIX, HP Unix, Linux, Sun
Solaris, and Windows.
1.2 Informix tools for developers

There is a significant list of tools available from Informix for developers. We focus
on the available application programming interface (API) options and language
possibilities. There are also other utilities and tools provided with the engine that
promote further ease of use. The 4GL developers edition includes the ACE
Report Writer, which can be used to quickly generate forms and reports. The
High Performance Loader, onload, onunload, and External table are bulk load
and unload utilities that provide fast flat file movements. Other tools include I-Spy
for auditing, MaxConnect, and the Open Administration Tool (OAT) for remote
7884ch01.fm
server administration and SQL. For more information on these or other tools,
consult with a sales representative or visit the Informix support website at
http://www-947.ibm.com/support/entry/portal/Overview/Software/Information_Manag
ement/Informix_Product_Family
1.2.1 I-Connect
I-Connect is the runtime version of Client SDK that comes with the Server
engine. You use Client SDK and related tools to develop your application. Once it
is ready for use in production, I-Connect is the tool used to deploy the
application. It is supplied with the Server engine software, and provides the
connectivity and runtime libraries which permit interaction between the engine
and the application. While some I-Connect development can be done without
Client SDK, the great majority should be handled through the Client SDK API.
1.2.2 Client SDK

IBM Informix Client Software Development Kit is a package of several APIs that
are optimized for developing applications for IBM Informix servers. Informix Client
SDK allows developers to write applications in the language they prefer and to
build applications that can access multiple IBM Informix databases. Here we
discuss the API packages included in Client SDK.
ODBC
Open Database Connectivity (ODBC) is a specification for a database API. It is
based on the Call Level Interface specifications from X/Open and the
International Standards Organization and International Electromechanical
Commission (ISO/IEC). ODBC supports SQL statements with a library of C
functions. An application calls these functions to implement ODBC functionality.
ODBC applications can perform the following operations:
Connect to and disconnect from data sources.
Retrieve information about data sources.
Retrieve information about the IBM Informix ODBC Driver.
Set and retrieve IBM Informix ODBC Driver options.
Prepare and send SQL statements.
Retrieve SQL results and process the results dynamically.
Retrieve information about SQL results and process the information
dynamically.
7884ch01.fm
ODBC lets you allocate storage before or after the SQL results are available.This
feature lets you determine the results and the action to take without the
limitations that predefined data structures impose. ODBC does not require a
preprocessor to compile an application program. ODBC supports Secure
Sockets Layer (SSL) connections. For information on using the SSL protocol, see
the Secure Sockets Layer Communication Protocol section of the IBM Informix
Version 11.5 Security Guide, SC23-7754-04.
The additional features and capabilities that Informix ODBC supports include:
Microsoft Transaction Server (MTS) environment. For more information on
MTS, see the MTS sections in the IBM Informix ODBC Driver Programmers
Manual, SC23-9423-03.
ODBC can handle extended data types such as:
Collection (LIST, MULTISET, and SET)

Distinct
Opaque (fixed, unnamed)
Row (named, unnamed)
Smart large objects (BLOB andCLOB)
Client functions that support extended data types
Long identifiers
Global Language Support (GLS) data types (NCHAR,NVARCHAR).
Support for Unicode and XA.
IPv6 internet protocol.
ODBC with the IBM Informix ODBC Driver can include the following
components:
Driver manager
An application can link to a driver manager that links to the driver specified by
the data source. The driver manager also checks parameters and transitions.
You can purchase the ODBC Driver Manager from a third-party vendor for
most UNIX platforms. On Microsoft Windows platforms, the ODBC Driver
Manager is a part of the Operating System.
IBM Informix ODBC Driver
This driver provides an interface to the Informix database server. Applications
can use the driver in the following configurations:
To link to the ODBC driver manager
To link to the Driver Manager Replacement & the driver
To link to the driver directly
Data sources
7884ch01.fm
The driver provides access to the following data sources:

Database management systems (DBMS) and database servers
Databases
Operating systems and network software required for accessing the
database
JDBC and SQLJ

JDBC is an application programming interface (API) that Java applications use to
access relational databases. IBM Informix database systems can be supported
by one of two APIs for client applications and applets written in Java:
IBM Data Server Driver
You can use the IBM Data Server Driver (also known as the IBM common
client driver) with either DB2 or Informix. This JDBC driver lets you write Java
applications to access a local Informix Server, DB2 data, or any remote
relational data on a server that supports DRDA. Using this API, you can
access these database systems using JDBC, SQLJ or pureQuery. SQLJ
provides support for embedded static SQL in Java applications.
Note: The IBM Data Server Driver is introduced with Client SDK beginning
with Client SDK 3.50.XC7. At this writing, the Client SDK distributed with
Informix database server is one (1) version less than the server version. As
an example, CSDK 3.50.XC6 is the version distributed with Server edition
11.50.XC7
The IBM Data Server Driver for JDBC and SQLJ is a single driver that
includes JDBC Type 2 and JDBC Type 4 behavior. For connections to IBM
Informix databases, only Type 4 behavior is supported. IBM Data Server
Driver for JDBC and SQLJ Type 4 driver behavior is also referred to as IBM
Data Server Driver for JDBC and SQLJ type 4 connectivity. For more
information on these APIs, see the Java discussions at
http://publib.boulder.ibm.com/infocenter/db2luw/v9r5/index.jsp?topic=/com.i
bm.db2.luw.apdv.java.doc/doc/rjvjdb2o.html
IBM Informix JDBC driver

IBM Informix JDBC driver is a native-protocol, pure- Java driver (JDBC type
4). This means that when you use IBM Informix JDBC Driver in a Java
program that uses the JDBC API to connect to an IBM Informix database,
your session connects directly to the database or database server, without a
middle tier.
When deciding which JDBC interface to use, it is important to understand the
differences in these two APIs. If there is a need to work with extended data types
10
7884ch01.fm
and features unique to IBM Informix databases, you will want to use the IBM
Informix JDBC driver.
To support DataSource objects, connection pooling, and distributed transactions,
IBM Informix JDBC Driver provides classes that implement interfaces and
classes described in the JDBC 3.0 API from Sun Microsystems.
Informix classes implementing Java interfaces

Table 1-2 lists the Java interfaces and classes and the Informix classes that
implement them.
Table 1-2 Java interfaces and classes
JDBC interface class
Informix class
java.io.Serializable
com.informix.jdbcx.IfxCoreDataSource
java.sql.Connection
com.informix.jdbc.IfmxConnection
javax.sql.ConnectionEventListener
com.informix.jdbcx.IfxConnectionEventListener
javax.sql.ConnectionPoolDataSource
com.informix.jdbcx.IfxConnectionPoolDataSource
javax.sql.DataSource
com.informix.jdbcx.IfxDataSource
javax.sql.PooledConnection
com.informix.jdbcx.IfxPooledConnection
javax.sql.XADataSource
com.informix.jdbcx.IfxXADataSource
java.sql.ParameterMetaData
com.informix.jdbc.IfxParameterMetaData
IBM Informix JDBC Driver, Version 3.0, and later implements the updateXXX()
methods defined in the ResultSet interface by the JDBC 3.0 specification. These
methods, such as updateClob, are further defined in the J2SDK 1.4.x API (and
later versions) and require that the ResultSet object can be updated. The
updateXXX methods allow rows to be updated using Java variables and objects
and extend to include additional JDBC types. These methods update JDBC
types implemented with locators, not the data designated by the locators.
Informix classes Java specification

To support the Informix implementation of SQL statements and data types, the
IBM Informix JDBC Driver provides classes that extend the JDBC 3.0 API. For
information on the Java classes and the Informix classes that application
programs can use to extend them, see the IBM Informix JDBC Driver
Programmer's Guide, v3.50, SC23-9421-02.
11
7884ch01.fm
Because IBM Informix have extended functionality, extra data types and smart
large objects, several Informix classes provide support for functionality that is not
present in the JDBC 3.0 specification. Table 1-3 lists these classes.
Table 1-3 Informix classes beyond the Java Specification
JDBC interface or class
Informix class
Provides support for
java.lang.object
UDTManager
Deploying opaque data types in the

database server
java.lang.object
UDTMetaData
Deploying opaque data types in the

database server
java.lang.object
UDRManager
Deploying user defined routines in

the database server
java.lang.object
UDRMetaData
Deploying user defined routines in

the database server
In releases prior to JDK Version 1.4, the UDTManager and UDRManager helper
classes included in ifxtools.jar were not accessible from a packaged class. As of
IBM Informix JDBC Driver 2.21.JC3, all these classes are in the udtudrmgr
package. For backwards compatibility, unpackaged versions of these classes are
also included. To access a packaged class, use the following import statements
in your program:
Import udtudrmgr.UDTManager;
Import udtudrmgr.UDRManager;
OLEDB
Microsoft OLE DB is a specification for a set of data access interfaces designed
to enable a variety of data stores to work together seamlessly. OLE DB
components are: data providers, data consumers, and service components.
Each data provider makes their data available to consumers using data in a
tabular form through virtual tables. Data consumers use the OLE DB interfaces to
access data. You can use the IBM Informix OLE DB Provider to enable client
applications, such as ActiveX Data Object (ADO) applications and Web pages, to
access data on an Informix server. Detailed information about the characteristics
of the IBM Informix OLE DB Provider is available in the IBM Informix OLE DB
Provider Programmer's Guide, Version 3.50, SC23-9424-00. The IBM OLEDB
provider works with any IBM Informix Server version greater than or equal to 7.3
(7.3, 8.x, 9.x,10.x or 11.x). For information about OLE DB architecture and
programming, go to the Microsoft website http://www.microsoft.com and search
for Introduction to OLE DB.
12
7884ch01.fm
ESQL/C
Informix ESQL/C is an application programming interface (API) that enables you
to embed SQL statements directly into a C program. It does this by means of the
Informix ESQL/C preprocessor, esql, which converts each SQL statement and
IBM Informix-specific code to C-language source code, and invokes the C
compiler to compile it.
The components of Informix ESQL/C include the following:
The Informix ESQL/C libraries of C functions, which provide access to the
database server, and all the Informix data types.
The Informix ESQL/C header files, which define the data structures,
constants, and macros useful to an Informix ESQL/C program.
The esql command, which processes the Informix ESQL/C source code to
create a C source file that it passes to the C compiler.
The finderr utility on the UNIX system and the Informix Error Messages
Windows-based utility that provides information on IBM Informix specific error
messages.
GLS locale and code set conversion files for locale specific information.
1.2.3 4GL
IBM Informix 4GL is a fourth-generation application development and production
environment that provides power and flexibility without the need for
third-generation languages such as C. The two package options, available for all
UNIX and Linux operating systems, include:
Informix 4GL Rapid Development System and Informix 4GL Interactive
Debugger. These provide a pseudo-compiled development environment for
applications.
Informix 4GL C Compiler. This provides the components needed to develop
and compile a high-performance application for a production environment.
1.2.4 Ruby on Rails

IBM Informix supports database access for client applications written in the
popular and dynamic open source programming language called Ruby. To
accomplish this, the developer must use a Ruby driver and a Rails adapter for the
standard Ruby framework Rails. IBM Informix offers two methods for using Ruby
and Ruby on Rails:
Ruby driver and Rails adapter for IBM Data Servers: This is supported on
Informix, Version 11.10 or later, using the DRDA protocol. This option requires
13
7884ch01.fm
the IBM Data Server Driver for ODBC and the Call Level Interface (CLI) which
are available as part of the Informix Client SDK.
Ruby/Informix and Rails Informix_adapter: This is specific for Informix
database server and works with Informix database connections. Supports all
versions of the IBM Informix database servers and requires the IBM Informix
Client SDK libraries for the communication with the database server.
You can download the both packages from the Rubyforge website at
http://rubyforge.org/projects/ruby-informix/.
1.2.5 Informix DataBlade Developers Kit

The IBM Informix DataBlade Developers Kit (DBDK) is an aid for developing
DataBlade modules. The kit runs on Windows and generates much of the code
you need for a DataBlade. DataBlades can be developed without a DBDK on
operating systems other than Windows; however, it is noteworthy that the
procedures for setting up the development environment on a UNIX system are
complex without help from DBDK.
The DataBlade Developers Kit provides several graphical user interfaces for
creating and working with Informix DataBlade modules.
BladeSmith: A tool for organizing a DataBlade module development project.
You can use BladeSmith to create a project and define the objects, such as
data types and routines, that belong to the DataBlade module. BladeSmith
generates source files, header files, make files, functional test files, SQL
scripts, messages files, and packaging files.
DBDK Visual C++ Add-In and IfxQuery: Tools for debugging a DataBlade
module using Microsoft Visual C++ on Windows. The add-in automates many
of the debugging tasks and calls the IfxQuery tool to run unit tests for
DataBlade module routines.
BladePack: A tool for creating a DataBlade module package. BladePack can
create a simple directory tree containing files to be installed or an installation
that includes an interactive user interface.
BladeManager: This is a command line tool on UNIX (or Windows) that
comes with the Informix Server. It is a utility that is needed for registering and
unregistering DataBlade modules in Informix databases.
1.2.6 Informix Spatial DataBlade

Many of the IBM Informix database servers support DataBlades. Among all these
DataBlades, there are a couple that are fairly popular because of their usefulness
and flexibility. We introduce one of these here to open your interest into an area
14
7884ch01.fm
that you may not have considered. Location based data is one of several features
and benefits that can be found in the IBM Informix Spatial DataBlade.
The Spatial DataBlade can transform both traditional and location-based data
into essential information:
It expands IBM Informix Server to provide SQL-based spatial data types and
functions that can be used directly through standard SQL queries or with
client-side Geographic Information Systems (GIS) software.
It delivers innovative spatial technology via a convenient no-charge download.
It generates vital business intelligence for a competitive edge.
It maximizes spatial data capabilities to enable critical business decisions.
It works in an enterprise replication environment which includes spatial data
types.
It enables organizations to manage complex geospatial information alongside
traditional data, without sacrificing the efficiency of the relational database
model.
It includes R-tree indexing:
R-tree is built into the database kernel and works directly with extended data
types to enable proper geospatial data management. Unlike standard indices,
the R-tree does not divide space into a full coverage of non- overlapping,
adjacent cells. Instead, it uses data partitioning, where each object is
automatically represented by a bounding box that is entirely determined by its
own shape. These bounding boxes may overlap and do not need to cover the
entire space. As a result there is no need to know the spatial extent of the
data in advance.
1.2.7 PHP on Informix

PHP, the powerful and popular server-side scripting language for creating Web
content, has become an important platform for Informix development. The
Informix Open Admin Tool (OAT), which provides the ability to administer multiple
Informix instances from a single location, is written entirely in the PHP language.
Informix supports database access for client applications written in the PHP
programming language by using a PDO (PHP Data Object) extension that
functions as a database extraction layer. The primary PHP driver available for
Informix is called PDO_IBM, and is supported on Informix Version 11.10, and
following. The other available PHP driver is PDO_INFORMIX. It is the older of
these two, and is the driver used for the Informix Open Admin Tool. You can find
the drivers for PHP on the PECL for PHP website at
http://pecl.php.net/package/PDO_IBM
15
7884ch01.fm
1.3 Informix overview

This section provides a description of the database architecture from a developer
perspective. There are entire manuals to describe Server side functions and
administration. We limit our overview to things that the developer should
consider. For more information about the details of the IBM Informix server
engine, consult the IBM Informix Administrators Guide at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/index.jsp?topic=/com.ibm.
adref.doc/adref.htm
1.3.1 Architecture overview

An understanding of how things work on the Informix database server is
beneficial for application development.
Client/Server architecture
IBM Informix Servers are based on a multi-threaded client-server architecture.
Whichever edition you use, the server can support a large number of client
connections while users are accessing data. A quick overview provides insights
into what aspects influence development, and what aspects may need DBA
attention.
The IBM Informix server edition environments consist of several parts that are
usually not directly visible to the developer. These parts include shared memory,
disk storage, and virtual processors (VPs). The virtual processors are divided up
into VP classes to manage specific tasks, such as SQL query processing (CPU),
physical and logical log processes (PIO and LIO), network connection processes
(NET), engine administration (ADM), miscellaneous (MSC), and disk I/O (AIO
and KAIO) processes. On UNIX and Apple Macintosh systems, VPs are visible
as virtual processes. On Windows environments, the VPs are seen as
threads.The program execution process is visible as oninit. Virtual processes
communicate by way of shared memory structures known as mutexes (mutually
exclusive). Since shared memory is also used to move database data in buffers,
it helps to know that some memory buffers are reserved for engine processing
and some buffers allocated (and managed dynamically) for data processing.
Network protocols and other connection types

Users can connect to the IBM Informix Server using a network connection,
shared memory, pipes, DRDA, or multiplexed connections. Shared memory and
pipe connections are only available on the same machine as the server instance.
Network connections are the most popular, since they can connect different
machines, and will use either a TCPIP or a socket connection, depending on the
16
7884ch01.fm
operating system. Whichever method you choose, it is important to minimize the

overhead costs associated with opening and closing connections. A developer
should keep sessions open only as long as it is prudent, and re-use connections
whenever possible.
Authentication and user connections

By default, authentication is determined for users connecting to a database by
the existence of a user name in the operating system environment. Other
authentication methods include lightweight directory access protocol (LDAP), or
single sign-on (SSO) through Kerberos.
If you need a secured authentication method, the IBM Informix Server also has
two secured connection methods by way of the encryppted Communication
Support Module (CSM) and single password CSM. You can use one or the other,
but not both at the same time. Neither the encrypted CSM not the single
password CSM method work over multiplexed connections. Enterprise
Replication (ER) and High Availability (HA) replication support encryption, but not
with CSM. Secure Sockets Layer (SSL) communications are an alternative to the
Informix-specific encryption CSMs. SSL encrypts data in end-to-end and secures
TCP/IP and Distributed Relational Database Architecture (DRDA) connections
between two points over a network. For setup information, see the
developerWorks article Protect your data with Secure Sockets Layer support in
Informix Dynamic Server, Part 1: Setting up SSL support in IDS at
http://www.ibm.com/developerworks/data/library/techarticle/dm-0912securesockets
1/index.html
Programming and user process considerations

Any SQL or stored procedure call is parsed and optimized when the engine first
receives the statement. Parsing and optimization methods require a little extra
running time. If you can minimize this time, you can speed up query execution.
Some factors that can speed up execution time are:
If you cannot optimize the query syntax in advance, have the DBA set the
onconfig parameter OPTCOMPIND to 2, in order to use distribution data
(determined by Update statistics usage).
If you can optimize all of your query syntax expressions in advance, you may
want your DBA to consider using OPTCOMPIND 1 or 2.
If the same statement is executed multiple times, it is placed in a procedure or
dictionary cache. Cache statements run faster.
If the statement is a prepared statement, the statement can be shared by
multiple users and will not need reparsing or optimizing.
When a statement execution thread is ready to run, it will be scheduled to run
on a CPU VP. On a machine with multiple processors where each CPU VP
17
7884ch01.fm
maps to a physical processor, multiple statements can run concurrently on

multiple CPU VPs. Take advantage of multiple CPUs whenever possible by
means of concurrent queries, parallel subqueries, and the degree of
parallelism whenever possible.
When a statement runs on the CPU VP, it is given a brief block of time to run.
At the end of the time unit, it is placed into a VP wait cycle and the next
statement thread gets a time block on the VP queue. This gives all executing
processes an opportunity to make progress, regardless of the number of
processes running.
Physical and logical logging

As each of the various CPU processes are running, the database engine is
keeping track of a number of operations. When data is retrieved from disk, the
data image is placed in a physical log buffer. Change process information is
placed in the logical log. When image data is affected by SQL processing, the
physical image is periodically modified and written back to disk. During this
process the physical log image and logical log information are also written into
log records. Depending on the commit protocols and isolation levels, this
information is available to the DBA through recovery operations if things should
suddenly go wrong. Automatic recovery features built into the engine can also be
used if the engine abruptly goes offline due to a power failure or crash.
Commit protocols
If transactions are made against a database that uses unbuffered logging, the
records in the logical-log buffer are guaranteed to be written to disk during
commit processing. When control returns to the application after the COMMIT
statement (and before the PREPARE statement for distributed transactions), the
logical-log records are on the disk. The database server flushes the records as
soon as any transaction in the buffer is committed (that is, a commit record is
written to the logical-log buffer).
If transactions are made against a database that uses buffered logging, the
records are held (buffered) in the logical-log buffer for as long as possible. They
are not flushed from the logical-log buffer in shared memory to the logical log on
disk until one of the following situations occurs:
The buffer is full.

A commit on a database with unbuffered logging flushes the buffer.
A checkpoint occurs.
The connection is closed.
If you use buffered logging and a failure occurs, you cannot expect the database
server to recover the transactions that were in the logical-log buffer when the
failure occurred. Thus, you could lose some committed transactions. In return for
18
7884ch01.fm
this risk of using buffered logging, performance during alterations improves

slightly. Only consider using buffered logging:
If the database is updated frequently (speed of updating is important).
If the application that is performing the transaction can continue as it is, and
you decide that the price (in time and effort) of returning the database to a
consistent state by either removing the effects or reapplying the transaction is
too high or you can re-create the updates in the event of failure.
If a transaction failure does occur, you can simply choose to leave the
database in its inconsistent state, if the transaction does not significantly
affect database data.
As you consider whether to use a buffered or unbuffered logging method,
remember that no automatic process or utility can perform a rollback of a
committed transaction or can commit part of a transaction that has been rolled
back. Without detailed knowledge of the application, messages are not enough
to determine what has happened. Based on your knowledge of your application
and your system, you may need to help the DBA determine when to roll back or
to follow though on interrupted transactions. For more information on this topic,
see the Guide to SQL:Tutorial section on Interrupted modifications at
sqlt.doc/ids_sqt_277.htm
1.3.2 Informix developer environment

The environment references you need will have some variations, depending on
whether you use the IBM common client API or the native Informix Client SDK,
and which operating system you use. The environment settings must include
identifying references to the components. This should include:
Setting the INFORMIXDIR environment variable to the installation directory of
the software install location.
Setting up the Informix server information (INFORMIXSERVER)
Adding the $INFORMIXDIR/bin to the PATH environment variable
On Windows, you only have to set the Informix database server information. This
is accomplished with the setnet32.exe program. On UNIX-type platforms, you
may want to set the environment variables so they are set in your user profile at
login, or a setup script. You also need to reference the sqlhosts file to identify the
Informix server information. In the chapters that follow, we will explain how to
install the server, and provide specific details for the relevant API environment
details.
19
7884ch01.fm
Before you begin to select an API or library, you should be aware that some
operating systems are not available for some of the tools, and server functionality
may not be accessible with some of the APIs. Here are some API selection
considerations:
In regard to IBM Informix Client SDK:
It is optimized for IBM Informix database servers.
It works with PHP.
It directly interfaces with the Informix server, no middle tier required.
It supports all the Informix data types, extended data types, and smart
BLOBs.
In regard to the IBM Data Server Client:
It is optimized for compatibility and easy development with Informix and
DB2 IBM database servers.
It works with PHP and Ruby.
No interface is available for Apple Mac machines.
The Data Server Client package includes support for C and Fortran
development.
There is no DataBlade API.
1.3.3 Informix capabilities

The IBM Informix database servers provide a broad spectrum of features that
allow resilience in response to fast changing systems and applications in a
modern business environment. If you are aware of what the product can already
do before you begin to develop, it aids and enables you to support applications
and easily grow to meet new demands. Here are the capabilities in the IBM
Informix:
Resilient:
High availability and fast recovery are standard aspects of an on-demand
environment. Informix database server engines have a variety of ways to
configure the instance, so there is little chance of the engine being
unavailable at critical points in time.
Reliable:
In addition to availability, there is a need for continuity. Informix has options to
protect a server environment, such as backup servers, full independent
copies of the processing environment for failover, and workload balancing at
alternate locations around the world.
20
7884ch01.fm
Available:
There is an Informix solution to fit any situation that requires availability:
Continuous availability feature
High Availability Replication with allowance for multiple secondary servers
Secure:
Informix supports open, industry-standard security mechanisms such as
roles, password-based authentication, and RDBMS schema authorizations.
These open standards ensure flexibility and security with easy validation and
verification. Column-level encryption and Pluggable Authentication Modules
(PAM) are also available. The Advanced Access Control Feature offers cell-,
column- and row-level label-based access control (LBAC). This means
access to data can be controlled down to the individual cell level.
Adaptable:
Informix is adaptable from the server side and from the client side. The engine
component can be stripped down, embedded, and will run with very little user
intervention. Using the developer tools which are discussed in the chapters
which follow, a developer can provide customized deployment by way of client
applications, as well as server-side processes such as stored procedures,
multiple triggers on tables and views, user defined functions, and DataBlades
Fast:
Informix is known for fast OLTP performance. Application performance is
helped by capabilities such as committed isolation level and non-blocking
checkpoints which provide maximum concurrency. Direct I/O calls to file
systems can result in performance similar to raw device I/O. SQL
performance can be improved through techniques or configuration options
which will redirect or focus the engines optimizer decisions. This can be done
by way of optimizer directives in the SQL, or by way of automated update
statistics collection which will inform the method used for running the query.
When the DBA and the developer both focus on performance, it will enable
things to run more smoothly and reduce infrastructure costs.
Flexible:
There are a number of application programmer interfaces (APIs) available,
both as specific programming language supplements and as interfaces that
extend the architecture of the server instance. In Informix 11 and following,
the Web Feature Service API allows developers to use location-based
services or location-enabled IT services. It is implemented in the Open
GeoSpatial Consortium Web Feature Service (OGC WFS) API. This API will
also interact with location-based data provided by the IBM Informix Spatial
and Geodetic DataBlade modules. There is a significant amount of
extensibility when using DataBlade technology.
21
7884ch01.fm
Hidden, behind-the- scenes tasking:

Business systems and applications need to work with minimum invasions into
the operations of a business environment. From the server side, applications
can run in an automated mode and perform self-maintenance. From an
application perspective, database administrative tasks can be controlled and
run from inside an application. The SQL Administration API allows you to do
tasks such as space management, monitoring and manipulating memory,
running scheduled tasks, and monitoring user sessions without developer
intervention after they are setup and running.
Customizable install footprint:
The installation of Informix can be automated, and the installation footprint
can be customized using the Deployment Wizard. In this way you can limit the
installation only to the data server functionality you need -- to reduce the size
and costs of a solution for your software deployment.
Affordable, reduced complexity:
When IBM adds features, or the developer adds features, the first impression
is that it will be more complex. With the features described up to this point, it is
a surprise to discover that administration can actually be reduced more than
ever before, by means of an application or one of the APIs and GUIs
explained in this book.
22
7884ch02.fm
Chapter 2.
Setting up an Informix
development environment
This chapter describes how to set up an environment for Informix application
development including the Installation and configuration of both Informix Servers
and Clients. We focus our discussion on the Informix Client products selection,
installation, and configuration. After the end of this chapter, you should be able
to:
Decide which Client product is suitable for your application.
Install and configure all Client products that can connect to Informix.
Know any special consideration for each Client in terms of connectivity.
In this chapter, Informix database server, Informix Server, and Informix all mean
the same and are used interchangeably. Also, Client products, Client, and
Informix Clients mean the same.
23
7884ch02.fm
2.1 Informix deployment and setup

In this section we describe the various methods of installing Informix Server. We
provide how you can plan for the installation depending upon your specific
requirement. We focus the discussion on the configuration required at the
database server side to enable the Informix Clients connectivity.
2.1.1 Installation planning

In general, an application developer does not involve in the Informix Server
product planning and installation. However, you should communicate your
application requirements with the DBA who plans, installs, and configures the
Informix database server for you, for example, the need for a smart blob space, a
reduced footprint of the server or DataBlades. The Informix installer provides
options to install only a small footprint of the server where you can add or remove
the features as you like. We discuss more about DataBlades in 10.3,
DataBlades and bladelets on page 346.
2.1.2 Informix Server installation

IBM provides the Informix Server editions, described in 1.1, Server options on
page 2 in zip, tar, or dmg format depends on the platforms. You can install
Informix Servers in many ways. The common ones are console, GUI, and silent
modes of installation. Since this book is more focused on the Informix Client
products for the application development, we describe only the silent mode of
Server installation on an UNIX platform. For other methods and platforms, refer
to the individual installation PDF file that comes with the product.
To perform silent install in UNIX perform, use the following steps:
1. Extract the compressed tar file into any directory using the following
command:
tar -xvpf IIF.11.50.tar
2. Create the installation ini file that specifies the installation options. You can
use the sample installation ini file, server.ini,as a template. The server.ini
file is located in the directory used to extract the server media package. Edit
the following in the server.ini and save it in a new name, say myserver.ini:
Change -P installLocation="/opt/IBM/informix" to the location of
directory where you want to install the product.
Change -G licenseAccepted= option to true to accept the software
license, otherwise, installation will not proceed.
24
7884ch02.fm
3. Start the silent install with the command:

./ids_install -silent -options myserver.ini
Performing these steps installs the Informix server in your specified directory
along with a demonstration instance by name demo_on.
For more information about Informix installation and creating new instance, refer
to the installation guide PDF that comes with the media file.
2.1.3 Configuration
The connectivity information allows a client application to connect to any IBM
Informix database server on the network. The connectivity data for a particular
database server includes the following:
The host name of the computer or node on which the database server runs
The type of connection that an application can use to connect to it
The service name or port number used by the database server
Database instance (also called database server name)
Even if the client application and the database server are on the same computer
or node, you may need to specify connectivity information for your client
application.
Connectivity on UNIX
On UNIX, the sqlhosts configuration file contains the connectivity information.
By default this file resides in the $INFORMIXDIR/etc directory. The
INFORMIXSQLHOSTS environment variable can be used to specify an alternative
location for the connectivity information. Edit this file according to the protocol
you are using or the port number to which your client application will connect. A
typical sqlhosts file contains the following fields:
Server Protocol Hostname Service/port Options
Here is an example:
demo_on onipcshm on_hostname on_servername k=0
demo_se seipcpip se_hostname sqlexec
Each line in the sqlhost file normally defines an Informix database server:
Server: Specifies the name of the IBM Informix database server. This value
normally correspond to the value in the INFORMIXSERVER environment
variable.
Protocol: Specifies the protocol used for the communication with the database
server. This must be the same protocol as used by the database server.
Chapter 2. Setting up an Informix development environment
25
7884ch02.fm
Hostname: Specifies the machine where the Informix database server is

running.
Service/port: Specifies the tcp port or service name, which must be defined in
the /etc/services file, used by the Informix database server to accept incoming
connections.
Options: Specify additional options for the communication, such as buffer size
or connection redirection rules.
APIs such as ESQL/C or ODBC relay on the sqlhosts file to obtain the
connection parameters for the database server.
Example 2-1 shows how a ESQL/C application connects to an Informix database
server.
Example 2-1 UNIX connection example
informix@kodiak:/work$ cat $INFORMIXDIR/etc/sqlhosts
#server
protocol
hostname
service/port
demo_on
onsoctcp
kodiak
demo_on_tcp
informix@kodiak:/work$ grep demo_on_tcp /etc/services
demo_on_tcp
9089/tcp
# Service for demo_on IDS server
informix@kodiak:/work$ cat connect.ec
#include <stdio.h>
int main()
{
EXEC SQL CONNECT TO 'stores_demo@demo_on';
printf("Connected\n");
}
informix@kodiak:/work$ esql connect.ec -o connect
informix@kodiak:/work$ ./connect
Connected
informix@kodiak:/work$
The definition for the demo_on Informix server specifies the communication
protocol used as onsoctcp; the machine where the database server is running is
kodiak; and the TCP port used by the server is demo_on_tcp. The service name
is defined as port 9089 in the /etc/services configuration file.
The ESQL/C example connects to the server and opens the stores_demo
database using the instruction:
EXEC SQL CONNECT TO 'stores_demo@demo_on';
For more information about the sqlhosts configuration file refer to the Installing
and Connecting to Client guide at:
26
7884ch02.fm
admin.doc/ids_admin_0158.htm
Connectivity on Windows
You can set the connectivity on Windows using the program setnet32.exe that
resides in $INFORMIXDIR/bin. This utility comes bundle with the Client SDK
package. Use this utility to create or change the server name, protocol, service
name, and host name. You also can use setnet32.exe to set any environment
variable used by the Informix products. Figure 2-1 on page 27 shows the Server
Information tab of setnet32.exe utility on Windows where you can specify the
server name, host IP address, network protocol, and service name.
Figure 2-1 setnet32 Server Information tab
Note: If you specify the name of the service (for example: sqlexec) rather than
the port number, make sure the service name used appears in the Windows
services file located in C:\WINDOWS\system32\drivers\etc
DRDA support
Since version 11.x, Informix supports the DRDA protocol. If you plan to use the
Data Server driver packages to connect to informix database server, you must
enable DRDA support.
27
7884ch02.fm
Windows
Use Setnet32 to configure an Informix instance to use DRDA protocol. Perform
the following steps:
1. Start setnet32 from any command window.
2. Go to Server Information tab (Figure 2-2 on page 28).
3. Select the required instance from the IBM Informix Server field. Our example
is demo_my.
4. Change the name of the Informix server and specify a new name for the
DRDA alias, for example demo_my_drda
5. HostName is pre-filled, if not, enter your host name or IP address.
6. Change the Protocolname from olsoctcp to drsoctcp to make it DRDA
compliant.
7. Change the Service Name specifying the service name or port number the
DRDA alias will use.
8. Click Apply and OK.
Figure 2-2 Configuring DRDA support
After adding all the connection details for the DRDA alias, the configuration file
for the Informix database server must be updated to link the new server definition
28
7884ch02.fm
with the existing Informix instance, in our example we are using demo_my for the
Informix database server and demo_my_drda for the DRDA instance.
The DBSERVERALIASES parameter inside the configuration file is used to
specify alias for a database server. Example 2-2 shows this parameter
Example 2-2 dbservername onconfig sample
...
DBSERVERNAME
demo_my
# Name of default Dynamic Server
#
DBSERVERNAME or DBSERVERALIASES that uses a
# DBSERVERALIASES - The list of up to 32 alternative dbservernames,
DBSERVERALIASES demo_my_drda
# List of alternate dbserver names
#
DBSERVERNAME or DBSERVERALIASES that uses ...
...
The database configuration file is located in etc directory of your Informix

database server. The environment variable ONCONFIG is used to specify the
name of this file. The complete path using environment variables for the
database configuration file is:
On a UNIX platform, $INFORMIXDIR\etc\$ONCONFIG
On a Windows platform, %INFORMIXDIR%\etc\%ONCONFIG%
The Informix database engine must be restarted for the changes to take effect.
After these step is done, your instance is now ready to accept connections as a
DRDA compliant server.
On windows, you can enable the DRDA support for the demonstration instance
during the installation if you use the Custom install. Figure 2-3 shows the
installation panel where you can enable the DRDA support.
29
7884ch02.fm
Figure 2-3 Enable DRDA support for the default instance
Linux and UNIX

On Linux and UNIX systems, you can enable the DRDA support by editing some
of the Informix configuration files:
Edit your sqlhosts file specified in the $INFORMIXSQLHOSTS environment
variable and add the following line:
<drda_name> drsoctcp <hostname> <drda_service_name>
drda_name, is the name for the drda alias

drsoctcp, protocol used by the new alias
hostname, name or IP address of the machine running the Informix server
drda_service_name, name of the TCP service or port number for the drda
instance
Edit the Informix server configuration file located in the etc directory of your
Informix server directory to include the new database definition as an alias for
the existing Informix server.
You can use the environment variables to access this file, for example,
$INFORMIXDIR/etc/$ONCONFIG . This is the file containing a list of configuration
30
7884ch02.fm
parameters for the informix server. Example 2-3 shows the configuration file
used in our database server.
Example 2-3 Linux onconfig sample
informix@irk:/usr3/11.50$ grep DBSERV $INFORMIXDIR/etc/$ONCONFIG
# DBSERVERNAME
- The name of the default database server
# DBSERVERALIASES - The list of up to 32 alternative dbservernames,
#
DBSERVERNAME or DBSERVERALIASES that uses a
DBSERVERNAME demo_my
DBSERVERALIASES demo_my_drda
The Informix server must be restarted for these changes to take effect.
You can find additional information about the sqlhost file at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.admin.doc/i
ds_admin_0158.htm
2.2 Client setup

Informix is a low-administration, easy-to-use, and embeddable database that
supports various application development languages and technologies such as
Java, .NET, PHP, and Web services. Informix can handle XML data easily and
can be extended to handle new data sets using DataBlades.
In this section, we describe how to install and configure the Informix Client
products to support the application development you need.
2.2.1 Client options

IBM offers the following Client products to develop and use an IBM Informix
database server:
IBM Informix Client Software Development Kit (Client SDK), includes several
APIs designed for develop against an Informix database.
IBM Informix Connect (IConnect), runtime version of Client SDK
IBM Informix JDBC driver, Java driver optimized for Informix databases
IBM Data Server Client. Includes all the API tools needed for development
against IBM databases such as DB2 and Informix plus functionality for
database administration and client/server configuration
IBM Data Server Runtime Client, runtime version of the IBM Data Server
Client.
31
7884ch02.fm
IBM Data Server Driver Package, lightweight version of the IBM Data Server
Runtime only including the interfaces and drivers.
IBM Data Server Driver for JDBC and SQLJ, Java driver for IBM Data
Servers.
The Informix Client SDK and IConnect products have distinct functionalities:
Client SDK contains a group of application-programming interfaces (APIs)
that developers can use to write applications for Informix, plus other client
components. Informix Client SDK must be installed on computers that
application programmers will use to write applications.
IConnect contains runtime libraries of the Client SDK APIs, plus other client
components. Install IConnect on computers that end-users will use to connect
to Informix database servers.
Same rule applies to the IBM Data Server products. Packages designed for
development like the IBM Data Server Client should not be used to deploy an
application. Some components such as the Informix JDBC driver do not have a
specific runtime version, so the same product can be used for both cases.
In addition to these products, there are some open source drivers such as PHP
or Ruby driver, which can be used to develop against an IBM Informix database.
Figure 2-4 on page 33 illustrates the options available to develop against an IBM
informix database.
32
7884ch02.fm
Figure 2-4 Available Client products for application development
2.2.2 Setting up IBM Informix Client Software Development Kit

IBM Informix Client Software Development Kit (Client SDK) is a collection of
components for developing and running client applications that connect to the
Informix database server. Client SDK bundles with the following components and
utilities:
ESQL/C with XA support
IBM Informix Object Interface for C++
IBM Informix OLE DB Provider (Windows only)
IBM Informix ODBC Driver with MTS support
IBM Informix .NET Provider (Windows only)
IBM Informix GLS
The Global Security Kit (GSKit)
Password Communication Support Modules (CSM)
Documentation Viewer
The finderr utility on UNIX systems and the Informix Error Messages utility
on Windows systems
The ILogin utility (Windows only)
33
7884ch02.fm
Starting from version 3.50.xC5 the IBM Data Server Driver Package is also
bundled with the Windows version of Informix Client SDK and IConnect.
Informix Client SDK provides access to all the IBM Informix database servers.
Table 2-1 lists the Informix database servers that supports Informix Client SDK.
Table 2-1 Supported database servers
Database server
Versions
IBM Informix
7.x, 9.x,10.x and 11.50
IBM Informix Extended Parallel Server
8.50 and higher
IBM Informix Standard Engine
7.25
IBM Informix Online
5.20 and higher
Some components such as Informix OLE DB Provider do not allow a connection

to an IBM Informix Standard Engine server. Refer to the component release
notes for more information
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.relnotes.do
c/relnotes_csdk350xc7.html
Installation
Before you install Client SDK, perform the following preliminary tasks:
Determine locations:
Media location: This is the directory location where all the media files
reside.
Installation location: This is the location where Installer will place all the
binaries and the configuration files. This directory is represented by the
environment variable $INFORMIXDIR. If $INFORMIXDIR is set, this will be the
default install location.
Java location: The script used during the install process requires the use
of a Java virtual machine (JVM). A JVM is bundled together with the Client
SDK package, but if required, It is possible to specify a different JVM for
the install process. The minimum version is Sun JRE is 1.4.2
Prepare the environment
You only need to check if you have the informix user and the informix group
on your computer, otherwise create them. For additional information about
this topic refer to the section Preparing to install IDS and client products at:
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.igul.do
c/ids_in_005x.htm
34
7884ch02.fm
Informix provides various methods for Client installation. Here we demonstrate

the GUI mode installation for both UNIX and Windows environments.
Installing Client SDK on UNIX using GUI

Perform these steps to install Client SDK installation in GUI mode:
1. Start the installation GUI with the following command from the root user.
./installclientsdk -gui
Remember to set the DISPLAY environment variable to the window which you
want to project the Installation. Figure 2-5 shows the Welcome panel.
Figure 2-5 Installation of Client SDK on UNIX
2. On the license agreement panel, select I accept the terms and the license
agreement.
3. Figure 2-6 on page 36 shows the next panel for specifying the directory to
where you want to install the Client. Enter the full path name.
35
7884ch02.fm
Figure 2-6 Directory where you like to install
4. On choosing setup type panel (Figure 2-7), select Typical or Custom. The
difference between Custom and Typical is that the Custom mode allows you
to select the features to be installed.
Figure 2-7 Do you want to install with the custom option or the typical one.
36
7884ch02.fm
5. Figure 2-8 shows the features that you can select on the Custom mode. If you
would like to install whatever the installer offers, select the Typical mode.
Figure 2-8 Custom options to choose
6. The installer presents the summary of available space and the components
you have chosen for installation, see Figure 2-9 on page 38. At this stage you,
can go back and change the installation selections. Click Next to start the
installation.
37
7884ch02.fm
Figure 2-9 Installation summary
7. Figure 2-10 shows that the Installation is complete without any errors.
Figure 2-10 Installation complete without any errors
At this stage the Client SDK installation is complete. If the Informix database
server is located on a different machine, you may need to setup the connection
information and modify the sqlhosts file, using the details of your Informix server.
38
7884ch02.fm
Installing Client SDK on Windows

Installation on Windows for Client is similar to UNIX installation except that at the
end, the Windows installation has the option to install IBM Data Server Driver
Package. If you select the option to install these drivers, refer to Installing IBM
Data Server Driver Package on page 41 for Installation details.
Configuration utility
On Windows, the Client SDK comes bundled with the setnet32.exe utility for
configuring the Client SDK. The Clint SDK has default values set and is ready to
use unless you want to change the protocol, server information, or any
environment variable under which the client application runs.
The Setnet32 utility sets environment variables and network parameters that IBM
Informix products use at run time. The environment variables and connectivity
parameters are stored in the Windows system registry and are valid for every
IBM Informix Client product that you install except the Informix JDBC driver.
The Setnet32 utility has the following tabs:
Environment tab allows you to set environment variables.
Server Information tab allows you to set database server network information.
Host Information tab allows you to set your host computer and login
information.
About Setnet32 tab provides the information about the Setnet32 utility.
For more information about these tabs, refer to the Client SDK installation PDF
file that comes with the Client SDK product.
Special consideration for 32/64 bit products

Beginning with version 3.50.FC4, you can install both 32-bit and 64-bit Client
SDK on the same Windows machine.
Before installing the version 3.50.FC4 product, you must uninstall the existing
version 3.50.FC3 or earlier Client SDK from your machine.
To uninstall your existing products, do one of the following:
Use the Add/Remove program, or
Run the setup.exe from the installed product media. When the Maintenance
Menu appears, select Remove.
Your new 32-bit and 64-bit products must be installed in separate directories.
The default location for the 32-bit Informix Client SDK is
C:\Program Files (x86)\IBM\Informix\Client-SDK
39
7884ch02.fm
The 64-bit Informix Client SDK default directory is

C:\Program Files\IBM\Informix\Client-SDK
The 64-bit product installation adds a suffix to the Program group folder and the
Add/Remove entry. This suffix provides an easy identification between the 32-bit
(no suffix) and 64-bit (suffix) product installations. For more information, refer to
the release notes.
2.2.3 Setting up IBM Data Server drivers

There are several types of IBM Data Server Clients and drivers available:
IBM Data Server Clients:
IBM Data Server Client, contains all the APIs and tools needed for
development against DB2 and Informix databases including tools for remote
administration.
IBM Data Server Runtime Client, contains all the APIs but only the minimum
number of utilities for setup and configure the development environment.
IBM Data Server Driver Package, contains only the APIs and drivers without
any additional tool.
The client products contain the APIs and drivers plus additional tools designed
for remote database administration and configuration of the client environment.
IBM Data Server Drivers:
IBM Data Server Driver for JDBC and SQLJ

IBM Data Server Driver for ODBC and CLI
IBM Data Server Driver for .NET
IBM Data Server Open Source drivers such as PHP or Ruby
Some of these components can also be download and installed as a separate

product.
Each IBM Data Server Client and driver provides a particular type of support:
For Java applications only, use IBM Data Server Driver for JDBC and SQLJ.
For applications using ODBC or CLI only, use IBM Data Server Driver for
ODBC and CLI
For applications using ODBC, CLI, .NET, OLE DB, PHP, Ruby, JDBC, or
SQLJ, use IBM Data Server Driver Package.
IBM Data Server Runtime Client and IBM Data Server Client includes all of
above drivers and some functions of DB2.
40
7884ch02.fm
In this section we discuss the installation and configuration of the following

drivers for UNIX and Windows platforms:
IBM Data Server Driver for ODBC and CLI
IBM Data Server Driver Package
For the installation and configuration of IBM Data Server Runtime Client and IBM
Data Server Client, refer to
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp?topic=/com.ibm.s
wg.im.dbclient.install.doc/doc/c0022615.html
Installing IBM Data Server Driver Package

The IBM Data Server Driver Package is a lightweight solution and the
recommended package for end user code deployment. It provides robust runtime
support for applications using ODBC, CLI, .NET, OLE DB, PHP, Ruby, JDBC, or
SQLJ without the need of installing Data Server Runtime Client or Data Server
Client. Here we discuss the installation of the drivers presented under the IBM
Data Server Driver Package.
Installation of driver package on UNIX and Linux

You can download the IBM Data Server Driver Package and individual drivers
such as IBM Data Server Driver for JDBC and SQLJ from IBM Support and
downloads website the at
http://www.ibm.com/support/docview.wss?rs=4020&uid=swg21385217
Uncompress and extract IBM Data Server Drivers Package to an empty directory.
after download.
If you want to install the complete IBM Data Server Driver Package, run the
installIDSDriver command. This driver package includes database drivers for
Java, ODBC/CLI, PHP, and Ruby on Rails, each of which is stored in its own
subdirectory. The Java and ODBC/CLI drivers are compressed.
If you like to install individual drivers contained in the driver package, follow these
steps (first two steps are common for all drivers):
1. Uncompress the Data Server Driver Package archive onto the target
machine.
2. For the Java and ODBC/CLI drivers, uncompress the driver file into the
chosen install directory.
3. Install the driver you need:
Java
41
7884ch02.fm
This driver provides support for client applications written in Java using
JDBC. The path and file name of the Java driver are
Path: jdbc_sqlj_driver/platform
File name: db2_db2driver_for_jdbc_sqlj.zip
For the additional details refer to Installing IBM Data Server Driver for
JDBC and SQLJ on page 47
ODBC and CLI
The open source drivers included in the IBM Data Server Driver Package
such as PHP or Ruby driver, require the use of the Data Server CLI driver
for connection to the database engine. If you plan to use any of these
driver you need to install the ODBC this driver.
Path: odbc_cli_driver/platform.
File name: ibm_data_server_driver_for_odbc_cli.tar.Z
For additional information refer to Installing IBM Data Server Driver for
ODBC and CLI on page 48
PHP
The IBM_DB2 and PDO_IBM are the PHP extensions for IBM Data
Servers including Informix. Installing the IBM_DB2 or PDO_IBM
extensions enables any application in the PHP environment to interact with
Informix database servers. The extensions included with the IBM Data
Server Driver Package might be of a lower version as compared to the one
available on the PHP repository. You should download the latest extension
from PHP repository page as it might contain important fixes and new
features available:
For IBM_DB2: Download from http://pecl.php.net/package/ibm_db2

For PDO_IBM: Download from http://pecl.php.net/package/pdo_ibm
The PHP driver path and file names in the IBM Data Server Driver
Packages are
Path: php_driver/platform/php32 or php_driver/platform/php64
Files: ibm_db2_n.n.n.so and pdo_ibm_n.n.n.so, where n represents

the version of the extension.
The following is installation information:
Prerequisite: The PHP drivers require the ODBC/CLI driver that is also
included in the IBM Data Server Driver Package to be installed.
Installation instructions:
Following links explain how to install PECL extensions:
Through the source code:
42
7884ch02.fm
http://www.php.net/manual/en/install.pecl.phpize.php
Through PECL command:

http://www.php.net/manual/en/install.pecl.pear.php
Installing IBM_DB2 extension:

http://www.php.net/manual/en/ibm-db2.installation.php
Installing PDO_IBM extension:

http://www.php.net/pdo_ibm
Ruby on Rails
The IBM_DB adapter and driver as a gem enables any application in the
Ruby environment, including Rails, to interact with IBM data servers. The
path, file name, and installation information for Ruby on Rails driver are
Path: ruby_driver/platform
File: ibm_db-0.10.0.gem
Prerequisite: The Ruby on Rails driver requires the ODBC/CLI driver

that is also included in the IBM Data Server Driver Package to be
installed.
To install the Ruby on Rails driver, from the location of the gem file, run
the following command
gem install ibm_db-0.10.0.gem
At this stage you are ready to use any of the driver mentioned above on UNIX or
Linux.
Installation of driver package on Windows

You can obtain the Data Server Driver Package for windows from the same site
as the Package for UNIX:
http://www.ibm.com/support/fixcentral/
The file for Windows is an .exe file which is the install file for the driver package.
The installations steps are as below:
1. Start the installation by executing the.exe file. Figure 2-11 on page 44 shows
the welcome window.
43
7884ch02.fm
Figure 2-11 Welcome panel for Data Server Driver Package
2. On the Software License Agreement panel, accept the license agreement.

3. Specify the installation directory in the Custom Setup panel (Figure 2-12). You
can change the installation directory to other location then the default.
Figure 2-12 Change the Install directory
44
7884ch02.fm
4. The IBM data server driver copy name is used to identify a location where
IBM Data Server Driver Package is installed on the computer. Enter the copy
name for the location you have chosen, the default is IBDBCL1. See
Figure 2-13 on page 45.
Figure 2-13 Set the IBM data server driver copy name
5. Verify the installation selections on the summary panel (Figure 2-14). If

everything is correct, proceed the installation by clicking Install.
45
7884ch02.fm
Figure 2-14 The summary of Install
6. When the Installer completes, you see the complete panel as shown in
Figure 2-15 on page 46. After clicking Finish, you are done with the
installation of Data Server driver for Informix, and are ready to use one.
Figure 2-15 Last panel of Installer
46
7884ch02.fm
Note: The Windows installer of the IBM Data Server Driver Package does not
provide any option for install individual components. All the driver and
interfaces included in the package are installed by default.
Installing IBM Data Server Driver for JDBC and SQLJ

The installation procedures is as follows:
1. Obtain Java SDK.
Before you install the IBM Data Server Driver for JDBC and SQLJ, you must
have an SDK for Java installed on your computer:
For JDBC 3.0 functions, you need Java SDK 1.4.2 or later.
For JDBC 4.0 functions, you need Java SDK 6 or later.
2. Download the zip file for the latest version of the IBM Data Server Driver for
JDBC and SQLJ from
http://www.ibm.com/software/data/support/data-server-clients/download.html.
3. Unzip the IBM Data Server Driver for JDBC and SQLJ zip file to the
installation location. The zip file contains the following files:
db2jcc.jar
db2jcc4.jar
sqlj.zip
sqlj4.zip
4. Modify the CLASSPATH environment variable to include the appropriate files,

You can set the CLASSPATH using the command:
java -classpath <dir>\<file>.jar
For JDBC:
Include db2jcc.jar in the CLASSPATH if you plan to use the version of the
IBM Data Server Driver for JDBC and SQLJ that includes only JDBC 3.0
and earlier functions.
Include db2jcc4.jar in the CLASSPATH if you plan to use the version of the
IBM Data Server Driver for JDBC and SQLJ that includes JDBC 4.0 and
earlier functions.
Important: Include either db2jcc.jar or db2jcc4.jar in the CLASSPATH.
Do not include both files.
For SQLJ
The steps are similar to the JDBC except that the files to be included in
CLASSPATH are sqlj.zip and sqlj4.zip.
47
7884ch02.fm
For more information regarding configuration and customizing of the IBM Data
Server Driver for JDBC and SQLJ, refer to
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.jccids.doc/
ids_jcc_0008.htm
Installing IBM Data Server Driver for ODBC and CLI

The IBM Data Server Driver for ODBC and CLI solution is designed mainly for
independent software vendor (ISV) deployments.To install the IBM Data Server
Driver for ODBC and CLI, perform the following steps:
1. Depending on your operating system, download the compressed file from
https://www14.software.ibm.com/webapp/iwm/web/preLogin.do?land=en_US&source
=swg-informixfpd
2. Uncompress or unzip the file to a installation directory of your choice.

3. On Mac OS X, set the DYLD_LIBRARY_PATH environment variable to the
clidriver/lib directory path.
For specific information about the setup and configuration of the ODBC refer to
3.2, Setup and configuration on page 54.
2.2.4 Setting up Informix JDBC

Informix JDBC is a platform independent, industry standard driver that provides
enhanced support for distributed transactions. It consists of a set of interfaces
and classes written in the Java programming language which can run on AIX,
HP-UX, Linux, Solaris, Windows, and all other platforms that support Java. It also
supports extensibility with a user-defined data type (UDT) routine manager that
simplifies the creation and use of UDTs.
JCC support is available through the IBM Data Server Driver package which has
a driver to run Java application using JCC.
This section describes the installation of JDBC driver for Informix that is bundled
with the Informix database server. There is a separate JDBC directory with a jar
file inside when you extract the Informix database server install media.
The IBM Informix JDBC Driver, Version 3.50 strives to be compliant with the Sun
Microsystems JDBC 3.0 specification. Nearly all the features that are required for
the Sun Microsystems JDBC 3.0 specification have the specified behavior. For
the Sun Microsystems JDBC 3.0 driver optional features, if the feature is
supported by IBM Informix Version 11.50, then it is supported by Informix JDBC
Driver, Version 3.50.
48
7884ch02.fm
IBM Informix JDBC Driver is a native-protocol, pure-Java driver (Type 4). This
means that when you use IBM Informix JDBC Driver in a Java program that uses
the JDBC API to connect to an Informix database, your session connects directly
to the database or database server, without a middle tier
Installation
The JDBC driver installation procedure is same for both Windows and UNIX
platforms.
To install the JDBC driver in console mode, use the following command:
java -cp dir/setup.jar run -console
For silent Install on both Windows and UNIX, run the following command from a
command prompt:
java -cp <dir>/setup.jar run -silent -P
product.installLocation=<destination-dir>
where:
<dir> is the location of setup.jar file.
<destination-dir> is the directory where you want to install the JDBC Driver.
The installation is complete when the command has finished executing. If you
want to log information during the installation, specify the -log parameter.
Perform these steps to install Informix JDBC Driver in GUI mode on Windows:
1. Start the GUI mode installation with the following command:
java -cp dir/setup.jar run
Figure 2-16shows the Welcome panel.
49
7884ch02.fm
Figure 2-16 Starting panel for JDBC informix Install
2. Accept the license agreement on the IBM License Acceptance Panel.

3. Specify the installation directory or use the default location as shown in
Figure 2-17.
Figure 2-17 Specify the Install directory
4. Review the information on the summary panel (Figure 2-18 on page 51) and
continue to complete the installation.
50
7884ch02.fm
Figure 2-18 Summary screen
Configuration
To use IBM Informix JDBC Driver in an application, you must set your
CLASSPATH environment variable to point to the driver files. The CLASSPATH
environment variable tells the Java Virtual Machine and other applications where
to find the Java class libraries used in a Java program-log parameter.
UNIX
There are two ways to set your CLASSPATH environment variable:
Add the full path name of ifxjdbc.jar to CLASSPATH:
setenv CLASSPATH /jdbcdriv/lib/ifxjdbc.jar:$CLASSPATH
Unpack ifxjdbc.jar and add its directory to CLASSPATH:

cd /jdbcdriv/lib
jar xvf ifxjdbc.jar
setenv CLASSPATH /jdbcdriv/lib:$CLASSPATH
Windows
Use one of the following methods to set your CLASSPATH environment variable:
Add the full path name of ifxjdbc.jar to CLASSPATH:
set CLASSPATH=c:\jdbcdriv\lib\ifxjdbc.jar;%CLASSPATH%
Unpack ifxjdbc.jar and add its directory to CLASSPATH:

cd c:\jdbcdriv\lib
jar xvf ifxjdbc.jar
set CLASSPATH=c:\jdbcdriv\lib;%CLASSPATH%
At this stage you are ready to develop a Java application using JDBC.
51
7884ch02.fm
52
7884ch03.fm
Chapter 3.
Working with the ODBC

driver
This chapter discusses the configuration and development of applications using
Open Database Connection (ODBC) interfaces to access an Informix database
server.
This chapter describes the following:
Types of ODBC drivers available

Configuration
Supported data types and mapping between SQL, C, and ODBC
Perform basic database operations
How to work with extended data types
Error handling
Diagnose typical problems
53
7884ch03.fm
3.1 ODBC and Informix

Open Database Connectivity (ODBC) is a software API method based on the
X/Open Call Level Interface to access data from a database management
system. ODBC allows developers to write database applications without having
to know a proprietary interface for a specific database. In an ODBC environment,
the ODBC driver is the component that directly communicates with the database
server. It receives calls from the application or the ODBC Driver Manager and
converts those calls into messages that the database server understand.
The available ODBC drivers for Informix are:
IBM Informix ODBC Driver
IBM Data Server Driver for ODBC and CLI (CLI Driver)
3.2 Setup and configuration

In this section, we discuss the setup and configuration parameters of the ODBC
driver.
3.2.1 IBM Informix ODBC Driver

The Informix ODBC Driver is installed by default as part of the Informix Client
SDK package. It is based on the Microsoft Open Database Connectivity (ODBC)
Version 3.0 standard.
Table 3-1 lists the Informix database server that support Informix ODBC.
Table 3-1 Supported databases
Database server
Versions
IBM Informix
10.0,11.10,11.50
8.50 and higher
IBM Informix Standard Engine
7.25
IBM Informix Online
5.20 and higher
Windows configuration
On a Windows machine the ODBC drivers is automatically registered within the
system during the Informix installation process.
54
7884ch03.fm
The default installation directory is

C:\Program Files\IBM\Informix\Client-SDK. The INFORMIXDIR environment
variable should point to the directory where the product was installed.
Note: If you install a 32-bit version of Informix Client-SDK on a Windows x64,
the default directory is C:\Program Files (x86)\IBM\Informix\Client-SDK
The ODBC driver requires loading additional libraries that are located in the
%INFORMIXDIR%\bin directory, make sure this directory is part of your PATH
variable.
The name of the ODBC driver depends on the version:
IBM INFORMIX ODBC DRIVER: For the Windows 32-bit driver, see
Figure 3-1.
Figure 3-1 Windows x86 ODBC driver administrator
IBM INFORMIX ODBC DRIVER (64-bit): for the Windows 64-bit driver. See
Figure 3-2 on page 56.
Chapter 3. Working with the ODBC driver
55
7884ch03.fm
Figure 3-2 Windows x64 ODBC driver administrator
To create an Informix ODBC Data Source (DSN), open the ODBC Administrator,
choose a DSN type, and select the Informix ODBC driver, the DSN-configuration
parameters are common to both drivers (32-bit and 64-bit).
Note: The default ODBC Data Source Administrator on a Windows x64
machine is the 64-bit version. If you want to create a 32-bit ODBC DSN you
must use the 32-bit version
C:\WINDOWS\SysWOW64\odbcad32.exe
Figure 3-3 on page 57 shows the Informix ODBC Connection tab.
56
7884ch03.fm
Figure 3-3 Connection parameters
Table 3-2 lists the required DSN parameters. If you have already defined an
Informix server using the Setnet32 tool, part of Informix Client SDK, most of
these values would be automatically updated when selecting an Informix server
from the dropdown box.
Table 3-2 Required DSN values
Parameter
Description
Data source name
Name of the DSN
Server name
Name of the Informix database server
Host name
Name or IP address of the computer where the database server

runs
Service
Name or value of the tcp service used by the database server
Protocol
Communication Protocol used by the database server
Database
Name of the database to which the DSN connects by default
You can set optional configuration parameters in the Environment and Advanced
tabs (Figure 3-4 on page 58).
57
7884ch03.fm
Figure 3-4 Environment parameters
Table 3-3 describe the parameters for the Environment tab.

Table 3-3 Environment parameters
58
Parameter
Description
Client Locale
Locale for the client machine
Database Locale
Locale used when the database was created
Use Server Database Locale
Using this option the Database Locale is automatically

retrieved from the database
Translation Library
Codeset conversion library
Translation Option
Options for a third party conversion library
Cursor Behavior
Close or preserve a cursor when a transaction is

resolved
VMB Character
How to report the length for Varying Multi byte

characters
Fetch Buffer Size
Size of the communication data package
Isolation Level
Default Isolation Level for the connection
7884ch03.fm
Note: The codeset value for the Client Locale parameter is determined by the
codeset used in the application which, in most cases, is the same as the
operating system. The default codeset on a Windows machine is CP1252.
The default codeset for an Informix database is en_us.8859-1 (ISO 8859-1).
Some versions of the IBM Informix database server (for example, Version 9.40
and Version 10.00) allow a connection from a client with a mismatched
DB_LOCALE value. This option is no longer available. A client application
must set the Database Locale parameter (DB_LOCALE) to the same value as
the locale of the database (locale used at creation time).
The typical syntax for an Informix locale is language_territory@codeset.
Figure 3-5 shows the Advanced configuration parameters.
Figure 3-5 Advanced parameters
Table 3-4 describes the parameters for the Advanced tab.

Table 3-4 Advanced parameters
Parameter
Description
Auto Commit Optimization
Optimize client-server communication deferring

the commit work after all the cursors are close.
Open-Fetch-Close Optimization
Optimize client-server communication

automatically closing the cursor after all the
rows have been retrieved.
59
7884ch03.fm
Parameter
Description
Insert Cursors
Buffers all the rows send by an insert cursor

and uses only one package to send the data to
the server.
Scrollable Cursors
Enable scrollable cursors.
Report KeySet Cursors
Report support for KeySet-driver cursors.
Report Standard ODBC Types Only
Report standard ODBC types Only. Extended

types are mapped to standard ODBC types.
Describe Decimal Floating Point as

SQL_REAL / SQL_DOUBLE
Decimal columns without a scale are reported

as SQL_REAL or SQL_DOUBLE.
Do Not Use Lvarchar
Do not report Lvarchar columns when using

SQL_VARCHAR.
Report Char columns as Wide Char

columns
Char columns described by SQLDescribeCol

are reported as Wide char columns
(SQLWCHAR)
length in Chars for

SQLGetDiagRecW
Returns the number of characters rather than

the number of bytes.
You can obtain more information about each one of these parameters by placing
the cursor over a parameter and pressing the F1, see Figure 3-6 on page 61.
60
7884ch03.fm
Figure 3-6 ODBC DSN Help
UNIX configuration
On UNIX platforms, the default installation directory for Informix Client SDK is
/opt/IBM/informix.
The INFORMIXDIR environment variable should point to the directory where the
product was installed.
On UNIX platforms, an ODBC driver manager is not normally supplied as part of
the operating system. Informix Client SDK does not include an ODBC driver
manager library, however, the Client SDK does have a Driver Manager
Replacement (DMR) library which provides most of the features of an ODBC
driver manager.
If the application does not use an ODBC driver manager such as UnixODBC
Driver Manager or Data Direct Driver Manager, the application must be linked
directly to the Informix ODBC libraries. The ODBC libraries are located at
$INFORMIXDIR/lib/cli.
Table 3-5 shows the ODBC libraries included with Client SDK.
61
7884ch03.fm

Table 3-5 UNIX ODBC libraries
Library
Description
libifcli.a or libcli.a
Static version for single (nonthreaded)
libifcli.so or iclis09b.so
Shared version for single (nonthreaded)
libthcli.a
Static version for multithreaded library
libthcli.so or iclit09b.so
Shared version for multithreaded library
libifdrm.so or idmrs09a.so
Shared library for DMR (thread safe)
The shared-library path environment variable specifies the library search path.
This variable should contains at least $INFORMIX/lib, $INFORMIXDIR/lib/esql,
and $INFORMIXDIR/lib/cli for the ODBC driver to work.
There are three configuration files for the ODBC driver:
sqlhosts
odbc.ini
odbcinst.ini
Note: the information in the sqlhosts file is also used by all the other Client
SDK components, as well as by the Informix database server. Be aware that
any changes in this file may have an impact in other clients and servers.
On Windows, instead of using a text file, this information is stored in the
registry through the Setnet32 tool.
The sqlhosts file

This text file contains most of the information required to connect to an IBM
Informix database server.
The default location for this file is $INFORMIXDIR/etc/sqlhosts. You can use the
INFORMIXSQLHOSTS environment variable to point to a different location.
Example 3-1 shows a simple sqlhost file.
Example 3-1 sqlhosts file
#server protocol hostname service/port
on_demo onsoctcp kodiak 9088
For more information about the sqlhosts file, refer to the Client/Server
Communications manual at:
62
7884ch03.fm
ds_admin_0158.htm
The odbcinst.ini file

This file contains a list of installed ODBC drivers on the UNIX machine and
specific attributes for each driver, such as the location of the shared library.
The default location of this file is under the home directory as
$HOME/.odbcinst.ini. A sample odbcinst.ini file is located in the
$INFORMIXDIR/etc directory.
Example 3-2 shows a simple odbcinst.ini file.
Example 3-2 odbcinst.ini
;--------------------------------------------------------------------------[ODBC Drivers]
IBM INFORMIX ODBC DRIVER=Installed
[IBM INFORMIX ODBC DRIVER]
Driver=/extra/informix/lib/cli/iclit09b.so
Setup=/extra/informix/lib/cli/iclit09b.so
APILevel=1
ConnectFunctions=YYY
DriverODBCVer=03.51
FileUsage=0
SQLLevel=1
smProcessPerConnect=Y
The odbinst.ini file has two sections:

ODBC Drivers: List the ODBC drivers installed in the system.
Driver Properties: List the properties for a specific ODBC driver.
Table 3-6 lists the parameters in the properties section.
Table 3-6 Parameters in the Driver Properties section
Keyword
Description
Driver
Location of the ODBC library
Setup
Location of the Setup library
APILevel
ODBC interface conformance level that the driver supports
ConnectFunctions
Support for SQLConnect, SQLDriverConnect, and

SQLBrowserConnect
DriverODBCVer
Supported version of the ODBC driver
63
7884ch03.fm
Keyword
Description
FileUsage
How handle File-system DSN
SQLLevel
Type of SQL-92 grammar supported
The odbc.ini file

The odbc.ini file is the data source configuration information. The default location
of this file is under the home directory as $HOME/.odbc.ini.
Alternatively, you can use the ODBCINI environment variable to point the odbc.ini
to a different location. A sample odbc.ini file is located in the $INFORMIXDIR/etc
directory.
Example 3-3 shows a simple odbc.ini file.
Example 3-3 odbc.ini
[ODBC Data Sources]
Infdrv1=IBM INFORMIX ODBC DRIVER
;
[Infdrv1]
Driver=/extra/informix/lib/cli/iclis09b.so
Description=IBM INFORMIX ODBC DRIVER
Database=stores_demo
LogonID=odbc
pwd=odbc
Servername=demo_on
CursorBehavior=0
CLIENT_LOCALE=en_us.8859-1
DB_LOCALE=en_us.8859-1
TRANSLATIONDLL=/extra/informix/lib/esql/igo4a304.so
;
[ODBC]
;UNICODE=UCS-4
;
; Trace file Section
;
Trace=0
TraceFile=/tmp/odbctrace.out
InstallDir=/extra/informix
TRACEDLL=idmrs09a.so
This file contains three sections:

ODBC Data Sources: DSN name and description of the driver used
Data Source Specification: Configuration parameters for this particular DSN
ODBC: Global options such as Unicode mode or Tracing
64
7884ch03.fm
Table 3-7 lists the available parameters for the odbc.ini configuration file.
Table 3-7 Parameters for the odbc.ini file
Keyword
Description
Driver
Location of the ODBC shared library
Description
DSN description
Database
Database name
LogonID
User Id
pwd
Password
Server
Database server
CLIENT_LOCALE
Locale used by the client application
DB_LOCALE
Locale of the database
TRANSLATIONDLL
Location of the codeset conversion library
CURSORBEHAVIOR
Close or preserve a cursor when a

transaction is resolved
DefaultUDTFetchType
Default type for a UDT (SQL_C_BINARY

or SQL_C_CHAR)
ENABLESCROLLABLECURSORS
Enable Scrollable cursors
ENABLEINSERTCURSORS
Optimize insert cursors process
OPTIMIZEAUTOCOMMIT
Defer the commit message after all cursor

are closed
NEEDODBCTYPESONLY
Extended types are mapped to standard

ODBC types
OPTOFC
Close cursors after all the rows have been

fetched
REPORTKEYSETCURSORS
Report support for keyset-driven cursors
FETCHBUFFERSIZE
Set the size of the fetch buffer
DESCRIBEDECIMALFLOATPOINT
Describe float types as SQL_REAL or

SQL_DOUBLE
USESERVERDBLOCALE
Use the server database locale
DONOTUSELVARCHAR
Dont report Lvarchar columns when using

SQL_VARCHAR
65
7884ch03.fm
Keyword
Description
REPORTCHARCOLASWIDECHARCOL
Char columns described by

SQLDescribeCol are reported as Wide
char columns (SQLWCHAR)
ISOLATIONLEVEL
Default isolation level
UNICODE
Type of Unicode (UCS-2, UCS-4)
TRACE
Trace enable or disable
TRACEFILE
Location of the trace file
TRACEDLL
Trace library name (idmrs09a.so)
3.2.2 IBM Data Server Driver for ODBC and CLI

The IBM Data Server Driver for ODBC and CLI (CLI Driver) is installed as part of
the IBM Data Server Driver Package or as a separate product. The IBM Data
Server Package is bundle with the Informix Client SDK product. It allows you to
connect to IBM Informix and IBM DB2 databases with the same set of libraries.
The communication protocol used is DRDA instead of the native Informix SQLI.
This driver is supported against Informix Version 11.10 and Version 11.50.
The default install directory for the IBM Data Server Driver is
C:\Program Files\IBM\IBM DATA SERVER DRIVER
The ODBC driver is registered in the Windows system during the installation of
the IBM Data Server Driver package. The name of the ODBC driver is IBM DB2
ODBC Driver.
Figure 3-7 shows the Drivers tab of the ODBC Data Source Administrator listing
both ODBC drivers:
IBM DB2 ODBC DRIVER
IBM INFORMIX ODBC DRIVER
66
7884ch03.fm
Figure 3-7 ODBC Data Source Administrator
To create an ODBC Data Source (DSN) using the IBM Data Server Driver, open
the ODBC Administrator, choose a DSN type, and select the IBM DB2 ODBC
Driver.
Figure 3-8 shows the Add Dialog.
Figure 3-8 DSN Add dialog
If there is no database alias defined, you must add one with the connection
details of your IBM Informix database server. See Figure 3-9 on page 68.
67
7884ch03.fm
Figure 3-9 DSN Settings dialog
Figure 3-10 shows the Advanced settings dialog box.
Figure 3-10 DSN Advanced Settings dialog
Table 3-8 on page 69 lists the minimum required parameters to connect to an

IBM Informix server.
68
7884ch03.fm

Table 3-8 CLI Parameters
CLI parameter
Description
Hostname
Name of the machine where the Informix server is running
Port
TCP Port used by the DRDA Informix alias
Protocol
Transport protocol (TCPIP)
Database
Database server
The configuration settings are stored as a File-DSN (text file contains all the
information required to connect to the database) in the
%USERPROFILE%\db2cli.ini file. Example 3-4 shows a simple db2cli.ini file.
Example 3-4 db2cli.ini
[test]
Protocol=TCPIP
Port=9089
Hostname=kodiak
For a complete description of all the CLI parameters, refer to the DB2 Information
Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp?topic=/com.ibm.d
b2.luw.apdv.cli.doc/doc/r0007964.html
UNIX configuration
On UNIX platforms, the default installation directory for IBM Data Server Client
package is /opt/IBM/db2/V9.7. The library for the ODBC driver is libdb2.a that is
located in the lib directory of your Data Server installation.
Same as on a Windows machine, the ODBC configuration is stored in the
db2cli.ini file which by default is included in the cfg directory of your Data
Server installation, or is pointed by the DB2CLIINI environment variable
Example 3-5 shows a simple db2cli.ini file.
[test]
Protocol=TCPIP
Port=9089
Hostname=kodiak
69
7884ch03.fm
For more information about the db2cli.ini file, refer to the db2cli.ini initialization
file section at
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp?topic=/com.ibm.d
b2.luw.apdv.cli.doc/doc/c0007882.html
3.2.3 Verifying connectivity

On a Windows system you can use the ODBC Data Source Administrator to
verify the connection with the Informix database server.
When using the IBM Informix ODBC driver, you can verify the connection using
the Apply & Test Connection option. See Figure 3-113-11.
Figure 3-11 Informix ODBC connection test
With the IBM Data Server Driver for ODBC, you can perform the connection test
using the Connect option. See Figure 3-12 on page 71.
70
7884ch03.fm
Figure 3-12 IBM Data Server Driver connection test
Informix Client SDK on UNIX does not include any tool to test an ODBC DSN.
You can use the C samples in $INFORMIXDIR\demo\cli to check if the ODBC DSN
is working.
The IBM Data Server Driver for ODBC includes a CLI client, db2cli, that can be
used for testing the connection. Example 3-6 shows a a typical output.
Example 3-6 db2cli output
$ db2cli
IBM DATABASE 2 Interactive CLI Sample Program
(C) COPYRIGHT International Business Machines Corp. 1993,1996
All Rights Reserved
Licensed Materials - Property of IBM
US Government Users Restricted Rights - Use, duplication or
disclosure restricted by GSA ADP Schedule Contract with IBM Corp.
> quickconnect 1 1 test informix password
SQLAllocEnv: rc = 0 (SQL_SUCCESS)
CLI henv = 1, Test Driver henv = 1
SQLAllocConnect: rc = 0 (SQL_SUCCESS)
CLI hdbc = 1, Test Driver hdbc = 1
SQLConnect: rc = 0 (SQL_SUCCESS)
>
71
7884ch03.fm
3.3 Developing an ODBC application

In this section we describe how to connect to a database server, type mapping
between standard ODBC and IBM Informix, and how to perform basic operations
with the ODBC driver.
3.3.1 Connecting to the database

These are the typical steps an ODBC application performs:
1. Connect to the database.
To connect to the database, the application must pass the connection details
to the ODBC driver. These details could be included directly in the connection
string or stored as a Data Source Name (DSN).
2. Process SQL statements.
The application sends SQL request to the ODBC driver to perform database
operations (insert, select, delete, update, and so on).
3. Resolves any open transaction.
If the application is using transaction, it should resolve (commit or rollback)
any open transactions.
4. Terminate the connection and free the allocated resources.
These steps are done using the ODBC API functions. See Figure 3-13 on
page 73 for a typical sequence of calls. The labels inside the gray boxes
correspond with the ODBC API functions used.
72
7884ch03.fm
Figure 3-13 Typical ODBC calls used by applications
Example 3-7 Shows a simple C program which connects to an ODBC DSN.

Example 3-7 test_connect.c sample
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <windows.h>
#include <conio.h>
#endif
#include "infxcli.h"
int main (long argc, char* argv[])
{
73
7884ch03.fm

SQLHDBC
SQLHENV
hdbc;
henv;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
if (argc != 2)
{
fprintf (stdout, "Please specify the name of a DSN!\n");
return(1);
}
else
{
if (strstr (argv[1],"DRIVER")==NULL)
sprintf((char *) connStrIn, "DSN=%s;", (char *)argv[1]);
else
sprintf((char *) connStrIn, "%s;", (char *)argv[1]);
}
rc = SQLAllocHandle (SQL_HANDLE_ENV, SQL_NULL_HANDLE, &henv);
if (rc != SQL_SUCCESS)
{
fprintf (stdout, "Environment Handle Allocation failed!\n");
return (1);
}
rc = SQLSetEnvAttr (henv, SQL_ATTR_ODBC_VERSION, (SQLPOINTER) SQL_OV_ODBC3, 0);
{
fprintf (stdout, "SQLSetEnvAttr failed!\n");
return (1);
}
rc = SQLAllocHandle (SQL_HANDLE_DBC, henv, &hdbc);
{
fprintf (stdout, "Connection Handle Allocation failed!\n");
return (1);
}
rc = SQLDriverConnect (hdbc, NULL, connStrIn, SQL_NTS, connStrOut, 1000,
&connStrOutLen, SQL_DRIVER_NOPROMPT);
{
fprintf (stdout, "Connectedion failed!\n");
return (1);
}
fprintf (stdout, "Connected\n");
SQLDisconnect (hdbc);
74
7884ch03.fm

SQLFreeHandle (SQL_HANDLE_DBC, hdbc);
SQLFreeHandle (SQL_HANDLE_ENV, henv);
return (rc);
}
You can compile this code on a Windows system using the following command
cl /DWIN32 -I%INFORMIXDIR%\incl\cli odbc32.lib test_connect.c
Example 3-8 shows how to compile and run the sample.

Example 3-8 test_connec.c output
c:\work>cl /DWIN32 /D_CRT_SECURE_NO_DEPRECATE -I%INFORMIXDIR%\incl\cli
odbc32.lib /nologo test_connect.c
test_connect.c
D:\work>test_connect
Please specify the name of a DSN!
D:\workt>test_connect dummy_dsn
Connection failed!
D:\work>test_connect test
Connected
D:\work>
You can use the sample C program to perform a DSN-less connection.

Example 3-9 shows the sample C program with a DSN-less connection string
using the Informix ODBC driver and the Data Server driver for ODBC.
Example 3-9 DSN-less connection
C:\work>test_connect "DRIVER={IBM INFORMIX ODBC
DRIVER};SERVER=demo_on;DATABASE=stores_demo;HOST=kodiak;PROTOCOL=onsoctcp;SERVI
CE=9088;UID=informix;PWD=password;";
Connected
C:\work>test_connect "driver={IBM DB2 ODBC
DRIVER};Database=stores_demo;hostname=kodiak;port=9089;protocol=TCPIP;
uid=informix; pwd=password"
Connected
C:\work>
75
7884ch03.fm
3.3.2 Type mapping

The data types used on the database differ from the data types used by your
application. This section shows the data type mapping needed when working
with specific Informix data types with the Informix ODBC driver.
When a query is executed by the application, the data returned by the Informix
server may be in a different format than what the application uses. The ODBC
driver converts the data passed between the application and the database
server. This process is invisible to the application. The only requirement for the
application is to specify the correct data types when calling the ODBC driver
functions.
The application sets ValueType and ParameterType when calling the bind
functions.
Example 3-10 shows the definition for one of the bind functions,
SQLBindParameter(). The ODBC driver converts the data type specified in
ValueType to the data type specified in ParameterType.
Example 3-10 SQLBindParameter definition
SQLBindParameter(
SQLHSTMT
SQLUSMALLINT
SQLSMALLINT
SQLSMALLINT
SQLSMALLINT
SQLUINTEGER
SQLSMALLINT
SQLPOINTER
SQLINTEGER
SQLINTEGER
StatementHandle,
ParameterNumber,
InputOutputType,
ValueType,
ParameterType,
ColumnSize,
DecimalDigits,
ParameterValuePtr,
BufferLength,
*StrLen_or_IndPtr);
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
hstmt */
ipar */
fParamType */
fCType */
fSqlType */
cbColDef */
ibScale */
rgbValue */
cbValueMax */
pcbValue */
Example 3-11shows a SQLBindParameter() call where the application is binding

a SQL_C_BINARY parameter with a SQL_INFX_UDT_FIXED value.
Example 3-11 Using SQLBindParameter()
SQLBindParameter (
hstmt,
1,
SQL_PARAM_INPUT,
SQL_C_BINARY,
SQL_INFX_UDT_FIXED,
loptr_size,
0,
loptr_buffer,
76
7884ch03.fm

loptr_size,
&loptr_valsize);
Table 3-9 shows the Informix specific data type mapping used with the Informix
ODBC driver.
Table 3-9 Mapping for specific Informix data types
Informix SQL
Informix ODBC driver
BIGINT
SQL_INFX_BIGINT
BIGSERIAL
SQL_INFX_BIGINT
BLOB
SQL_IFMX_UDT_BLOB
BOOLEAN
SQL_BIT
BYTE
SQL_LONGVARBINARY
CLOB
SQL_IFMX_UDT_CLOB
DATETIME
SQL_TIMESTAMP
DISTINCT
Any
IDSSECURITYLABEL
Built-in DISTINCT OF VARCHAR(128)
INT8
SQL_BIGINT
INTERVAL DAY
SQL_INTERVAL_DAY
INTERVAL DAY TO HOUR
SQL_INTERVAL_DAY_TO_HOUR
INTERVAL DAY TO MINUTE
SQL_INTERVAL_DAY_TO_MINUTE
INTERVAL DAY TO SECOND
SQL_INTERVAL_DAY_TO_SECOND
INTERVAL HOUR
SQL_INTERVAL_HOUR
INTERVAL HOUR TO MINUTE
SQL_INTERVAL_HOUR_TO_MINUTE
INTERVAL HOUR TO SECOND
SQL_INTERVAL_HOUR_TO_SECOND
INTERVAL MINUTE
SQL_INTERVAL_MINUTE
INTERVAL MINUTE TO SECOND
SQL_INTERVAL_MINUTE _TO_SECOND
INTERVAL MONTH
SQL_INTERVAL_MONTH
INTERVAL SECOND
SQL_INTERVAL_SECOND
INTERVAL YEAR
SQL_INTERVAL_YEAR
INTERVAL YEAR TO MONTH
SQL_INTERVAL_YEAR_TO_MONTH
77
7884ch03.fm
Informix SQL
Informix ODBC driver
LIST, MULTISET, SET
Any
LVARCHAR
SQL_VARCHAR
MONEY
SQL_DECIMAL
NCHAR
SQL_CHAR
NVARCHAR
SQL_VARCHAR
OPAQUE (fixed)
SQL_INFX_UDT_FIXED
OPAQUE (varying)
SQL_INFX_UDT_VARYING
ROW
Any
SERIAL
SQL_INTEGER
SERIAL8
SQL_BIGINT
TEXT
SQL_LONGVARCHAR
For a complete list of Data Type mapping, refer to the IBM Informix ODBC Driver
Guide at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.odbc.doc/si
i04979050.htm#sii04979050
If your application requires only Standard ODBC data types, you can enable the
Report Standard ODBC Types option in your DSN or connection string. When this
option is enabled, the ODBC driver handle Smart Large Objects (BLOB and
CLOB) as if they were simple large objects (byte and text), the driver
automatically generates the smart-large-object calls (ifx_lo_open,
ifx_lo_write, and so on).
This option also changes the mapping for user define types (including
MULTISET, SET, ROW, and LIST) to SQL_C_CHAR.
These options are also available as the ODBC attributes:
SQL_INFX_ATTR_ODBC_TYPES_ONLY
SQL_INFX_ATTR_LO_AUTOMATIC
SQL_INFX_ATTR_DEFAULT_UDT_FETCH_TYPE
78
7884ch03.fm
3.3.3 Performing database operations

In this section we show samples of how to use the ODBC driver to perform basic
operations with an Informix database. Most of these tasks are common to any
ODBC application so we do not explain them in detail.
This section includes:
Simple SQL statements (INSERT, UPDATE, and DELETE)

Fetching data
Using parameters
Calling SQL routines
Using transactions
Simple SQL statements

You can run SQL statements directly using the SQLExecDirect() function. This
function is commonly use when there is no need to run the SQL statement more
than once. The ODBC driver would prepare the statement, execute it, and free
the resources in one operation.
Example 3-12 shows a sample of using SQLExecDirect() to set the LOCK WAIT
time out.
Example 3-12 SQLExecDirect() sample. Simple_sql.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
int
SQLCHAR
sqllen;
*sqlstmt;
79
7884ch03.fm
if (argc != 3)
{
fprintf (stdout, "Please specify the name of a DSN and the SQL Statement to
run!\n");
return(1);
}
else
{
else
sqllen = strlen((char *)argv[2]);

sqlstmt = (SQLCHAR *) malloc (sqllen + sizeof(char));
strcpy((char *)sqlstmt, (char *)argv[2]);
}
{
fprintf (stdout, "Environment Handle Allocation failed!\n");
return (1);
}
{
fprintf (stdout, "SQLSetEnvAttr failed!\n");
return (1);
}
{
fprintf (stdout, "Connection Handle Allocation failed!\n");
return (1);
}
{
fprintf (stdout, "Connection failed!\n");
return (1);
}
rc = SQLAllocHandle (SQL_HANDLE_STMT, hdbc, &hstmt );
80
7884ch03.fm

{
fprintf (stdout, "Statement Handle Allocation failed!\n");
return (1);
}
rc = SQLExecDirect (hstmt, sqlstmt, SQL_NTS);
if (rc != SQL_SUCCESS && rc != SQL_SUCCESS_WITH_INFO)
{
fprintf (stdout, "SQLExecDirectW() failed!\n");
return (1);
}
fprintf (stdout, "Executed SQL Statement:\n%s\n",sqlstmt);
return (rc);
}
Example 3-13 shows the result of running the sample program with different SQL
statements. Here we input the SQL statement directly from the command line, in
a real application, the SQL is normally constructed based on the customer input.
In this simple example, we also did not provide error information if the execution
fails. You can obtain error information using the SQLGetDiagRec() function which
is cover in the 3.3.5, Error handling on page 108.
Example 3-13 Output of SQLDirectExec sample
c:\work>cl /DWIN32 /D_CRT_SECURE_NO_DEPRECATE -I%INFORMIXDIR%\incl\cli
odbc32.lib /nologo simple_sql.c
simplesql.c
c:\work>simples_sql.exe test "set lock mode to wait 10"
Connected
Executed SQL Statement:
set lock mode to wait 10
c:\work>simple_sql.exe test "create temp table temp1(c1 int)"
Connected
create temp table temp1(c1 int)
c:\work>simple_sql.exe test "invalid_SQL"
Connected
SQLExecDirectW() failed!
81
7884ch03.fm

c:\work>
Fetching data
Example 3-14 demonstrates how to run a SELECT statement to retrieve data
from the server. To make the code clear we have removed all the error handling.
Example 3-14 Select.c sample
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
SQLCHAR
SQLCHAR
SQLCHAR
SQLLEN
SQLLEN
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
sqlstmt[100];
code[2+1];
sname[15+1];
lcode=0;
lsname=0;
sprintf((char *) connStrIn, "DSN=demo_on");

sprintf((char *) sqlstmt, "SELECT code, sname FROM state where code<'CA'");
SQLAllocHandle (SQL_HANDLE_ENV, SQL_NULL_HANDLE, &henv);
SQLSetEnvAttr (henv, SQL_ATTR_ODBC_VERSION, (SQLPOINTER) SQL_OV_ODBC3, 0);
SQLAllocHandle (SQL_HANDLE_DBC, henv, &hdbc);
SQLDriverConnect (hdbc, NULL, connStrIn, SQL_NTS, connStrOut, 1000, &connStrOutLen,
SQL_DRIVER_NOPROMPT);
SQLAllocHandle (SQL_HANDLE_STMT, hdbc, &hstmt );
SQLExecDirect (hstmt, sqlstmt, SQL_NTS);
SQLBindCol (hstmt, 1, SQL_C_CHAR, &code, 3, &lcode);
82
7884ch03.fm

SQLBindCol (hstmt, 2, SQL_C_CHAR, &sname, 16, &lsname);
while (SQLFetch(hstmt)!=SQL_NO_DATA_FOUND)
fprintf (stdout, "Fetched row: %s, %s\n",code, sname);
SQLCloseCursor(hstmt);
return (rc);
}
Example 3-15 shows the output of this sample.

Example 3-15 Output of Select.c
C:\work>cl /DWIN32 /D_CRT_SECURE_NO_DEPRECATE -I%INFORMIXDIR%\incl\cli
odbc32.lib /nologo select.c
select.c
C:\work>select
Connected
SELECT code, sname FROM state where code<'CA'
Fetched row: AK, Alaska
Fetched row: AL, Alabama
Fetched row: AR, Arkansas
Fetched row: AZ, Arizona
C:\work>
Using parameters
In most of the cases an application would run the same SQL statement several
times, so it makes sense if the SQL statement is prepared and then used with
different values.
Example 3-16 demonstrates how to run a simple SQL SELECT statement using
an input parameter. The SQL statement is prepared and then executed with one
parameter.
Example 3-16 Example of a parametrized query. Select_param.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
83
7884ch03.fm

#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
SQLCHAR
SQLCHAR
SQLCHAR
SQLCHAR
SQLLEN
SQLLEN
SQLSMALLINT
SQLUINTEGER
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
sqlstmt[100];
inputcode[2+1];
code[2+1];
sname[15+1];
lcode = 0;
lsname = 0;
datatype, decimaldigits, nullable;
paramsize;

sprintf((char *) sqlstmt, "SELECT code, sname FROM state where code < ?");
sprintf((char *) inputcode, "CA");
SQLDriverConnect (hdbc, NULL, connStrIn, SQL_NTS, connStrOut, sizeof(connStrOut),
SQLPrepare (hstmt, sqlstmt, strlen(sqlstmt));
SQLDescribeParam(hstmt, 1, &datatype, &paramsize, &decimaldigits, &nullable);
SQLBindParameter(hstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, datatype, paramsize,
decimaldigits, inputcode, 0, NULL);
SQLExecute (hstmt);
fprintf (stdout, "Executed SQL Statement:\n%s\nUsing: '%s'\n",sqlstmt,inputcode);
SQLBindCol (hstmt, 1, SQL_C_CHAR, &code, sizeof(code), &lcode);
SQLBindCol (hstmt, 2, SQL_C_CHAR, &sname, sizeof(sname), &lsname);
while ((rc = SQLFetch(hstmt))!=SQL_NO_DATA_FOUND)
fprintf (stdout, "Fetched: %s, %s\n",code, sname);
84
7884ch03.fm

return (rc);
}
Example 3-17 shows the SQL statement the program prepares and the
parameter used for the placeholder.
Example 3-17 Output for Select_param.c
C:\work>cl /DWIN32 /D_CRT_SECURE_NO_DEPRECATE -ID:\infx\csdk350tc7\incl\cli
odbc32.lib /nologo select_param.c
select_param.c
C:\work>select_param
Connected
SELECT code, sname FROM state where code < ?
Using: 'CA'
Fetched: AK, Alaska
Fetched: AL, Alabama
Fetched: AR, Arkansas
Fetched: AZ, Arizona
C:\work>

In this section we demonstrate how to call a SQL routine from an ODBC
application. The following is a simple stored procedure that returns the user
name and session ID.
create procedure get_sid(user char(20)) returning char(100);
return 'user= '||trim(user)|| ' session= '||dbinfo('sessionid');
end procedure;
We use the ODBC standard syntax for calling a database routine:

{? = call client_routine(?, ?,...)}
The first placeholder (?) is used only when the first parameter of the routine is
an output parameter. If the first parameter is not an output parameter, you can
run the routine (stored procedure or SQL function) using the following syntax:
{call client_routine(?, ?, ?, ?)}
In Example 3-18 on page 86 we illustrate how to call the get_sid() SQL function.
85
7884ch03.fm

Example 3-18 SPL Function sample. Function.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
SQLCHAR
SQLCHAR
SQLCHAR
SQLLEN
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
sqlstmt[100];
errorn[20+1];
result[100+1];
lresult = 0;

sprintf((char *) sqlstmt, "{? = call get_sid(?)}");
sprintf((char *) errorn, "informix");
SQLBindParameter(hstmt, 1, SQL_PARAM_OUTPUT, SQL_C_CHAR, SQL_CHAR,sizeof(result),0,
result, sizeof(result), &lresult);
SQLBindParameter(hstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, strlen(errorn),0,
errorn, 0, NULL);
SQLFetch(hstmt);
fprintf (stdout, "Executed SQL Statement:\n%s\nUsing: '%s'\n",sqlstmt,errorn);
fprintf (stdout, "Value returned: %s\n",result);
86
7884ch03.fm

return (rc);
}
Example 3-19 shows the input parameter passed to the get_sid() procedure
and the return values.
Example 3-19 Output of Function.c
odbc32.lib /nologo function.c
function.c
C:\work>function
Connected
{? = call get_sid(?)}
Using: 'informix'
Value returned: user: informix session: 213
C:\work>
Local transactions
The default transaction mode for the Informix ODBC driver is auto-commit. This
means that the transaction is automatically committed if the SQL statement is run
successfully.
You can switch to manual-commit by setting the SQL_AUTOCOMMIT_OFF attribute
using the SQLSetConnectAttr() function. When an application sets
SQL_AUTOCOMMIT_OFF, the next SQL statement automatically starts a transaction
that will remain open until the application calls SQLEndTran().
In manual mode, all the statements executed by the application are committed or
rolled back when the application calls SQLEndTran().
Having auto-commit set may be more convenient when running simple SQL
statements because there is less tasks to care about in the application code.
However, it gives less control to the developer than using the manual-commit
mode. When using auto-commit, there is no option of rolling back a particular
change in the database, neither to insert a multiple rows as a batch operation,
which may improve the performance of the application.
Example 3-20 on page 88 illustrates how to run a local transaction in manual
mode.
87
7884ch03.fm

Example 3-20 Transaction.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
SQLCHAR
SQLINTEGER
SQLSMALLINT
SQLUINTEGER
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
sqlstmt[100];
cnum = 101;
paramsize;

sprintf((char *) sqlstmt, "INSERT INTO orders(order_num,order_date,customer_num)
VALUES (0,current,?)");

SQLSetConnectAttr (hdbc, SQL_ATTR_AUTOCOMMIT, SQL_AUTOCOMMIT_OFF, (SQLINTEGER)
NULL);
SQLPrepare (hstmt, sqlstmt, strlen(sqlstmt));
SQLDescribeParam(hstmt, 1, &datatype, &paramsize, &decimaldigits, &nullable);
SQLBindParameter(hstmt, 1, SQL_PARAM_INPUT, SQL_C_SHORT, datatype, paramsize,
decimaldigits, &cnum, 0, NULL);
while (cnum<105)
{
88
7884ch03.fm
SQLExecute (hstmt);
fprintf (stdout, "Executed SQL Statement:\n%s\nUsing: '%d'\n",sqlstmt,cnum);
cnum++;
}
rc = SQLEndTran (SQL_HANDLE_DBC, hdbc, SQL_ROLLBACK);
fprintf (stdout, "Transaction Rolled back\n");
return (rc);
}
Example 3-21 shows the output of the previous example. A local transaction is
created, five rows are inserted in the orders table and then, the transaction is
rolled back using SQLEndTran().
Example 3-21 Output of Trasaction.C
odbc32.lib /nologo transact.c
transact.c
C:\work>del pp*
C:\work>transact.exe
Connected
INSERT INTO orders(order_num,order_date,customer_num)
Using: '101'
Using: '102'
Using: '103'
Using: '104'
Transaction Rolled back
VALUES (0,current,?)
C:\work>
Distributed transactions
On Windows you can run a distributed transaction using the Microsoft
Distributed Transaction Coordinator (MSDTC) service.
89
7884ch03.fm
Example 3-22 demonstrate how to run a distributed transaction. The example

opens two connections to two different database servers and enlists both
connections into the same distributed transaction.
Example 3-22 Transaction_dtc.cpp
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#include <txdtc.h>
#include <xolehlp.h>
#endif
{
SQLHDBC
hdbc, hdbc2;
SQLHENV
henv;
SQLHSTMT
hstmt, hstmt2;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
SQLCHAR
SQLINTEGER
SQLSMALLINT
SQLUINTEGER
rc = 0;
connStrIn[100],connStrIn2[100];
connStrOut[1000];
connStrOutLen;
sqlstmt[100];
cnum = 101;
paramsize;
ITransactionDispenser *pTransactionDispenser = NULL;

ITransaction * pITransaction;
sprintf((char *) connStrIn , "DSN=server_1");
sprintf((char *) connStrIn2, "DSN=server_2");
sprintf((char *) sqlstmt, "INSERT INTO orders(order_num,order_date,customer_num)
VALUES (0,current,101)");
SQLAllocHandle (SQL_HANDLE_DBC, henv, &hdbc2);
// Get DTC
DtcGetTransactionManager(NULL, NULL,IID_ITransactionDispenser, 0, 0, NULL, (void**)
&pTransactionDispenser);
90
7884ch03.fm

fprintf (stdout, "Connected to %s\n",connStrIn);
SQLDriverConnect (hdbc2, NULL, connStrIn2, SQL_NTS, connStrOut, sizeof(connStrOut),
fprintf (stdout, "Connected to %s\n",connStrIn2);
//Start the transaction on DTC
pTransactionDispenser->BeginTransaction
(NULL,ISOLATIONLEVEL_READCOMMITTED,0,NULL,&pITransaction);
//Enlist the connections in distributed transaction
SQLSetConnectAttr(hdbc, SQL_ATTR_ENLIST_IN_DTC,
(SQLPOINTER)pITransaction,SQL_IS_INTEGER);
SQLSetConnectAttr(hdbc2, SQL_ATTR_ENLIST_IN_DTC,
(SQLPOINTER)pITransaction,SQL_IS_INTEGER);
fprintf (stdout, "Executed SQL Statement:\n%s\nUsing: '%d'\n",sqlstmt,cnum);
SQLAllocHandle (SQL_HANDLE_STMT, hdbc2, &hstmt2 );
if (SQLExecDirect (hstmt2, sqlstmt, SQL_NTS)!=SQL_SUCCESS)
{
fprintf(stdout, "Rolling back global transaction\n");
pITransaction->Abort(NULL,FALSE,FALSE);
}
else
{
pITransaction->Commit( 0, XACTTC_SYNC_PHASEONE, 0 );
pITransaction->Release();
fprintf(stdout, "Transaction Committed\n");
}
SQLFreeHandle(SQL_HANDLE_STMT,hstmt2);
SQLFreeHandle(SQL_HANDLE_STMT,hstmt);
SQLDisconnect (hdbc2);
SQLFreeHandle (SQL_HANDLE_DBC, hdbc2);
return (rc);
}
Example 3-23 shows how to compile the sample and the output of the program.
Example 3-23 Output of Transact_dtc.cpp
odbc32.lib xoleHlp.lib /nologo transaction_dtc.cpp
transaction_dtc.cpp
91
7884ch03.fm
C:\work>transaction_dtc
Connected to DSN=server_1
Connected to DSN=server_2
INSERT INTO orders(order_num,order_date,customer_num) VALUES (0,current,101)
Using: '101'
Transaction Committed
C:\work>
Note: Remember that the DSN must be accessible to the MSDTC service. In
Example 3-22 on page 90, test_1 and test_2 are both created as
System-DSN.
3.3.4 Handling special data types

This section describe how to work with the Informix specific data types such as
Smart Large Objects and complex data types.
Smart Large Object

Smart Large Objects are a type of large objects supported by Informix. Smart
Large Objects are logically stored in a table column but physically stored in a
specific type of dbspaces called Smart Blob Space (SBS).
The information stored in the table column is structure which contains information
about the large object, like the pointer to location in the Smart Blob Space
containing the data or special attributes. Informix has two types of smart large
objects.
BLOB: Stores binary data.
CLOB: Stores character data.
A client application uses these data structures to perform random I/O operations
on the large data, such as open, read, or write operations. There are two options
for accessing Smart Large Objects from an ODBC application:
Use the smart-large-object ODBC API.
If the application requires random access to the large data, it must use the
smart-large-object functions. These functions give the application a greater
control over the smart-large-object data in terms of object properties,
concurrency access, and logging.
Use Smart Large Object Automation.
92
7884ch03.fm
The ODBC driver automatically uses the ODBC API for handling large
objects. The application can access Smart Large Objects as standard ODBC
data types (SQL_LONGVARBINARY and SQL_LONGVARCHAR).
To enable Smart Large Object Automation use the Report Standard ODBC
Types Only DSN option under the DSN Advanced tab or set the
SQL_INFX_ATTR_LO_AUTOMATIC connection attribute.
Data structures for Smart Large Objects

Table 3-10 describes the data structures used by the smart-large-object
functions.
Table 3-10 Smart Large Object data structures
Data structure
Name
Description
lofd
file descriptor
File descriptor to access

smart-large-object data
loptr
pointer structure
Contains the security information and

pointer for a Smart Large Object. This is
the data stored with the table columns.
lospec
specification structure
Contains the storage characteristics for a

Smart Large Object
lostat
status structure
contains the status information for a Smart

Large Object
Client functions to access Smart Large Objects

Table 3-11 lists the SQL functions that you can use in your ODBC program to
access the Informix smart large objects.
Table 3-11 Smart Large Object client functions
Function
Description
ifx_lo_alter(loptr, lospec)
Alters the storage attributes like Logging or

Last-access
ifx_lo_close(lofd)
Closes a Smart Large Object
ifx_lo_col_info(colname, lospec)
Updates column-level storage characteristics
ifx_lo_create(lospec, flags, loptr, lofd)
Creates and opens a new Smart Large Object
ifx_lo_def_create_spec(lospec)
Creates a smart-large-object specification

structure
ifx_lo_open(lofd, loptr, flags)
Opens a Smart Large Object
93
7884ch03.fm
94
Function
Description
ifx_lo_read(lofd, buf)
Reads data from an Smart Large Object
ifx_lo_readwithseek(lofd, buf, offset,

whence)
Reads starting at a specific location
ifx_lo_seek(), ifx_lo_seek(lofd, offset,

whence, seek_pos)
Sets the file position for the next operation
ifx_lo_specget_estbytes(lospec,
estbytes)
Returns the estimated size
ifx_lo_specget_extsz(lospec, extsz)
Returns the allocation extent size
ifx_lo_specget_flags(lospec, flags)
Returns create-time flags
ifx_lo_specget_maxbytes(lospec,
maxbytes)
Returns the maximum number of bytes
ifx_lo_specget_sbspace(lospec,
sbspace)
Returns the sbspace name
ifx_lo_specset_estbytes(lospec,
estbytes)
Sets the estimated number of bytes
ifx_lo_specset_extsz(lospec, extsz)
Sets the allocation extent size
ifx_lo_specset_flags(lospec, flags)
Sets the create-time flags
ifx_lo_specset_maxbytes(lospec,
maxbytes)
Sets the maximum number of bytes
ifx_lo_specset_sbspace(lospec,
sbspace)
Sets the sbspace name
ifx_lo_stat(lofd, lostat)
Initializes a smart-large-object status structure
ifx_lo_stat_atime(lostat, atime)
Returns the last-access time
ifx_lo_stat_cspec(lostat, lospec)
Returns the specification structure
ifx_lo_stat_ctime(lostat, ctime)
Returns the last-time change
ifx_lo_stat_refcnt(lostat, refcount)
Returns the number of references
ifx_lo_stat_size(lostat, size)
Returns the size
ifx_lo_tell(lofd, seek_pos)
Returns the current file position
ifx_lo_truncate(lofd, offset)
Truncates a Smart Large Object at a given

position
ifx_lo_write(lofd, buf)
Writes data
7884ch03.fm
Function
Description
ifx_lo_writewithseek(lofd, buf, offset,

whence)
Writes data at a specific location
Calling client functions from ODBC

An application must use the standard ODBC syntax to call the smart-large-object
functions. The syntax is as follows:
{? = call function_name (?, ?,...)}
The following is the example of calling the ifx_lo_open() function:

{? = call ifx_lo_open(?, ?, ?)}
Creating a Smart Large Object

To create a Smart Large Object, you must create a BLOB descriptor and then
insert the BLOB into the database.
Use the following steps to create a new Smart Large Object:
1. Allocate memory for the specification structure using lospec.
2. Create a lospec using ifx_lo_def_create_spec().
3. Initialize the specification structure using lospec.
4. Allocate memory for the pointer structure using loptr.
5. Create the object using ifx_lo_create().
6. Write the data into the object using ifx_lo_write().
7. Perform a SQL operation (INSERT, UPDATE, and so on).
8. Close the Smart Large Object using ifx_lo_close().
Example 3-24 shows how to insert a BLOB into the database.
Example 3-24 Create_lo.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#include <fcntl.h>
#include <sys/stat.h>
#endif
95
7884ch03.fm

{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
rc = 0;
SQLCHAR
connStrIn[1000], connStrOut[1000], sqlstmt[100];
SQLSMALLINT
connStrOutLen;
SQLINTEGER
cnum = 101;
SQLSMALLINT
SQLUINTEGER
paramsize;
SQLCHAR
colname[25] = "catalog.advert_descr";
SQLINTEGER
colname_size = SQL_NTS;
SQLCHAR*
file_name = (SQLCHAR *) "create_lo.c";
SQLCHAR*
blob_data;
SQLSMALLINT
blob_size;
SQLINTEGER
blob_wsize, mode = LO_RDWR, cbMode = 0;
int
fd=0;
struct stat
statbuf;
/* BLOB file descriptor */
SQLINTEGER
lofd;
SQLINTEGER
lofd_valsize = 0;
/* BLOB pointer structure */
SQLCHAR*
loptr_buffer;
SQLSMALLINT
loptr_size;
SQLINTEGER
loptr_valsize = 0;
/* BLOB specification structure */
SQLCHAR*
lospec_buffer;
SQLSMALLINT
lospec_size;
SQLINTEGER
lospec_valsize = 0;

sprintf((char *) sqlstmt, "INSERT INTO catalog(catalog_num,advert_descr) VALUES
(0,?)");
/* Read the file to insert in the BLOB */
stat(file_name,&statbuf);
blob_size = statbuf.st_size;
blob_data = malloc (blob_size + 1);
fd = _open (file_name, O_RDONLY);
_read (fd, blob_data, blob_size);
_close(fd);
blob_data[blob_size] = '\0';
blob_wsize = blob_size;
96
7884ch03.fm

/* Obtain the size of the lospec and allocate memory */
SQLGetInfo (hdbc, SQL_INFX_LO_SPEC_LENGTH, &lospec_size, sizeof(lospec_size), NULL);
lospec_buffer = malloc (lospec_size);
/* Call ifx_lo_def_create_spec() to create a lospec */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT_OUTPUT, SQL_C_BINARY,
SQL_INFX_UDT_FIXED, (UDWORD)lospec_size, 0, lospec_buffer, lospec_size,
&lospec_valsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_def_create_spec(?)}", SQL_NTS);
SQLFreeStmt (hstmt, SQL_RESET_PARAMS);
/* Call ifx_lo_col_info() to initialise the lospec */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, sizeof(colname),
0, colname, sizeof(colname), &colname_size);
lospec_valsize = lospec_size;
SQL_INFX_UDT_FIXED, (UDWORD)lospec_size, 0, lospec_buffer, lospec_size,
&lospec_valsize);
rc=SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_col_info(?, ?)}", SQL_NTS);
/* Obtain the size of the smart loptr and allocate memory */
SQLGetInfo (hdbc, SQL_INFX_LO_PTR_LENGTH, &loptr_size, sizeof(loptr_size), NULL);
loptr_buffer = malloc (loptr_size);
/* Call ifx_lo_create() to create the BLOB */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_UDT_FIXED,
(UDWORD)lospec_size, 0, lospec_buffer, lospec_size, &lospec_valsize);
rc=SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_SLONG, SQL_INTEGER, (UDWORD)0,
0, &mode, sizeof(mode), &cbMode);
loptr_valsize = loptr_size;
SQL_INFX_UDT_FIXED, (UDWORD)loptr_size, 0, loptr_buffer, loptr_size, &loptr_valsize);
SQLBindParameter (hstmt, 4, SQL_PARAM_OUTPUT, SQL_C_SLONG, SQL_INTEGER, (UDWORD)0,
0, &lofd, sizeof(lofd), &lofd_valsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_create(?, ?, ?, ?)}", SQL_NTS);
/* Call ifx_lo_write to write the BLOB data */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_SLONG, SQL_INTEGER, (UDWORD)0, 0,
&lofd, sizeof(lofd), &lofd_valsize);
SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR,
(UDWORD)blob_size, 0, blob_data, blob_size, &blob_wsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_write(?, ?)}", SQL_NTS);
loptr_valsize = loptr_size;
/* Call SQLExecDirect to do the INSERT
*/
97
7884ch03.fm

(UDWORD)loptr_size, 0, loptr_buffer, loptr_size, &loptr_valsize);
/* Call ifx_lo_close() to close the BLOB */
rc = SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_LONG, SQL_INTEGER,
(UDWORD)0, 0, &lofd, sizeof(lofd), &lofd_valsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_close(?)}", SQL_NTS);
free (lospec_buffer);
free (loptr_buffer);
free (blob_data);
SQLFreeStmt (hstmt, SQL_CLOSE);
SQLFreeHandle (SQL_HANDLE_STMT, hstmt);
return (rc);
}
Accessing a Smart Large Object

Reading a Smart Large Object from the database involves the following steps:
1. Allocate memory for the smart-large-object pointer structure (loptr).
2. Perform the SELECT statement
3. Bind the smart-large-object pointer structure variable with loptr retrieved from
the database.
4. Call ifx_lo_open() to open the smart-large-object with the loptr fetched from
the database. This step open the smart-large-object on the database.
5. Obtain the size of the smart-large-object status structure (lostat) and allocate
enough memory.
6. Call ifx_lo_stat() to obtain the BLOB status structure, so we can retrieve
the estimated size of the BLOB in the next step.
7. Call ifx_lo_stat_size() to retrieve the size of the BLOB data and allocate
enough memory.
8. Call ifx_lo_read() to retrieve the BLOB data.
9. Perform any operation with the BLOB data.
10.Call ifx_lo_close() to close the BLOB.
Example 3-25 on page 99 shows a simple application that select a CLOB column
from the database.
98
7884ch03.fm
Example 3-25 Select_lo.c

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#include <fcntl.h>
#endif
{
SQLHDBChdbc;
SQLHENVhenv;
SQLHSTMThstmt;
/* BLOB file descriptor */
SQLLENlofd;
SQLLENlofd_valsize = 0;
/* BLOB pointer structure */
SQLCHAR*loptr_buffer;
SQLSMALLINTloptr_size;
SQLLENloptr_valsize = 0;
/* BLOB status structure */
SQLCHAR*lostat_buffer;
SQLSMALLINTlostat_size;
SQLLENlostat_valsize = 0;
/* BLOB data */
SQLCHAR*lo_data;
SQLLENlo_data_valsize = 0;
SQLRETURNrc = 0;
SQLCHAR*dsn = "demo_on";
SQLCHAR*selectStmt = (SQLCHAR *) "SELECT advert_descr FROM catalog where catalog_num
= 10075";
SQLINTEGERmode = LO_RDONLY;
SQLLENlo_size;
SQLLENcbMode = 0, cbLoSize = 0;

SQLConnect (hdbc, dsn, SQL_NTS, (SQLCHAR *) "", SQL_NTS, (SQLCHAR *) "", SQL_NTS);
/* Get the size of the loptr */
SQLGetInfo (hdbc, SQL_INFX_LO_PTR_LENGTH, &loptr_size, sizeof(loptr_size), NULL);
99
7884ch03.fm

loptr_buffer = malloc (loptr_size);
SQLExecDirect (hstmt, selectStmt, SQL_NTS);
/* Bind the loptr with the resulset column */
SQLBindCol (hstmt, 1, SQL_C_BINARY, loptr_buffer, loptr_size, &loptr_valsize);
SQLFetch (hstmt);
SQLCloseCursor (hstmt);
/* Call ifx_lo_open() using the loptr fetched from the database */
SQLBindParameter (hstmt, 1, SQL_PARAM_OUTPUT, SQL_C_LONG, SQL_INTEGER, (UDWORD)0, 0,
(UDWORD)loptr_size, 0, loptr_buffer, loptr_size, &loptr_valsize);
SQLBindParameter (hstmt, 3, SQL_PARAM_INPUT, SQL_C_LONG, SQL_INTEGER, (UDWORD)0, 0,
&mode, sizeof(mode), &cbMode);
SQLExecDirect (hstmt, (SQLCHAR *) "{? = call ifx_lo_open(?, ?)}", SQL_NTS);
/* Get the size of the lostat and allocate enough memory */
SQLGetInfo (hdbc, SQL_INFX_LO_STAT_LENGTH, &lostat_size, sizeof(lostat_size), NULL);
lostat_buffer = malloc(lostat_size);
/* Call ifx_lo_stat() to get the BLOB lostat */
SQL_INFX_UDT_FIXED, (UDWORD)lostat_size, 0, lostat_buffer, lostat_size,
&lostat_valsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_stat(?, ?)}", SQL_NTS);
/* Call ifx_lo_stat_size to get the size of the BLOB data and allocate enough memory*/
(UDWORD)lostat_size, 0, lostat_buffer, lostat_size, &lostat_valsize);
SQLBindParameter (hstmt, 2, SQL_PARAM_OUTPUT, SQL_C_LONG, SQL_BIGINT, (UDWORD)0, 0,
&lo_size, sizeof(lo_size), &cbLoSize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_stat_size(?, ?)}", SQL_NTS);
lo_data = malloc (lo_size + 1);
/* Call ifx_lo_read() to get the BLOB data */
SQLBindParameter (hstmt, 2, SQL_PARAM_OUTPUT, SQL_C_CHAR, SQL_CHAR, lo_size, 0,
lo_data, lo_size, &lo_data_valsize);
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_read(?, ?)}", SQL_NTS);
lo_data[lo_size] = '\0';
/* Call ifx_lo_close() to close the BLOB */
SQLExecDirect (hstmt, (SQLCHAR *) "{call ifx_lo_close(?)}",SQL_NTS);
100
7884ch03.fm
fprintf(stdout,"%s",lo_data);
free (loptr_buffer);
free (lostat_buffer);
free (lo_data);
return (rc);
}
Use Smart Large Object automation

Developing an application to handle Smart Large Object using the Smart Large
Object automation method does not require any special attention, just use
SQL_LONGBINARY and SQL_LONGVARCHAR as data type.
Example 3-26 Illustrates how to insert a large-object with the
SQL_INFX_ATTR_LO_AUTOMATIC enabled
Example 3-26 Lo_auto.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#include <fcntl.h>
#endif
int main (long
{
SQLHDBC
SQLHENV
SQLHSTMT
SQLCHAR*
SQLCHAR
SQLCHAR
SQLLEN
argc, char* argv[])
hdbc;
henv;
hstmt;
dsn="demo_on";
sqlstmt[128];
str[128];
str_len = SQL_NTS;
101
7884ch03.fm

SQLSMALLINT
SQLINTEGER
SQLRETURN
BufferLength = 128;
int_val;
rc = 0;

/* Set SQL_INFX_ATTR_ODBC_TYPES_ONLY connection attribute to FALSE*/
SQLSetConnectAttr(hdbc, SQL_INFX_ATTR_ODBC_TYPES_ONLY, (SQLPOINTER) SQL_FALSE,
SQL_IS_UINTEGER);
/* Set SQL_INFX_ATTR_LO_AUTOMATIC connection attribute to TRUE*/
SQLSetConnectAttr(hdbc, SQL_INFX_ATTR_LO_AUTOMATIC, (SQLPOINTER) SQL_TRUE,
SQL_IS_UINTEGER);
/* Preare the SQL statement */
sprintf(sqlstmt, "INSERT INTO catalog(catalog_num,advert_descr) VALUES (0,?)");
SQLPrepare (hstmt, sqlstmt, SQL_NTS);
/* Bind the input parameter and Execute the SQL */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_LONGVARCHAR,
BufferLength, 0, &str, 0, &str_len);
SQLExecute (hstmt);
return (rc);
}
Complex data types

Rows and Collections are composite values that consist of one or more
elements. You can use the execute SQL statements to access an entire row or
collection. However, you can not access individual elements within a row or
collection.
To access composite values from the ODBC driver you must use the Client
functions for Rows and Collections.
Client functions for Rows and Collections

The ODBC functions for handling Rows and Collections start with the prefix
ifx_rc_ . Table 3-12 on page 103 lists the available functions.
102
7884ch03.fm

Table 3-12 Row and Collection Client functions
Function
Description
ifx_rc_count()
Returns the number of elements or fields that are in a row or

collection
ifx_rc_create()
Creates a buffer for a row or collection
ifx_rc_delete()
Deletes an element from a collection
ifx_rc_describe()
Returns information about the complex data type or and individual

element
ifx_rc_fetch()
Returns the value of an element that is in a row or collection
ifx_rc_free()
Frees a row or collection handle
ifx_rc_insert()
Inserts a new element into a collection
ifx_rc_isnull()
Evaluate if a complex type is NULL or not
ifx_rc_setnull()
Set a complex type to NULL
ifx_rc_typespec()
Returns the type specification for a row or collection
ifx_rc_update()
Updates the value for an element that is in a row or collection
Creating a complex data type

Example 3-27 shows how to perform an INSERT operation using the Informix
ODBC functions ifx_rc_xxx. We are using a simple table customer_rc defined as
follows:
CREATE ROW TYPE name_t (
fname
VARCHAR(15),
lname
VARCHAR(15)
);
CREATE TABLE customer_rc (
customer_num
SERIAL,
customer_name name_t,
contact_dates LIST(DATETIME YEAR TO DAY NOT NULL)
);
Example 3-27 Create_rc.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
103
7884ch03.fm

#include <conio.h>
#include <fcntl.h>
#endif
{
SQLHDBChdbc;
SQLHENVhenv;
SQLHSTMThstmt;
/* Row variables */
HINFX_RChrow;
HINFX_RChlist;
SQLCHAR*dsn = "demo_on";
SQLRETURNrc = 0;
SQLINTEGERi, in;
SQLLENdata_size = SQL_NTS;
SQLSMALLINTposition = SQL_INFX_RC_ABSOLUTE;
SQLSMALLINTjump;
SQLCHARrow_data[2][15] = {"Ludwig", "Pauli"};
SQLLENrow_data_size = SQL_NTS;
SQLCHARlist_data[2][25] = {"2008-08-16","2008-08-16"};
SQLLENlist_data_size = SQL_NTS;
SQLCHAR*insertStmt = (SQLCHAR *) "INSERT INTO customer_rc VALUES (0, ?, ?)";
SQLLENcbHrow = 0, cbHlist = 0, cbPosition = 0, cbJump = 0;
/* Connection */
/* Call ifx_rc_create() to allocate a row handle */
SQLBindParameter (hstmt, 1, SQL_PARAM_OUTPUT, SQL_C_BINARY, SQL_INFX_RC_ROW,
sizeof(HINFX_RC), 0, &hrow, sizeof(HINFX_RC), &cbHrow);
SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 0, 0, (SQLCHAR *)
"ROW(fname VARCHAR(15), lname VARCHAR(15))", 0, &data_size);
SQLExecDirect (hstmt, (SQLCHAR *) "{? = call ifx_rc_create(?)}", SQL_NTS);
/* Call ifx_rc_create() to allocate a list handle */
SQLBindParameter (hstmt, 1, SQL_PARAM_OUTPUT, SQL_C_BINARY, SQL_INFX_RC_LIST,
sizeof(HINFX_RC), 0, &hlist, sizeof(HINFX_RC), &cbHlist);
"LIST (DATETIME YEAR TO DAY NOT NULL)", 0, &data_size);
/* call ifx_rc_update() to insert elements into the row */
for (i=0; i<2; i++)
104
7884ch03.fm

{
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_ROW,

sizeof(HINFX_RC), 0, hrow, sizeof(HINFX_RC), &cbHrow);
SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 25, 0,
row_data[i], 0, &row_data_size);
SQLBindParameter (hstmt, 3, SQL_PARAM_INPUT, SQL_C_SHORT, SQL_SMALLINT, 0, 0,
&position, 0, &cbPosition);
jump = i + 1;
&jump, 0, &cbJump);
SQLExecDirect (hstmt, (SQLCHAR *)"{call ifx_rc_update(?, ?, ?, ?)}", SQL_NTS);
}
/* Call ifx_rc_insert() to insert elements into the list */
for (i=0; i<2; i++)
{
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_LIST,
sizeof(HINFX_RC), 0, hlist, sizeof(HINFX_RC), &cbHlist);
SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR,SQL_DATE, 25, 0,
list_data[i], 0, &list_data_size);
&position, 0, &cbPosition);
jump = i + 1;
&jump, 0, &cbJump);
SQLExecDirect (hstmt, (SQLCHAR *)"{call ifx_rc_insert( ?, ?, ?, ? )}", SQL_NTS);
}
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_COLLECTION,
SQLBindParameter (hstmt, 2, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_COLLECTION,
SQLExecDirect (hstmt, insertStmt, SQL_NTS);
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_ROW,
SQLExecDirect(hstmt, (SQLCHAR *)"{call ifx_rc_free(?)}", SQL_NTS);
/* Free the list handle */
SQLBindParameter (hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_LIST,
return (rc);
}
105
7884ch03.fm
Example 3-28 shows the information inserted into the customer_rc table after the
execution of the example.
Example 3-28 customer_rc table
> select * from customer_rc
> ;
customer_num 1
customer_name ROW('Ludwig','Pauli')
contact_dates LIST{'2008-08-16','2008-08-16'}
1 row(s) retrieved.
>
Accessing a complex data type

Example 3-29 demonstrates how to select a Row type with the ODBC driver.
Example 3-29 Select_row.c.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#include <fcntl.h>
#endif

{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt, hstmt_row;
HINFX_RC
hrow;
SQLCHAR
*dsn="demo_on";
SQLRETURN
rc = 0;
SQLINTEGER
i;
SQLSMALLINT
position = SQL_INFX_RC_ABSOLUTE, jump = 0;
SQLLEN
cbHrow = 0, cbPosition = 0, cbJump = 0, cbRCData = 0;
SQLLEN
data_size = SQL_NTS, cbrow = 0;
106
7884ch03.fm
SQLCHAR*
sqlstmt = (SQLCHAR *) "SELECT customer_name FROM customer_rc WHERE
customer_num = 1";
SQLCHAR
row_data[200];
/* Connection */
SQLAllocHandle (SQL_HANDLE_STMT, hdbc, &hstmt);
SQLAllocHandle (SQL_HANDLE_STMT, hdbc, &hstmt_row);
/* Call ifx_rc_create() to allocate a row handle */
SQLBindParameter (hstmt, 1, SQL_PARAM_OUTPUT, SQL_C_BINARY, SQL_INFX_RC_ROW,
sizeof(HINFX_RC), 0, &hrow, sizeof(HINFX_RC), &cbHrow);
"row", 0, &data_size);
/* Bind the row handle with the resulset column */
SQLBindCol (hstmt, 1, SQL_C_BINARY, (SQLPOINTER) hrow, sizeof(HINFX_RC), &cbHrow);
SQLFetch (hstmt);
/* Get the row data */
fprintf(stdout, "Row data:\n");
for (i=0; i<2; i++)
{
strcpy((char *) row_data, "<null>");
SQLBindParameter (hstmt_row, 1, SQL_PARAM_OUTPUT, SQL_C_CHAR, SQL_CHAR, 0, 0,
row_data, 200, &cbRCData);
SQLBindParameter (hstmt_row, 2, SQL_PARAM_INPUT, SQL_C_BINARY,
SQL_INFX_RC_COLLECTION, sizeof(HINFX_RC), 0, hrow, sizeof(HINFX_RC), &cbHrow);
SQLBindParameter (hstmt_row, 3, SQL_PARAM_INPUT, SQL_C_SHORT, SQL_SMALLINT, 0,
0, &position, 0, &cbPosition);
jump = i + 1;
SQLBindParameter (hstmt_row, 4, SQL_PARAM_INPUT, SQL_C_SHORT, SQL_SMALLINT, 0, 0,
&jump, 0, &cbJump);
/* Call ifx_rc_fetch() to fetch individual elements from the row */
SQLExecDirect (hstmt_row, (SQLCHAR *) "{ ? = call ifx_rc_fetch( ?, ?, ? ) }",
SQL_NTS);
SQLFreeStmt (hstmt_row, SQL_RESET_PARAMS);
fprintf(stdout, "\t\t%s\n", row_data);
}
/* Call ifx_rc_free() to free the row handle */
SQLBindParameter(hstmt, 1, SQL_PARAM_INPUT, SQL_C_BINARY, SQL_INFX_RC_ROW,
sizeof(HINFX_RC), 0, hrow, sizeof(HINFX_RC), &cbrow);
107
7884ch03.fm

return (rc);
}
Example 3-30 shows the output of Example 3-29 on page 106.

Example 3-30 Output of Select_row.c
odbc32.lib /nologo select_row.c
select_row.c
C:\work>select_row
Row data:
Ludwig
Pauli
C:\work>
3.3.5 Error handling

In this section we show how to obtain additional information when an ODBC
function fails.
Error handling sample

When an ODBC function fails, it stores information in the diagnostic record. This
record contains information such as error messages, warning, and status
information about the success or failure of the ODBC call.
An ODBC application can retrieve the diagnostic record or individual fields using
the SQLGetDiagRec() and SQLGetDiagField() ODBC functions.
Example 3-31 shows a typical usage of SQLGetDiagRec().
Example 3-31 Connect_error.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
108
7884ch03.fm

#include <conio.h>
#endif
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLRETURN
rc = 0;
SQLCHAR
connStrIn[1000];
SQLCHAR
connStrOut[1000];
SQLSMALLINT
connStrOutLen;
UCHAR
SqlState[200] = "", ErrorMsg[200] = "";
SQLINTEGER
IsamError = 0;
SDWORD
NativeError = 0L;
SWORD
ErrorMsgp = 0;
SQLSMALLINT
recnum = 1;
if (argc != 2)
{
fprintf (stdout, "Please specify the name of a DSN!\n");
return(1);
}
else
{
else
}

{
SQLGetDiagRec(SQL_HANDLE_DBC, hdbc, recnum, SqlState, &NativeError, ErrorMsg,
199, &ErrorMsgp);
fprintf(stdout, "SqlState = %s\n Native Error = %d\n Error Message = %s\n",
SqlState, NativeError, ErrorMsg);
fprintf (stdout, "Connection failed!\n");
return (1);
}
fprintf (stdout, "Connection Successful\n");
return (rc);
109
7884ch03.fm

}
Example 3-32 shows the output of Example 3-31 on page 108. In addition to the
ODBC SQLSTATE error we return the Informix native error and the description
for the error.
Example 3-32 Output of Connect_error.c
odbc32.lib /nologo connect_error.c
connect_error.c
C:\work>connect_error wrongDNS
SqlState = IM002
Native Error = 0
Error Message = [Microsoft][ODBC Driver Manager] Data source name not found
and no default driver specified
Connection failed!
C:\work>connect_error demo_on;UID=wronguser
SqlState = 28000
Native Error = -951
Error Message = [Informix][Informix ODBC Driver][Informix]Incorrect password
or user wronguser@localhost is not known on the database server.
Connection failed!
C:\work>
Because this function is called quite often, it make sense to have a function to
display the error message. Example 3-33 Illustrates a typical error handling
function.
Example 3-33 Simple_select_werr.c
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef WIN32
#include <io.h>
#include <conio.h>
#endif
void CheckDiag (SQLSMALLINT handle_type, SQLHANDLE handle, char *text)
{
RETCODErc = SQL_SUCCESS;
UCHARSqlState[200] = "", ErrorMsg[200] = "";
110
7884ch03.fm

SQLINTEGERIsamError = 0;
SDWORDNativeError = 0L;
SWORDErrorMsgp = 0;
SQLSMALLINTrecnum = 1;
fprintf (stdout,"Error in %s \n",text);

while (rc != SQL_NO_DATA_FOUND) {
rc = SQLGetDiagRec(handle_type, handle, recnum, SqlState, &NativeError,
ErrorMsg, 199, &ErrorMsgp);
if (rc != SQL_NO_DATA_FOUND) {
SQLGetDiagField(handle_type, handle, recnum, SQL_DIAG_ISAM_ERROR,
&IsamError, SQL_IS_INTEGER, NULL);
fprintf (stdout," SqlState = %s\n Native Error = %d\n Error Message = %s\n
ISAM Error = %d\n", SqlState, NativeError, ErrorMsg, IsamError);
}
recnum++;
}
}
{
SQLHDBC
hdbc;
SQLHENV
henv;
SQLHSTMT
hstmt;
SQLRETURN
SQLCHAR
SQLCHAR
SQLSMALLINT
rc = 0;
connStrIn[1000];
connStrOut[1000];
connStrOutLen;
int
SQLCHAR
sqllen;
*sqlstmt;
if (argc != 3)
{
fprintf (stdout, "Please specify the name of a DSN and the SQL Statement to
run!\n");
return(1);
}
else
{
else
sqllen = strlen((char *)argv[2]);
sqlstmt = (SQLCHAR *) malloc (sqllen + sizeof(char));
strcpy((char *)sqlstmt, (char *)argv[2]);
}
111
7884ch03.fm

SQLDriverConnect (hdbc, NULL, connStrIn, SQL_NTS, connStrOut, 1000, &connStrOutLen,
SQL_DRIVER_NOPROMPT);
rc = SQLExecDirect (hstmt, sqlstmt, SQL_NTS);
if (rc != SQL_SUCCESS && rc != SQL_SUCCESS_WITH_INFO)
{
CheckDiag(SQL_HANDLE_STMT, hstmt,"SQLExecDirect()");
fprintf (stdout, "SQLExecDirectW() failed!\n");
return (1);
}
return (rc);
}
We create a function called CheckDiag() to retrieve the diagnostic information:

void CheckDiag (SQLSMALLINT handle_type, SQLHANDLE handle, char *text)
Example 3-34 shows the output of the previous sample.

Example 3-34 Output of Simple_select_werr.c
C:\work>simple_sql_werror demo_on "wrong sql"
Connected
Error in SQLExecDirect()
SqlState = 42000
Native Error = -201
Error Message = [Informix][Informix ODBC Driver][Informix]A syntax error has
occurred.
ISAM Error = 0
C:\work>simple_sql_werror demo_on "DELETE from wrongtable"
Connected
SqlState = 42S02
Native Error = -206
Error Message = [Informix][Informix ODBC Driver][Informix]The specified table
(wrongtable) is not in the database.
ISAM Error = -111
112
7884ch03.fm

C:\work>
Note: If an error is generated by the database server, the error message

contains the string [Informix] before the description of the error. For example:
SqlState = 42S02
Native Error = -206
Error Message = [Informix][Informix ODBC Driver][Informix]The specified
table (test) is not in the database.
ISAM Error = -111
ODBC SQLSTATE errors with Informix ODBC Driver

The error returned by the ODBC function is based on the X/Open standard
SQLSTATE errors. For the SQLSTATE codes, refer to the Informix ODBC Driver
Guide at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.odbc.doc/si
i091033222.htm#sii091033222
3.3.6 Troubleshooting
In this section we discuss typical problems when using the Informix ODBC
drivers and available traces.
Environment
Make sure the setup of your Informix Client is correct. If your application fails to
load the ODBC driver, check if the Informix Client or Data Server Client libraries
are accessible to you application.
The environment variable PATH should contain the bin directory of your Client
package.
The shared library PATH variable on an UNIX machine should contain the
directory where the shared libraries are located.
Make sure that the INFORMIXDIR variable is set correctly (as an environment
variable or stored in the registry using the Setnet32 tool).
32-bit applications requires 32-bit drivers to work. Make sure the drivers you
have install are of the same type as your application.
If your application is failing to connect, make sure the connection details are
valid.
113
7884ch03.fm
Informix Client SDK on Windows contains a simple connection tool called ilogin
which can be used to check if you have connection with the database server. If
ilogin fails to connect, all the other drivers would also fail.
Check that the values stored in the registry with Setnet32 are valid.
If you are using 32-bit and 64-bit drivers on the same machine remember that
on Windows, there are two registry hives to stored the connectivity
information, so make sure both are correct.
Tracing
When developing an application using the ODBC driver, most problems are
caused by passing incorrect parameters to the ODBC functions.
ODBC Trace
You can generate a trace file of all the calls to the ODBC driver using the ODBC
Trace facility.
On a Windows machine, you can enable ODBC Trace using the ODBC Data
Source Administrator. Figure 3-14 shows the Tracing tab.
Figure 3-14 Enabling trace
Note: Remember to select Machine-Wide tracing for all user identities if

your application runs as a service or with a different user.
Example 3-35 on page 115 shows some of the entries in an ODBC trace file.
114
7884ch03.fm

Example 3-35 ODBC trace
...
rccreate
(SQL_SUCCESS)
99c-db8 EXIT
HSTMT
UWORD
SWORD
SWORD
SWORD
SQLULEN
SWORD
PTR
SQLLEN
SQLLEN *
rccreate
SQLBindParameter
with return code 0
01D51DC0
2
1 <SQL_PARAM_INPUT>
1 <SQL_C_CHAR>
1 <SQL_CHAR>
25
0
0x002CFEC8
0
0x002CFF50 (-3)
99c-db8 ENTER SQLBindParameter

HSTMT
01D51DC0
UWORD
3
SWORD
1 <SQL_PARAM_INPUT>
SWORD
5 <SQL_C_SHORT>
SWORD
5 <SQL_SMALLINT>
SQLULEN
0
SWORD
0
PTR
0x002CFF68
SQLLEN
0
SQLLEN *
0x002CFF54
...
On a UNIX machine, ODBC trace is active if the Trace parameter in the [ODBC]
section of the odbc.ini configuration file is set to 1.
Example 3-36 shows the Trace settings for the odbc.ini file.
Example 3-36 Trace in odbc.ini
;
; Trace file Section
;
Trace=1
TraceFile=/tmp/odbctrace.out
InstallDir=/opt/IBM/informix
TRACEDLL=idmrs09a.so
SQLIDEBUG
All the Informix clients uses the SQLI protocol to communicate with the database
server.
115
7884ch03.fm
In addition to the ODBC trace, you can generate a SQLIDEBUG trace which
contains all the message between the client application and the database server.
You can enable SQLIDEBUG trace at the Client side by defining the environment
variable as follows:
SQLIDEBUG=2:path_to_trace_files
Or at the server side by running the following command:

onmode -p 1 sqli_dbg
Note: On a Windows machine, when using the server side tracing, the
SQLIDEBUG files are created in C:\temp\sqli. This directory must exists
before enabling the trace.
On a UNIX machine the trace files are created in /tmp/sqli
Example 3-37 demonstrates how to set SQLIDEBUG trace on the client.
Example 3-37 SQLIDEBUG sample
C:\work>set SQLIDEBUG=2:sqli_trace
C:\work>simple_sql_werror demo_on "wrong_sql"
Connected
SqlState = 42000
Native Error = -201
Error Message = [Informix][Informix ODBC Driver][Informix]A syntax error has
occurred.
ISAM Error = 0
C:\work>dir sqli_trace_2940_1016_1f428b8
Volume in drive C is W2003
Volume Serial Number is 50DA-70D7
Directory of C:\work
22/06/2010
18:39
348 sqli_trace_2940_1016_1f428b8
1 File(s)
348 bytes
0 Dir(s) 76,628,099,072 bytes free
C:\work>
The SQLIDEBUG trace files contains the SQLI packages between the server and
the client. To obtain an readable output, you must run the sqliprt tool.
116
7884ch03.fm
Example 3-38 demonstrates how to run sqliprt.

Example 3-38 sqliprint output
C:\work>sqliprt -o trace.txt sqli_trace_2940_1016_1f428b8
C:\work>type trace.txt
SQLIDBG Version 1
C->S (4)
Time: 2010-06-22 18:39:05.43700

SQ_INTERNALVER
Internal Version Number: 316
C->S (14)
SQ_PROTOCOLS
SQ_EOT
Time: 2010-06-22 18:39:05.43700
S->C (14)
SQ_PROTOCOLS
SQ_EOT
Time: 2010-06-22 18:39:05.43700
C->S (90)
SQ_INFO
Time: 2010-06-22 18:39:05.43700

INFO_ENV
Name Length = 12
Value Length = 34
"DBTEMP"="C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp"
"SUBQCACHESZ"="10"
"OPTOFC"="0"
INFO_DONE
SQ_EOT
S->C (2)
Time: 2010-06-22 18:39:05.43700

SQ_EOT
C->S (4)
Time: 2010-06-22 18:39:05.43700

SQ_BEGIN
SQ_EOT
S->C (10)
SQ_XACTSTAT
SQ_EOT
Time: 2010-06-22 18:39:05.43700
C->S (4)
Time: 2010-06-22 18:39:05.43700

SQ_CMMTWORK
SQ_EOT
S->C (10)
SQ_XACTSTAT
SQ_EOT
Time: 2010-06-22 18:39:05.43700
C->S (22)
SQ_PREPARE
# values: 0
Time: 2010-06-22 18:39:05.43700
117
7884ch03.fm

CMD.....: "wrong_sql" [9]
SQ_NDESCRIBE
SQ_WANTDONE
SQ_EOT
S->C (12)
SQ_ERR
Time: 2010-06-22 18:39:05.43700

SQL error..........:
ISAM/RSAM error....:
Offset in statement:
Error message......:
-201
0
1
"" [0]
SQ_EOT
C:\work>
Note: On a UNIX machine, the equivalent of the sqliprt tool is sqliprint.
DRDADEBUG
All the IBM Data Server Drivers use the DRDA protocol to communicate with the
database server.
Similar to SQLIDEBUG, you can enable DRDADEBUG trace at the server side
by running the following onmode command:
onmode -p 1 drda_dbg
The DRDA trace files are created in /tmp/drda for a UNIX machine and
C:\temp\drda for a Windows. You need to use the drdaprint tool to convert the
trace files to a readable format.
Example 3-39 shows the use of drdaprint.
Example 3-39 DRDAPRINT
C:\temp\drda>dir
Directory of C:\temp\drda
22/06/2010
22/06/2010
22/06/2010
18:50
<DIR>
.
18:50
<DIR>
..
18:50
2,269 drda.47
1 File(s)
2,269 bytes
2 Dir(s) 76,628,135,936 bytes free
C:\temp\drda>drdaprint
Usage: drdaprint [-f] [-o outfile] inpfile
-f: format hex dump
118
7884ch03.fm
C:\temp\drda>drdaprint -o trace.txt drda.47

C:\temp\drda>
Example 3-40 shows a section of the DRDA trace file.

Example 3-40 DRDATrace file
C:\temp\drda>type trace.txt
DRDADBG Version 1
1
data IDS DRDA Communication Manager

function sqljcIntReceive()
bytes 270
time 2010-06-22 18:50:13.31200
RECEIVE BUFFER(AS):
0000
0010
0020
0030
0040
0050
0060
0070
0080
0090
00A0
00B0
00C0
EXCSAT RQSDSS
0 1 2 3 4 5 6 7 8 9 A B C D E F
00C3D041000100BD 10410080115EA289
949793856DA29893 6DA685999996994B
85A7F0F6F1F4F0C4 F9F4F0F0F0000000
0000000000000000 0000000000000000
0000000000000000 000000000060F0F0
F0F1C1C4D4C9D5C9 E2E3D9C1E3D6D940
4040404040404040 4040404040404040
E2E3D6D9C5E26DC4 F0C4C2F240404040
4040404040404040 40F0001814041403
000A2407000A1474 0005240F00081440
0009000B1147D8C4 C2F261D5E3000A11
6DC4E4C2C9E3D600 0C115AE2D8D3F0F9
F0F7F0
(ASCII)
0123456789ABCDEF
...A.....A...^..
....m...m......K
................
................
.............`..
...............@
@@@@@@@@@@@@@@@@
......m.....@@@@
@@@@@@@@@.......
..$....t..$....@
.....G....a.....
m.........Z.....
...
(EBCDIC)
0123456789ABCDEF
.C}..........;si
mple_sql_werror.
ex06140D94000...
................
.............-00
01ADMINISTRATOR
STORES_D0DB2
0......
...............
......QDB2/NT...
_DUBITO...]SQL09
070
119
7884ch03.fm
120
7884ch04.fm
Chapter 4.
Working with ESQL/C

This chapter discusses about ESQL/C, an SQL Application Programming
Interface (API) that enables you to embed Structured Query Language (SQL)
statements directly into a C program. ESQL/C is bundled with the Informix Client
Software Development Kit (Client SDK). The ESQL/C preprocessor, esql,
converts each SQL statement and all IBM Informix-specific code to C-language
source code and invokes the C compiler to compile it. The advantage of using
ESQL/C is that it supports all the data types as well as extended data types of
Informix and is optimized for an IBM Informix database. If you are going to use
only Informix as your database server then this might be your best choice.
We discuss the basic elements of an Informix ESQL/C application and show how
to perform database operations within an ESQL/C program. We end the chapter
by looking at various methods in handling exceptions and troubleshooting.
121
7884ch04.fm
4.1 Informix ESQL/C

Informix ESQL/C consists of the following software components:
The Informix ESQL/C libraries of C functions, which provide access to the
database server.
The Informix ESQL/C header files, which provide definitions for the data
structures, constants, and macros useful to the Informix ESQL/C program.
The esql command, which processes the Informix ESQL/C source code to
create a C source file that it passes to the C compiler.
The finderr utility on the UNIX system and the Informix Error Messages
Windows-based Informix error messages utility that enable you to obtain
information about IBM Informix-specific error messages.
On Windows platforms, Informix provide the following additional utilities:
The Setnet32 utility is a Windows-based utility that enables you to set
configuration information.
The ILOGIN utility is a demonstration program that displays a dialog box with
fields for the connection parameters, for testing a connection to the database
server (uses the stores7 database).
Figure 4-1 gives an overview of the relationship between ESQL/C and the native
C and how the transfer of control happens. Once you have created an Informix
ESQL/C program file, you run the esql command on that file. By default, the
Informix ESQL/C preprocessor runs first and translates the embedded SQL
statements in the program into Informix ESQL/C function calls that communicate
with the database server. The Informix ESQL/C preprocessor produces a C
source file and calls the C compiler. The C compiler then compiles your source
file and links any other C source file, object file, or library file the same way as
any other C program. If esql does not encounter errors in one of these steps, it
generates an executable file. You can run the compiled Informix ESQL/C
program as you would run any C program. When the program runs, it calls the
Informix ESQL/C library procedures; the library procedures set up
communications with the database server to carry out the SQL operations.
Figure 4-1 Relationship between Informix ESQL/C and C
122
7884ch04.fm

Informix ESQL/C is installed, by default, as a part of the Client SDK package.
The release notes of Client SDK contains the information about which versions of
the IBM Informix database server are supported. Make sure that you install the
Client SDK that is supported by the database you are going to work with.
On a Windows machine, the default Client SDK installation directory is
C:\Program Files\IBM\Informix\Client-SDK. Make sure that the INFORMIXDIR
environment variable is pointing to the directory where the product was
installed.The PATH environment variable should contain the following directories:
The path to the bin directory under the installation directory, that is, the PATH
variable should have $INFORMIXDIR/bin.
The path at where native C compiler is located. (IBM Informix Client products
are certified with the Microsoft Visual C++ 2005 SP1.
UNIX configuration
On UNIX platforms, the default installation directory for Informix Client SDK is
/opt/IBM/informix. The INFORMIXDIR environment variable should point to the
directory where the product was installed. Add the following directories to the
PATH environment variable:
The bin directory under the installation directory, $INFORMIXDIR/bin.
The path to the native C compiler.
The sqlhosts configuration file contains the information required to connect to an
IBM Informix database server. You must set this file to include an entry of your
Informix database server name, ESQL/C required protocol, host name, and port
number. By default the sqlhosts file is under the $INFORMIXDIR/etc/ directory.
You can use the INFORMIXSQLHOSTS environment variable to point to a different
location.
Example 4-1 shows a simple sqlhosts file.
Example 4-1 sqlhosts file
#server protocol hostname service/port
demo_on onsoctcp kodiak 9088
For more information about the sqlhosts file, refer to the Client/Server
Communications manual at:
Chapter 4. Working with ESQL/C
123
7884ch04.fm

ds_admin_0158.htm
4.3 Developing an ESQL/C application

In this section, we describe about connecting to database, performing basic
database operations such as delete, insert, and updates using ESQL/C. We also
look into the handling of extended data types by ESQL/C.
We use the sample instance and database that Informix server provides to
demonstrate ESQL/C application development. The sample instance, demo_on, is
created when the Informix server is installed. On windows the sample instance is
ol_svr_custom. You can create a sample database, stores_demo, under the
demo_on instance and populate it with tables and data by running the utility
dbaccessdemo residing in $INFORMIXDIR/bin/.
4.3.1 Creating an ESQL/C application

The steps involved in building an Informix ESQL/C application are:
1. Develop a C program with the embedded Informix SQL statements and name
it with .ec or .ecp extension. The SQL statement are qualified with EXEC SQL
keyword or the $ symbol, for example:
EXEC SQL include sqlca;
EXEC SQL SELECT fname into :c1 FROM customer WHERE customer_num=:i1
$SET ISOLATION DIRTY READ;
2. Preprocess the Informix ESQL/C source file with the esql command. The esql
command also invoke the C compiler to compile the program into object code.
As necessary, correct errors reported by the preprocessor and the compiler
and repeat step 2.
3. Link the object code into one or more executable files using the esql
command. The executable files have .exe extension.
4. Run the application.
You run the compiled Informix ESQL/C program as you would any C program.
When the program runs, it calls the Informix ESQL/C library procedures; the
library procedures set up communications with the database server to carry
out the SQL operations.
Example 4-2 on page 125 shows a simple program, customer.ec, that connects
to the database and retrieves some data. All the SQL related statements are
embedded in the C program with EXEC SQL keyword.
124
7884ch04.fm

Example 4-2 Simple ESQL/C program
#include <stdio.h>
EXEC SQL include sqltypes;
int main()
{
EXEC SQL BEGIN DECLARE SECTION;
int i1=101;
char c1[50];
char c2[50];
EXEC SQL END DECLARE SECTION;
int i2;
EXEC SQL connect to 'stores_demo';
EXEC SQL select fname,lname into :c1, :c2 from customer where
customer_num=:i1
;
if (SQLCODE == 100)
{
printf("SQLCODE=%d\n",SQLCODE);
printf("Data not found\n");
}
else
{
printf("Data found \n");
printf("Last Name \t%s\n",c1);
printf("First Name \t%s\n",c2);
}
}
Compiling ESQL/C programs

The esql command translates Informix ESQL/C code to C code and then calls the
C compiler to compile and link the C code.
The C compiler takes the following actions:
Compiles the C language statements to object code.
Links to Informix ESQL/C libraries as well as any other files or libraries you
specify.
Creates an executable file.
The general syntax for an esql to compile looks like below:
esql <options> <source.ec> <options> <outfile>
125
7884ch04.fm
By default, esql creates an executable by name a.out in the current directory. You
can explicitly specify the name of the executable file with the -o option. For
example, the following command compile the customer.ec shown in Example 4-2
on page 125 and produce the executable file customer.exe.
esql -o customer.ec customer.exe
If esql is running on a Windows operating system, the name of the target file
defaults to the name of the first Informix ESQL/C source file on the esql
command line. The extension is changed to either .exe or .dll depending on the
type of target being generated.
you can use a compiler other than the native C compiler by setting the
INFORMIXC environment variable. Table 4-1 lists the native C compilers on
various platforms.
Table 4-1 Native C compiler
Platforms
Native compiler
Solaris
CC
HP
aC++
Windows
VC++ / VS2008/ VS2005
AIX
xlc
Open Source
gcc/g++
If you want to pass C compiler options that have the same names as Informix
ESQL/C processor options, precede them with the -cc option. For example, the
following esql command passes the -od and -g options to the C compiler but
uses the -db2 option itself:
esql -cc -od -g demo1.ec -db2
Shared libraries
IBM Informix products use the Informix general libraries for interactions between
the client SQL application programming interface (API) products (IBM Informix
ESQL/C and IBM Informix ESQL/COBOL) and the database server. You can
choose between the following types of Informix general libraries to link with your
Informix ESQL/C application:
Static Informix general libraries
To link a static library, the linker copies the library functions to the executable
file of your Informix ESQL/C program. The static Informix general libraries
126
7884ch04.fm
allow an Informix ESQL/C program on computers that do not support shared

memory to access the Informix general library functions.
To link static libraries use option -static, for example:
esql -static file.ec -o file.exe
Shared Informix general libraries

To link a shared library, the linker copies information about the location of the
library to the executable file of your Informix ESQL/C program. The shared
Informix libraries allow several applications to share a single copy of these
libraries, which the operating system loads, just once, into shared memory.
Thread-safe versions of static and shared Informix general libraries
The thread-safe versions of Informix general libraries allow an Informix
ESQL/C application that has several threads to call these library functions
simultaneously. The thread-safe versions of Informix libraries are available as
both static libraries and shared libraries.
There are some platform specific considerations when you link shared Informix
general libraries to an ESQL/C module. The environment variable that specifies
the search path at run time is different for platform, see Table 4-2.
Table 4-2 Environment variable for shared libraries.
Platform
Environment variable
AIX
LIBPATH
Solaris
LD_LIBRARY_PATH
HP-UX
SHLIB_PATH
Mac OS X
DYLD_LIBRARY_PATH
Linux
LD_LIBRARY_PATH
Windows
LIB
Set $INFORMIXDIR/lib and any of its subdirectories to specify the shared-library

path. For example, on Linux, set the LD_LIBRARY_PATH environment variable
as follows:
Bourne shell:
LD_LIBRARY_PATH=$INFORMIXDIR/lib:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH
C shell:
setenv LD_LIBRARY_PATH $INFORMIXDIR/lib:$LD_LIBRARY_PATH
127
7884ch04.fm
On Windows use the following command:

set LIB = %INFORMIXDIR%\lib\;%LIB%
To link shared Informix general libraries with an Informix ESQL/C module, you do
not need to specify a command-line option. Informix ESQL/C links shared
libraries by default. The following command compiles the file.ec source file with
shared Informix libraries:
esql myfile.ec -o myfile.exe
Choosing between shared and static library versions

Shared libraries are most useful in multiuser environments where only one copy
of the library is required for all applications. Shared libraries bring the following
benefits to your Informix ESQL/C application:
Shared libraries reduce the sizes of executable files because these library
functions are linked dynamically, on an as-needed basis.
At runtime, a single copy of a shared library can be linked to several
programs, which results in less memory use.
The effects of shared libraries in an Informix ESQL/C executable are
transparent to the end user.
Although shared libraries save both disk and memory space, when an Informix
ESQL/C application uses them, it must perform the following overhead tasks:
Dynamically load the shared library into memory for the first time
Perform link-editing operations
Execute library position-independent code
These overhead tasks can incur runtime penalties and are not necessary when
you use static libraries.

In this section, we discuss how to perform basic database operations in an
ESQL/C application:
Database connections
Simple SQL statements (SELECT, INSERT, UPDATE, and DELETE)
Static and dynamic SQL
Calling SQL routines (stored procedures)
Using transactions
In the program we try to make use of most commonly used data types, for other
data types, refer to
128
7884ch04.fm
esqlc.doc/sii-03-sourceforchaptitle.htm
When an Informix ESQL/C application begins execution, it has no connections to
any database server. For SQL statements to run, however, such a connection
must exist. To establish a connection to a database server, the Informix ESQL/C
program must take the following actions:
Use an SQL statement to establish a connection to the database server.
Specify, in the SQL statement, the name of the database server to which to
connect.
The client application connects to the default database server when the
application does not explicitly specify a database server for the connection. You
must set the INFORMIXSERVER environment variable even if the application does
not establish a connection to the default database server.
if user name and password are not explicitly specified using the InetLogin
structure or the USER clause, the default user ID is used to attempt the
connection. The default user ID is the login name of the user running the
application.
The Example 4-3 is a statement that connects to the database stores_demo
under the instance demo_on.
Example 4-3 Database connection statement
EXEC SQL connect to 'stores_demo@demo_on';
EXEC SQL connect to stores_demo user :username using :password;
For more information on database connections in ESQL/C, refer to

http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.esqlc.doc/s
ii12164286.htm#sii12164286

There are several way to execute SQL statement from a ESQL/C application.
The application must choose an appropriate method based on the nature of the
SQL statement.
SQL statements which do not return a result set of data such as INSERT,
DELETE, UPDATE or Data Definition Language (DDL) can be executed using
the EXEC SQL command (same as the $ prefix) or the EXEC SQL execute
immediate statement. Both methods produce the same C code.
EXEC SQL execute immediate :cmdstring;
129
7884ch04.fm

$CREATE TABLE my_customer (fname char(20));
If the SQL statement is going to be executed more than one time during the
life of the application, it can be prepared with the EXEC SQL PREPARE command
and then execute it when required using EXEC SQL EXECUTE.
EXEC SQL prepare d_id from :stmt_buf;
EXEC SQL execute d_id;
SQL statements returning one row can be executed using the EXEC SQL
EXECUTE INTO command, normally referred as singleton statements.
EXEC SQL SELECT specs into :mspecs FROM my_customer WHERE customer_num=2;
SQL statements returning more than one row should be executed using a
Select Cursor.
EXEC SQL declare cursor1 cursor for SELECT fname into :var1 from customer;
EXEC SQL open cursor1;
EXEC SQL fetch cursor1;
In the Example 4-4, we demonstrate how to execute an SQL statement using

some of the methods previously described.
Example 4-4 Simple SQL statements (INSERT, UPDATE and DELETE)
#include
#include
EXEC SQL
EXEC SQL
<stdio.h>
<string.h>
include sqlca;
include sqltypes;
int main()
{
char cmdstring[4096];
string *fname[5]
= { "Ludwig","Carole", "Philip", "Anthony", "Raymond" };
string *lname[5]
= { "Pauli", "Sadler", "Currie", "Higgins", "Vector" };
string *company[5]
= {"All Sports Supplies","Sports Spot", "Phil's Sports", "Play
Ball!", "Los Altos Sports" };
char pref='t';
char *specs[5]
= { "This is just any string ", "This is just another string ",
"This is the third string", "This is one more string", "This is the last string"};
lvarchar mspecs[250];
char *order_date[5]
float ship_charge[5]
char *ship_duration[5]
= {"08/01/77","08/02/77","08/03/77","08/04/77","08/05/77"};
= {10000.59,590000.32,345577.12,987098.32,876893.22};
= {"10:10:10","11:11:11","22:22:22","33:33:33","44:44:44"};

$WHENEVER ERROR STOP;
int i;
$ CREATE TABLE my_customer (
130
7884ch04.fm

customer_num
fname
lname
company
preferred
specs
);
bigserial,
char(15),
char(15),
char(30),
boolean,
lvarchar
$ CREATE TABLE my_orders (

order_num
bigserial,
order_date
date ,
customer_num
bigint,
ship_charge
money(10,2),
ship_duration
INTERVAL hour to second
);
for (i = 0; i < 5 ; i++)
{
$INSERT INTO my_customer VALUES (
0,
:fname[i],
:lname[i],
:company[i],
:pref,
:specs[i]
);
}
for (i = 0; i < 5 ; i++)
{
sprintf(cmdstring, "INSERT INTO my_orders VALUES (10000001,'%s',%d,%f,'%s' );",
order_date[i], i,ship_charge[i], ship_duration[i]);
}
sprintf(cmdstring, "UPDATE my_customer SET specs='This is a new spec' WHERE customer_num
= 2");
if (SQLCODE != 0)
sprintf(cmdstring, "DELETE FROM my_customer where customer_num=3");
if (SQLCODE != 0)
EXEC SQL SELECT specs into :mspecs FROM my_customer WHERE customer_num=2;
if (SQLCODE != 0)
else
printf("The spec for the customer is :%s\n", mspecs);
}
131
7884ch04.fm
For additional information regarding the execution of SQL statements refer to the
IBM Informix ESQL/C Programmer's Manual at
http://publib.boulder.ibm.com/infocenter/idshelp/v10/index.jsp?topic=/c
om.ibm.esqlc.doc/esqlc298.htm
Static and dynamic SQL

The SQL statements contained in Example 4-4 on page 130 are all static SQL
statements. A static SQL statement is one for which all the information is known
at compile time. However, in some applications the programmer does not know
the contents, or possibly even the types, of SQL statements that the program
needs to execute. For example, a program may prompt the user to enter a select
statement, so that the programmer has no idea what columns will be accessed
when the program is run. Such applications require dynamic SQL. Dynamic SQL
allows an IBM Informix ESQL/C program to build an SQL statement at runtime,
so that the contents of the statement can be determined by user input.
Example 4-5 shows how to use dynamic SQL in a program.
Example 4-5 Dynamic SQL statements
#include <stdio.h>
int main()
{
int i1;
char c1[50];
char c2[50];
int i2;
printf("Enter the customer number [101-128] to see the names:");
scanf("%d", &i1);
sprintf (cmdstring, "SELECT fname,lname from customer where customer_num = ?");
EXEC SQL prepare ex_id from :cmdstring;
EXEC SQL execute ex_id into :c1,:c2 using :i1;
;
if (SQLCODE == 100)
{
}
else
{
132
7884ch04.fm

printf("Data found \n");
printf("Last Name \t%s\n",c1);
printf("First Name \t%s\n",c2);
}
}
OUTPUT:
Enter the customer number [101-128] to see the names:102
SQLCODE=0
Data found
Last Name
Carole
First Name
Sadle
The SELECT statement shown in Example 4-5 on page 132 is called a singleton
SELECT as it returns only 1 row. For SELECT statements returning multiple rows
you have to use a SELECT cursor. Example 4-6 below shows how to use cursor
for the SELECT statements returning multiple rows.
Example 4-6 Using cursor for SELECT statements returning multiple rows
#include <stdio.h>
int main()
{
int i;
char c1[50];
char c2[50];
printf("Enter customer number [103 - 128] :");
scanf("%d", &i);
sprintf (cmdstring, "SELECT fname,lname from customer where customer_num < ?;");
EXEC SQL declare ex_cursor cursor for ex_id;
EXEC SQL open ex_cursor using :i;
/* Print out what DESCRIBE returns*/
for (;;)
{
EXEC SQL fetch ex_cursor into :c1,:c2;
if (strncmp(SQLSTATE, "00", 2) != 0)
break;
if (SQLCODE == 100)
{
133
7884ch04.fm

}
else
{
printf("%s\t%s\n",c1,c2);
}
}
OUTPUT:
Enter customer number [103 - 128] : 108
Ludwig Pauli
Carole Sadler
Philip Currie
Anthony Higgins
Raymond Vector
George Watson
Charles Ream
For more details on dynamic SQL and the cursors, refer to

ii-sc3dysq-35492.htm#sii-sc3dysq-35492

In this section we demonstrate how to call a SQL routine from an ESQL/C
application. You can accomplish a wide range of objectives with SQL routines,
including improving database performance, simplifying writing applications, and
limiting or monitoring access to data. Because an SQL routine is stored in an
executable format, you can use it for repeated tasks to improve performance.
When you invoke an SPL routine, rather than straight SQL code, you can bypass
repeated parsing, validity checking, and query optimization.
Example 4-7 shows a simple program containing a SPL routine that receives
arguments price, percent tax, and returns price after adding the tax.
Example 4-7 Calling SQL routines
#include <stdio.h>
$ INCLUDE sqlca.h;
$ INCLUDE sqltypes.h;
int main()
{
float n1,n2;
int i=2;
EXEC SQL whenever sqlerror stop;
134
7884ch04.fm

EXEC SQL CONNECT TO 'stores_demo';
sprintf(cmdstring, "CREATE FUNCTION inc_price(n1 money(8,2), n2 int) RETURNING

money(8,2) RETURN(n1 + (n1 *n2)/ 100); END FUNCTION;");
sprintf(cmdstring, "SELECT FIRST 5 total_price, inc_price(total_price,?) from items ;");
EXEC SQL declare ex_cursor cursor for ex_id;
EXEC SQL open ex_cursor using :i;
/* Print out what DESCRIBE returns*/
for (;;)
{
EXEC SQL fetch ex_cursor into :n1,:n2;
if (strncmp(SQLSTATE, "00", 2) != 0)
break;
if (SQLCODE == 100)
{
}
else
{
printf("%f\t%f\n",n1,n2);
}
}
}
OUTPUT:
250.000000
960.000000
240.000000
20.000000
840.000000
255.000000
979.200012
244.800003
20.400000
856.799988
Using transactions
A transaction is a collection of SQL statements that are treated as a single unit of
work. All the SQL statements that you issue in an ANSI-compliant database are
automatically contained in transactions. With a database that is not ANSI
compliant, transaction processing is an option.
In a database that is not ANSI compliant, a transaction is enclosed by a BEGIN
WORK statement and a COMMIT WORK or a ROLLBACK WORK statement. In an
ANSI-compliant database, the BEGIN WORK statement is unnecessary, because all
statements are automatically contained in a transaction. You only need to
indicate the end of a transaction with a COMMIT WORK or ROLLBACK WORK statement.
Example 4-8 on page 136 shows an example using transactions.
135
7884ch04.fm

Example 4-8 Using transactions
#include
$include
$include
$include
<stdio.h>
"sqlca.h";
"sqlhdr.h";
"sqltypes.h";
main()
{
int n1;
char c1[20];
EXEC SQL CREATE DATABASE itso WITH LOG;
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
SQL
SQL
SQL
SQL
SQL
SQL
SQL
SQL
SQL
SQL
CREATE TABLE t1(num serial, name char(20));

BEGIN WORK;
INSERT INTO t1 VALUES (0,'name1');
ROLLBACK;
BEGIN WORK;
COMMIT;
EXEC SQL declare ifx_cursor cursor for select num,name into :n1,:c1 from t1;
EXEC SQL open ifx_cursor ;
for (;;)
{
EXEC SQL fetch ifx_cursor;
if (sqlca.sqlcode!=0)
break;
printf("%d\t%s\n",n1,c1);
}
EXEC SQL close ifx_cursor;
EXEC SQL free ifx_cursor;
}
OUTPUT:
4
name4
5
name5
136
7884ch04.fm
4.3.3 Data types mapping

This section contains information about the correspondence data types between
SQL and C and how to handle data types in an IBM Informix ESQL/C program.
When a query is executed by the application, the data returned by the Informix
server may be in a different format than what the application uses. The ESQL/C
converts the data passed between the application and the database server. This
process is transparent to the application. The only requirement for the application
is to specify the correct data types in ESQL/C.
Table 4-3 lists a few ESQL/C data type mapping as examples.
Table 4-3 Informix data type mapping
Informix SQL
ESQL/C
BOOLEAN
boolean
BYTE
loc_t
LVARCHAR
lvarchar
NCHAR(n)
fixchar [n] or string [n+1]
NVARCHAR(m)
varchar[m+1] or string [m+1]
SERIAL8
int8 or ifx_int8_t
TEXT
loc_t
BLOB
ifx_lo_t
CLOB
ifx_lo_t
LIST(e)
collection
MULTISET(e)
collection
Opaque data type
lvarchar, fixed binary, or var binary
ROW(...)
row
SET(e)
collection
For a complete list of all the Informix SQL to ESQL/C data type mapping refer to
the IBM Informix ESQL/C Programmers Manual at
esqlc.doc/sii03147680.htm
137
7884ch04.fm
4.3.4 Handling special data types

This section describe how to work with Informix specific data types. We discuss
the following data types:
Smart Large Objects (BLOB and CLOB)
Collection data types (LIST, MULTISET, and SET)
ROW data type
To store large objects inside database, you can use data types such as TEXT,
BYTE, BLOB, and CLOBS. Informix supports simple large objects and smart
large objects. Simple large objects are the TEXT and BYTE types which are
primarily there for compatibility with earlier versions of Informix applications. The
smart large objects are the BLOB and CLOBs. When you write new applications
that need to access large objects, use smart large objects to hold character
(CLOB) and binary (BLOB) data. Table 4-4 summarizes the advantages that
smart large objects present over simple large objects.
Table 4-4 Advantages of smart blobs over simple blobs
Large-Object feature
Simple large objects
Smart large objects
Maximum size of data
2 gigabytes
4 terabytes
Data accessibility
No random access to
data
Random access to data
Reading the large

object
The database server

reads a simple large
object on an all or
nothing basis.
Library functions
provide access that is
similar to accessing an
operating-system file.
You can access
specified portions of the
smart large object.
Writing the large object
The database server

updates a simple large
object on an all or
nothing basis.
The database server

can rewrite only a
portion of a smart large
object.
Data logging
Data logging is always

on.
Data logging can be

turned on and off
Example using CLOB and BLOBs

Informix ESQL/C supports the SQL data types CLOB and BLOB with the ifx_lo_t
data type. Because of the potentially huge size of smart-large-object data, the
Informix ESQL/C program does not store the data directly in a host variable.
Instead, the client application accesses the data as a file-like structure. To use
138
7884ch04.fm
smart-large-object variables in an Informix ESQL/C program, take the following

actions:
Declare a host variable with the ifx_lo_t data type.
Access the smart large object with a combination of the following three data
structures:
The LO-specification structure, ifx_lo_create_spec_t
The LO-pointer structure, ifx_lo_t
An integer LO file descriptor
Smart Large Objects are logically stored in a table column but physically stored
in a specific type of dbspaces called Smart Blob Space (SBS). you have to
create the Smart Blob Space in server before you can run this example. For the
details about creating the BLOB space, refer to
ds_admin_0491.htm
Example 4-9 illustrates how to work with BLOB and CLOB data types using
ESQL/C functions. The function filetoclob and filetoblob are used to enter
the CLOB and BLOB data respectively. If your files are in the client machine then
mention client as the second argument to the filetoblob function. If they are
present at the server machine then mention server as the second argument.
The output will be a part of the file which you insert as a BLOB. In this example
we used a README.txt file.
Example 4-9 ifx_lo_sample.ec
#include
$include
$include
$include
<stdio.h>
"sqlca.h";
"sqlhdr.h";
"sqltypes.h";
main()
{
int error, ic1, oflags, cflags, extsz, imsize, isize, iebytes;
time_t time;
struct tm *date_time;
char col_name[300]="test", sbspc[129];
fixed binary 'blob' ifx_lo_t c2;
char srvr_name[256];
ifx_lo_create_spec_t *cspec;
ifx_lo_stat_t *stats;
ifx_int8_t size, c1, estbytes, maxsize;
int lofd;
long atime, ctime, mtime, refcnt;
139
7884ch04.fm

EXEC SQL create table t2 (c1 int, c2 blob);
EXEC SQL insert into t2 values (1,filetoblob ('/tmp/README.txt', 'server','t2','c2'));
EXEC SQL declare ifxcursor cursor for select c1,c2 into :c1,:c2
from t2 for update;
EXEC SQL open ifxcursor;

for (;;)
{
EXEC SQL fetch ifxcursor;
if (sqlca.sqlcode!=0)
break;
lofd = ifx_lo_open(&c2, LO_RDWR, &error);
ifx_lo_read(lofd, srvr_name, 256, &error);
printf("Value: %s\n",srvr_name);
ifx_lo_write(lofd,col_name,5,&error);
ifx_lo_close(lofd);
}
EXEC SQL close ifxcursor;
EXEC SQL free ifxcursor;
ifx_lo_close(lofd);
}
More details on the ESQL/C functions to create, alter and access BLOB and
CLOB data are available at
esqlc.doc/sii-03-sourceforchaptitle.htm
Example using collection data types

Collection data types (SET, LIST, and MULTISET) enable you to store and
manipulate collections of data within a single row of a table. A collection data
type has two components:
A type constructor, which determines whether the collection type is a SET,
MULTISET, or LIST.
An element type, which specifies the type of data that the collection can
contain.
The elements of a collection can be of any data type. The elements of a
collection are the values that the collection contains. In a collection that contains
the values {'blue', 'green', 'yellow', and 'red'}, blue represents a single
element in the collection. Every element in a collection must be of the same type.
140
7884ch04.fm
For example, a collection whose element type is INTEGER can contain only
integer values.
Informix ESQL/C uses collection variables to access collection data types. A
collection variable stores the elements from a collection column as if they were
rows in a table. This virtual table can be used as part of a cursor declaration to
fetch the individual elements of the collection column.
Example 4-10 shows the ESQL/C commands used to retrieve an element from a
collection column. The parents column is a collection of INT values. The
collection is stored into the host variable hv1, and used to open a cursor using the
variable as a virtual table, table(:hv1).
Example 4-10 collection_sample
EXEC
EXEC
EXEC
EXEC
EXEC
EXEC
SQL
SQL
SQL
SQL
SQL
SQL
client collection hv1;

int parent_id;
select parents into :hv1 from grade12_parents where class_id = 1;
declare cur1 cursor for select id from table(:hv1);
open cur1;
fetch cur1 into :parent_id;
The INT elements of the collection are retrieved from the host variable using the
EXEC SQL fetch cur1 command.
Example 4-11 shows a complete ESQL/C program illustrating how to insert and
select into tables containing the collection data types.
Example 4-11 Example using collection types
#include <stdio.h>
static void print_collection(
const char *tag,
parameter client collection c
)
{
int4
value;
mint
item = 0;
EXEC SQL WHENEVER ERROR STOP;
printf("COLLECTION: %s\n", tag);
EXEC SQL DECLARE c_collection CURSOR FOR
SELECT * FROM TABLE(:c);
EXEC SQL OPEN c_collection;
while (sqlca.sqlcode == 0)
141
7884ch04.fm

{
EXEC SQL FETCH c_collection INTO :value;
if (sqlca.sqlcode != 0) break;
printf("\tItem %d, value = %d\n", ++item, value);
}
EXEC SQL CLOSE c_collection;
EXEC SQL FREE c_collection;
}
mint main(int argc, char **argv)
{
client collection list
(integer not null) lc1;
client collection set
(integer not null) sc1;
client collection multiset (integer not null) mc1;
char *dbase = "stores_demo";
mint seq;
char *stmt1 =
"INSERT INTO t_collections VALUES(0, "
"'LIST{-1,0,-2,3,0,0,32767,249}', 'SET{-1,0,-2,3}', "
"'MULTISET{-1,0,0,-2,3,0}') ";
if (argc > 1)
dbase = argv[1];
EXEC SQL WHENEVER ERROR STOP;
printf("Connect to %s\n", dbase);
EXEC SQL connect to :dbase;
EXEC SQL CREATE TEMP TABLE t_collections
(
seq
serial not null,
l1 list
(integer not null),
s1 set
(integer not null),
m1 multiset(integer not null)
);
EXEC SQL EXECUTE IMMEDIATE :stmt1;

EXEC SQL ALLOCATE COLLECTION :lc1;
EXEC SQL ALLOCATE COLLECTION :mc1;
EXEC SQL ALLOCATE COLLECTION :sc1;
EXEC SQL DECLARE c_collect CURSOR FOR
SELECT seq, l1, s1, m1 FROM t_collections;
EXEC SQL OPEN c_collect;
EXEC SQL FETCH c_collect INTO :seq, :lc1, :sc1, :mc1;
EXEC SQL CLOSE c_collect;
EXEC SQL FREE c_collect;
print_collection("list/integer", lc1);
print_collection("set/integer", sc1);
print_collection("multiset/integer", mc1);
142
7884ch04.fm
EXEC SQL DEALLOCATE COLLECTION :lc1;

EXEC SQL DEALLOCATE COLLECTION :mc1;
EXEC SQL DEALLOCATE COLLECTION :sc1;
puts("OK");
return 0;
}

Example 4-12 Collection example output
Connect to stores_demo
COLLECTION: list/integer
Item 1, value = -1
Item 2, value = 0
Item 3, value = -2
Item 4, value = 3
Item 5, value = 0
Item 6, value = 0
Item 7, value = 32767
Item 8, value = 249
COLLECTION: set/integer
Item 1, value = -1
Item 2, value = 0
Item 3, value = -2
Item 4, value = 3
COLLECTION: multiset/integer
Item 1, value = -1
Item 2, value = 0
Item 3, value = 0
Item 4, value = -2
Item 5, value = 3
Item 6, value = 0
OK
4.3.5 Exception handling

The applications which you write require that the database server processes your
SQL statements successfully as you intend. If a query fails and you do not know
it, you might display meaningless data to the user.
To handle such errors, an Informix ESQL/C program must check that every SQL
statement executes as you intend.
This section discusses the two widely used exception handling in Informix
applications development:
143
7884ch04.fm
Using SQLSTATE variable and the GET DIAGNOSTICS statement to check

for runtime errors and warnings that your Informix ESQL/C program might
generate.
Using SQLCODE variable and the SQL Communications Area (sqlca) to
check for runtime errors and warnings that your Informix ESQL/C program
might generate.
For a complete information on ESQL/C Exception handling, refer to
esqlc.doc/sii-11-40709.htm
Exception handling with SQLSTATE

After the database server executes an SQL statement, it sets SQLSTATE with a
value that indicates the success or failure of the statement. From this value, your
program can determine if it needs to perform further diagnostics. If SQLSTATE
indicates a problem, you can use the GET DIAGNOSTICS statement to obtain more
information.
Example 4-13 shows us how to handle the exceptions using SQLSTATE and the
GET DIAGNOSTICS statement. The GET DIAGNOSTICS statement returns information
that is held in the fields of the diagnostics area. The diagnostics area is an
internal structure that the database server updates after it executes an SQL
statement. Each application has one diagnostics area.
Example 4-13 Error handling in ESQL/C
#include <stdio.h>
#include <stdio.h>
void error_chk()
{
mint excp_count;
char overflow[2];
mint excp_num=1;
char message[255];
mint mlen;
char sql_state_code[6];
mint i=1;
printf("SQLSTATE: %s\n",SQLSTATE);
printf("SQLCODE: %d\n", SQLCODE);
printf("\n");
EXEC SQL get diagnostics :excp_count = NUMBER, :overflow = MORE;
EXEC SQL get diagnostics exception :i :sql_state_code = RETURNED_SQLSTATE, :message =
MESSAGE_TEXT, :mlen = MESSAGE_LENGTH;
printf("EXCEPTION %d: SQLSTATE=%s\n", i, sql_state_code);
144
7884ch04.fm

message[mlen-1] = '\0';
printf("MESSAGE TEXT: %s\n", message);
}
mint main(int argc, char **argv)
{
char *dbase = "stores_demo";
if (argc > 1)
dbase = argv[1];
EXEC SQL WHENEVER sqlerror CALL error_chk ;
printf("Connect to %s\n", dbase);
EXEC SQL connect to :dbase;
EXEC SQL SELECT province FROM customer WHERE customer_num=1;
return 0;
}
OUTPUT:
Connect to stores_demo
SQLSTATE: IX000
SQLCODE: -217
EXCEPTION 1: SQLSTATE=IX000
MESSAGE TEXT: Column (province) not found in any table in the query (or SLV is
undefined).
This section we provide a few frequently seen ESQL/C errors and discuss how to
diagnose a problem using the trace facility.
The following are typical errors seen in ESQL/C application:
Locale mismatch
The most common error you might encounter when writing ESQL/C
application is a locale mismatch error:
-23197
Database locale information mismatch
The straight forward solution is to match all the locale related environment
variables namely CLIENT_LOCALE, DB_LOCALE, and SERVER_LOCALE. The default
values for CLIENT_LOCALE are en_us.8859-1 for UNIX and en_us.1252 for
Windows. You can set these by using export (ksh) or setenv (csh) commands
depending on which shell you are using. On windows, you can set the locale
from setnet32.
145
7884ch04.fm
To know more about these locales and related material, refer to

http://publib.boulder.ibm.com/infocenter/idshelp/v115/index.jsp?topic=/com.
ibm.glsug.doc/ids_gug_035.htm
Compile and linking errors

The error message is:
error while loading shared libraries: libifsql.so: cannot open shared
object file: No such file or directory
This error is generally due to wrong setting of the shared library path, refer to
4.3.1, Creating an ESQL/C application on page 124 for correct settings.
Database connection errors
The error message is
-25596 The INFORMIXSERVER value is not listed in the sqlhosts file or the
Registry.
This error is due to incorrect setting of INFORMIXSERVER variable. Check if

it is set to the server instance to which you want to connect.
SQLIDEBUG
ESQL/C uses the SQLI protocol to communicate with the Informix database
server. The SQLIDEBUG trace allows to trace all the messages between client
and server and can be used to diagnose problems such as SQL errors,
unexpected return values, or performance issues.
You can enable SQLIDEBUG on both the client and the server side.
On the server side, you can enable SQLIDEBUG using the onmode -p 1
sqli_dbg parameter.
To set SQLIDEBUG on the client side, create the environment variable
SQLIDEBUG with the following format
sqlidebug=2:path_trace_file
where path_trace_file corresponds to the location and name of the trace file.
Example 4-14 shows how to create the variable on an UNIX platform and how to
decode the SQLI trace file using the sqliprint utility included as part of Informix
Client SDK.
Example 4-14 sqlidebug trace
informix@irk:/work$ export SQLIDEBUG=2:/tmp/sqlitrace
informix@irk:/work$
informix@irk:/work$ sqliprint /tmp/sqlitrace_17008_0_8c819d0 | more
...
146
C->S (16)
SQ_DBOPEN
7884ch04.fm
Time: 2010-07-06 20:21:58.05185
"stores7" [7]
NOT EXCLUSIVE
SQ_EOT
S->C (28)
SQ_DONE
Time: 2010-07-06 20:21:58.05487

Warning..:
# rows...:
rowid....:
serial id:
0x15
0
0
0
SQ_COST
estimated #rows: 1
estimated I/O..: 1
SQ_EOT
C->S (56)
Time: 2010-07-06 20:21:58.05595
SQ_PREPARE
# values: 1
CMD.....: "SELECT code, sname FROM state WHERE code = ?" [44]
SQ_NDESCRIBE
SQ_WANTDONE
...
147
7884ch04.fm
148
7884ch05.fm
Chapter 5.
Working with the JDBC

drivers
In this chapter, we describe how to develop a Java application using the two
available Java Database Connectivity (JDBC) drivers. This chapter contains the
following:
Application software requirements

Setting up the development environment
Verifying connectivity
Performing typical database operations
Working with special Informix data types
Troubleshooting
149
7884ch05.fm
5.1 JDBC drivers for an Informix database

Java Database Connectivity (JDBC) is the JavaSoft specification of a standard
application programming interface (API) that allows Java programs to access
database management systems.
The JDBC API consists of a set of interfaces and classes written in the Java
programming language. Using these standard interfaces and classes,
programmers can write applications that connect to databases, send queries
written in structured query language (SQL), and process the results.
Since JDBC is a standard specification, one Java program that uses the JDBC
API can connect to any database management system (DBMS), as long as a
driver exists for that particular DBMS.
For more information about the JDBC API, refer to JDBC documentation at
http://java.sun.com/javase/6/docs/technotes/guides/jdbc/
5.1.1 JDBC drivers for Informix database

There are two types of JDBC drivers that you can use from a Java application to
process data in an IBM Informix database.
IBM Informix JDBC driver
Both JDBC drivers are developed as pure-Java drivers (Type 4). This means
that, when you use a Type 4 driver from a Java application, your session
connects directly to the database or database server without a middle layer.
IBM Informix JDBC Driver

IBM Informix JDBC Driver is a platform-independent, industry-standard Type 4
driver that provides enhanced support for distributed transactions.
Informix JDBC Driver follows the JDBC 3.0 specifications providing support for
the following IBM Informix database engines:
IBM Informix 7.x, 9.4x, 10.0, 11.10 and 11.50

IBM Informix Extended Parallel Server (XPS) 8.5x
IBM Informix Standard Engine (SE) 7.x
IBM Informix OnLine Version 5.x
To use IBM Informix JDBC Driver Version 3.50.JC6, you must use a JDK 1.4.2 or
later package on your platform.
150
7884ch05.fm
You can download the Informix JDBC Driver from here:

http://www14.software.ibm.com/webapp/download/search.jsp?go=y&rs=ifxjdbc
Example 5-1 shows the contents of the Informix JDBC Driver package
Example 5-1 Informix JDBC directory
demo
doc
javadoc
release
lib
ifxjdbc.jar
ifxjdbcx.jar
ifxlang.jar
ifxlsupp.jar
ifxsqlj.jar
ifxtools.jar
license
proxy
This is a brief explanation of the jar files in the driver:

ifxjdbc.jar: This is the optimized implementations of the JDBC API interfaces,
classes, and methods.
ifxjdbcx.jar: This jar file includes the implementation of datasource,
connection pooling, and XA-related class files
ifxlang.jar: This is the localized versions of all message text supported by the
driver.
ifxlsupp.jar: This jar file includes support functions for the ifxlang.jar package.
ifxsqlj.jar: This jar includes the classes for runtime support of SQLJ programs.
ifxtools.jar: This jar includes utilities ClassGenerator, lightweight directory
access protocol (LDAP) loader, and others

IBM Data Server Driver for JDBC and SQLJ, formerly known as IBM Driver for
JDBC and SQLJ is a JDBC driver that uses the DRDA protocol to communicate
with IBM database servers. The use of a common communication protocol such
as DRDA means that the IBM Data Server Driver for JDBC and SQLJ allows you
to write client applications that can use both DB2 and Informix database servers.
The IBM Data Server Driver for JDBC and SQLJ is compliant with the JDBC 3.0
and JDBC 4.0 specifications. This driver is included as part of the IBM Data
Server Driver Package, which is bundled with the IBM Informix Client SDK
Chapter 5. Working with the JDBC drivers
151
7884ch05.fm
product. IBM Data Server Driver for JDBC and SQLJ is also available as a
separate download at:
http://www-01.ibm.com/support/docview.wss?rs=71&uid=swg24026929
Example 5-2 shows a listing of the files included in the IBM Data Server Driver for
JDBC and SQLJ install package highlighting some important files.
Example 5-2 Data Server Driver for JDBC directory (snippet)
db2jcc.jar
db2jcc4.jar
sqlj.zip
sqlj4.zip
jdbc4_LI_en
jdbc4_LI_en.rtf
jdbc_LI_en
The Data Server Driver for JDBC and SQLJ includes the following JDBC drivers:
db2jcc.jar : Use db2jcc.jar in the CLASSPATH if you plan to use the version of
the IBM Data Server Driver for JDBC and SQLJ that includes only JDBC 3.0
and earlier functions.
db2jcc4.jar: Use db2jcc4.jar in the CLASSPATH if you plan to use the version
of the IBM Data Server Driver for JDBC and SQLJ that includes only JDBC
4.0 and earlier functions.
sqlj.zip: Provides support for SQLJ java applications. SQLJ is used to embed
SQL statements inside Java applications.
sqlj4.zip: Type 4 driver for SQLJ applications.

In this section, we discuss the setup and configuration parameters of the Informix
JDBC drivers. We also discuss how to verify the connection with the Informix
database server.
5.2.1 Configuration
The Data Server Client Driver for JDBC requires the use of a DRDA connection,
this means the Informix database server must have an alias configured for DRDA
communications. Informix JDBC Driver uses the native Informix protocol or SQLI
and requires no specific alias on the Informix database server when using the
Informix JDBC Driver.
152
7884ch05.fm
On the database server, check if the sqlhost file contains the correct listeners for
the Informix database server.
Example 5-3 shows the sqlhosts file of the Informix server used by our examples.
Example 5-3 SQLHOSTS
demo_on
demo_on_tcp
demo_on_drda
onipcshm
onsoctcp
drsoctcp
kefka
kefka
kefka
demo_on
9088
9089
For information regarding how to setup and configure a DRDA alias on the
Informix server refer to 2.1.3, Configuration on page 25.
On the Client system, You must have one of the two JDBC drivers installed. for
the installation and configuration details, refer to 2.2.3, Setting up IBM Data
Server drivers on page 40 and 2.2.4, Setting up Informix JDBC on page 48.
To use any JDBC driver in an application, you must set your CLASSPATH
environment variable to point to the driver files. The CLASSPATH environment
variable tells the Java virtual machine (JVM) and other applications where to find
the Java class libraries used in a Java program.
Example 5-4 shows the environment variable CLASSPATH containing the default
location of both JDBC drivers.
Example 5-4 CLASSPATH example
set CLASSPATH=C:\Program Files\IBM\IBM DATA SERVER
DRIVER\java\db2jcc.jar;C:\Program
Files\IBM\Informix_JDBC_Driver\lib\ifxjdbc.jar;C:\Program
Files\IBM\Informix_JDBC_Driver\lib\ifxjdbcx.jar;%CLASSPATH%;.;
Products such as IBM Data Studio or IBM WebSphere Application Server have
specific location and configuration files for the JDBC drivers. Always refer to each
individual documentation for setup details.
You can find additional information regarding the configuration of the
CLASSPATH environment variable in refer to 2.2.3, Setting up IBM Data Server
drivers on page 40 and 2.2.4, Setting up Informix JDBC on page 48.
5.2.2 Verify Connectivity with the Informix JDBC driver

In this section we demonstrate how to verify a correct setup and configuration of
the Informix JDBC Driver.
153
7884ch05.fm
To load IBM Informix JDBC Driver, the application can use the Class.forName()
Java method, passing the name of the Informix JDBC Driver as argument.
Class.forName("com.informix.jdbc.IfxDriver");
Example 5-5 contains a basic Java program that can be used to verify the
connectivity with the Informix JDBC driver. The parameters used in the
connection string are:
The IBM Informix JDBC Driver identifier, jdbc:informix-sqli.
The database server host name, kefka.lenexa.ibm.com. You can also specify
the its IP address.
The port number of the SQLI listener of the Informix server, 9088.
The database name, stores_demo.
Informix instance identified by INFORMIXSERVER.
User and password for connecting to the database server.
Example 5-5 SimpleConnection.java
import java.sql.*;
public class SimpleConnection {
public static void main(String[] args) {
String url =
"jdbc:informix-sqli://kefka.lenexa.ibm.com:9088/stores_demo:INFORMIXSERVER=demo
_on;user=informix;password=Ifmx4you";
Connection conn = null;
System.out.println("URL = \"" + url + "\"");
try {
} catch (Exception e) {
System.out.println("FAILED: failed to load Informix JDBC driver.");
}
try {
conn = DriverManager.getConnection(url);
} catch (SQLException e) {
System.out.println("FAILED: failed to connect!");
}
try {
System.out.println("Connected ...");
DatabaseMetaData md = conn.getMetaData();
System.out.println("Driver name: " + md.getDriverName());
System.out.println("Driver version: " + md.getDriverVersion());
System.out.println("Database product name: "
154
7884ch05.fm

+ md.getDatabaseProductName());
System.out.println("Database product version: "
+ md.getDatabaseProductVersion());
System.out.println("FAILED: failed to connect!");
}
try {
conn.close();
System.out.println("FAILED: failed to close the connection!");
}
System.out.println("Done!");
}
}
Example 5-6 shows the compile line and output of the SimpleConnect.java
program.
Example 5-6 Running SimpleConnection.java
C:\RedBook>javac SimpleConnection.java
C:\RedBook>java SimpleConnection
URL =
"jdbc:informix-sqli://kefka:9088/stores_demo:INFORMIXSERVER=demo_on;user=i
nformix;password=Ifmx4you"
Connected ...
Driver name: IBM Informix JDBC Driver for IBM Informix Dynamic Server
Driver version: 3.50.JC6W1
Database product name: Informix Dynamic Server
Database product version: 11.50.FC7
Done!
5.2.3 Verify Connectivity with the Data Server Driver

You can use the same Java program (Example 5-5 on page 154) to verify the
connection with the Data Server Driver for JDBC with minor change. The
changes required:
Load the Data Server JDBC driver class using the Class.forName() method:
Class.forName("com.ibm.db2.jcc.DB2Driver");
The name of the Data Server JDBC driver in the connection string should be
jdbc:ids.
155
7884ch05.fm
The port number. The port number to use the Data Server Driver client is
different from the Informix JDBC client. A common mistake is to specify the
SQLI port in the URL that leads to an error like:
com.ibm.db2.jcc.am.io: [jcc][t4][2030][11211][3.58.82] A communication
error occurred during operations on the connection's underlying socket,
socket input stream,or socket output stream. Error location: Reply.fill().
Message: Insufficient data. ERRORCODE=-4499, SQLSTATE=08001
at com.ibm.db2.jcc.am.ed.a(ed.java:319)
at com.ibm.db2.jcc.t4.a.a(a.java:416)
Example 5-7 shows the changes done to the SimpleConnection.java source

code. There is no need to specify the INFORMIXSERVER when using the IBM
Data Server Driver for JDBC.
Example 5-7 SimpleConnection.java with Data Server JDBC driver
...
} catch (ClassNotFoundException cnfe) {
System.out.println("No such class available.");
return;
}
try {
con = DriverManager
.getConnection("jdbc:ids://kefka:9089/stores_demo",
"informix", "Ifmx4you");
System.out.println("connected");
...
Example 5-8 shows how to compile and run the program. Notice that the
metadata information retrieved by the JDBC driver is slightly different than with
the IBM Informix JDBC Driver.
Example 5-8 Output of SimpleConnection.java using the Data Server JDBC driver
C:\RedBook>javac SimpleConnection.java
C:\RedBook>java SimpleConnection
connected
Driver name: IBM DB2 JDBC Universal Driver Architecture
Driver version: 3.58.82
Database product name: IDS/UNIX64
Database product version: IFX11500
connected
Done
156
7884ch05.fm
5.3 JDBC type mapping

Mapping is the way of specifying data type equivalents. Because there are
variations between the SQL data types supported by each database vendor, the
JDBC API defines a set of generic SQL data types that should be used on the
Java application. In addition to these data types, defined by the JDBC driver, the
Java language itself has its own data types that may differ from the SQL types
used by the database vendor. When writing a Java application, the programmer
uses Java data types to manipulate the data. The application developers should
understand what is the equivalent data type in JDBC and the equivalent data
types on the database server.
Table 5-1 lists a few of the data type mapping required when working with an IBM
Informix database as examples.
Table 5-1 Data Type Mapping from JDBC to Informix basic data types
Java Type
JDBC Type
Informix Type
long
BIGINT
INT8, BIGINT, BIGSERIAL
byte[]
BINARY, VARBINARY
BYTE
boolean
BIT
BOOLEAN
java.sql.Date
DATE
DATE
java.math.BigDecimal
DECIMAL
DECIMAL
byte[]
LONGVARBINARY
BYTE or BLOB
java.lang.String
LONGVARCHAR
TEXT or CLOB
java.math.BigDecimal
NUMERIC
MONEY
float, java.lang.Float
REAL
SMALLFLOAT
java.sql.Time
TIME
DATETIME HOUR TO
SECOND
java.sql.Timestamp
TIMESTAMP
DATETIME YEAR TO
FRACTION(5)
For the complete list of data type mappings between the Informix SQL data types
and the Informix JDBC driver, refer to the JDBC Programmers Guide at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.jdbc_pg.doc
/sii-xc-21122.htm#sii-xc-21122
157
7884ch05.fm
For more information regarding the data types used by the IBM Data Server
Driver for JDBC and SQLJ refer to the JDBC and SQLJ, refer to:
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.jccids.doc/
com.ibm.db2.luw.apdv.java.doc/doc/rjvjdata.htm
5.4 Performing database operations

In this section we provide examples of how to use the IBM Informix JDBC Driver
and the IBM Data Server Driver for JDBC for typical database operations such as
connection to the Informix database and manipulate data.
5.4.1 Connection to the database

A Java application can obtain a connection with an Informix database in two
ways:
Use the DriverManager class. This involves loading the JDBC driver using the
Class.forName() method and obtain a connection by calling the
getConnection() method.
con = DriverManager.getConnection("jdbc:ids://...");
Use a DataSource object through the javax.sql extensions.

DB2ConnectionPoolDataSource ds = new DB2ConnectionPoolDataSource();
PooledConnection poolconn = ds.getPooledConnection();
con = poolconn.getConnection();
The DriverManager class was implemented in the original JDBC 1.0 API. This
class provides direct access to all the JDBC driver features, however, it requires
the application to load the JDBC driver manually and to use a hard-coded
connection string with the connection details.
The DataSource interface was introduced in the JDBC 2.0 API as the preferred
method to obtain a JDBC connection. The application has no need to load a
JDBC driver class or provide hard-coded connection details. The interface only
needs to know the name of the DataSource it want to use. The details for the
database connection are stored within the DataSource definition outside the
application code.
Connecting using the DriverManager

When using the DriverManager class to connect to an IBM Informix database, a
connection string must be passed to the getConnection() method with the
details for the database server. These details include information such as the
158
7884ch05.fm
machine where the database server is running and the port number needed for
the connection.
Example 5-9 shows the syntax of the connection string used with the Informix
JDBC Driver and the IBM Data Server Driver for JDBC.
Example 5-9 DriverManager connection string syntax
jdbc:[jdbc-driver]://[{ip-address|host-name}:{port-number|service-name}][/dbnam
e]:INFORMIXSERVER=servername[{;user=user;password=password]
The parameters used when connecting to an Informix database are:

jdbc-driver: Identifies a particular JDBC driver. Informix JDBC Driver uses
informix-sqli, IBM Data Server Driver for JDBC uses ids or db2.
ip-address or host-name: Specify the machine running the database server.
port-number|service-name: Specify the communication port used by the
database server.
dbname: Name of the database to open after a successful connection with
the server.
INFORMIXSERVER: Identifies the name of the Informix database server to
connect. This parameter is not required when using the Data Server JDBC
driver.
User and password: Provide the authentication credentials for the connection.
In addition to these standard details, common to all JDBC drivers, the connection
string is also used to specify properties only relevant to a specific JDBC driver or
database server. Table 5-2 lists some of the properties specific to the Informix
JDBC driver.
Table 5-2 Informix JDBC Driver connection string.
Parameter
Description
DB_LOCALE
Locale of the database in the database server.
CLIENT_LOCALE
Locale used by the Client application
INFORMIXCONTIME
Time-out value for a connection with the server
SQLIDEBUG
Enable the trace of the SQLI protocol
STMT_CACHE
Enables SQL statement cache on the server
IFX_ISOLATION_LEVEL
Specify the default Isolation Level for the session
159
7884ch05.fm
For a complete list of the supported properties by the Informix JDBC Driver refer
to
/sii-02conect-66368.htm#sii-02conect-66368
The IBM Data Server Driver for JDBC supports most of the Informix JDBC Driver
properties. A list of the properties supported when connecting to an Informix
database server can be found at
jdbc_pg.doc/sii-02conect-66368.htm
Example 5-10 shows how to compile a basic Java code used to connect to the
Informix database. The driver name and connection string are been passed as
parameters for the program.
Example 5-10 connect.java
C:\work>set CLASSPATH=ifxjdbc.jar:db2jcc.jar:.
C:\work>type connect.java
import java.sql.*;
public class connect{
public static void main( String [] args ) {
Statement is = null;
if (args.length<2) {
System.out.println("Need the JDBC driver and Connection_String as parameters");
System.out.println("
java connect \"com.informix.jdbc.IfxDriver\"
\"jdbc:informix-sqli://kodiak:9088/stores_demo
:INFORMIXSERVER=demo_on;user=informix;password=password;\" ");
return;
}
try {
Class.forName(args[0]);
conn = DriverManager.getConnection(args[1]);
System.out.println("Connected to "+conn.getMetaData().getDatabaseProductName());
conn.close();
}
catch ( Exception e ) {
System.err.println(e);
}
}
}
C:\work>javac connect.java
C:\work>java connect "com.informix.jdbc.IfxDriver"

"jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;
160
7884ch05.fm

"
Connected to Informix Dynamic Server
C:\work>java connect "com.ibm.db2.jcc.DB2Driver"

"jdbc:ids://kodiak:9089/stores_demo:user=informix;password=password";
Connected to IDS/NT64
C:\work>
Connecting using the DataSource extensions

When using a DataSource object, the information for the database connection is
stored as properties of the DataSource object. These properties are usually
stored in a Java Naming and Directory (JNDI) service managed by the Java
runtime or the application server.
The application requires no knowledge about the specific connection details. It
makes use of the DataSource object to get a connection with the database
server.
In addition to the method implemented by the JDBC driver to set standard
properties, both the Informix JDBC Driver and the Data Server Driver for JDBC
provide extra methods to set and retrieve properties specific to the Informix
database. These properties have the same effect on the database session as the
Informix environment variables used by other Informix APIs, such as ODBC or
.NET.
Example 5-11 shows a DataSource object created using the Informix JDBC
Driver and how to set some of the DataSource properties.
Example 5-11 Datasource sample
IfxDataSource ds = new IfxDataSource();
ds.setServerName("demo_on");
ds.setDatabaseName("stores_demo");
ds.setPortNumber(9089);
ds.setIfxIFXHOST("kodiak");
ds.setIfxCLIENT_LOCALE("en_US.CP1252");
ds.getConnection("informix", "password");
You can use these Informix specific properties to modify the behavior of the
JDBC driver or to control particular features of the database server. Table 5-3 on
page 162 lists a portion of the DataSource properties implemented in the
Informix JDBC Driver as examples.
161
7884ch05.fm

Table 5-3 DataSource properties
Informix variable
Method
CLIENT_LOCALE
public String getIfxCLIENT_LOCALE()

public void setIfxCLIENT_LOCALE()
DB_LOCALE
public String getIfxDB_LOCALE()

public void setIfxDB_LOCALE()
IFX_ISOLATION_LEVEL
public String getIfxIFX_ISOLATION_LEVEL()

public void setIfxIFX_ISOLATION_LEVEL (l)
IFX_LOCK_MODE_WAIT
public int getIfxIFX_LOCK_MODE_WAIT()

public void setIfxIFX_LOCK_MODE_WAIT ()
SQLIDEBUG
public String getIfxSQLIDEBUG ()

public void setIfxSQLIDEBUG ()
STMT_CACHE
public String getIfxSTMT_CACHE()

public void setIfxSTMT_CACHE()
USEV5SERVER
public boolean isIfxUSEV5SERVER()

public void setIfxUSEV5SERVER()
For a complete list of all the available properties, refer to the Getting and Setting
Informix Connection Properties section of the IBM Informix JDBC Driver
Programmers Guide at
jdbc_pg.doc/sii-xb-13590.htm
Because the IBM Data Server Driver for JDBC has support for both Informix and
DB2 database servers, the DataSource properties it provides is different than the
Informix JDBC Driver. Refer to the Properties for the IBM Data Server Driver for
JDBC and SQLJ for a complete list at:
jccids.doc/com.ibm.db2.luw.apdv.java.doc/doc/rjvdsprp.htm
5.4.2 Manipulating data

In this section we provide basic examples of how to perform typical data
manipulation tasks using a JDBC driver with an IBM Informix database.

A Java application can use a Statement object to run basic SQL statements
against the database server. The Statement interface provides two methods:
162
7884ch05.fm
executeQuery(), returns a ResultSet value. Use this method to run SQL

statements that return data, for example, a SELECT SQL statement.
executeUpdate(). returns the number of rows effected by the SQL statement.
This method is normally used to perform INSERT, DELETE, and UPDATE
operations or to run the Data Definition Language (DDL) statements.
Example 5-12 demonstrates how to perform a DELETE statement using the
executeUpdate() method.
Example 5-12 executeUpdate sample
C:\work>type delete.java
import java.sql.*;
public class delete{
try {
conn =
DriverManager.getConnection("jdbc:informix-sqli://kodiak:9088/stores_demo:INFOR
MIXSERVER=demo_on");
is = conn.createStatement();
int rc=is.executeUpdate("DELETE FROM state WHERE code='"+args[0]+"'");
System.out.format("Deleted %d rows",rc);
conn.close();
}
}
}
}
C:\work>
C:\work>java delete AZ
Deleted 1 rows
C:\work>
Prepare SQL statement with parameters

The PreparedStatement object can be used to run the prepared SQL statement
against the database server. Use PreparedStatement when the application use a
163
7884ch05.fm
SQL statement repeatedly. The SQL statement can be run multiple times by the
application with different values, saving the time the application takes to process
and optimize a SQL statement.
Example 5-13 demonstrates how to use aa PreparedStatement object. The
example inserts a row in the state table using the parameters supplied through
the command line.
Example 5-13 insert.java
C:\work>cat insert.java
import java.sql.*;
import java.util.*;
import java.text.*;
public class insert{

try {
conn =
PreparedStatement pstmt=conn.prepareStatement("INSERT INTO state
VALUES(?,?,?)");
pstmt.setString(1, args[0]);
int rc=pstmt.executeUpdate();
System.out.format("Inserted %d rows",rc);
conn.close();
}
}
}
}
C:\work>java insert AZ Arizona 0
Inserted 1 rows
C:\work>
164
7884ch05.fm
Retrieving data from a database table

To retrieve data from a database table or a function that returns more than one
row, a JDBC application must use the ResultSet object. When the
executeQuery() method from the Statement or PreparedStatement object is call,
the data is returned in the form of a ResultSet object. This object allows fetching
rows from the database.
Example 5-14 demonstrates how to run an SQL SELECT statement to retrieve
data using the PreparedStatement and the ResultSet object. The result of the
prepare statement is stored in the ResultSet object, dbRes. After that, the
dbRes.next() method is used to scroll through the ResultSet data.
Example 5-14 select.java
C:\work>cat select.java
import java.sql.*;
public class select {
ResultSet dbRes = null;
try {
conn =
MIXSERVER=demo_on;);
PreparedStatement pstmt=conn.prepareStatement("SELECT * FROM state where
code<?");
pstmt.executeQuery();
dbRes = pstmt.getResultSet();
while (dbRes.next()) {
System.out.format("%s,",dbRes.getString(1));
System.out.format("%s,",dbRes.getString(2));
System.out.format("%f\n",dbRes.getDouble(3));
}
dbRes.close();
conn.close();
}
165
7884ch05.fm

}
}
}
C:\work>java select CA
AL,Alabama
,0.040000
AR,Arizona
,0.000000
AZ,Arizona
,0.055000
C:\work>
Transactions
Local transaction are controlled directly with the connection object. Methods
such as Connection.commit() and Connection.rollback() are used to resolved
a transaction.
By default, all the connections created by the Informix JDBC Driver Connection
object are in AutoCommit mode. This means that every SQL statement send to
the Informix database server is automatically committed. If the control over the
transaction is required, you can turn off the AutoCommit mode by using the
Connection.setAutoCommit() method.
When using the Informix JDBC driver in a XA environment (XA is a standard
specification for distributed transactions), the AutoCommit feature is always
disable. The transaction manager is the only component that has control over the
transaction.
Example 5-15 demonstrates how to disable the AutoCommit mode using
setAutoCommit(false).
Example 5-15 localtrans.java
C:\work>cat localtrans.java
import java.sql.*;
import java.util.*;
import java.text.*;
public class localtrans{

try {
166
7884ch05.fm
conn =
conn.setAutoCommit(false);
is = conn.createStatement();
int rc = is.executeUpdate("DELETE FROM state WHERE code='UK'");
System.out.format("Deleted %d rows\n",rc);
System.out.format("Aborting transaction\n");
conn.rollback();
conn.close();
}
e.printStackTrace();
}
}
}
C:\work>java localtrans
Deleted 1 rows
Aborting transaction
C:\work>
Running a user-define routine

The CallableStatement object provides a way to run user-defined routines
(UDRs) in a standard way, common to all the IBM database servers.
The results from the execution of the UDRs are returned as a result set or as an
OUT parameter. The SQL syntax for calling a user-defined routine using the
CallableStatement interface is:
"{? = call function_name (?, ?,...)}";
The placeholders identify the IN, OUT and INOUT parameters for the UDR.
The application can set and retrieve the value for the routine parameters using
the registerOutParameter() and getxxx() methods.
Example 5-16 shows a basic UDR defined with two OUT parameters.
Example 5-16 state_tax SQL routine
CREATE FUNCTION state_tax(OUT vtax percent,
167
7884ch05.fm

vcode CHAR(2),
OUT vsname CHAR(20))
RETURNS BOOLEAN;
SELECT sales_tax,upper(sname)
INTO vtax, vsname FROM state
WHERE code=vcode;
RETURN 't';
END FUNCTION;
We use CallableStatement to invoke this UDR from a JDBC application.

Example 5-17 demonstrates how to set the IN and OUT for the UDR. The
BOOLEAN value is returned directly by the SQL function and retrieved using a
ResultSet object. To retrieve the OUT parameters, the example uses the
getDouble() and GetString() methods.
Example 5-17 Callable.java
C:\work>cat callable.java
import java.sql.*;
public class callable {
try {
conn =
MIXSERVER=demo_on;");
CallableStatement cstmt = conn.prepareCall ("{? = call state_tax(?, ?,
?)}");
cstmt.registerOutParameter(1, Types.DOUBLE);
cstmt.registerOutParameter(3, Types.CHAR);
cstmt.setString(2, "CA");
ResultSet dbRes = cstmt.executeQuery();
// Retrieve OUT parameters from the function
while (dbRes.next())
System.out.format("UDR returns = %s\n", dbRes.getBoolean(1));
// Retrieve OUT parameters from the function
System.out.format("OUT tax
%s\n",cstmt.getDouble(1));
System.out.format("OUT sname %s\n",cstmt.getString(3));
168
7884ch05.fm

dbRes.close();
cstmt.close();
conn.close();
}
}
}
}
C:\work>java callable
UDR returns = true
OUT tax
0.0825
OUT sname CALIFORNIA
C:\work>
5.5 Informix additional features

This section demonstrates the use of additional features of IBM Informix, such
as:
Batch Inserts
Resultset Metadata
Informix BIGSERIAL data type
Informix Smart Large Objects
Secure Socket Layer
5.5.1 Batch inserts, updates, and resultset metadata

Here we demonstrate the following:
Batch Inserts
When performing multiple inserts (or updates) using a prepared statement, it
is more efficient to do all the inserts in bulk. The method is to add all the
inserts in a batch and then run the batch.
Using ResultSet metadata
When you run a query from a Java program that generates a result set, there
are additional information available apart from the result set itself. These
additional information about the data is returned by the ResultSet metadata.
An example of using the ResultSet metadata is for a tool that will obtain table
definition information from one database to another database. You can
dynamically find out all the details about each column that returns data.
169
7884ch05.fm
Example 5-18 shows the batch Insert and how to obtain the column name and
column type from ResultSet metadata. The schema for the gentemp table is:
CREATE TABLE gentemp (id INT, name VARCHAR(15) NOT NULL)
We use con.prepareStatement() method to prepare the SQL INSERT

statement, then we execute the statement using the pstmt.addBatch() method
which makes the INSERT statement to be part of an batch operation. After
adding 10 rows to the table, the batch operation is executed using the
pstmt.executeBatch() method.
The metadata information for the table is retrieved using the rs.getMetaData()
method.
Example 5-18 Listing for IfxBatchDemo.java
import java.sql.*;
import java.io.*;
public class IfxBatchDemo {
public static void main(String args[]) throws SQLException, IOException,Exception {
String IfxURL =
"jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on";
Connection con = null;
int cnt = 0;
System.out.println("Start");
con = DriverManager.getConnection(IfxURL);
con.setAutoCommit(false);
Statement stmt = null;
PreparedStatement pstmt = null;
ResultSet rs = null;
ResultSetMetaData rsmd = null;
pstmt = con.prepareStatement("INSERT INTO gentemp (id,name) VALUES (?,?)");
// Fill in 10 rows
for (int i = 0; i < 10; i++) {
pstmt.setInt(1, i + 1000);
pstmt.setString(2, "String #" + i);
pstmt.addBatch();
}
int[] rows = pstmt.executeBatch();
System.out.println(" Inserted data. Rows = " + rows.length);
con.commit();
pstmt.close();
pstmt = con.prepareStatement("SELECT * FROM gentemp");
rs = pstmt.executeQuery();
rsmd = rs.getMetaData();
170
7884ch05.fm
" + "Col Names " + rsmd.getColumnName(1) + ","
+ rsmd.getColumnName(2));
" + "Col Types " + rsmd.getColumnTypeName(1)
+ "," + rsmd.getColumnTypeName(2));
int row = 1;
while (rs.next()) {
Row " + row + " = "
+ rs.getInt(rsmd.getColumnName(1)) + ","
+ rs.getString(rsmd.getColumnName(2)));
row++;
}
pstmt.close();
con.commit();
con.close();
}
}
5.5.2 BIGSERIAL data type

IBM Informix version 11.50 implements the ANSI standard SQL data type
BIGINT and the BIGSERIAL data types:
BIGINT is mapped to the JDBC standard BIGINT data type, so it can be used
from a Java application as any other Informix data type.
BIGSERIAL, (like SERIAL and SERIAL8) do not have an obvious mapping to
any JDBC data type. This means the application must use Informix specific
JDBC methods to retrieved the value of these columns.
The methods to access Informix serial data types are included in the Informix
JDBC implementation of the JAVA java.sql.Statement interface, IfxStatement.
You can use the methods getSerial(), getSerial8(), and getBigSerial() to
retrieve the last value inserted on a serial column.
Example 5-19 shows how to retrieve a BIGSERIAL value using the
IfxStatement.getBigSerial() method. The schema for the table used in this
example is:
CREATE TABLE tempbs(id BIGSERIAL, name CHAR(10));
Example 5-19 bigserial.java sample
import java.sql.*;
import com.informix.jdbc.*;
public class bigserial {
171
7884ch05.fm

long insertedserial = 0;
String
url="jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;";
try {
conn = DriverManager.getConnection(url);
IfxStatement stmt= (IfxStatement) conn.createStatement();
stmt.executeUpdate("INSERT INTO tempbs VALUES (0,'test');");
insertedserial =stmt.getBigSerial();
System.out.println("Last serial: \t"+insertedserial);
stmt.executeQuery("SELECT FIRST 1 * FROM tempbs ORDER BY rowid DESC");
dbRes = stmt.getResultSet();
while (dbRes.next()) {
System.out.format("Last row: \t%d,",dbRes.getLong(1));
System.out.format("%s\n",dbRes.getString(2));
}
dbRes.close();
conn.close();
}
}
}
}
Example 5-19 on page 171 inserts a row in the tempbs table and stores the last
BIGSERIAL value in the insertedSerial variable. Example 5-20 shows how to
compile the example and its output.
Example 5-20 output of bigserial.java
C:\work>javac bigserial.java
C:\work>java bigserial
Last serial:
14
Last row:
14,test
C:\work>
172
7884ch05.fm
5.5.3 Informix Smart Large Objects

Smart Large Objects is a type of large objects supported by Informix. Smart
Large Objects is logically stored in a table column but physically stored in a
specific type of dbspaces called Smart Blob Space.
The data stored in the table column is a structure which contains information
about the large object, such as special attributes or pointers to location in the
Smart Blob Space containing the data. Informix has two types of smart large
objects.
BLOB: Stores binary data.
CLOB: Stores character data
There are two methods to manipulate Smart Large Objects with the Informix
JDBC Driver.
Standard JDBC 3.0 API.
Using standard JDBC methods such as getString(), setAsciiStream(), or
getBinaryStream() allow an application to handle Smart Lager Objects as
standard Java data types.
Informix smart-large-object extensions.
If an application requires random access to the large data, it must use the
Informix smart-large-object extensions. These extensions to the JDBC API
give the application a greater control over the smart-large-object data in terms
of object properties, concurrency access, and logging.
The IfxSmartBlob interface implements most of the smart-large-object
extensions, methods such as IfxLoCreate(), IfxLoOpen(), IfxLoRead(), and
IfxLoWrite() can be used to access the smart large object structures and
manipulate the data of the object.
For a complete description of all the Informix smart-large-objects extensions,
refer to the Informix JDBC Driver Guide at:
http://publib.boulder.ibm.com/infocenter/idshelp/v115/index.jsp?topic=/com.
ibm.jdbc_pg.doc/sii-04data-36421.htm
Note: The IBM Data Server Driver for JDBC does not support the Informix
Smart Large Object extensions, only the standard JDBC API is available for
working with Smart Large Objects
Using the standard JDBC API

Working with Smart large Objects using the JDBC API does not require any
specific Java code. Example 5-21 on page 174 demonstrates how to retrieve a
CLOB column from the Informix database using the getString() method.
173
7884ch05.fm

Example 5-21 getclob.java sample
C:\work>cat getclob.java
import java.sql.*;
public class getclob {
try {
conn = DriverManager.getConnection(
"jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;");
PreparedStatement pstmt=conn.prepareStatement("SELECT * FROM catalog WHERE
catalog_num=?");
dbRes = pstmt.getResultSet();
dbRes.next();
System.out.println("Advert description: "+dbRes.getString("advert_descr"));
dbRes.close();
conn.close();
}
}
}
}
C:\work>javac getclob.java
C:\work>java getclob 10001
Advert description: Brown leather. Specify first baseman's or infield/outfield
style. Specify right- or left-handed.
C:\work>
Using the Informix extensions

The smart-large-object extensions allows a Java application to have direct
control over the large object structures and data pointers used to describe the
large object.
The following interfaces provide the methods needed to handle smart large
objects:
174
7884ch05.fm
IfxLobDescriptor, stores the internal characteristics for a smart large object.

The application must create an IfxLobDescriptor object to insert a new large
object in the database.
IfxLocator, used to identify a particular large object. An IfxLocator object can
be created or retrieved from the database to perform any I/O operations with
the large object.
IfxSmartBlob, represent a Smart Large Object within the Informix JDBC
Driver. Provides all the methods necessary to create, open, read, and write to
a large object.
IfxloStat, stores statistical information about a smart large object such as the
size, last access time, last modified time and last status change.
IfxBblob and IfxCblob, add extended functionality to the standard JDBC 3.0
BLOB and CLOB classes.
Example 5-22 demonstrate how to read data from a smart large object using the
IfxSmartBlob method. The code performs the following steps to select the CLOB
column advert_descr:
1. Stores the advert_descr column as an IfxCblob object using the getClob()
method.
2. Gets the IfxLocator from the IfxCblob object through the
cblob.getLocator() method.
3. Creates an IfxSmartBlob object and uses the IfxLocator to open the smart
large object in smb.IfxLoOpen().
4. Reads the first 200 bytes of the CLOB using the smb.IfxLoRead() method.
5. Close the large object and release the IfxLocator.
Example 5-22 loext.java
import java.sql.*;
import com.informix.jdbc.*;
public class loext{
byte[] buffer = new byte[200];
try {
conn = DriverManager.getConnection(
"jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;");
PreparedStatement pstmt=conn.prepareStatement("SELECT * FROM catalog WHERE
catalog_num=?");
175
7884ch05.fm

ResultSet dbRes = pstmt.getResultSet();
dbRes.next();
IfxCblob cblob = (IfxCblob) dbRes.getClob("advert_descr");
IfxLocator loPtr = cblob.getLocator();
IfxSmartBlob smb = new IfxSmartBlob(conn);
int loFd = smb.IfxLoOpen(loPtr, smb.LO_RDONLY);
int bytesReaded = smb.IfxLoRead(loFd, buffer, buffer.length);
smb.IfxLoClose(loFd);
smb.IfxLoRelease(loPtr);
System.out.println("Advert description: "+new String(buffer).trim());
dbRes.close();
conn.close();
}
}
}
}
Example 5-23 shows how to compile and the output of the loext.java sample
Example 5-23 output of loext.java
C:\work>javac loext.java
C:\work>java loext 10001
Advert description: Brown leather. Specify first baseman's or infield/outfield
C:\work>
For more information regarding the use of Smart Large Object with the Informix
JDBC driver refer to the Informix JDBC Driver Guide at
/sii-04data-36421.htm#sii-04data-36421
5.5.4 Secure Socket Layer (SSL)

In addition to the encrypted communications provided by the Communication
Support Module (CSM), the Informix database server supports the use of Secure
Socket Layer (SSL) communications for the encryption of the packages between
client and server.
176
7884ch05.fm
The Informix JDBC Driver supports only CSM encryptions, you must use the IBM
Data Server Driver for JDBC to connect to an SSL-enabled Informix database
server.
Before you can use SSL encryption with an Informix database, you must
configure both server and client machines.
Preparing the Informix Server for SSL

To enable SSL on the Informix server, create an Informix alias using the drsocssl
protocol.
Example 5-24 shows the configuration and commands used to enable SSL
communications with an IBM Informix server.
Example 5-24 SSL configuration on the Informix server
$gsk7capicmd -keydb -create -db kodiak.kdb -pw password -type cms -stash
$gsk7capicmd -cert -create -db kodiak.kdb -pw password -label testlabel -dn
"CN=bedfont.uk.ibm.com,O=ibm,C=UK" -size 1024 -default_cert yes
$gsk7capicmd -cert -extract -db kodiak.kdb -format ascii -label testlabel -pw
password -target testlabel.cert
$ pwd
/usr3/11.50/ssl
$ ls -a
.
kodiak.crl kodiak.rdb testlabel.cert
..
kodiak.kdb kodiak.sth
$ onstat -c | grep ssl
DBSERVERALIASES kodiak_shm,kodiak_drda,kodiak_ssl
$ grep kodiak_ssl $INFORMIXSQLHOSTS
kodiak_ssl
drsocssl
kodiak
9191
$
For detail information regarding the configuration of SSL with an Informix server,
refer to the section Configuring a Server Instance for Secure Sockets Layer
Connections on Information Center at:
sec.doc/ids_ssl_002.htm
Preparing the Informix Client for SSL

To enable SSL encryption both client and server machine must use the same
certificated file. A certified file can be imported into the client machine using the
keytool utility included with the Java SDK framework.
177
7884ch05.fm
Example 5-25 demonstrates how to import the testlabel.cert certificate file

created in Example 5-24 on page 177 to enable a trusted relationship between
client and server. A keystore file is created to stored the keys for the SSL
encrypted connection.
Example 5-25 SSL Client side configuration
C:\work>keytool -importcert -file testlabel.cert -keystore .keystore
Enter keystore password:
Re-enter new password:
Owner: CN=bedfont.uk.ibm.com, O=ibm, C=UK
Issuer: CN=bedfont.uk.ibm.com, O=ibm, C=UK
Serial number: -14b0e6b89751eab1
Valid from: Thu Jul 15 08:10:31 PDT 2010 until: Sat Jul 16 08:10:31 PDT 2011
Certificate fingerprints:
MD5: 66:FA:CF:44:9F:F3:38:40:7B:9D:93:D6:D6:1C:DB:C5
SHA1: CF:38:6F:CA:C9:A1:54:43:FC:64:AD:F6:DF:5F:CA:65:01:58:DE:DE
Signature algorithm name: SHA1withRSA
Version: 3
Trust this certificate? [no]: yes
Certificate was added to keystore
C:\work>dir .key* /b
.keystore
C:\work>
Java client
A Java application can connect to the SSL Informix server by performing the
following operations:
Set the System property javax.net.ssl.truststore to point to the keystore
created. In our example, it is C:\work\.keystore.
Set the System property javax.net.ssl.trustStorePassword to the
password used for the certificate.
Get a datasource object.
Set the port number to the SSL port, 9191.
Set the datasource property setSslConnection to true.
Example 5-26 shows a basic Java code that connects to a SSL Informix server.
Example 5-26 Listing conssl.java
import java.sql.*;
import javax.sql.*;
178
7884ch05.fm

import java.io.*;
import java.util.*;
import com.ibm.db2.jcc.*;
public class conssl {

public static void main(String args[]) throws SQLException, IOException,
Exception {
System.setProperty ("javax.net.ssl.trustStore","c:/work/.keystore");
System.setProperty ("javax.net.ssl.trustStorePassword","password");
DB2ConnectionPoolDataSource ds = new DB2ConnectionPoolDataSource();
ds.setUser("informix");
ds.setPassword("password");
ds.setServerName("kodiak");
ds.setDatabaseName("stores_demo");
ds.setPortNumber(9191);
ds.setDriverType(4);
ds.setSslConnection(true); // SSL
PooledConnection poolconn = ds.getPooledConnection();
Connection con = poolconn.getConnection();
DatabaseMetaData md = con.getMetaData();
System.out.println("Driver name: " + md.getDriverName());
System.out.println("Connected to "+ md.getDatabaseProductName());
System.out.println("Database product version: " +
md.getDatabaseProductVersion());
con.close();
}
}
After a successful connection, the code prints metadata information such as

driver name and database version. Example 5-27 shows the output.
Example 5-27 output of conssl.java
C:\work>javac conssl.java
C:\work>java conssl
Driver name: IBM DB2 JDBC Universal Driver Architecture
Connected to IDS/UNIX32
Database product version: IFX11500
C:\work>
179
7884ch05.fm
5.5.5 Typical errors

Errors result from the application development environment without proper
configuration and SQL syntax errors are the common seen problem in the Java
applications for Informix.
Class not found errors

The following shows an class not found error message:
C:\RedBook>java IfxSimpleConnection
Exception in thread "main" java.lang.NoClassDefFoundError:
IfxSimpleConnection
Caused by: java.lang.ClassNotFoundException: IfxSimpleConnection
...
These types of error are usually from either failing to load the JDBC driver or
failing to load the application class because the environment is not configured
properly. To resolve this problem:
Make sure that you have the JDBC driver jar files in you CLASSPATH.
Make sure that you have a . (dot) to include your current directory in the
CLASSPATH.
Connectivity errors
When an application fails to connect to the server, you see error similar to this:
Exception in thread "main" java.sql.SQLException:
com.informix.asf.IfxASFException: Attempt to connect to database server
(demo_on) failed.
at com.informix.jdbc.IfxSqliConnect.<init>(IfxSqliConnect.java:1319)
...
To resolve this problem:

Check if the server is up and running and the ports are configured correctly
on the server.
Check the port number specified in the application. Remember, Informix
JDBC driver connects to a SQLI port and the IBM Data Server driver
connects to a DRDA port.
Check that there is no firewall between the client and server. Doing a telnet
from the client to the server box can help to check this.
Syntax errors
The exception with SQLCODE -201 is a syntax error in a SQL statement you are
trying to run. Apart from eyeballing the offending SQL, it can be helpful to know
180
7884ch05.fm
the location of the SQL that the server is complaining about. You can achieve this
by obtaining the SQL statement offset as follows:
try {
stmt.execute( SQL );
}
catch(Exception e) {
System.out.println ("Error Offset :"+((IfmxConnection
conn).getSQLStatementOffset() );
System.out.println(e.getMessage() );
}
5.5.6 Tracing
Tracing the communication with the Informix database server may be required for
further diagnostics. The method to enable tracing depends on the Informix JDBC
driver used.
IBM Informix JDBC Driver

The Informix JDBC uses the Informix SQLI protocol for the communication with
the database server. A trace file with all the SQLI messages can be generated by
enabling the SQLITRACE feature within the JDBC driver.
Depending on the method used for the connection, the SQLIDEBUG can be
activated using a connection string keyword or a DataSource method:
Setting trace for Informix JDBC driver using the DriverManager
DriverManager.getConnection("jdbc:informix-sqli://kodiak:9088/sysmaster:INF
ORMIXSERVER=demo_on;user=;password=;SQLIDEBUG=/tmp/jdbctrace");
Setting trace for Informix JDBC driver using a Datasource

IfxDataSource ifxds = new IfxDataSource();
ifxds.setIfxSQLIDEBUG("c:/temp/sqli.trc");
You can run the trace through the sqliprint utility to render it to text form.
IBM Data Server Driver for JDBC

The IBM Data Server Driver for JDBC uses DRDA as the network protocol and
produces a DRDA trace. This is a straight text output trace which can be
examined directly.
You can enable DRDA trace with the JDBC using the following methods:
Setting trace for IBM Data Server driver using the DriverManager:
DriverManager.getConnection("jdbc:ids://kodiak:9089/sysmaster:user=;passwor
d=;traceFile=/jcc.trc;TraceLevel=TRACE_ALL;");
181
7884ch05.fm
Setting trace for IBM Data Server driver using a datasource:

DB2SimpleDataSource ds = new DB2SimpleDataSource();
ds.setTraceFile("/jcc.trc");
ds.setTraceLevel(com.ibm.db2.jcc.DB2BaseDataSource.TRACE_ALL);
ds.setTraceFileAppend(false);
Both SQLIDEBUG and DRDADEBUG traces can be activated at the server side
rather than on the client machine. This is done using the Informix utility onmode
included with the Informix database server. You can find more information
regarding both traces on the Tracing on page 114.
182
7884ch06.fm
Chapter 6.
IBM Informix with Hibernate

This chapter describes how to use the HIbernate Java package with an IBM
Informix database. we discuss the following topics:
Description of the Hibernate
Setup and configuration
Application development using the Hibernate and the Informix JDBC drivers
183
7884ch06.fm
6.1 Hibernate for Java

This section describes the concepts of Hibernate and the Java Persistence API
(JPA) programming model.
6.1.1 Hibernate
Hibernate is an open source Java object-relational mapping (ORM) and
persistence framework that allows you to map plain old Java objects to relational
database tables using XML configuration files. Hibernate also provides a data
query language, Hibernate Query Language (HQL) and retrieval facilities that
help reduce development time spent on manual data handling in JDBC calls or
SQL statements.
Persistence data refers to any data that must be stored in the database and has
to exist after the application has finished running. What Hibernate provides is a
way to persist not only the typical relational data but also Java objects used by
the application.
A relational database stores the information in a tabular way, tables and columns.
A relational database ensures the integrity of the data using SQL objects such as
constraints and referential integrity.
An Object-orientated programming (OOP) language such as Java or C++ do not
have the same data representation as relational databases. The data is
represent by objects, with attributes and methods. The relationship between
these objects is implemented with concepts like inheritance or polymorphism that
do not exists in a relational database.
Object-relational mapping (ORM) technologies such as Hibernate tries to solved
the limitations of object-orientated languages when using relational databases.
object-relational mapping allows developers to focus on the business logic of the
application, without the need for dealing with the data access layer. A developer
only needs to load the Customer object and not care about how the information
for that particular customer is stored in the database. Object-relational mapping
solutions produce more robust and portable code which means faster
development time.
For more information about Hibernate refer to the Hibernate Reference
Documentation at
http://docs.jboss.org/hibernate/stable/core/reference/en/html/
184
7884ch06.fm
6.1.2 Hibernate concepts

The main concepts in the Hibernate framework are:
Persistence
Persistence, in Hibernate terms, refers to the concept of storing the state of
an Java object in the database, so it can be restored later on.
Mapping
Mapping is the process to map SQL tables to Java objects. Mapping is done
using XML mapping files or through the use of Java annotations (metadata
added to the application source) in the Java code.
Object processing
Object processing refers to the use of the mapped objects from the
application code. How to save and load the state of an Java object using the
Hibernate API or the specific Hibernate Query Language (HQL).
In a simplified words, Hibernate for Java is a set of Java APIs that allows store
and retrieve the state of Java objects into a database through the use of a JDBC
driver.
Figure 6-1 illustrates the typical components of a Hibernate application.
Figure 6-1 Hibernate components
Chapter 6. IBM Informix with Hibernate
185
7884ch06.fm
The components in Figure 6-1 on page 185 are:

Application: This represents the Java application.
Java object: It is the object the application wants to persist (keep it in the
database).
Hibernate properties: This is the configuration file for the Hibernate
framework, named hibernate.cfg.xml that contains information such as the
connection details for the database server or the SQL dialect to use.
XML Mapping: This is an XML file which contains the mappings between Java
Objects and database Objects.
JDBC: This is the JDBC driver used for the database connection. Both
Informix JDBC driver, IBM Informix JDBC Driver, and IBM Data Server Driver
for JDBC can be used with Hibernate.
Informix database: This is the IBM Informix database server.
In addition to the object-relational mapping, Hibernate also provides a connection
management and a transaction management services to be used with a Java
application. These topics are not cover in this section. For information regarding
the use and configuration of the Connection and Transaction management
services refer to Transactions and Concurrency section of the Hibernate
Reference Documentation at
http://docs.jboss.org/hibernate/core/3.3/reference/en/html/transactions.html

This section describes the installation and configuration of Hibernate to be used
against an IBM Informix database.
6.2.1 Installation
Hibernate is an open source project which can be downloaded through the
SourceForge website at
http://sourceforge.net/projects/hibernate/files
The latest version of the Hibernate API is 3.5.3 that fully implements the Java
Persistence API 2.0 (JPA). The zip package name is
hibernate-distribution-3.5.3-Final-dist.zip
Example 6-1 on page 187 shows the contents of the Hibernate zip package.
186
7884ch06.fm

Example 6-1 Hibernate distribution content
changelog.txt
documentation
javadocs
manual
hibernate-testing.jar
hibernate3.jar
hibernate_logo.gif
lgpl.txt
lib
bytecode
jpa
hibernate-jpa-2.0-api-1.0.0.Final.jar
optional
c3p0
ehcache
infinispan
jbosscache
oscache
proxool
swarmcache
required
antlr-2.7.6.jar
commons-collections-3.1.jar
dom4j-1.6.1.jar
javassist-3.9.0.GA.jar
jta-1.1.jar
slf4j-api-1.5.8.jar
project
The package contains most of the Java libraries required to run a Hibernate
application including the Hibernate documentation and the source code project
files for the Hibernate libraries.
hibernate3.jar: Hibernate Core library for Relational Persistence
hibernate-jpa-2.0-api-1.0.0.Final.jar: Hibernate definition of the Java
Persistence 2.0 API
antlr-2.7.6.jar: Framework for grammatical descriptions containing Java
commons-collections-3.1.jar: Types that extend and augment the Java
Collections Framework
dom4j-1.6.1.jar: XML framework for Java
javassist-3.9.0.GA.jar: JAVA programming ASSISTant
jta-1.1.jar: The javax.transaction package
slf4j-api-1.5.8.jar: API for the Simple Logging Facade for Java
187
7884ch06.fm
In addition to the libraries supplied by the Hibernate package, the jar packages
containing the libraries for the JDBC driver that must be included in the
CLASSPATH:
ifxjdbc.jar: IBM Informix JDBC Driver
db2jcc.jar: IBM Data Server Driver for JDBC
Note: The slf4j-api-1.5.8.jar package does not contain the complete set of
libraries for the Simple Logging Facade for Java (SLFJ). A package
providing a implementation for the SLFJ logger must be used in
conjunction with Hibernate API.
There are several implementations of the SLFJ logger available at:
http://www.slf4j.org/download.html
In our examples we use slf4j-simple-1.6.1.jar and slf4j-nop-1.6.1.jar that

provide simple-logging and discarded-logging.
The Hibernate package is also available as a Maven 2 artifact. Maven is an open
source software project management that uses XML files based on the Project
Object model (POM) to manage all the attributes and dependencies files of a
Java project. For more information about Maven refer to the Apache Maven
Project documentation at
http://maven.apache.org/what-is-maven.html
6.2.2 Configuration
In this section we discuss the configuration setting required to develop Informix
application with Hibernate.
CLASSPATH
CLASSPATH is an environment variable that tells the Java compiler and the Java
Virtual Machine (JVM) where to look for Java class files and Java libraries.
To be able to use Hibernate with a Java program, the core library, hibernate3.jar,
and all the libraries in the lib/required directory must be included in the
application CLASSPATH. An implementation of the SLFJ logger and the JDBC
driver is also required for Hibernate to work.
Example 6-2 shows an UNIX script used to create the CLASSPATH environment
variable. We use discarded-logging slf4j-nop-1.6.1.jar and the Informix JDBC
Driver ifxjdbc.jar for our examples.
Example 6-2 Hibernate classpath script
export CLASSPATH=$CLASSPATH:.
188
7884ch06.fm

export
export
export
export
export
export
export
export
CLASSPATH=$CLASSPATH:/work/lib/commons-collections-3.1.jar
CLASSPATH=$CLASSPATH:/work/lib/dom4j-1.6.1.jar
CLASSPATH=$CLASSPATH:/work/lib/hibernate-jpa-2.0-api-1.0.0.Final.jar
CLASSPATH=$CLASSPATH:/work/lib/hibernate3.jar
CLASSPATH=$CLASSPATH:/work/lib/javassist-3.9.0.GA.jar
CLASSPATH=$CLASSPATH:/work/lib/jta-1.1.jar
CLASSPATH=$CLASSPATH:/work/lib/slf4j-api-1.6.1.jar
CLASSPATH=$CLASSPATH:/work/lib/antlr-2.7.6.jar
export CLASSPATH=$CLASSPATH:/work/lib/slf4j-nop-1.6.1.jar
export CLASSPATH=$CLASSPATH:/work/lib/ifxjdbc.jar
Hibernate configuration file

Hibernate provides two alternative methods for specify configuration parameters:
hibernate.properties: A standard java properties text file
hibernate.cfg.xml: An XML formatted file
Both files contains the same configuration details for the Hibernate service, If
both files exist, the XML configuration file overrides the settings in the properties
file.
The configuration file (hibernate.properties or hibernate.cfg.xml) contains the
information Hibernate needs to connect to the database server. Details such as
name of the JDBC driver class, ConnectionString, authentication details are kept
in this file.
The Hibernate libraries require the configuration file (hibernate.properties or
hibernate.cfg.xml) to be located in the root directory of the CLASSPATH.
Example 6-3 shows a hibernate.cfg.xml file using the Informix JDBC Driver.
Example 6-3 hibernate.cfg.xml sample
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE hibernate-configuration PUBLIC
"-//Hibernate/Hibernate Configuration DTD 3.0//EN"
"http://hibernate.sourceforge.net/hibernate-configuration-3.0.dtd">
<hibernate-configuration>
<session-factory name="Informix_Session">

<property name="connection.driver_class">
com.informix.jdbc.IfxDriver
</property>
<property name="connection.url">
jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;
</property>
<property name="hibernate.connection.username">
189
7884ch06.fm

informix
</property>
<property name="hibernate.connection.password">
password
</property>
<property name="dialect">
org.hibernate.dialect.InformixDialect
</property>
<property name="hibernate.show_sql">false</property>
<property name="hbm2ddl.auto">update</property>

<mapping resource="State.hbm.xml"/>
</session-factory>
</hibernate-configuration>
In addition to the JDBC information, the configuration file also contains

configuration details for the Hibernate service.
The property dialect is used to specify the SQL dialect for the database server.
A Hibernate dialect is a Java class containing specific details regarding the SQL
syntax needed to communicate with a particular database.
When using the IBM Informix JDBC Driver with Hibernate, the dialect must be set
to orb.hibernate.dialect.InformixDialect. The dialect for the IBM Data
Server Driver for JDBC is org.hibernate.dialect.DB2Dialect.
These classes are included in the hibernate3.jar package and contains a basic,
non-optimized implementation of the Informix dialect.
Note: A patch that adds Informix SQL optimizations to the original Hibernate
dialect for Informix databases can be found on the International Informix Users
Group website at:
http://www.iiug.org/opensource/files/hibernate-3.3.2_informix.tar.gz
The XML configuration file allows the inclusion of XML mapping files. The
property mapping resource is used to specify the XML file containing the
mapping between the SQL table and the Java object. In Example 6-3 on
page 189 we include the State.hbm.xml mapping file as part of the hibernate
configuration.
Properties such as hibernate.show_sql or hbm2ddl.auto controls the behavior

of the Hibernate service. hibernate.show_sql is used to dump all the SQL
190
7884ch06.fm
statements to the console which may be useful for debugging purposes. Setting
hbm2ddl.auto to update specifies that the schema for the SQL table is
automatically updated if differs from the definition in the XML definition.
For more information regarding all the supported properties for the Hibernate
configuration files, refer to the Hibernate Reference Documentation at:
http://docs.jboss.org/hibernate/core/3.3/reference/en/html/session-configuratio
n.html
XML mapping file

The XML mapping file is used to specify the relations between a Java object and
an SQL database object. These mapping definitions, are used to provided
Hibernate with the information needed to persist the Java objects into the
relational database. They also provide support features, such as creating the
database schema and relationship between the objects.
XML mapping files are not required when using Java annotations, the mapping
information is added through using annotations in the Java source code.
Example 6-4 shows a basic XML file used to map the States table to the State
Java object.
Example 6-4 State.hbm.xml
<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC
"-//Hibernate/Hibernate Mapping DTD//EN"
"http://hibernate.sourceforge.net/hibernate-mapping-2.0.dtd">
<hibernate-mapping>
<class
name="State"
table="States">
<id
name="id"
column="id">
<generator class="increment"/>
</id>
<property
name="code"
column="code"/>
<property
name="sname"
column="sname"/>
</class>
</hibernate-mapping>
191
7884ch06.fm
The definition includes information such as the table name and column to be
used as identifier (ID), and specific attributes for the elements like the generator
class which defines a column as an identifier automatically generated by the Java
libraries. These details are used by the Hibernate libraries to automatically
generate SQL statements, including Data Definition Language (DDL) and Data
Manipulation Language (DML).
For a complete list of all the attributes available in an XML mapping file, refer to
section Basic O/R Mapping of the Hibernate Reference Documentation at:
http://docs.jboss.org/hibernate/core/3.3/reference/en/html/mapping.html#mapping
-declaration
6.3 Using Hibernate with an Informix database

In this section we demonstrate how to perform basic operations using the
Hibernate API with an IBM Informix database server.
6.3.1 Components of a Hibernate application

To develop a Hibernate application, the following tasks must be perform:
Create the Java objects.
Create the XML Mapping files for the Java objects.
Create the configuration file for the Hibernate framework.
Create the HibernateUtil helper class that provides access to the Hibernate
session.
Create the application that would use the Java persistence objects.
Java object
Hibernate supports the use of Plain Old Java Objects (POJO) for the definition of
a persistence object. This means that there is special requirement when writing
the Java class that represents the object.
Example 6-5 shows the contents of the State.java file used to define the State
object. The State Java object is created as any normal Java object. The object
have three properties (id, code, and sname) and the usual methods to set and get
those properties such as setcode() or setsname(). The id attribute is used to
uniquely identify the Java persistence object.
Example 6-5 State.java
public class State {
192
7884ch06.fm

private Long id;
private String code;
private String sname;
public State() {}
public State(String text, String text2) {
this.code = text;
this.sname = text2;
}
public Long getId() {
return id;
}
private void setId(Long id) {
this.id = id;
}
public String getcode() {
return code;
}
public void setcode(String text) {
this.code = text;
}
public String getsname() {
return sname;
}
public void setsname(String text) {
this.sname = text;
}
}
Example 6-6 shows the CLASSPATH used in the environment and how to
compile the State.java source code.
Example 6-6 Compile line for State.java
C:\work>set classpath
CLASSPATH=C:\work\lib\commons-collections-3.1.jar;C:\work\lib\dom4j-1.6.1.jar;C
:\work\lib\hibernate-jpa-2.0-api-1.0.0.Final.jar;C:\work\lib\hibernate3.jar;C:\
work\lib\ifxjdbc.jar;C:\work\lib\javassist-3.9.0.GA.jar;C:\work\lib\jta-1.1.jar
;C:\work\lib\slf4j-api-1.6.1.jar;C:\work\lib\slf4j-nop-1.6.1.jar;C:\work\lib\an
tlr-2.7.6.jar;.
C:\work>javac State.java
C:\work>
XML mapping file

A XML mapping file is required to link the Java object with an Informix database
object. The States table would keep all the instances of the Java State object
193
7884ch06.fm
The common convention for the name of the XML mapping file is
objectname_hbm.xml.
Example 6-7 shows the contents of the State_hbm.xml.
Example 6-7 State_hbm.xml file
<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC
"-//Hibernate/Hibernate Mapping DTD//EN"
"http://hibernate.sourceforge.net/hibernate-mapping-2.0.dtd">
<hibernate-mapping>
<class
name="State"
table="States">
<id
name="id"
column="id">
<generator class="increment"/>
</id>
<property
name="code"
column="code"/>
<property
name="sname"
column="sname"/>
</class>
</hibernate-mapping>
The name of the class is State and is mapped to the Informix SQL table States.
The State class contains two types of elements:
<id>: Defines the mapping from that property to the primary key column. It
usually contains the generator element which is used to generate unique
identifiers for the ID property.
The generator class supports different ways to generate identifiers. increment
is the most basic; sequence allows to use an SQL sequence to obtain the
identifier value. For a list of all the methods implemented in the generator
interface, refer to the Hibernate Reference Documentation at:
http://docs.jboss.org/hibernate/core/3.3/reference/en/html/mapping.html#map
ping-declaration-id
<property name>: Defines a property for the object including the name of the
property and the name of the mapped database table column.
194
7884ch06.fm
The elements of the XML mapping file may have additional attributes to define
specific characteristics of the columns it maps. Attributes like type and not-null
can be used to specify the data type and the nullability of a column.
Hibernate configuration file

The Hibernate configuration file must reside in the root directory of the
CLASSPATH. It contains configuration values for the Hibernate service and
connection details for the database server.
Example 6-8 shows the contents of the hibernate.cfg.xml file used in by our
sample. The connection details correspond to the IBM Informix JDBC Driver.
Example 6-8 hibernate.cfg.xml
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE hibernate-configuration PUBLIC
"-//Hibernate/Hibernate Configuration DTD 3.0//EN"
"http://hibernate.sourceforge.net/hibernate-configuration-3.0.dtd">
<hibernate-configuration>
<session-factory name="Informix_Session">

com.informix.jdbc.IfxDriver
</property>
jdbc:informix-sqli://kodiak:9088/stores_demo:INFORMIXSERVER=demo_on;
</property>
<property name="hibernate.connection.username">
informix
</property>
<property name="hibernate.connection.password">
password
</property>
org.hibernate.dialect.InformixDialect
</property>
<property name="hibernate.show_sql">false</property>
<property name="hbm2ddl.auto">update</property>

</session-factory>
Example 6-9 on page 196 shows the attributes required for an IBM Data Server
Driver for JDBC connection. The configuration file includes the XML mapping
State.hbm.xml for the State object.
195
7884ch06.fm

Example 6-9 Data Server Driver hibernate.cfg.xml file
org.hibernate.dialect.DB2Dialect
</property>
com.ibm.db2.jcc.DB2Driver
</property>
jdbc:ids://kodiak:9089/stores_demo;
</property>
HibernateUtil helper class

The HibernateUtil class is used to interact with the Hibernate service. It
performs the operations related to the Hibernate SessionFactory classes that
provide an convenient way for the application to access the Hibernate session.
The Java file containing the HibernateUtil helper class is HibernateUtil.java.
Example 6-10 shows a typical helper class.
Example 6-10 HiberanteUtil.java
import org.hibernate.SessionFactory;
import org.hibernate.cfg.Configuration;
public class HibernateUtil {
private static final SessionFactory sessionFactory;
static {
try {
// Creates the SessionFactory from hibernate.cfg.xml
sessionFactory = new Configuration().configure().buildSessionFactory();
}
catch (Throwable ex) {
System.err.println("SessionFactory creation failed." + ex);
throw new ExceptionInInitializerError(ex);
}
}
public static SessionFactory getSessionFactory() {
return sessionFactory;
}
public static void shutdown() {
// Close caches and connection pools
getSessionFactory().close();
}
196
7884ch06.fm

}
This class creates a Hibernate session using the parameters specified in the
Hibernate configuration file. Create the class by compiling the
HibernateUtil.java java file as follows:
javac HibernateUtil.java
Java application
The only task required by a Java application to use the Hibernate persistence
object is to create a Hibernate Session instance using the HibernateUtil helper
class. After that, the application can create persistence objects in the same way
as it does for any other Java object.
6.3.2 Working with a Hibernate object

Developers must follow an object-orientated methodology when writing
applications using the Hibernate API. With Hibernate, the data is represented by
the status and properties of an object, not by tables and rows in a relational
database.
Storing
When a persistent object is created and saved or stored, an insert operation is
performed on the database.
Example 6-11 shows a basic Java code that creates a new State object using
the parameters supplied in the command line. The example creates a Hibernate
Session object using the helper class, and then creates a new State object
called myState. The myState.setcode() and myState.setsname() are used to
stored the values passed from the command line.
Example 6-11 Create.java
import java.util.*;
import org.hibernate.*;
import javax.persistence.*;
public class create {
Session Session = HibernateUtil.getSessionFactory().openSession();
Transaction Transaction = Session.beginTransaction();
State myState = new State();
myState.setcode(args[0]);
myState.setsname(args[1]);
197
7884ch06.fm

Long stateId = (Long) Session.save(myState);
System.out.println("Stated added: "+myState.getcode()+", "
+myState.getsname());
Transaction.commit();
Session.close();
// Shutting down the application
HibernateUtil.shutdown();
}
}
Example 6-12 shows how the compile line and the output of the Create.java
sample.
Example 6-12 Create.java
C:\work>javac create.java
C:\work>java create AZ Arizona
Stated added: AZ, Arizona
C:\work>
The Hibernate property hbm2ddl.auto is set to update in the configuration file,

which means that if the table does not exists in the database, it is automatically
created using the table model defined in the XML mapping file.
Example 6-13 shows the schema of the created States table and the new row
added to the table.
Example 6-13 States table schema
D:\Infx\ids1150>dbaccess stores_demo Database selected.
> INFO COLUMNS FOR states;
Column name
Type
id
int8
code
varchar(255)
sname
varchar(255)
> SELECT * FROM states;
198
Nulls
no
yes
yes
7884ch06.fm

id
code
sname
1
AZ
Arizona
1 row(s) retrieved.
>
The Java object is made persistence, which means saving the state of the object
in the database, using the Session.save() Hibernate method and committing the
unit of work with Transaction.commit().
Loading
To retrieve a persistence object from the database server, the application must
create a Hibernate session and load the state of the object into the current
session.
The process of loading a persistence object can be achieved using several
methods:
Use Session.load() and Session.get() to retrieve the state of an object from
the database using the object identifier as reference. For example, to load the
State object with ID equal to 1, the following code is needed:
State mystateobj = (State) Session.load(State.class,1);
The only difference between the load() and get() methods is the returned
value. If the object is not found, get() returns a null and load() throws an
exception.
Use a SQL or HQL query. The Hibernate Query Language which is similar to
SQL but optimized for an object-orientated environment. Methods such as
Session.createQuery() and Session.createCiteria() can be used to load
the state of one or multiple objects from the database server. The following
command loads all State objects into a list.
List<State> states = session.createQuery("from state").list();
Example 6-14 demonstrates how to load a single object from the database using
the Session.load() method:
Example 6-14 Load.java
import
import
import
import
import
java.util.List;
org.hibernate.HibernateException;
org.hibernate.Session;
org.hibernate.Transaction;
org.hibernate.criterion.*;
199
7884ch06.fm

public class load {
Long stateId = null;
stateId=Long.parseLong(args[0]);
State mystate = (State) Session.load(State.class,stateId);
System.out.println(mystate.getcode() +", " +
mystate.getsname() );
Session.close();
}
}
Example 6-15 shows how to compile and run the previous example. The
application loads the State object identified by the id value passed through the
command line.
Example 6-15 Output of load.java
C:\work>javac load.java
C:\work>java load 1
AZ, Arizona
C:\work>java load 2
CA, California
C:\work>
The application can use a HQL query to load all the objects, or entities, of a
particular type. All the State objects are kept in the States table. Example 6-16
shows how to retrieve a list of the State entities.
Example 6-16 List.java
C:\work>cat list.java
import java.util.*;
import org.hibernate.*;
public class list {
200
7884ch06.fm

Session newSession = HibernateUtil.getSessionFactory().openSession();

Transaction newTransaction = newSession.beginTransaction();
List states = newSession.createQuery("from State order by id asc").list();
for ( Iterator iter = states.iterator();
iter.hasNext(); ) {
State state = (State) iter.next();
System.out.println(state.getId() +", "+ state.getcode()
+", " + state.getsname() );
}
newTransaction.commit();
newSession.close();
}
}
C:\work>javac list.java
C:\work>java list
1, AZ, Arizona
2, CA, California
C:\work>
Updating
Updating an persistence object is the process of changing the state of the object.
There is no specific operation required for this task, the application must load the
object, change the object properties, and save it.
Example 6-17 demonstrates how to update individual objects. It uses the
parameters from the command line to change the code and sname properties of a
State object.
Example 6-17 Update.java
C:\work>cat update.java
import
import
import
import
java.util.List;
org.hibernate.HibernateException;
org.hibernate.Session;
org.hibernate.Transaction;
public class update {

201
7884ch06.fm

mystate.setcode(args[1]);
mystate.setsname(args[2]);
Session.save(mystate);
System.out.println("new values: "+ mystate.getcode()+", "+
Session.close();
}
}
C:\work>javac update.java
C:\work>java update 1 AZ ARIZONA
new values: AZ, ARIZONA
C:\work>
Criteria
Criteria queries are a feature of the Hibernate Query Language that allows to
build complex queries using an object-orientated API.
Example 6-18 shows how to update a president object using a HQL criteria to
retrieve the object from the database. The HQL Criteria is created using the
name of the class for the object we want to select, State.class. A HQL
Restriction, Restriction.eq("code", args[0]), is used to specify a filter for the
criteria. This restriction specifies that the query should return the objects with a
specific code value.
Example 6-18 Update2.java
import
import
import
import
java.util.*;
org.hibernate.*;
javax.persistence.*;
org.hibernate.criterion.*;
public class update2 {

202
7884ch06.fm

Criteria crit = Session.createCriteria(State.class);
crit.add( Restrictions.eq( "code", args[0]) );
List states = crit.list();
for ( Iterator iter = states.iterator(); iter.hasNext(); ) {
State mystate = (State) iter.next();
mystate.setsname(args[1]);
Session.flush();
Long msgId = (Long) Session.save(mystate);
System.out.println("new values: "+ mystate.getcode()+", "+
}
Session.close();
}
}
Example 6-19 shows the output of the example.

Example 6-19 Update2.java output
C:\work>javac update2.java
C:\work>java update2 AZ Arizona
new values: AZ, Arizona
C:\work>
Delete
In the Hibernate framework, delete a persistence object means to make the
object transient. A transient object is a typical Java object but the state of the
object is not stored anywhere, which means that the object will only exists during
the life of the application.
An object can be deleted using the Session.delete() method. Example 6-20
demonstrates how to use the Session.delete() method:
Example 6-20 delete.java
C:\work>cat delete.java
import java.util.List;
import org.hibernate.HibernateException;
203
7884ch06.fm

import org.hibernate.Session;
import org.hibernate.Transaction;
public class delete {
Session.delete(mystate);
System.out.println( "Object deleted");
Session.close();
}
}
C:\work>javac delete.java
C:\work>java delete 2
Object deleted
C:\work>java list
1, AZ, Arizona
C:\work>
One of the main features of using Hibernate is that abstracts the JDBC layer from
the application. This means that applications written using the Hibernate API are
not tied to a specific JDBC driver or relational database server.
All the examples presented in this section can be run using any of the two
Informix JDBC drivers available, IBM Informix JDBC Driver or IBM Data Server
Driver for JDBC. Refer to the Hibernate configuration file on page 189 for
information regarding how to switch between JDBC drivers.
For more information regarding how to develop using the Hibernate API, refer to
the Hibernate Reference Documentation at:
http://docs.jboss.org/hibernate/core/3.3/reference/en/html/objectstate.html
204
7884ch06.fm
6.3.3 Using annotations

Java annotations is a feature added into the Java 5 Software Development Kit
(SDK) that allows the inclusion of special annotations within the Java source
code to express metadata relating to program objects.
Hibernate supports Java annotations through the use of the Hibernate
Annotations Extensions. These extensions allow to include the definition of
specific Hibernate properties such as configuration properties or object
mappings properties directly into the Java code.
With annotations, there is no need for a specific XML mapping file to link the Java
objects with the database objects.
Example 6-21 shows the definition of the State object using Hibernate
annotations. Annotations such as @Entity, @Table or @Column define the
mapping information needed to persist the Java objects into the database.
Example 6-21 State.java with annotations
@Entity
@Table(name="States")
public class State {
private Long id;
private String code;
private String sname;
public State() {}
public State(String text, String text2) {
this.code = text;
this.sname = text2;
}
@Id
@GeneratedValue(strategy=GenerationType.IDENTITY )
@Column(name="id")
public Long getId() {
return id;
}
private void setId(Long id) {
this.id = id;
}
@Column(name="code", length=2, nullable=false)
public String getcode() {
return code;
}
public void setcode(String text) {
205
7884ch06.fm

this.code = text;
}
@Column(name="sname", length=15, nullable=false)
public String getsname() {
return sname;
}
public void setsname(String text) {
this.sname = text;
}
}
Because all the information required to map the objects is specify as annotations,
there is no need to include a <mapping> section in the Hibernate configuration
file, hibernate.cfg.xml.
When using annotations, the HibernateUtil helper class, HibernateUtil.java,
uses the AnnotationConfiguration() interface to retrieve the Hibernate
properties and the metadata definition from the object class. Example 6-22
shows the HibernateUtil.java file used to create a SessionFactory for the
State.class object.
Example 6-22 HibernateUtil.java for annotations
import org.hibernate.SessionFactory;
import org.hibernate.cfg.AnnotationConfiguration;
public class HibernateUtil {
private static final SessionFactory sessionFactory;
static {
try {
// Create the SessionFactory from hibernate.cfg.xml
sessionFactory = new AnnotationConfiguration()
.configure()
.addAnnotatedClass(State.class)
.buildSessionFactory();
} catch (Throwable ex) {
System.err.println("Initial SessionFactory creation failed." + ex);
throw new ExceptionInInitializerError(ex);
}
}
public static SessionFactory getSessionFactory() {
return sessionFactory;
}
public static void shutdown() {
206
7884ch06.fm
getSessionFactory().close();
}
}
There is no need to change the code for the Java applications that use the
persistence object, the process for manipulate the objects is the same as when
using the XML mapping files than when using Hibernate annotations.
A description of all the Hibernate annotation extensions can found in the
Hibernate Reference Documentation at
http://docs.jboss.org/hibernate/stable/annotations/reference/en/html_single/#en
tity-hibspec
207
7884ch06.fm
208
7884ch07.fm
Chapter 7.
Working with the IBM

Informix OLE DB Provider
This chapter describes the IBM Informix OLE DB provider. we discuss the
following topics:
Description of the Informix OLE DB provider

Setup and Configuration
Application development using the OLE DB provider
Typical errors and how to enable tracing
209
7884ch07.fm
7.1 IBM Informix OLE DB Provider

Informix OLE DB Provider is a Universal Data Access component that enables
IBM Informix Server access from OLE DB consumers.
Microsoft OLE DB is a specification for a set of interfaces designed to expose
data from a variety of sources (relational and non-relational). OLE DB uses the
Component Object Model (COM) to accomplish this.
You can use IBM Informix OLE DB Provider to enable client applications, such as
ActiveX Data Object (ADO) applications and Web pages, to access data on an
IBM Informix database server.
Table 7-1 shows the name and COM class ID of the IBM Informix OLE DB
Provider.
Table 7-1 COM class ID
Name
DLL
CLSID
ifxoledbc
iifxoledbc.dll
{A6D00422-FD6C-11D0-8043-00A0C90F1C59}
Table 7-2 lists the Informix database servers that support IBM Informix OLE DB
Provider.
Table 7-2 Supported databases
Database Server
Versions
IBM Informix
10.0,11.10,11.50
8.50 and higher
IBM Informix Online
5.20 and higher

In this section we discuss how to set up the OLE DB provider and how to perform
basic connectivity tests.
7.2.1 Installation and setup

The IBM Informix OLE DB provider is only included in the Windows version of
IBM Informix Client SDK. The OLE DB provider is selected by default in the
210
7884ch07.fm
Client SDK installation. During the install process, the provider is automatically
registered on the Windows registry as a command component.
The default install directory is C:\Program Files\IBM\Informix\Client-SDK.
The INFORMIXDIR environment variable should point to the directory where the
product was installed. The provider shared library, ifxoledbc.dll, is located in
%INFORMIXDIR%\bin directory.
Because the OLE DB provide is registered, by default, during the Client SDK
installation, manual registration is usually not required. However, in cased
needed, you can manually register the OLE DB provider using the Microsoft
regsvr32.exe tool by running the following command in a command prompt:
regsvr32.exe %INFORMIXDIR%\bin\ifxoledbc.dll
Note: On a Windows x64 (64-bit) there are two versions of the regsvr32.exe
tool:
The 32-bit version is located in C:\WINDOWS\SysWOW64
The 64-bit version is located in C:\WINDOWS\System
Use the correct version when registering the OLE DB provider, otherwise, the
application would fail to load the shared library due to a mismatched version.
After installation, you must run the script coledbp.sql on the database server
against the sysmaster database as user informix to add the tables and functions
required by the provider to work.
The script coledbp.sql is located in the %INFORMIXDIR%\etc directory. If you want
to remove the support functions and tables, use the doledbp.sql script located in
the same directory.

Informix Client SDK does not contain any specific tool to test the OLE DB
provider.
Internally, the IBM Informix OLE DB Provider uses the same Informix connection
libraries as ESQL/C or ODBC. You can test the basic connection details for your
database server using the Ilogin tool included in the %INFORMIXDIR%\bin
directory.
Chapter 7. Working with the IBM Informix OLE DB Provider
211
7884ch07.fm
Testing using VBScript

VBScript is a scripting language included as part of Windows operating system.
It provide access to ADO objects through COM and can be use to test if the IBM
Informix OLE DB provider is properly configured.
VBScripts are text files with a vbs extension which are automatically loaded and
executed by the Visual Basic runtime.
Example 7-1 shows a simple VBS script which loads the Informix OLE DB
provider and selects a single row from the database.
Example 7-1 Connect.vbs
' ---- Test_Ifx.vbs ---On Error Resume Next
args = WScript.Arguments.Count
If args < 1 then
WScript.Echo "usage: connect.vbs Connection_string"
WScript.Echo " e.g.: connect.vbs ""Data Source=stores_demo@demo_on;User
ID=informix;Password=password;"""
WScript.Quit
end If
set conn=createobject("ADODB.Connection")
conn.provider = "Ifxoledbc"
conn.connectionstring = WScript.Arguments.Item(0)
conn.open
If Err then
WScript.Echo "Error!! "+conn.Errors(0).Description
Else
WScript.Echo "Connected"
conn.close
End If
' ---- Test_Ifx.vbs ----
Example 7-2 shows how run the Connect.vbs script with a data source string as
the parameter to test the database connectivity.
Example 7-2 Output of Connect.vbs
c:\work>connect.vbs
usage: connect.vbs Connection_string
e.g.: connect.vbs "Data Source=stores_demo@demo_on;User
ID=informix;Password=password;"
c:\work>connect.vbs "Data Source=stores_demo@demo_on;User
ID=informix;Password=password;"
212
7884ch07.fm
Connected
c:\work>connect.vbs "Data Source=stores_demo@demo_on;User
ID=informix;Password=invalid;"
Error!! EIX000: (-951) Incorrect password or user informix@dubito is not known
on the database server.
Rowset Viewer
Another way to test the Informix OLE DB provider is using the Rowset Viewer.
The Rowset Viewer is included with the Microsoft Data Access Components
(MDAC) 2.8 Software Development Kit (SDK) which can be download from the
following location
http://www.microsoft.com/downloads/details.aspx?FamilyID=6c050fe3-c795-4b7d-b03
7-185d0506396c
To test the OLE DB provider using the Rowset tool, perform the following steps:
Run the Rowset tool .
Select File Full Connect.
In the Provider field choose Ifxoledbc.
In the DataSource field, enter the database as server name you want to
connect to, for example, stores_demo@demo_on.
Press the OK.
Figure 7-1 shows the Full Connect dialog box.
Figure 7-1 Rowset Full Connect dialog
213
7884ch07.fm
You also can use Rowset Viewer to run SQL statements against the database
server. To execute a SQL statement, write the SQL text in the Rowset pane and
click
.
Figure 7-2 shows the Rowset Viewer running the SQL statement select * from
state.
Figure 7-2 Rowset Viewer SQL Statement
7.3 Developing an OLE DB application

This section describes the interfaces implemented in the Informix OLE DB
Provider and demonstrates how to perform basic database operations.
In this section we discuss:
214
Supported interfaces in the Informix OLE DB Provider

Parameters used in the connection string
Data type mapping
Typical database operations
7884ch07.fm
7.3.1 Supported interfaces

A OLE DB application creates objects based on ADO interfaces to perform
operations against a database server. Table 7-3 lists a few ADO interfaces
implemented in the IBM Informix OLE DB Provider as examples. For a complete
list, refer to
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.oledb.doc/s
ii-xa-21954.htm#sii-xa-21954
Table 7-3 OLE DB supported interfaces
Interface
Description
IAccessor
Provides methods for accessor management
IColumnsInfo
Information about columns of a rowset or prepared command
ICommand
Execute commands
IDBCreateCommand
Obtain a new command.
IDBCreateSession
Obtain a new session.
IDBDataSourceAdmin
Create, destroy, and modify a data source objects
IDBProperties
Gets and sets the values of properties on the data source

object or enumerator and to get information about all
properties supported by the provider.
IErrorLookup
Used by OLE DB error objects to determine the values of the

error message, source, Help file path, and context ID based
on the return code and a provider-specific error number.
IGetDataSource
Obtaining an interface pointer to the data source object.
IRowsetIdentity
Indicates row instance identity is implemented on the rowset

and enables testing for row identity.
ISessionProperties
Returns information about the properties a session supports

and the current settings of those properties.
ITransaction
Used to commit, abort, and obtain status information about

transactions.
7.3.2 Connecting to database

In this section we describe the connection string options with the IBM Informix
OLE DB Provider.
215
7884ch07.fm
The connection details for the database server are passed to the OLE DB
provider using a connection string. If the Provider attribute of the Connection
object is not set, you must include the provider keyword as part of the connection
string to specify the name of the Informix OLE DB provide, ifxoledbc. For
example:
connStr="Provider=Ifxoledbc;Data Source=stores_demo@demo_on"
Table 7-4 describes the specific connection string attributes for the Informix OLE
DB Provider.
Table 7-4 Connection String attributes
Keyword
Description
Data Source
Database and Informix server to connect to.

The syntax for the Data Source parameter is
database@server
If @server is not specified, the default database
server is used (corresponding to the value specified
by the clients INFORMIXSERVER registry entry or
environment variable).
User ID
The user ID used to connect to the Informix server
Password
The password for the user ID
Persist Security Info
Specifies whether the data source can keep

authentication information such as the password
Client_locale
The client locale for the application which

correspond to the locale used by the Windows OS
Db_locale
The database locale. Locale used when the

database was created
UNICODE
Indicates whether to use IBM Informix GLS Unicode

Controls how the codeset conversion to Unicode is
done
decasr8
If set, floating point numbers with a scale greater

than 30 are returned as DBTYPE_R8
RSASWS or
REPORTSTRINGASWSTRING
Enables you to control the data mapping for wide

strings
FBS or FETCHBUFFERSIZE
The size in bytes of the buffer size used to send data

to or from the database. Minimum value is 4096,
Maximum 32767. Default is 4096
A typical connection string for the IBM Informix OLE DB Provider looks like:
216
7884ch07.fm
Data Source=stores_demo@demo_on; User ID=informix; Password=password; Persist

Security Info=True; CLIENT_LOCALE=en_US.CP1252; DB_LOCALE=en_US.819
The connection details about the Informix server must be stored in the registry
using the Setnet32 tool included in Informix Client SDK.
7.3.3 Type mapping

The Informix OLE DB provider supports all the Informix data types, both build-in
and extended.
Table 7-5 lists the mappings between the standards OLE DB data types and
specific Informix OLE DB types.
Table 7-5 Informix specific type mapping table
Informix SQL
BIGINT
DBTYPE_I8
BIGSERIAL
DBTYPE_I8
BLOB
DBTYPE_BYTES
BOOLEAN
DBTYPE_BOOL
BYTE
DBTYPE_BYTES
CLOB
DBTYPE_STR
DATETIME
DBTYPE_DBDATE, DBTYPE_DBTIME,
DBTYPE_DBTIMESTAMP
DECIMAL
DBTYPE_NUMERIC
DISTINCT
Same as underlying type
FLOAT
DBTYPE_R8
INT8
DBTYPE_I8
INTERVAL
DBTYPE_STR if mapped as a string

DBTYPE_Ix (8,4,2 or 1) if mapped as a numeric
LIST
DBTYPE_VARIANT
LVARCHAR
DBTYPE_STR
MONEY (p<=19 s<=4
DBTYPE_CY
MONEY (p>19 s<>4)
DBTYPE_NUMERIC
217
7884ch07.fm
Informix SQL
MULTISET
DBTYPE_VARIANT
Named ROW
DBTYPE_VARIANT
NCHAR
DBTYPE_STR
OPAQUE
DBTYPE_BYTES
SERIAL
DBTYPE_I4
SERIAL8
DBTYPE_I8
SET
DBTYPE_VARIANT
TEXT
DBTYPE_STR
Unnamed ROW
DBTYPE_VARIANT
There are three OLE DB Provider date and time data types that map to one
DATETIME Informix SQL data type. Which one to use depends on the precision
of the data type desired. See Table 7-6 for a summary.
Table 7-6 DATETIME OLE DB mapping table
Date and time precision
DBTYPE_DBDATE
Day to day, month to day, month to month, year to day,

year to month, and year to year
DBTYPE_DBTIME
Hour to hour, hour to minute, hour to second, minute to

minute, minute to second, and second to second
DBTYPE_DBTIMESTAMP
Day to fraction, day to hour, day to minute, and day to

second
Fraction to fraction, hour to fraction, and minute to
fraction
Month to fraction, month to hour, month to minute, and
month to second
Second to fraction
Year to fraction, year to hour, year to minute, and year
to second
For a complete list of all the data type mapping, refer to the Data Type Mappings
section of the IBM Informix OLE DB Guide at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.oledb.doc/u
sing990839.htm#using990839
218
7884ch07.fm
7.3.4 Cursors
An ADO application uses cursors to move among the data returned by a
recordset. ADO defines four types of cursors:
Forward-Only (adOpenForwardOnly):
Provides a copy of the records at the time the Recordset was created. This is
the default cursor type of ADO.
Static (adOpenStatic):
Provides a static copy of the records.
Dynamic (adOpenDynamic):
Provides a real-time copy of the records, new and altered records by other
users are also included.
Keyset (adOpenKeyset):
Provides a updateable copy of the records at the time the Recordset was
created. Only existing records are included.
The IBM Informix OLE DB Provider supports all these cursor types with some
limitations regarding the use of extended data types and location of the cursor
(Client or Server side). For example:
Updates of tables with extended data types are not allowed with Client-side
scrollable cursors.
Server-side scrollable cursors are not supported with simple large objects
(BYTE and TEXT) or collections.
ROWIDs (internal columns added by the Informix server on non fragmented
tables) are required on tables bookmarks and updates.
Refer to the IBM Informix OLE DB Provider Programmer's Guide for a complete
list of these caveats:
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.oledb.doc/s
ii-using-29878.htm#sii-using-29878
7.3.5 Typical database operations

In this section we provide examples of how to use the IBM Informix OLE DB
Provider to perform typical database operations against an Informix database.
We cover operations such as executing SQL statements, selecting data, or how
to handle errors coming from the OLE DB provider.
for the general information about OLE DB architecture and programming, refer to
the Microsofts OLE DB Programer's Guide at
219
7884ch07.fm

http://msdn.microsoft.com/en-us/library/ms713643%28v=VS.85%29.aspx
Command
The execution of the SQL commands using an OLE DB provider is accomplished
through the ICommand interface.
The application has to perform the following steps in order to execute a SQL
statement:
1. Call QueryInterface for IDBCreateCommand to check if commands are
supported in the session.
2. Call IDBCreateCommand::CreateCommand to create the command.
3. Call ICommandText::SetCommandText to specify the SQL statement for the
command.
4. Call ICommand::Execute to execute the SQL statement.
Example 7-3 on page 221 demonstrates how to connect to the database and
execute an UPDATE statement. In this example, we use the following operations
and interfaces:
Connect to the database
a. Initialize common property options (prompt, DSN, user, and password).
b. Get an Interface for properties:
InterfacepIDBInitialize->QueryInterface(IID_IDBProperties)
pIDBProperties->SetProperties()
c. Get an Interface for creating session object:

pIDBInitialize->QueryInterface(IID_IDBCreateSession)
d. Create a session object:

pCreateSession->CreateSession()
Create a command object:

e. Get an Interface for creating command object:
pSession->QueryInterface(IID_IDBCreateCommand)
f. Create command object and get ICommandText

Interface:pCreateCommand->CreateCommand()
Set the CommandText for the command object:
pCommandText->SetCommandText()
Execute the command:

pCommandText->Execute()
Clean up.
220
7884ch07.fm
Example 7-3 Command.cpp

#define UNICODE
#define _UNICODE
#include <oledb.h>
#include <oledberr.h>
#include <msdaguid.h>
#include <msdadc.h>
#include <comdef.h>
#include <stdio.h>
// CLSID For IBM-Informix Client Side OLE DB Provider
const GUID CLSID_IFXOLEDBC= {0xa6d00422, 0xfd6c, 0x11d0,
{0x80, 0x43, 0x0, 0xa0, 0xc9, 0xf, 0x1c, 0x59}};
#define CHECK(hr) if (((HRESULT)(hr)) < 0) {printf( "Error"); return(hr);}
int main()
{
HRESULT
hr = S_OK;
IDBInitialize
*pIDBInitialize;
IDBCreateSession
*pCreateSession;
IDBCreateCommand
*pCreateCommand;
IUnknown
*pSession;
ICommandText
*pCommandText;
IDBProperties
*pIDBProperties;
DBPROP
InitProperties[4];
DBPROPSET
rgInitPropSet;
_bstr_t bstrDsnName = "stores_demo@demo_on";
_bstr_t bstrUserName = "informix";
_bstr_t bstrPassWord = "password";
_bstr_t bstrCommand = (WCHAR *) L"UPDATE state SET sname = 'California'

WHERE code = 'CA'";
CoInitialize( NULL );
// Instantiate a data source object
CHECK( hr = CoCreateInstance((REFCLSID) CLSID_IFXOLEDBC, NULL,
CLSCTX_INPROC_SERVER, IID_IDBInitialize, (void **) &pIDBInitialize))
// Set all Properties, Prompt, DSN, User and Password
// Initialize common property options.
for (ULONG i = 0; i < 4; i++ )
{
VariantInit(&InitProperties[i].vValue);
InitProperties[i].dwOptions = DBPROPOPTIONS_REQUIRED;
InitProperties[i].colid = DB_NULLID;
InitProperties[1].vValue.vt = VT_BSTR;
}
InitProperties[0].dwPropertyID = DBPROP_INIT_PROMPT;
221
7884ch07.fm

InitProperties[0].vValue.vt = VT_I2;
InitProperties[0].vValue.iVal = DBPROMPT_NOPROMPT;
InitProperties[1].dwPropertyID = DBPROP_INIT_DATASOURCE;
InitProperties[1].vValue.bstrVal = bstrDsnName;
InitProperties[2].dwPropertyID = DBPROP_AUTH_USERID;
InitProperties[2].vValue.bstrVal = bstrUserName;
InitProperties[3].dwPropertyID = DBPROP_AUTH_PASSWORD;
InitProperties[3].vValue.bstrVal = bstrPassWord;
rgInitPropSet.guidPropertySet = DBPROPSET_DBINIT;
rgInitPropSet.cProperties = 4;
rgInitPropSet.rgProperties = InitProperties;
// Get initialization properties.Interface
CHECK (hr = pIDBInitialize->QueryInterface(IID_IDBProperties,
(void**) &pIDBProperties))
CHECK(hr = pIDBProperties->SetProperties( 1, &rgInitPropSet))
CHECK( hr = pIDBProperties->Release())
// Connect to the Database Server
CHECK(hr = pIDBInitialize->Initialize())
// Get an Interface for creating Session Object
CHECK( hr = pIDBInitialize->QueryInterface(IID_IDBCreateSession,
(void **) &pCreateSession))
// Create a Session Object
CHECK( hr = pCreateSession->CreateSession(NULL, IID_IUnknown,
(IUnknown **) &pSession))
CHECK( hr = pCreateSession->Release())
// Create Command Object

// Get an Interface for creating Command Object
CHECK( hr = pSession->QueryInterface( IID_IDBCreateCommand,
(void **) &pCreateCommand))
// Create Command Object and get ICommandText Interface
CHECK( hr = pCreateCommand->CreateCommand(NULL, IID_ICommandText,
(IUnknown **) &pCommandText ))
CHECK( hr = pCreateCommand->Release())
// Set the CommandText
CHECK( hr = pCommandText->SetCommandText(DBGUID_DBSQL, bstrCommand))
// Executing the Command
CHECK( hr = pCommandText->Execute( NULL, IID_NULL, NULL, NULL, NULL))
printf( "Row Updated");
// Cleanup
pCommandText->Release();
pCommandText = NULL;
pSession->Release();
pSession = NULL;
pIDBInitialize -> Uninitialize();
pIDBInitialize -> Release();
pIDBInitialize = NULL;
222
7884ch07.fm
CoUninitialize();
return(0);
}
Example 7-4 describes how to compile and run Example 7-3 on page 221. The
connection string is constructed inside the program, so there are not passing any
command line parameters.
Example 7-4 Output of Command.cpp
C:\work>cl /EHsc /nologo command.cpp
command.cpp
C:\work>command
Row Updated
C:\work>
Rowset
A Rowset is a set of rows and each row has a number of columns of data.
Rowsets are the main OLE DB objects used to exposes data from a data source.
You can create rowsets object with the following methods:
Explicitly create a rowset by calling IOpenRowset::OpenRowset().
Execute an SQL statement such as SELECT that returns rows with a
ICommand:Execute method.
Execute any method that returns a rowset or a schema rowset, for example,
ColumnsRowset::GetColumnsRowset or IDBSchemaRowset::GetRowset.
Example 7-5 on page 224 demonstrates how to create a rowset object and
retrieve metadata information. The operations and interfaces to initialize the
session and connect to the database are always the same in a OLE DB
application. This example also does the following operations:
Connect to the database
Create a Command Object
Create a OpenRowSet Object
pCreateCommand->QueryInterface(IID_IOpenRowset)
pIOpenRowset->OpenRowset()
Obtain access to the IColumnsInfo interface, from the rowset object

pRowset->QueryInterface(IID_IColumnsInfo)
Retrieve the Column Information
223
7884ch07.fm

pColumnsInfo->GetColumnInfo()
Cleanup
Example 7-5 Rowset.cpp
#define UNICODE
#define _UNICODE
#include <oledb.h>
#include <msdadc.h>
#include <comdef.h>
#include <stdio.h>
{0x80, 0x43, 0x0, 0xa0, 0xc9, 0xf, 0x1c, 0x59}};
int main()
{
HRESULT
hr = S_OK;
IDBInitialize
IUnknown
ITransactionJoin
ICommandText
IOpenRowset
IColumnsInfo
*pIDBInitialize;
*pSession;
*pITransactionJoin;
*pCommandText;
*pIOpenRowset;
*pColumnsInfo;
IDBProperties*
pIDBProperties;
DBPROP
InitProperties[4];
DBPROPSET
rgInitPropSet;
IDBCreateSession
*pCreateSession = NULL;
IDBCreateCommand
*pCreateCommand = NULL;
IRowset
*pRowset = NULL;
DBID TableID;
DBPROPSET rgPropSets[1];
DBCOLUMNINFO *pDBColumnInfo;
WCHAR *pStringsBuffer;
ULONG lNumCols;
int i=0;
_bstr_t
_bstr_t
_bstr_t
bstrDsnName = "stores_demo@demo_on";
bstrUserName = "informix";
bstrPassWord = "password";
_bstr_t
TableName = L"customer";
224
7884ch07.fm

for (ULONG i = 0; i < 4; i++ )
{
}
CHECK(hr = pIDBInitialize->Initialize())
// Get an Interface for creating Session Object
CHECK( hr = pIDBInitialize->QueryInterface(IID_IDBCreateSession,
(void **) &pCreateSession))
// Create a Session Object
CHECK( hr = pCreateSession->CreateSession(NULL, IID_IUnknown,
(IUnknown **) &pSession))
CHECK( hr = pCreateSession->Release())
// Create Command Object

// Get an Interface for creating Command Object
CHECK( hr = pSession->QueryInterface( IID_IDBCreateCommand,
(void **) &pCreateCommand))
// Create Command Object and get ICommandText Interface
CHECK( hr = pCreateCommand->CreateCommand(NULL, IID_ICommandText,
(IUnknown **) &pCommandText ))
CHECK( hr = pCreateCommand->Release())
// Create a OpenRowSet
CHECK( hr = pCreateCommand->QueryInterface( IID_IOpenRowset,
225
7884ch07.fm

(void **)&pIOpenRowset ))
TableID.eKind = DBKIND_NAME;
TableID.uName.pwszName = TableName;
CHECK( hr = pIOpenRowset->OpenRowset( NULL, &TableID, NULL, IID_IRowset,
0, rgPropSets, (IUnknown **)&pRowset
))
// Obtain access to the IColumnsInfo inteface, from the Rowset object
CHECK( hr = pRowset->QueryInterface(IID_IColumnsInfo, (void ** )
&pColumnsInfo))
// Retrieve the Column Information

CHECK( hr = pColumnsInfo->GetColumnInfo( &lNumCols, &pDBColumnInfo,
&pStringsBuffer))
wprintf( L"\nTable : %s\nColumns :%d\nName\t\tType\tLength",
TableID.uName.pwszName, lNumCols);
for ( i=0; i<(int)lNumCols; ++i)
{
wprintf( L"\n%s\t%d\t%ld", (pDBColumnInfo+i)->pwszName,
(pDBColumnInfo+i)->wType, (pDBColumnInfo+i)->ulColumnSize);
}
// Free the Column Information Interface
CHECK( (hr = pColumnsInfo->Release()))
// Cleanup
pCommandText->Release();
pCommandText = NULL;
pSession->Release();
pSession = NULL;
CoUninitialize();
return(0);
}
Example 7-6 shows the name, type, and length columns of the customer table.
Example 7-6 Output of Rowset.cpp
C:\work>cl /EHsc /nologo rowset.cpp
rowset.cpp
C:\work>rowset
Table : customer
Columns :7
Name
Type
226
Length

customer_num
customer_type
customer_name
customer_loc
contact_dates
cust_discount
credit_status
C:\work>
3
129
129
3
129
131
129
7884ch07.fm
4
1
32767
4
32767
19
1
Large objects
Large object (simple and smart) do not require any special handling when using
the Informix OLE DB provider.
Smart Large Objects (BLOB and TEXT) support the ADO GetChunk() and
AppendChunk() methods.
Example 7-7 demonstrates how to retrieve a CLOB column using GetChunk()
from the ADO library in C++. This example fetches the first three rows of the
catalog table from the sample stores_demo database and displays the data for
the catalog_number and advert_descr columns.
Example 7-7 Select.cpp
#include <stdio.h>
#include <afxdisp.h>
#import "c:\program files\common files\system\ado\msado15.dll" rename
("EOF","adoEOF") no_namespace
#define CREATEiNSTANCE(sp,riid) { HRESULT _hr =sp .CreateInstance( __uuidof( riid
) ); \
if (FAILED(_hr)) _com_issue_error(_hr); }
#define RsITEM(rs,x) rs->Fields->Item[_variant_t(x)]->Value
#define UC (char *)
struct InitOle {
InitOle() { ::CoInitialize(NULL); }
~InitOle() { ::CoUninitialize();
}
} _init_InitOle_;
void main(){
_RecordsetPtr
_ConnectionPtr
_variant_t
long
int
spRS;
spCON;
varBLOB;
lDataLength = 0;
nrows=3;
227
7884ch07.fm

try{
CREATEiNSTANCE(spCON,Connection);
spCON->Provider="ifxoledbc";
spCON->ConnectionString = L"Data Source=stores_demo@demo_on;"
L"User ID=informix; Password=password;";
// Connect to the database

spCON->Open( "", "", "", -1 );
CREATEiNSTANCE(spRS,Recordset)
spRS->PutRefActiveConnection( spCON );
spRS->Open(_bstr_t("catalog"),vtMissing, adOpenForwardOnly,
adLockOptimistic,adCmdTable);
while( (spRS->adoEOF == false) && nrows>0 ){
nrows--;
printf("catalog_num = %s\n", UC _bstr_t(RsITEM(spRS,"catalog_num")));
// Get the size of the large object
lDataLength = spRS->Fields->Item["advert_descr"]->ActualSize;
if(lDataLength > 0) {
// Call GetChunk to retrieve the Blob data

VariantInit(&varBLOB);
varBLOB = spRS->Fields->Item["advert_descr"]->GetChunk(lDataLength);
printf("advert_descr = %s\n", UC _bstr_t(varBLOB));
}
// Move the cursor to the next row
spRS->MoveNext();
}
spRS->Close();
spCON->Close();
}
catch( _com_error &e){
_bstr_t bstrSource(e.Source());
_bstr_t bs = _bstr_t(" Error: ") + _bstr_t(e.Error()) +
_bstr_t(" Msg: ") + _bstr_t(e.ErrorMessage()) +
_bstr_t(" Description: ") + _bstr_t(e.Description());
printf("Error %s\n", bs);
}
}
#undef UC
Example 7-8 shows the content of the output.

Example 7-8 Output of Select.cpp
C:\work>cl /EHsc /nologo select.cpp /DWINVER=0x0600
select.cpp
C:\work>select
catalog_num = 10001
228
7884ch07.fm
advert_descr = Brown leather. Specify first baseman's or infield/outfield

catalog_num = 10002
catalog_num = 10003
C:\work>
Errors
The Informix OLE DB Provider supports the ISupportErrorInfo interface.
Applications can retrieve information about an OLE DB error using this interface.
Methods like IErrorRecords->GetErrorInfo(), IErrorInfo->GetDescription()
and IErrorInfo->GetSource() are fully supported by the Informix OLE DB
provider.
Example 7-9 Illustrates how to use these methods. This example retrieves the
error information directly after the IDBInitialize->Initialize() function call.
Because error handling should be done after every use of an interface function, it
is a good practice to create an error handle routine to avoid code redundancy.
Example 7-9 Errorinfo.cpp
#define UNICODE
#define _UNICODE
#include <oledb.h>
#include <msdadc.h>
#include <comdef.h>
#include <stdio.h>
{0x80, 0x43, 0x0, 0xa0, 0xc9, 0xf, 0x1c, 0x59}};
HRESULT GetDetailedErrorInfo(
HRESULThresult,
IUnknown *pBadObject,
GUID IID_BadIntereface);
int main()
{
HRESULT
hr = S_OK;
IDBInitialize
*pIDBInitialize;
229
7884ch07.fm

IUnknown
*pSession;
IDBProperties*
DBPROP
DBPROPSET
IDBCreateSession
pIDBProperties;
InitProperties[4];
rgInitPropSet;
*pCreateSession = NULL;
IErrorInfo
IErrorInfo
IErrorRecords
ISupportErrorInfo
*pErrorInfo = NULL;
*pErrorInfoRec = NULL;
*pErrorRecords = NULL;
*pSupportErrorInfo = NULL;
ULONG
BSTR
BSTR
i,ulNumErrorRecs;
bstrDescriptionOfError = NULL;
bstrSourceOfError = NULL;
_bstr_t
_bstr_t
_bstr_t
bstrDsnName = "wrong_db@demo_on"; // WRONG DATABASE

bstrUserName = "informix";
bstrPassWord = "password";
for (ULONG i = 0; i < 4; i++ )
{
}
230
7884ch07.fm

hr = pIDBInitialize->Initialize();
if (hr < 0) {
pIDBInitialize->QueryInterface(IID_ISupportErrorInfo,
(LPVOID FAR*)&pSupportErrorInfo);
pSupportErrorInfo->InterfaceSupportsErrorInfo(__uuidof(pIDBInitialize));
GetErrorInfo(0,&pErrorInfo);
//Get the IErrorRecord interface and get the count of error recs.
pErrorInfo->QueryInterface(IID_IErrorRecords,(LPVOID FAR*)&pErrorRecords);
pErrorRecords->GetRecordCount(&ulNumErrorRecs);
//Get the error record, (we only get the first one)
pErrorRecords->GetErrorInfo(0, GetUserDefaultLCID(), &pErrorInfoRec);
pErrorInfoRec->GetDescription(&bstrDescriptionOfError);
pErrorInfoRec->GetSource(&bstrSourceOfError);
printf("ERROR!\nResult of 0x%0x (%ld) returned\n",(long)hr,(long)hr);
printf("Error Source: %S\n",bstrSourceOfError);
printf("Error Description: %S\n", bstrDescriptionOfError);
pErrorInfo->Release();
pErrorRecords->Release();
pSupportErrorInfo->Release();
pErrorInfoRec->Release();
SysFreeString(bstrSourceOfError);
SysFreeString(bstrDescriptionOfError);
} // if
/* ... some useful code here .... */
CoUninitialize();
return(0);
}
Example 7-10 shows the output of the Example 7-9 on page 229. We use a non
existent database for the connection, so an error is expected during initialization.
Example 7-10 Output of Errorinfo.cpp
C:\work>cl /EHsc /nologo errorinfo.cpp
errorinfo.cpp
C:\work>errorinfo
ERROR!
Result of 0x80004005 (-2147467259) returned
231
7884ch07.fm

Error Source: Ifxoledbc
Error Description: EIX000: (-329) Database not found or no system permission.
C:\work>
7.4 Visual Basic, ADO.NET, and SQL Server

ActiveX and COM objects are a core component of any Windows operating
system. For this reason, OLE DB providers are extensively used by many
Windows applications. In this section we describe how to use the IBM Informix
OLE DB Provider with some of the Microsoft technologies and applications such
as ADO.NET or Microsoft SQL Server.
7.4.1 OLE DB with Visual Basic

Visual Basic (VB) is a programming language developed by Microsoft focused on
the use of the COM objects. Visual Basic is design to be easy to learn and use
allowing the development of database applications with far less effort than C or
C++ languages. ADO and the Informix OLE DB provider is widely used with
Visual Basic.
In this section we demonstrate how to use the OLE DB provider to access an
Informix database using VBScript that is based on Visual Basic.
Select data
Example 7-11 shows how to query the state table using a recordset with a
VBScript file. We open the recordset using a static server cursor (adUseServer=2
and adOpenStatic=3).
Example 7-11 Select.vbs
' Create the ADO objects
set cx=createobject("ADODB.Connection")
set cr=createobject("ADODB.Recordset")
' Set the connection string
cx.provider="Ifxoledbc"
cx.connectionstring="Data Source=stores_demo@demo_on;"
' Open the Connection
cx.open
set cr.activeconnection=cx
cr.cursorlocation=2
' Open the Recordset
cr.open "SELECT * FROM state WHERE code='CA'", cx, 2, 3
232
7884ch07.fm
WScript.Echo cr.fields("sname")
Example 7-12 shows the output of the this script.

Example 7-12 Output of Select.vbs
c:\work>select.vbs
California
c:\work>
Comparing this example with Example 7-5 on page 224 written on C++ you can
clearly see how easy it is to use OLE DB with Visual Basic.
Example 7-13 demonstrates how to scroll through a recordset.
Example 7-13 Scroll.vbs
set cx=createobject("ADODB.Connection")
set cr=createobject("ADODB.Recordset")
set cm=createobject("ADODB.Command")
cx.provider = "Ifxoledbc.2"
cx.Open "Data Source=stores_demo@demo_on"
' Set the Command SQL text
cm.ActiveConnection = cx
cm.CommandText = "SELECT * FROM state"
' Open the cursor
cr.CursorLocation = 3 adUseClient
cr.Open cm
WScript.Echo "First row: " & cr.fields("sname")
' Scroll to the last element in the recordset
cr.MoveLast
WScript.Echo "Last row : " & cr.fields("sname")
We use the MoveLast() method to position the cursor in the last record, see the
output in Example 7-14.
Example 7-14 Output of Scroll.vbs
c:\work>scroll.vbs
First row: Alaska
Last row : Puerto Rico
c:\work>
233
7884ch07.fm
Add new data

In Example 7-15 we demonstrate how to add a new record to a table using the
AddNew() method. After adding a new record to the table, the code opens a
cursor and position at the end of the recordset to retrieve the last inserted row.
Example 7-15 Addnew.vbs
set cx = CreateObject("ADODB.Connection")
set cr = CreateObject("ADODB.Recordset")
cx.provider = "Ifxoledbc"
cx.connectionstring = "Data Source=stores_demo@demo_on;"
cx.Open
Set cr.activeconnection = cx
' Open the recordset
cr.Open "SELECT * FROM state", cx, 3, 2
' Add a new record to the recordset
cr.AddNew
cr.fields("code") = "UN"
cr.fields("sname") = "Unkown"
cr.fields("sales_tax") = "0.0"
' Update the database table
cr.Update
cr.Close
' Retrieve the new inserterd row
cr.open "SELECT * FROM state", cx, 3, 3
cr.MoveLast
WScript.Echo cr.fields("sname")
Example 7-16 shows the value of the sname column after the update operation
Example 7-16 Output of Addnew.vbs
c:\work>addnew.vbs
Unkown
c:\work>
7.4.2 ADO.NET and the OLEDB bridge

In addition to ADO, developers can also use the IBM Informix OLE DB Provider
with ADO.NET. Similar to ADO but within the .NET Framework, ADO.NET is a set
of .NET components used to retrieve, manipulate, and update data from a data
source.
234
7884ch07.fm
One of the .NET data providers included with ADO.NET is the Microsoft OLE DB
Provider for .NET. This .NET provider acts as bridge between the .NET
framework and standard OLE DB providers so developers can use the .NET
technology with a database server even if there is no specific .NET data provider
for that particular database.
IBM has specific .NET providers to access an Informix database: IBM Informix
.NET Provider and IBM Data Server Provider for .NET. However, the developers
can use the Informix OLE DB Provider through the Microsoft OLE DB .NET
Provider.
Example 7-17 demonstrates how to use the Informix OLE DB with a .NET
application. The connection string for the Informix OLE DB provider is the same,
refer 7.3.2, Connecting to database on page 215 for the complete syntax.
Example 7-17 Getoledb.cs
using System;
using System.Data;
using System.Data.OleDb;
class sample {
static void Main(string[] args) {
OleDbConnection con = new OleDbConnection();
con.ConnectionString = "Provider=Ifxoledbc;"
+" Data Source=stores_demo@demo_on; User ID=informix;"
+" Password=password";
DataSet ds = new DataSet();
string sql = "SELECT first 3 * FROM STATE";
OleDbDataAdapter da = new OleDbDataAdapter(sql,con);
da.Fill(ds,"state");
foreach(DataRow dr in ds.Tables[0].Rows) {
Console.WriteLine(dr["code"]);
Console.WriteLine(dr["sname"]);
Console.WriteLine(dr["sales_tax"]);
}
con.Close();
con.Dispose();
}
}
Example 7-18 on page 236 demonstrates how to compile and run this .NET
example.
235
7884ch07.fm

Example 7-18 Output of Getoledb.cs
c:\work>csc.exe /nologo getoledb.cs
c:\work>getoledb
AK
Alaska
0.00000
HI
Hawaii
0.04000
CA
California
0.08250
c:\work>
7.4.3 SQL Server

Using the IBM Informix OLE DB Provider you can select data from an Informix
database directly from Microsoft SQL Server.
SQL Server uses OLE DB to create Linked Servers, and retrieve data from any
OLE DB data source.
There are no specific steps to configure or to use the Informix OLE DB Provider
with SQL Server as a linked server. If the Informix OLE DB Provider is correctly
installed and configured, SQL Server can use it to connect to an IBM Informix
database server.
We describe how to setup a Linked Server using the Informix provider here.
Figure 7-3 on page 237 shows the New Linked Server dialog box where the
connection details for the OLE DB provider must be set.
236
7884ch07.fm
Figure 7-3 New Linked Server Dialog
The following fields must be set

Linked Server: Specify the name for the SQL Server to link.
Provider: Choose IBM Informix OLE DB Provider from the dropdown list.
Product name: Specify the name of the Informix provider which is ifxoledbc.
Data source: Specify the name of the datasource as database@server.
Provider string: Specify any additional Connection String parameters used by
the provider.
The user id for the Informix database server may differ from the one used with
SQL Server. If this is the case, a remote user mapping must be set. Figure 7-4 on
page 238 shows a user mapping between the sa and informix users.
237
7884ch07.fm
Figure 7-4 Linked Server Security options
You can also create linked servers using only SQL statements. In Example 7-19
we illustrate how to create a linked server to connect to the demo_on Informix
instance. We use the same connection details shown in Figure 7-3 on page 237.
Example 7-19 Linked Server SQL script
EXEC master.dbo.sp_addlinkedserver
@server = N'demo_on',
@srvproduct=N'ifxoledbc',
@provider=N'ifxoledbc',
@datasrc=N'stores7@demo_on',
@provstr=N''
EXEC sp_addlinkedsrvlogin 'demo_on',false,'sa','informix','password'
SELECT * FROM demo_on.stores7.informix.customer
238
7884ch07.fm
Figure 7-5 shows the output for the previous SQL script.
Figure 7-5 Linked Servers Select output
The last SQL statement in the script uses the linked server to retrieve the data
from the customer table
SELECT * FROM demo_on.stores7.informix.customer
For more information about Linked Servers, refer to the Microsoft documentation
at
http://msdn.microsoft.com/en-us/library/ms188279.aspx
7.5 Troubleshooting and tracing

In this section we discuss typical errors seen when using the IBM Informix .NET
Provider and how to enable the tracing facility to collect diagnostic information.
239
7884ch07.fm
7.5.1 Typical errors

Most of the errors encountered when using the Informix OLE DB provider are
due to an incorrect setup or a invalid connection string been passed to the
provider.
The Informix OLE DB provider uses the same connectivity libraries as other
components of the Client SDK package. It also uses the information stored in the
registry through the Setnet32 tool for the database server connection.
It is always a good practise to test this basic connectivity using the Ilogin tool
included in the Client SDK directory.
Provider not found

Here we explain the reason and the solution for the provider not found error.
Reason
This error appears if Windows fails to locate the OLE DB provider class specified
by the application.
Solution
During the installation of Client SDK, the OLE DB provider is automatically
registered within the Windows registry. You can manually register the Informix
OLE DB using the regsvr32 tool as follows:
regsvr32.exe %INFORMIXDIR%\bin\ifxoledbc.dll
If you are running a 64-bit version of Windows, make sure that you register the
provider using the correct version of the regsvr32 binary. Otherwise, it may
happen that the 32-bit provider is registered as a 64-bit one (or vice versa) and
Windows would fail to find and load the provider.
You can check if the provider library is correctly registered by examining the
Windows registry. Example 7-20 shows the registry entry 32-bit and 64-bit
Informix OLE DB providers. Remember that the name of the Informix provider is
ifxoledbc.
Example 7-20 Informix OLE DB registry keys
C:\work>c:\windows\syswow64\reg query
"HKEY_CLASSES_ROOT\CLSID\{A6D00422-FD6C-11D0-8043-00A0C90F1C59}\InprocServer32"
HKEY_CLASSES_ROOT\CLSID\{A6D00422-FD6C-11D0-8043-00A0C90F1C59}\InprocServer32
(Default)
REG_SZ
C:\Program Files
(x86)\IBM\Informix\Client-SDK\bin\ifxoledbc.dll
ThreadingModel
REG_SZ
Both
240
7884ch07.fm
C:\work>c:\windows\system32\reg query
"HKEY_CLASSES_ROOT\CLSID\{A6D00422-FD6C-11D0-8043-00A0C90F1C59}\InprocServer32"
HKEY_CLASSES_ROOT\CLSID\{A6D00422-FD6C-11D0-8043-00A0C90F1C59}\InprocServer32
(Default)
REG_SZ
C:\Program
Files\IBM\Informix\Client-SDK\bin\ifxoledbc.dll
ThreadingModel
REG_SZ
Both
C:\work>
Failed to load the Ifxoledbc.dll

Here we explain the reason and the solution for the Failed to load the
Ifxoledbc.dll error.
Reason
The application fails to load the Informix OLE DB provider or one of its libraries.
Solution
The value of INFORMIXDIR (as a environment variable or in the registry) should
be the directory where Informix Client SDK was installed. Make sure the
environment variable PATH contains the %INFORMIXDIR%\bin directory. For
example:
PATH=C:\Program Files\IBM\Informix\Client-SDK\bin;%PATH%
When an application loads the Informix OLE DB provider, the Client SDK libraries
that the provider needs for communication and registry access are loaded from
the directory used in the registration process.
On 64-bit version of Windows, make sure PATH and INFORMIXDIR variables are
correctly set for all users. Some applications such as SQL Server are executed
under different credentials as the logged user, and may have different settings for
this variables.
Failed to connect to the database

Here we explain the reason and the solution for the Failed to connect to the
database error.
Reason
Most common reason for these errors is due to the invalid parameters in the
connection string or the invalid connectivity information.
241
7884ch07.fm
Solution
If the application is getting an Informix OLE DB error, it means it has managed to
load the Informix OLE DB provider, but it is failing during connection.
For any connection error, always check if the basic connectivity works. Test if the
client machine has access to the database using Ilogin would help to narrow
down the problem.
If Ilogin fails, verify the connection parameters for the database (server name,
port number, host name, and so on) are correct. Use setnet32 to check if the
information stored in the registry is valid.
The user credentials (user and password) can also be set in the registry using
setnet32. Verify both of them are valid.
The majority of the connection string parameters must be included exactly as
they appear in Table 7-4 on page 216. Invalid parameters are ignored and may
cause an error connection. See Table 7-7 for same examples.
Table 7-7 Valid parameters
Valid
Invalid
Data Source
DataSource, DSN
User ID
UserID, UID
Password
PWD
If you are using GLS variables (Client_Locale or Db_locale), make sure they are
correctly set. If UNICODE is not used, The codeset part of Client_Locale should
be the same as your operating system. Db_Locale should be your database
locale.
Failure when handling database data

Here we explain the reason and the solution for the Failed when handling
database data error.
Reason
These types of errors are caused by using the wrong data type when accessing
Informix tables.
Solution
The Informix OLE DB Provider requires additional functions and tables in the
sysmaster database to correctly handle Informix data types. These objects are
created when running the coledbp.sql SQL script.
242
7884ch07.fm
This script is located in the %INFORMIXDIR%\etc directory and must be executed

against the sysmaster database as the informix user.
If you are upgrading from an old version of Informix Client SDK, remove these
objects by running the doledbp.sql script from the old client before using the new
coledbp.sql.
7.5.2 Tracing
Tracing is useful when diagnosing application problems such as SQL errors or
unexpected behaviors and can even be used for performance analysis.
Tracing
There are two types of tracing that a developer can set when using the Informix
OLE DB Provider:
Informix OLE DB Trace
SQLIDEBUG
Informix OLE DB Trace

Informix OLE DB Trace is specific to the Informix OLE DB Provider. You can set
the IFXOLEDBCTRACE environment variable to point to the location of the trace file,
for example:
IFXOLEDBCTRACE=oledbtrace.txt
The trace file generated contains the calls to all the OLE DB interfaces used by
the provider.
Example 7-21shows a typical Informix OLE DB trace file
Example 7-21 OLE DB trace
Wed Jun 30 13:59:01 2010
A44:
53C Enter Clsfact::QueryInterface
Object Address: 0x032C2CB8
A44:
53C Enter Clsfact::AddRef
A44:
53C Exit
Object Address: 0x032C2CB8 hr=2
qi: 032C2CB8 (pif 032C2CB8)
A44:
53C Enter Clsfact::Release
A44:
53C Exit
A44:
243
7884ch07.fm

A44:
53C Exit
A44:
53C Enter Clsfact::Release
A44:
53C Exit
A44:
53C Enter Clsfact::QueryInterface
A44:
A44:
53C Exit
...
In addition to the OLE DB interfaces, the trace facility collects the return value for
any method used during the OLE DB session. Example 7-22 illustrates this.
Example 7-22 Entry and Exit points OLE DB trace
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
A44:
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
53C
Enter
Enter
Enter
Exit
Enter
Exit
Enter
Enter
Exit
Enter
Enter
Exit
Enter
Enter
Exit
Enter
Enter
Exit
Enter
Enter
Datasrc::QueryInterface
Datasrc::AddRef
Object Address: 0x032CC370 hr=2
Datasrc::Release
Datasrc::XIDBProperties::SetProperties
Object Address: 0x032CC38C hr=0
Datasrc::XIDBProperties::SetProperties
Datasrc::XIDBInitialize::Initialize
Datasrc::XIDBProperties::GetProperties
Datasrc::XIDBCreateSession::CreateSession
SQLIDEBUG
SQLI is the communication protocol used by Informix Client SDK. SQLIDEBUG
is a trace facility of all the messages between an Informix client and a Informix
database server. This trace is common to all the Informix Client SDK
components (ESQL/C, ODBC, OLE DB, and so on).
244
7884ch07.fm
You can find a description and how to enable SQLIDEBUG in 3.3.6,

Troubleshooting on page 113.
When diagnosing a performance problem, sometimes, it is useful to collect a
SQLIDEBUG trace to have an idea of where the delay occurs. Using the summary
option when processing the sqlidebug file with sqliprint that produces a detail
description of all the message between client and server. See Example 7-23.
Example 7-23 Sqliprint Summary Output
>>>>>>>>>>>>>>>>>> SUMMARY INFORMATION <<<<<<<<<<<<<<<<<
>>>>>>>>>>TOTAL ELAPSED CLOCK TIME (sec): 0.047000
>>>>>>>>>>CLIENT ELAPSED CLOCK TIME (sec): 0.047000
>>>>>>>>>>SERVER+NETWORK CLOCK TIME (sec): 0.000000
FROM C->S
Msg
occured
Total
Avg
Min
Max
-----------------------------------------------------------------SQ_PREPARE
1
0.000000
0.000000
0.000000
0.000000
SQ_ID
4
SQ_CLOSE
1
SQ_RELEASE
2
SQ_EOT
7
SQ_NDESCRIBE
1
SQ_SFETCH
1
0.000000
0.000000
0.000000
0.000000
SQ_WANTDONE
1
SQ_EXIT
1
SQ_INFO
1
SQ_RET_TYPE
1
SQ_INTERNALVER
1
SQ_PROTOCOLS
1
-----------------------------------------------------------------FROM S->C
Msg
occured
Total
Avg
Min
Max
-----------------------------------------------------------------UNKNOWN
1
SQ_DESCRIBE
1
0.000000
0.000000
0.000000
0.000000
SQ_EOT
6
SQ_TUPLE
1
SQ_TUPID
1
SQ_PROTOCOLS
1
-----------------------------------------------------------------=======================================================================
COMMAND TEXT INFORMATION (first 99 are listed)
a(
b) CMD[##]= ' first 60 bytes of cmd text '
where a = 'P'repare, or 'N'ot prepared, and b = length of command string
----------------------------------------------------------------------P(
28) CMD[ 0]='SELECT rowid, * FROM state;'
245
7884ch07.fm

----------------------------------------------------------------------=======================================================================
Fetch Array feature not used
OPTOFC feture not used
Number of open/reoptimzation encountered = 0
Number of C->S message send = 22
Number of S->C message send = 21
Number
Number
Number
Number
Number
of prepare statements encountered = 1

of execute statements encountered = 0
of singleton select encountered = 0
of
open cursor encountered = 0
of
close cursor encountered = 1
Number
Number
Number
Number
of non-blob put
of non-blob fetch
of
blob put
of
blob fetch
=
=
=
=
0,
1,
0,
0,
averge
averge
averge
averge
size
size
size
size
of
of
of
of
each
put is 0.000000
each fetch is 25.000000
each
put is 0.000000
each fetch is 0.000000
>>>>>>>>>>>>>>> END SUMMARY INFORMATION <<<<<<<<<<<<<<<
246
7884ch08.fm
Chapter 8.
Working with the .NET Data

Providers
This chapter describes the different .NET Providers available for working with an
We cover the following topics:
.NET data Provider setup and configuration

Basic connectivity
Simple operations with the database
Using the .NET extensions for special data types
Problem diagnosis
Tracing
247
7884ch08.fm
8.1 Informix and .NET data providers

A .NET data provider is a .NET assembly that lets the .NET applications access
and manipulate data in a database. It does this by implementing several
interfaces that follow the Microsoft ADO.NET architecture. A .NET data provider
can be used by any application that can be executed by the Microsoft .NET
Framework.
IBM provides two .NET data providers to work with an IBM Informix database:
IBM Infomix .NET Provider
IBM Data Server Provider for .NET
Both providers, Informix .NET and Data Server, rely on internal calls to the
ODBC and CLI drivers, so they are not 100% managed-code providers.

This chapter describes the .NET providers available to connect to an IBM
Informix database. We explain how to configure and test the connectivity with
both IBM Infomix .NET Provider and IBM Data Server Provider for .NET.
8.2.1 IBM Informix .Net Provider

The Informix .NET Provider is part of the IBM Informix Client SDK. IBM Informix
.NET Provider has been certified to work on both 32-bit and 64-bit editions of
Windows XP, Windows Vista, Windows Server 2003, Windows Server 2008 and
Windows 7.
The provider requires that Microsoft .NET Framework SDK 1.1 or higher are
already installed on the computer. The provider is installed by default when
performing a complete install of the Informix Client SDK package.
Informix .NET Data Provider supports the following Informix database servers:
IBM Informix 10.00, 11.10, and 11.50
IBM Informix Extended Parallel Server 8.50 and later
The assembly name of the Informix .NET provider is IBM.Data.Informix.
The assembly files are located in the Client SDK directory. The provider for a
.NET 1.1 application is in %INFORMIXDIR%\netf11. For a .NET 2.0 and above, use
the provider located in %INFORMIXDIR\netf20.
248
7884ch08.fm
The installation process registers two strong-named assemblies in the global

assembly cache (GAC).
Table 8-1 shows the Informix .NET Provide assembly versions.
Table 8-1 Informix.NET Provider assembly versions
.NET framework
Assembly version
1.1
2.81.0.0
2.0 and above
3.0.0.2
The assemblies are registered in the common section of the GAC (GAC_MSIL).
Because of this, only one version of the provider (32-bit or 64-bit) can be stored
in the GAC at one time.
Note: There is no 64-bit version of the .NET 1.1 Framework, so the .NET 1.1
provider is not installed on a Windows 64-bit platform.
8.2.2 IBM Data Server Provider for .NET

The IBM Data Server Provider for .NET is included in the IBM Data Server
Package. The Provides support multiple IBM data servers including IBM Informix
(11.x) and DB2.
Same as the Informix .NET provider, the Data Server Provider is not 100%
managed code, this means that it requires the DB2 CLI component for the
communication with the database.
There are two Data Server providers for an Informix database:
IBM.Data.DB2: This is the preferred .NET provider when developing new
applications.
IBM.Data.Informix: This .NET Provider assembly is normally used to migrate
an application developed using the IBM Informix .NET Provider.
The assembly files are located in the netf20 subdirectory under the Data Server
Client directory, by default is C:\Program Files\IBM\IBM DATA SERVER
DRIVER\netf20.
Note: The 64-bit version of the Data Server Client Package also includes the
32-bit drivers.
Chapter 8. Working with the .NET Data Providers
249
7884ch08.fm
Both providers had the same assembly version, 9.0.0.2. Table 8-2 shows the
Data Server Client .NET Provider assembly version.
Table 8-2 Data Server Client Assembly versions
.NET Provider
Assembly version
IBM.Data.DB2
9.0.0.2
IBM.Data.Informix
9.0.0.2

The Informix .NET provider uses communication libraries Internally as the other
drivers included in the Informix Client SDK, this means that it uses the same
connectivity information normally stored in the registry through the use of
Setnet32.
When there is no specific tool to verify the connectivity with the Informix .NET
provider, you can use a simple .NET application to test it.
The Microsoft .NET Framework SDK is required to use the .NET Provider that
contains a .NET language compiler.
Example 8-1 shows a simple C# code to verify the connection with the database
server. This example takes the connection string as first argument and uses it to
connect to the database.
Example 8-1 Connect.cs
using System;
using System.Data;
using IBM.Data.Informix;
class Connect
{
static void Main(string[] args)
{
IfxConnection conn;
if (args.Length>0) {
try {
conn = new IfxConnection(args[0]);
conn.Open();
Console.WriteLine("Connected");
Console.WriteLine(String.Format("Server Type: {0}, Server
Version: {1}", conn.ServerType, conn.ServerVersion));
250
7884ch08.fm
Console.WriteLine(String.Format("Database: {0}",
conn.Database));
}
catch (Exception e) {
Console.WriteLine(e.Message);
}
}
else
Console.WriteLine("Need a connection string as argument\n e.g:
\"Host=kodiak;Service=9088;Server=demo_on;Database=stores_demo;User
ID=informix;password=;\"\n");
}
}
Example 8-2 shows how to compile and run Example 8-1 on page 250.
Example 8-2 Output of Connect.cs
C:\work>csc.exe /R:%INFORMIXDIR%\bin\netf20\IBM.Data.informix.dll /nologo
connect.cs
C:\work>connect "Server=demo_on;Database=stores_demo"
Connected
Server Type: Informix, Server Version: 11.50.0000 FC6
Database: stores_demo
C:\work>
If the connection information for the Informix server is stored in the registry, the
only parameter required in the connection string is the name of the database
server, everything else such as Host or Service is taken from the registry.
At compile time, you must pass the assembly as a reference to the csc.exe
compiler using the /R parameter. For example:
/R:%INFORMIXDIR%\bin\netf20\IBM.Data.Informix.dll
If you want to use the Data Server provider, the reference path should be
/R:C:\Program Files (x86)\IBM\IBM DATA SERVER DRIVER\bin\netf20\IBM.Data.Informix.dll
Example 8-3 shows how to compile and execute the sample with the Data Server
.NET provider.
Example 8-3 Data Server Connect.cs sample
C:\work>csc.exe /R:"C:\Program Files (x86)\IBM\IBM DATA SERVER
DRIVER\bin\netf20\IBM.Data.Informix.dll" /nologo connect.cs
251
7884ch08.fm

C:\work>connect "Server=kodiak:9089;Database=stores_demo;User
ID=informix;password=password;"
Connected
Server Type: IDS/NT64, Server Version: 11.50.0000
Database: STORES_DEMO
C:\work>
The Data Server provider includes a connection tool called testconn20.exe which
can be used to test the connection with the database. Example 8-4 shows a
typical output of the testconn tool.
Example 8-4 Output of testconn20.exe
C:\work>testconn20.exe "server=kodiak:9089; database=stores_demo; uid=informix;
pwd=password;"
Step 1: Printing version info
.NET Framework version: 2.0.50727.3603
DB2 .NET provider version: 9.0.0.2
DB2 .NET file version: 9.7.2.2
Capability bits: ALLDEFINED
Build: 20100514
Factory for invariant name IBM.Data.DB2 verified
VSAI assembly version: 9.1.0.0
VSAI file version: 9.7.1.53
Elapsed: 1.015625
Step 2: Validating db2dsdriver.cfg against db2dsdriver.xsd schema file
C:\PROGRA~1\IBM\IBMDAT~2\cfg\db2dsdriver.cfg against C:\Program
Files\IBM\IBM DATA SERVER DRIVER\cfg\db2dsdriver.xsd
Elapsed: 0.015625
Step 3: Connecting using "server=kodiak:9089; database=stores_demo;
uid=informix; pwd=password;"
Server type and version: IDS 11.50.0000
Elapsed: 0.46875
Step 4: Selecting rows from informix.systables to validate existance of
packages
SELECT * FROM informix.systables
Elapsed: 0.171875
Step 5: Calling GetSchema for tables to validate existance of schema functions
Elapsed: 0.21875
Test passed.
252
7884ch08.fm
C:\work>
8.3 Developing a .NET application

This section describes the connection string attributes for two Informix .NET
providers and provides samples of basic database operations.

In this section we describe the connection string attributes for the IBM Informix
.NET Provider and the Data Server Provider for .NET
Informix .NET connection attributes

Table 8-3 lists the connection attributes that the Informix .NET provider supports.
Table 8-3 Informix .NET connection string attributes.
Attribute
Description
Default
Client Locale,
Client_Locale
Locale used on the application
en_us.1252
Connection Lifetime
Time in seconds that a connection is

allowed in the pool
Database, DB
Database to connect to
Database Locale,
DB_LOCALE
Locale of the database
en_US.819
DELIMIDENT
If set, any string within double quotes (")

is treated as an identifier, and any string
within single quotes () is treated as a
string literal
Enlist
Enables or disables automatic enlistment

in a distributed transaction
true
Exclusive, XCL
if set the database is open in exclusive

mode
No
Host
Name or IP address of the machine on

which the Informix server is running
localhost
Max Pool Size
Maximum number of connections

allowed in the pool
100
253
7884ch08.fm
Attribute
Description
Default
Min Pool Size
Minimum number of connections allowed

in the pool
Optimize OpenFetchClose,
OPTOFC
Reduces the number of client-server

messages when using cursors
Packet Size, Fetch Buffer

Size, FBS
Size in bytes of the buffers used to send

data to or from the server
32767
Password, PWD
Password associated with the User ID
If set the security-sensitive information,

such as the password, is returned as part
of the connection string
false
Pooling
Enable Connection Pooling
true
Protocol, PRO
Communication protocol
Server
Name or alias of the Informix server
Service
Service name or port number used by the

server for incoming connection
Skip Parsing
if set, there is no internal SQL parsing it

increases performance but the SQL must
be valid
UserDefinedTypeFormat
Changes the mapping of UDTs to either

DbType.String or DbType.Binary
Leave Trailing Spaces
If set disable the automatic trailing of

spaces done in a varchar column
User ID, UID
Login account.
false
false
Data Server Client for .NET connection attributes

Table 8-4 lists the most common connection attributes that the IBM Data Server
.NET Provider uses.
Table 8-4 IBM Data Server connection string attributes
254
Attribute
Description
Default Value
ConnectTimeout,
Connect Timeout
The time (in seconds) to wait for the

database connection to be established.
Connection Lifetime
Time in seconds that a connection is

allowed in the pool.
60
7884ch08.fm
Attribute
Description
Default Value
Database, DB
Database to connect to.
Enlist
Enables or disables automatic enlistment

in a distributed transaction.
true
Max Pool Size
Maximum number of connections

allowed in the pool.
Min Pool Size
Minimum number of connections allowed

in the pool.
Password, PWD
Password associated with the User ID
If set the security-sensitive information,

such as the password, is returned as part
of the connection string.
false
Pooling
Enable Connection Pooling within the

.NET Provider
true
Query Timeout
Time to wait for a SQL query response
Server
Name or alias of the Informix server

followed by service name or the port
number
User ID, UID
Login account
The configuration parameters for Data Server .NET provider and most of the
components in the Data Server Client are stored in the db2sdriver.cfg file. Refer
to the following website for a complete list of configuration parameters:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/topic/com.ibm.swg.im.dbcli
ent.config.doc/doc/c0054698.html
If the connection string does not contain all the required information, the provider
takes the missing arguments from the db2sdriver.cfg configuration file.
The majority of the connection string attributes of Data Server Provider for .NET
have the similar meaning and format with the Informix .NET Provider except the
the Server attribute.
For the Informix .NET Provider the keyword Server contains the name (or alias)
of the Informix server which normally is the same as the value of the
INFORMIXSERVER variable.
For the Data Server .NET Provider, Server contains the host name where the
Informix server is running, and the service name or port number used to listen for
DRDA client connections.
255
7884ch08.fm
Example 8-5 shows the Server attribute format. In this example, we connect to
an Informix database server running in a machine named kodiak.ibm.com. The
Informix server DRDA alias is using the port number 9089.
Example 8-5 Server attribute for the Data Server provider
C:\work>connect
"Server=kodiak.ibm.com:9089;Database=stores_demo;UID=informix;password=password
;"
Connected
Server Type: IDS/NT64, Server Version: 11.50.0000
Database: STORES_DEMO
C:\work>
8.3.2 Data type mapping

The data types used by the .NET Framework differ from the data type used by
the IBM Informix database. In this section we describe the optimal data types to
use when accessing data in an Informix database. The optimal data type
depends on .NET method used to access the data.
You can use types other than those shown, for example, you can use the
IfxDataReader.GetString method to obtain any Informix data type. The types
mentioned here are the most efficient and least likely to change in value in data
conversion.
Informix .NET type mapping

Table 8-5 shows the mapping for the specific IBM Informix data types.
Table 8-5 Type mapping for Informix specific data
256
Informix data type
For DataReader
For DataSet
BIGINT
Int64
Int64
BIGSERIAL
Int64
Int64
BLOB
IfxBlob
Byte[]
BOOLEAN
Boolean
Boolean
BYTE
Byte[]
Byte[]
CLOB
IfxClob
Byte[]
COMPLEX (ROW, LIST)
String
String
7884ch08.fm
Informix data type
For DataReader
For DataSet
DECIMAL(p<=28) fixed scale
IfxDecimal
Decimal
DECIMAL(p<=28) floating point
IfxDecimal
Double
DECIMAL (p>28)
IfxDecimal
String
IDSSECURITYLABEL
Int64[]
Int64[]
INT8
Int64
Int64
INTERVAL, year-month
IfxMonthSpan
String
INTERVAL, day-fraction
IfxTimeSpan
TimeSpan
LVARCHAR
String
String
MONEY
IfxDecimal
Decimal with same precision
NCHAR
String
String
SERIAL
nt32
Int32
SERIAL8
Int64
Int64
TEXT
String
String
Data Server Client type mapping

Table 8-6 shows the type mapping with the Data Server provider when using
specific Informix data types.
Table 8-6 Type mapping for Informix specific data
Informix Data Type
Optimal for DataReader
Optimal for DataSet
LVARCHAR
IfxString
String
BLOB, BYTE
IfxBlob
Byte[]
CLOB, TEXT
IfxClob
String
BOOLEAN, SMALLINT
IfxInt16
Int16
BIGINT, BIGSERIAL, INT8,

SERIAL8
IfxInt64
Int64
DECIMAL(p<=28) fixed scale
IfxDecimal
Decimal
DECIMAL(p<=28) floating point
IfxDouble
Double
DECIMAL (p>28)
IfxDouble
Double
257
7884ch08.fm
Informix Data Type
Optimal for DataReader
Optimal for DataSet
MONEY
IfxDecimal
Decimal

This section describes the main classes in the .NET Provider and demonstrate
how to use them with basic examples.
IfxConnection
This object represents a unique connection with the database server. You can
specify the connectivity details about the database when creating the connection
object or by setting the ConnectionString property later on.
Because the .NET Provider uses native resources (not managed by the .NET
CLR), you should always close or dispose any open IfxConnection object when
it is not longer needed.
Table 8-7 lists the public method of the IfxConnection class.
Table 8-7 IfxConnection public methods
Method
Description
void Open()
Opens a database connection
void Close()
Closes the connection to the database
void ChangeDatabase(String)
Changes the current database
void EnlistTransaction()
Enlist the connection in a DTC transaction
IfxTransaction
BeginTransaction()
BeginTransaction(IsolationLevel)
Begins a database transaction
IfxCommand CreateCommand()
Creates an instance of a IfxCommand object

associated with this IfxConnection
IfxBlob GetIfxBlob()
Creates an instance of a IfxBlob object associated

with this connection
IfxClob GetIfxClob()
Creates an instance of a IfxClob object associated

with this connection
We described the connection string attributes in 8.3.1, Connecting to database

on page 253.
258
7884ch08.fm
Example 8-6 demonstrates how to connect to the database

Example 8-6 connect_sample.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
try {
conn = new IfxConnection("Server=demo_on;database=stores_demo");
conn.Open();
Console.WriteLine("Connected to "+conn.Database);
conn.Close();
}
}
}
}
Example 8-7 shows the output of Example 8-6.

Example 8-7 Output of Connect_sample.cs
C:\work>csc.exe /R:%INFORMIXDIR%\bin\netf20\IBM.Data.Informix.dll /nologo
connect_sample.cs
C:\work>connect_sample
Connected to stores_demo
C:\work>
IfxCommand
IfxCommand represents a SQL statement or stored procedure to execute against
a database server.
Table 8-8 on page 260 shows methods that the IfxCommand class provides for
executing a SQL statement.
259
7884ch08.fm

Table 8-8 IfxCommand public methods
Method
Description
Int32 ExecuteNonQuery()
Executes an SQL statement against the

IfxConnection object.
IfxDataReader ExecuteReader()
ExecuteReader(DataCommandBehavior
behavior)
Executes the command in the

CommandText property against the
IfxConnection object and builds an
IfxDataReader object.
Object ExecuteScalar()
Executes the query, and returns the first

column of the first row.
void Cancel()
Attempts to cancel the execution of a

command.
IfxParameter CreateParameter()
Creates a new instance of an

IfxParameter object.
void Prepare()
Creates a prepared (or compiled) version

of the command against the database.
Example 8-8 demonstrates how to run a simple DELETE SQL statement.

Example 8-8 Command_sample.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
int drows=0;
try {
conn.Open();
IfxCommand cmmd = conn.CreateCommand();
cmmd.CommandText = "DELETE FROM customer WHERE customer_num = 103";
drows = cmmd.ExecuteNonQuery();
Console.WriteLine("Deleted rows: "+drows.ToString());
conn.Close();
}
}
260
7884ch08.fm

}
}

Example 8-9 Output of command_sample.cs
command_sample.cs
C:\work>command_sample
Deleted rows: 1
C:\work>
IfxDataAdapter
The IfxDataAdapter class represents a set of data commands that are used to
communicate between a DataSet and the database. A DataSet is a copy of the
database data stored in memory.
Table 8-9 lists the methods that IfxDataAdapter provides for accessing the data.
Table 8-9 IfxDataAdapter public methods
Method
Description
Int32 Fill(DataSet)
Adds or refreshes rows in the DataSet.
DataTable FillSchema(DataSet,
SchemaType)
Adds a DataTable to the DataSet and

configures the schema to match that in the
database based on the specified SchemaType.
IDataParameter GetFillParameters()
Returns the parameters set by the user when

executing a SELECT statement
Int32 Update(DataSet)
Execute the SQL statement associated with

the InsertCommand, UpdateCommand or
DeleteCommand for each inserted, updated,
or deleted row in the specified DataSet
Example 8-10 demonstrates how to retrieve data from the database using the
IfxDataAdapter class. In this example, we retrieve the data rows directly from the
DataSet.
Example 8-10 Dataadapter_sample.cs
using System;
261
7884ch08.fm

using System.Data;
class sample {
IfxConnection conn;
conn.Open();
cmmd.CommandText = "SELECT * FROM state WHERE code='CA'";
IfxDataAdapter dadap = new IfxDataAdapter();
DataSet dset = new DataSet();
dadap.SelectCommand = cmmd;
dadap.Fill(dset);
foreach(DataRow dr in dset.Tables[0].Rows) {
Console.WriteLine(String.Format("\tCode\tState\n"));
Console.WriteLine(String.Format("\t{0}\t{1}", dr["code"], dr["sname"]));
}
conn.Close();
}
}
Example 8-11 shows the data returned by the DataSet object.

Example 8-11 Output of Dataadapter_sample.cs
dataadapter_sample.cs
C:\work>dataadapter_sample.exe
Code
State
CA
California
C:\work>
IfxDataReader
The IfxDataReader class provides fast read-only, forward-only access to a set of
rows, similar to SQL cursors data. To create an IfxDataReader object the
application calls the ExecuteReader() method of the IfxCommand object.
262
7884ch08.fm
IfxDataReader inheres from the System.Data.Common.DbDataReader. It provides

all common DbDataReader methods as well as additional methods to handle
Informix specific data types.
Table 8-10 contains the additional public methods of the IfxDataReader class.
Table 8-10 Additional methods of IfxDataReader
Method
Description
Boolean GetBoolean()
Gets the value of the specified column as a Boolean.
TimeSpan GetTimeSpan()
Gets the time span value of the specified field.
IfxBlob GetIfxBlob()
Gets the IfxBlob value of the specific field.
IfxClob GetIfxClob()
Gets the IfxClob value of the specific field.
IfxDateTime GetIfxDateTime()
Gets the IfxDateTime value of the specific field.
IfxDecimal GetIfxDecimal
Gets the GetIfxDecimal value of the specific field.
IfxMonthSpan
GetIfxMonthSpan
Gets the IfxMonthSpan value of the specific field.
IfxTimeSpan GetIfxTimeSpan
Gets the IfxTimeSpan value of the specific field.
Refer to the IBM Informix .NET Provider Guide for a complete list of all the
IfxDataReader public methods at
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.netpr.doc/n
etpr031023156.htm#netpr031023156
Note: Even if the IfxDataReader.Read() method returns only one row each
time it is called, the provider may retrieve more than one row from the
database. The .NET provider does this to increase performance. The
application can set the number of rows the provider fetches from the database
using the IfxCommand property RowFetchCount.
Example 8-12 demonstrates how to use a IfxDataReader to access a database
table.
Example 8-12 Datareader_sample.cs
using System;
using System.Data;
class sample {
263
7884ch08.fm

IfxConnection conn;
conn.Open();
cmmd.CommandText = "SELECT * FROM state WHERE code='CA'";
IfxDataReader drdr;
drdr = cmmd.ExecuteReader();
while (drdr.Read()) {
Console.WriteLine(String.Format("\tCode\tState\n"));
Console.Write(String.Format("\t{0}", drdr.GetString(0)));
Console.WriteLine(String.Format("\t{0}", drdr.GetString(1)));
}
drdr.Close();
conn.Close();
}
}

Example 8-13 Output of Datareader_sample.cs
datareader_sample.cs
C:\work>datareader_sample.exe
Code
State
CA
California
C:\work>
IfxError
The IfxError class collects information relevant to a warning or error returned by
the database. You can use IfxError and IfxErrorCollection to retrieve additional
information when an error occurs.
An application can access this information using the properties of the IfxError
object listed in Table 8-11 on page 264.
Table 8-11 IfxError properties
264
Property
Type
Description
Message
String
Text message
7884ch08.fm
Property
Type
Description
NativeError
Int32
IBM Informix error code
SQLState
String
ANSI SQL error code
Example 8-14 desmonstrates a typical use of the IfxError class. The

IfxException class contains one or more IfxError objects.
Example 8-14 Error_sample.cs
using System;
using System.Data;
class Connect {
IfxConnection conn;
int drows=0;
try {
conn = new IfxConnection("Server=demo_on;database=wrong_db");
conn.Open();
Console.WriteLine("Connected");
conn.Close();
}
catch (IfxException e) {
Console.WriteLine("----------------------------");
if (e.Errors.Count > 0) {
IfxError ifxErr = e.Errors[0];
Console.WriteLine("Message :" + ifxErr.Message);
Console.WriteLine("Native error :" + ifxErr.NativeError);
Console.WriteLine("SQL state
:" + ifxErr.SQLState );
}
Console.WriteLine(e.StackTrace);
Console.WriteLine("----------------------------");
}
}
}
IfxParameter
The IfxParameter class is used to pass parameters to the IfxCommand method.
IfxParameters are stored as a collection in a IfxParameterCollection object.
265
7884ch08.fm
Table 8-12 contains the public attributes of a IfxParameter object.

Table 8-12 IfxParameter attributes
Property
Type
Description
DbType
DbType
Defines the DbType of the parameter.
Direction
ParameterDirection
Defines the direction of the parameter

(Input, Output or both).
IfxType
IfxType
Defines the IfxType of the parameter.
IsNullable
Boolean
Specifies whether the parameter is

allowed to be null.
ParameterName
String
Name of the parameter. Unique

reference in the parameter collection.
SourceColumn
String
Specifies the column that is mapped to

the DataSet.
SourceVersion
DataRowVersion
Defines the DataRowVersion to use

when you load Value.
Value
Object
value of the parameter.
Example 8-15 shows how to use IfxParameter to perform an UPDATE operation.

Example 8-15 Parameter_sample.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
int urows=0;
try {
conn.Open();
cmmd.CommandText = "UPDATE state SET sname = ? where code = ?";
IfxParameter pcode = new IfxParameter(code, DbType.String);
IfxParameter psname = new IfxParameter(sname, DbType.String);
pcode.Value="CA";
psname.Value="CALIFORNIA";
266
7884ch08.fm
cmmd.Parameters.Add(psname);
cmmd.Parameters.Add(pcode);
urows = cmmd.ExecuteNonQuery();
Console.WriteLine("Updated rows: "+urows.ToString());
conn.Close();
}
}
}
}
You can see the output in Example 8-16

Example 8-16 Output of Parameter_sample.cs
parameter_sample.cs
C:\work>parameter_sample.exe
Updated rows: 1
C:\work>
IfxTransaction
The IfxTransaction class represents an SQL transaction to be made at a
database. The application creates an transaction by calling the
BeginTransaction method of the IfxConnection object.
IfxConnection.BeginTransaction() returns a IfxTransaction object. When the
application decides how to resolve the transaction, it uses the Commit and
Rollback methods of the IfxTransaction object.
Any IfxCommand involved in a transaction must have the transaction property set
to the IfxTransaction object.
Example 8-17 on page 267 demonstrates how to create a transaction.
Example 8-17 Creating a transaction using lfxTransaction
using System;
using System.Data;
class sample {
267
7884ch08.fm

IfxConnection conn;
int drows=0;
try {
conn.Open();
IfxTransaction trans = conn.BeginTransaction();
cmmd.Transaction = trans;
cmmd.CommandText = "DROP TABLE state";
drows = cmmd.ExecuteNonQuery();
trans.Rollback();
conn.Close();
}
}
}
}
You can specify the following Isolation levels with the BeginTransaction()
method:
ReadUncommitted
ReadCommitted
RepeatableRead
Serializable
The default Isolation level is ReadCommited.

Example 8-18 demonstrates how to run a distributed transaction.
Example 8-18 Transact_cts.cs
using
using
using
using
System;
System.Data;
IBM.Data.Informix;
System.Transactions;
class sample {
IfxConnection conn;
268
7884ch08.fm
try {
// Transaction options
TransactionOptions tsopt = new TransactionOptions();
tsopt.IsolationLevel =
System.Transactions.IsolationLevel.RepeatableRead;
tsopt.Timeout = new TimeSpan(0, 60, 0);
using (TransactionScope tscope = new TransactionScope(
TransactionScopeOption.RequiresNew, tsopt,
EnterpriseServicesInteropOption.Full))
{
conn.Open();
cmmd.CommandText = "DROP TABLE state";
cmmd.ExecuteNonQuery();
// Rollback the distributed transaction not calling Complete()
//
tscope.Complete();
}
}
}
}
}
8.3.4 Handling Informix specific data types

In this section we cover how to work with IBM Informix specific data types such
as Smart Large Objects or Decimal using the .NET Provider.
IfxBlob and IfxClob

CLOB and BLOB are Informix data types used to stored large amount of
character or binary data.The application can handle these Smart Large Objects
using the IfxClob and IfxBlob classes.
The method implemented on these classes allows random access of its contents.
The application can read or write to certain positions in the large object without
reading or writing through all of the data up to that position.
BLOBs and CLOBs are both smart large object type. Both types share many of
the same methods. BLOBs differ from CLOBs in that the data in a CLOB is
269
7884ch08.fm
treated as text characters but the data in a BLOB is not. The data in a BLOB is
considered to be binary data and no translation or conversion of any kind is
performed on it when it is moved to or from the database server.
Properties and methods

Table 8-13 lists the IfxBlob and IfxClob public properties.
Table 8-13 Properties of lfxBlob and lfxClob
Property
Type
Description
EstimatedSize
Int64
Estimated final size of the large object
ExtentSize
Int32
Next extent size for this large object (disk space)
Flags
Int32
Flags for this large object
IsNull
Boolean
Checks if the large object is NULL
IsOpen
Boolean
Checks if the large object is open
LastAccessTime
Int32
Last time the large object was accessed.
LastChangeTime
Int32
Last time the status was changed
LastModificationTime
Int32
Last time the large object was modified
MaxBytes
Int64
Maximum size for the large object.
Position
Int64
Current position on the large object.
ReferenceCount
Int32
Number of records in the database that currently

contain a reference to this large object.
SBSpace
String
Sbspace in which the large object is stored.
Size
Int64
Size of the large object in byte.
Table 8-14 lists the available public methods from IfxClob and IfxBlob classes.
Table 8-14 Public methods of lfxClob and lfxBlob
270
Method
Description
void Open(mode)
Opens a large object in a specific mode
IfxSmartLOBLocator GetLocator()
Returns the IfxSmartLOBLocator

associated with this instance.
Int64 Read(buff), Read(buff, buffOffset,

numBytesToRead, sLOBOffset, whence )
Reads the complete data or a portion of a

large object as a Byte[] or Char[]
7884ch08.fm
Method
Description
Int64 Write(buff), Write( buff, buffOffset,

numBytesToWrite, sLOBOffset, whence)
Writes the complete buffer or a portion into

a large object.
void Truncate(offset)
Truncates everything in the large object

past the position offset.
Int64 Seek(offset, whence)
Changes the current position within the

large object.
void FromFile(filename, appendTosLOB,

fileLocation)
Reads a operating system file and write

the complete content into the large object.
String ToFile(filename, mode, fileLocation)
Writes the contents of the large object to

an operating system file.
void Lock(sLOBOffset, whence, range,

lockMode)
Locks the complete large object or only a

portion.
void Unlock(sLOBOffset, whence, range)
Unlock a large object.
void Release()
Frees database server resources
void Close()
Closes the large object
Creating a Smart Large Object

You can use the GetIfxClob() and GetIfxBlob() method of the
IfxConnection() object to create a large object. Perform the following steps to
create a large object:
1. Create an instance of a the large object, IfxClob or IfxBlob.
2. Open the large object.
3. Write data into the large object.
4. Execute the SQL Statement using the IfxClob or IfxBlob as a parameter for
the statement.
5. Close the large object.
Example 8-19 on page 271 demonstrate how to update a CLOB column. The
code reads a text file sample_blob.cs and use the contents for the CLOB data
Example 8-19 Lo_create.cs
using
using
using
using
System;
System.IO;
System.Data;
IBM.Data.Informix;
class sample {
271
7884ch08.fm

IfxConnection conn;
int urows=0;
IfxClob vclob;
char[] vclobBuff;
try {
conn = new IfxConnection("Server=demo_on;Database=stores_demo");
conn.Open();
// Create and Open the Clob
vclob = conn.GetIfxClob();
vclob.Open(IfxSmartLOBOpenMode.ReadWrite);
// Read a text file and insert into the clob
vclobBuff = File.ReadAllText("lo_create.cs").ToCharArray();
vclob.Write(vclobBuff);
// Create the UPDATE Ifxcommand
cmmd.CommandText = "UPDATE catalog set advert_descr = ? WHERE" +
" catalog_num = ?;";
// Bind the IfxClob value and execute the UPDATE statement
IfxParameter padvt_desc = new IfxParameter(null, vclob);
IfxParameter pcatalog_n = new IfxParameter(null, "10072");
cmmd.Parameters.Add(padvt_desc);
cmmd.Parameters.Add(pcatalog_n);
// Close the IfxClob object
vclob.Close();
conn.Close();
}
}
}
}
If the application requires more control over the smart large object, it can use the
specific properties of the ifxClob and IfxBlob classes to set up values such as
in which smart BLOB space the large object should be created, or the maximum
size.
Example 8-20 Illustrates how to use the properties of the IfxBlob classes.
272
7884ch08.fm
Example 8-20 Large Object Extended properties

...
// Create and Open the Clob
vclob = conn.GetIfxClob();
vclob.EstimatedSize = 5000;
vclob.ExtentSize = 1000;
vclob.Flags = (int) IfxSmartLOBCreateTimeFlags.DontKeepAccessTime |
(int) IfxSmartLOBCreateTimeFlags.NoLog;
vclob.MaxBytes = 10000;
vclob.SBSpace = "sbspace";
...
Selecting a Smart Large Object

Reading a large object using the Large Object extensions require the following
steps:
1.
2.
3.
4.
Execute the SQL Statement and retrieve the large object column.
Open the large object.
Read from the large object into an application buffer.
Close the large object.
Example 8-21 shows how to read a large object column from a database table.
Example 8-21 Lo_select.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
IfxClob vclob;
int maxSize = 2000;
char[] vclobBuff = new char[maxSize];
try {
conn.Open();
// Select the large object from the database
cmmd.CommandText = "SELECT * FROM catalog WHERE catalog_num = 10072";
IfxDataReader drdr;
273
7884ch08.fm

while (drdr.Read()) {
// Get the large object. The clob is the 6th column
vclob = drdr.GetIfxClob(5);
// Open the large object
vclob.Open(IfxSmartLOBOpenMode.ReadOnly);
vclob.Read(vclobBuff, 0, maxSize, 0, IfxSmartLOBWhence.Current);
Console.WriteLine(vclobBuff);
// Close the large object
vclob.Close();
}
conn.Close();
}
}
}
}
Random access
One of the benefits of the Smart Large Objects against a normal large object is
the option of reading partial data.
The Read(buff, buffOffset, BytesToRead, loOffset, whence) method allows
you to position the data pointer anywhere in the large object data and read from
there.
The code in Example 8-22 shows how to read the last 10 bytes of a large object.
Example 8-22 Partial large object Read
...
// Read only the last 10 bytes of large object
int bytesToRead = 10;
vclob.Read(vclobBuff, 0, bytesToRead, (-1) * bytesToRead,
IfxSmartLOBWhence.End);
...
Smart Large Objects support random IO access. Using the Seek(offset,

whence) method you can position anywhere in the large object and perform IO
operations with the data.
Example 8-23 demonstrates how to update only a portion of a large object using
the Seek() method.
274
7884ch08.fm
Example 8-23 Lo_seek.cs

using System;
using System.Data;
class sample {
IfxConnection conn;
IfxClob vclob;
int urows=0;
int maxSize = 100;
char[] vclobBuff = new char[maxSize];
try {
conn.Open();
// Select the large object from the database
cmmd.CommandText = "SELECT * FROM catalog WHERE catalog_num = 10072";
IfxDataReader drdr;
drdr.Read();
// Get the large object. The clob is the 6th column
vclob = drdr.GetIfxClob(5);
// Open the large object
// Move the pointer 15 bytes from the beginning of the large object.
vclob.Seek(15,IfxSmartLOBWhence.Begin);
// Update the large object.
vclobBuff="//------------//".ToCharArray();
vclob.Write(vclobBuff);
drdr.Close();
// Update the column in the table
cmmd.CommandText = "UPDATE catalog set advert_descr = ? WHERE" +
" catalog_num = ?;";
// Bind the IfxClob value and execute the UPDATE statement
IfxParameter padvt_desc = new IfxParameter(null, vclob);
IfxParameter pcatalog_num = new IfxParameter(null, "10072");
cmmd.Parameters.Add(padvt_desc);
cmmd.Parameters.Add(pcatalog_num);
275
7884ch08.fm
// Close the large object

vclob.Close();
conn.Close();
}
}
}
}
Note: The IBM Data Server Provider for .NET does not support random I/O
access.
IfxDateTime
The IfxDateTime class represents the Informix DATETIME data type. The
IfxDateTime class provides support for all the precisions allowed in an Informix
DATETIME.
The Informix DATETIME data type is made of these time units:
Year
Month
Day
Hour
Minute
Second
Fractions of a second
In an Informix database, the maximum precision for a DATETIME column is YEAR

TO FRACTION(5). A DATETIME column can be defined with any precision from a
year to fraction of a second allowing any subset of these units.
Table 8-15 lists the public properties of the IfxDataTime class.
Table 8-15 Public properties of IfxDataTime
276
Property
Type
Description
Date
IfxDateTime
A Year to Day IfxDateTime instance
Day
Int32
The day portion of the value
EndTimeUnit
IfxTimeUnit
The end time unit of the instance
Hour
Int32
The hour portion of the value
MaxValue
IfxDateTime
Maximum value allowed for this IfxDateTime
7884ch08.fm
Property
Type
Description
Millisecond
Int32
Millisecond unit in this IfxDateTime
MinValue
IfxDateTime
Smalles value allowed for this IfxDateTime
Minute
Int32
Minute unit in this IfxDateTime
Month
Int32
Month unit in this ifxDateTime
Now
IfxDateTime
Current date with a range of Year to Fraction (5)
Second
Int32
Second unit of this IfxDateTime.
StartTimeUnit
IfxTimeUnit
Start time unit of this IfxDateTime
Ticks
Int64
Ticks from midnight on 1 Jan 0001
Today
IfxDateTime
Current time with a range of Year to Day
Year
Int32
Year unit of the value
Table 8-16 lists the the public methods of the IfxDateTime class.
Table 8-16 Public methods of IfxDateTime
Method
Description
Add(IfxTimeSpan or IfxMonthSpan)
Current value plus by object passed
AddDays(days)
Current value plus days passed
AddMilliseconds(milliseconds)
Current value plus milliseconds passed
AddMinutes(minutes)
Current value plus minutes passed
AddMonths(months)
Current value plus months passed
AddSeconds(seconds)
Current value plus seconds passed
AddYears(years)
Current value plus years passed
Compare(ifxDT1, ifxDT1)
Compare two IfxDataTime instances
CompareTo(obj)
Compare two IfxDataTime instances
DaysInMonth(year, month)
Number of days in the month of the year
Equals(ifxDT1, ifxDT2)
True if ifxDT1 is equal to ifxDT2
GreaterThan(ifxDT1, ifxDT2)
True if ifxDT1 is later than ifxDT2
GreaterThanOrEqual(ifxDT1, ifxDT2)
True if ifxDT1 is later or equal than ifxDT2
LessThan(ifxDT1, ifxDT2)
True if ifxDT1 is earlier than ifxDT2
277
7884ch08.fm
Method
Description
LessThanOrEqual(ifxDT1, ifxDT2)
True if ifxDT1 is earlier or equal than ifxDT2
NotEquals(ifxDT1, ifxDT2)
True if ifxDT1 is different than ifxDT2
Parse(dateTimeStr)
Parse(dateTimeStr, start, end)
Parse(dateTimeStr, format, start, end)
New IfxDateTime with a value based on

dateTimeStr
ToString(), ToString (format)
Calue of the instance as a string
To access a DATETIME column in the database, the application uses the

IfxGetDateTime() method from the IfxDataReader() object.
Example 8-24 demostrate how to select a DATETIME data type from a table.
Example 8-24 Datetime.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
IfxDateTime vdtime;
try {
conn.Open();
// Select the datetime from the database
cmmd.CommandText = "SELECT * FROM cust_calls WHERE "
+ "customer_num = 106";
IfxDataReader drdr;
drdr.Read();
// Get the IfxDateTime from the recordset, 2nd columnd
vdtime = drdr.GetIfxDateTime(1);
Console.WriteLine("Call:\t"+vdtime);
Console.WriteLine("Hour :\t "+vdtime.Hour);
Console.WriteLine("Minute:\t"+vdtime.Minute);
conn.Close();
}
278
7884ch08.fm
}
}
}
In Example 8-25 you can see the Hour and Minute units extracted from the
IfxDateTime object.
Example 8-25 Output of Datetime.cs
datetime.cs
C:\work>datetime
Call:
2008-06-12 08:20
Hour :
8
Minute: 20
C:\work>c
Arithmetic operations between two IfxDateTime objects may return a

lfxDateSpan() object or a IfxMonthSpan() object depending on the precision used
by the IfxDateTime objects.
Example 8-26 shows the usage of the IfxDateSpan() class.
Example 8-26 Datetime2.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
IfxDateTime vdtime,vrestime;
IfxTimeSpan vtimetaken;
try {
conn.Open();
cmmd.CommandText = "SELECT * FROM cust_calls WHERE "
+ "customer_num = 106";
IfxDataReader drdr;
279
7884ch08.fm

drdr.Read();
// Get the IfxDateTime from the recordset
vdtime = drdr.GetIfxDateTime(1);
vrestime = drdr.GetIfxDateTime(5);
// Stores the difference in an IfxTimeSpan object.
vtimetaken=vrestime-vdtime;
Console.WriteLine("Initiated at:\t"+vdtime);
Console.WriteLine("Resolved at:\t"+vrestime);
Console.WriteLine("Minutes taken:\t"+vtimetaken.Minutes);
conn.Close();
}
}
}
}
IfxDecimal
An IfxDecimal object represents the Informix decimal data type. The DECIMAL
data type on an Informix database can take two forms: DECIMAL(p) floating point
and DECIMAL(p,s) fixed point. The Informix .NET IfxDecimal supports both
versions.
Table 8-17 contains the description of the public properties of a IfxDecimal
object.
Table 8-17 Public properties of IfxDecimal
280
Property
Type
Description
IfxDecimal
Irrational number e.
IsFloating
Boolean
Whether is a floating number.
IsNull
Boolean
Whether is NULL.
IsPositive
Boolean
Whether is positive
MaxPrecision
Byte
Maximum precision supported (32).
MaxValue
IfxDecimal
Largest value allowed.
MinusOne
IfxDecimal
-1.
MinValue
IfxDecimal
Smallest value allowed.
Null
IfxDecimal
Null.
One
IfxDecimal
1.
7884ch08.fm
Property
Type
Description
Pi
IfxDecimal
Irrational number pi.
Zero
IfxDecimal
0.
Table 8-18 contains the description for the IfxDecimal public methods.
Table 8-18 Public methods of IfxDecimal
Method
Description
Abs(IfxDec)
Absolute value of IfxDec
Add(IfxDec1, IfxDec2)
Sum of IfxDec1 and IfxDec2
Ceiling(IfxDec)
Smallest integer that is not less than IfxDec
Clone()
Creates a duplicate of this instance
Compare(IfxDec1, IfxDec2)
Compares two IfxDecimal values
CompareTo(obj)
Compares current instace with obj
Divide(Dividend, Divisor)
Dividing result for Dividend by Divisor
Equals(obj)
Equal
Equals(IfxDec1, IfxDec2)
True if IfxDec1 is the same as IfxDec2
Floor(IfxDec)
Largest integer not larger than IfxDec
GreaterThan(IfxDec1, IfxDec2)
True if IfxDec1 > IfxDec2
GreaterThanOrEqual(IfxDec1,IfxDec
2)
True if IfxDec1 >= IfxDec2
LessThan(IfxDec1, IfxDec2)
True if IfxDec1 < IfxDec2
LessThanOrEqual(IfxDec1, IfxDec2)
True if IfxDec1 <= IfxDec2
Max(IfxDec1, IfxDec2)
Returns whichever is larger, IfxDec1 or IfxDec2
Min(IfxDec1, IfxDec2)
Whichever is smaller, IfxDec1 or IfxDec2
Modulo(a, b)
Returns the remainder
Multiply(IfxDec1, IfxDec2)
Returns IfxDec1 times IfxDec2
Negate(IfxDec)
Current value negated
NotEquals(IfxDec1, IfxDec2)
True if IfxDec1<>IfxDec2
Parse(DecString)
Returns a new IfxDecimal based on DecString
281
7884ch08.fm
Method
Description
Remainder(a, b)
Remainder of the integer division of a by b
Round(IfxDec1, FractionDigits)
Value of IfxDec1 rounded to FractionDigits
Subtract(IfxDec1, IfxDec2)
Returns IfxDec1 minus IfxDec2
ToString(), ToString (format)
Returns the current value as a string
Truncate(IfxDec1, FractionDigits)
Round() with truncation
An application uses the GetIfxDecimal() method from the IfxDataReader()

class to access a DECIMAL column from the database.
Example 8-27 demonstrates how to retrieve a DECIMAL column
Example 8-27 Decimal.cs
using System;
using System.Data;
class sample {
IfxConnection conn;
IfxDecimal vtax;
try {
conn.Open();
cmmd.CommandText = "SELECT * FROM state WHERE "
+ "code = 'CA'";
IfxDataReader drdr;
drdr.Read();
// Get the IfxDecimal from the recordset
vtax = drdr.GetIfxDecimal(2);
Console.WriteLine("Value Sales Tax :\t "+vtax);
Console.WriteLine("Ceiling Sales Tax :\t " + IfxDecimal.Ceiling(vtax));
Console.WriteLine("Negative Sales Tax :\t " + IfxDecimal.Negate(vtax));
conn.Close();
}
282
7884ch08.fm
}
}
}
Example 8-28 shows the output of the previous example.

Example 8-28 Decimal.cs output
C:\work>csc.exe /R:%INFORMIXDIR%\bin\netf20\IBM.Data.Informix.dll /platform:x86
decimal.cs
C:\work>decimal
Value Sales Tax :
Ceiling Sales Tax :
Negative Sales Tax :
0.0825
1.0
-0.0825
C:\work>
In this section we discuss typical errors seen when using the .NET Provider and
how to enable the tracing facility to collect diagnostic information.
Typical errors
If the application is failing to connect to the Informix database server, check if
connection string details are correct for the server and version of the .NET
provider that you are using.
Both .NET providers use the ODBC or CLI layer to communicate with the
database server. You can test the connection with these components before
using the .NET provider.
When using the Informix .NET provider, always test the connectivity first using
the Ilogin tool included in the Informix Client SDK package.
The Data Server Driver for .NET package includes the Testconn11.exe and
Testconn20.exe executables which can be used to diagnose connectivity and
setup problems when using the Data Server provider.
Table 8-19 contains some typical errors and how to resolved them.
283
7884ch08.fm

Table 8-19 Typical errors
Error
Reason
Solution
ERROR [IM009] [Informix

.NET provider]
Unable to load translation
shared library (DLL)
The Informix .NET

provider failed to load
the ODBC shared
library.
Check the values for the

INFORMIXDIR environment
variable and PATH are correct.
INFORMIXDIR should point to
your Client SDK install
directory, PATH should
contain the
%INFORMIXDIR%\bin
directory.
System.BadImageFormat
Exception: An attempt was
made to load a program
with an incorrect format
A 32-bit application
loading a 64-bit
IBM.Data.Informix
assembly.
Register the proper version of

the .NET assembly into the
GAC using the gacutil.exe tool
Unhandled Exception:
System.EntryPointNotFou
ndException: Unable to
find an entry point named
'InterlockedIncrement' in
DLL 'kernel32.dll'.
A 64-bit application
loading a 32-bit
IBM.Data.Inforimx.dll
assembly.
Register the proper version of

the .NET assembly into the
GAC using the gacutil.exe tool
Unhandled Exception:
System.DllNotFoundExce
ption: Unable to load DLL
'IfxDotNetIntrinsicModule.
dll': The specified module
could not be found.
Failed to load
IfxDotNetIntrinsicModu
le assembly required by
the 64-bit version of the
Informix .NET provider
When using the 64-bit version

of the .NET provider, the
environment variable PATH
should contains the
%INFORMIXDIR%\bin and
%INFORMIXDIR%\bin\netf20
directories
Tracing
Your application can set two types of tracing when using the Informix .NET
provider:
Informix .NET Trace
SQLIDEBUG
Informix .NET trace

This trace is specific to the Informix .NET provider. To use this feature the
application must use the Trace version of the IBM.Data.Informix assembly
located in the %INFORMIXDIR%\netf20 directory. The name of shared library is
IBM.Data.Trace.dll.
Register the assembly in the GAC and enable the trace by setting the
environment variables IFXDOTNETTRACE and IFXDOTNETTRACEFILE.
284
7884ch08.fm
Example 8-29 demonstrates how to register the trace library and generate a
.NET trace file.
Example 8-29 Trace setup
C:\work>gacutil /i %INFORMIXDIR%\bin\netf20\IBM.Data.IfxTrace.dll /nologo
Assembly successfully added to the cache
C:\work>set IFXDOTNETTRACE=2
C:\work>set IFXDOTNETTRACE=trace.txt
C:\work>decimal
Value Sales Tax :
Ceiling Sales Tax :
Negative Sales Tax :
0.0825
1.0
-0.0825
C:\work>dir trace.txt
Directory of C:\work
25/06/2010
13:52
13,172 trace.txt
1 File(s)
13,172 bytes
0 Dir(s) 76,259,311,616 bytes free
C:\work>
The .NET trace file contains information about all the .NET classes used and
their return values. Example 8-30 shows some of the entries in the .NET trace
file.
Example 8-30 .NET Trace file.
...
3532:1 Entry: IfxDecimal.ToString()
3532:1 Exit: IfxDecimal.ToString
IfxDecimal 0.0825
3532:1 Entry: IfxDecimal.ToString()
532:1 Exit: IfxDecimal.ToString
IfxDecimal -1.0
)
3532:1 Entry: IfxDecimal.get_IsNull()
3532:1 Exit: IfxDecimal.get_IsNull
532:1 Exit: IfxDecimal.Multiply
3532:1 Exit: IfxDecimal.Negate
3532:1 Entry: IfxConnection.Close()
3532:1 Entry: IfxConnection.GetLatch
285
7884ch08.fm

(
String
)
...
IfxConnection.Close
The level of tracing is determine by the value of IFXDOTNETTRACE variable

0, No tracing
1, Tracing of API entry and exit, with return code
2, Tracing of API entry and exit, with return code, plus tracing of parameters to
the API
Levels 3 and 4 are for internal use only.
SQLIDEBUG and DRDADEBUG

Refer to the SQLIDEBUG section in 3.3.6, Troubleshooting on page 113 for the
description and use of the SQLIDEBUG.
8.4 Visual Studio Add-In for Visual Studio

This section gives an overview of the Visual Studio Add-in. The IBM Database
Add-Ins for Visual Studio are a collection of features that integrate into Microsoft
Visual Studio development environment so that you can work with IBM Data
servers.
Visual Studio Add-in Version 9.5 is compatible with Visual Studio 2005 and it
uses the IBM Informix .NET Provider to connect to the database server.
Version 9.7 is compatible with Visual Studio 2005 and Visual Studio 2008 and
uses the IBM Data Server for .NET.
Although it is not included in the Client SDK bundle, the VS Add-ins is available
as a separate product:
https://www.software.ibm.com/webapp/iwm/web/reg/download.do?source=swg-vsai
The Visual Studio Add-In provides tools and wizards that simplify the .NET
development for common tasks such as:
Explore catalog information of your database server.
Generate DDL for your database objects.
Drag and drop your server objects onto your .NET applications components
to automatically generate the required ADO.NET code.
286
7884ch08.fm
The first step to use the Visual Studio Add-In is to define the IBM Informix server
into the Data Connections group of your Visual Studio environment. Figure 8-1
shows the Add Connection selection for adding a new server definition.
Figure 8-1 Server Explorer Add Connection
Figure 8-2 shows the Connection details dialog for a new server
Figure 8-2 Connection Dialog
Figure 8-3 shows two of the stores7 database tables dragged into a .NET
DataSet component.
287
7884ch08.fm
Figure 8-3 Visual Studio Add-in
For more details about using the IBM Database Add-Ins and the Data Server
Provider for .NET for application development, refer to the IBM Information
Management and Visual Studio .NET zone at
http://www.ibm.com/developerworks/data/zones/vstudio
288
7884ch09.fm
Chapter 9.
Working with PHP

PHP is a widely-used general-purpose scripting language that is especially
suited for web development and can be embedded into HTML. This chapter
discusses the use of PHP with Informix as the database server. We discuss
various PHP database extension which you can use to connect to informix.
Following that is the discussion about other components such as Apache Web
server and OAT (OpenAdmin Tool) and how to bring these together. At the end
we describes, in detail, the usage of programming interfaces provided by the
selected PHP database extensions through the use of various examples.
We discuss the following SQL and database concepts:
Basic database operations like Insert, Delete, Update.
Handling special data types (BLOB, CLOB, TEXT and BYTE)
Exception handling
Troubleshooting
289
7884ch09.fm
9.1 Informix and PHP extensions

The PHP Data Objects (PDO) extension defines a lightweight, consistent
interface for accessing databases in PHP. Each database driver that implements
the PDO interface can expose database-specific features as regular extension
functions.
Informix database can be accessed by PHP using two PDO drivers:
PDO_INFORMIX
This is also called PHP Driver for IBM Informix. To compile and use the driver,
you must install the Informix Client Software Development Kit. You can
download this driver from
http://pecl.php.net/package/PDO_INFORMIX
PDO_IBM
This driver is also know as PHP Driver for Data Server clients. The PHP
Driver for Data Server clients requires the IBM Data Server Driver for ODBC
and CLI to be installed on the same computer. This driver can be accessed
from the site
http://pecl.php.net/package/PDO_IBM
In addition to these two drivers, there are two PHP extensions which allows you
to access an IBM Informix database:
PHP_INFORMIX
This extension only provides support for standard Informix data types. This
extension is developed by the open source community and is available at
http://cvs.php.net/viewvc.cgi/pecl/informix/
IBM_DB2
This extension is available under the IBM Data Server Driver Package, and is
written, maintained, and supported by IBM. It supports the Informix and DB2
databases.
Choosing the driver

Your application development environment, including hardware and software,
should determine which PHP drivers and extensions to use. If the platform of
your choice is not supported by IBM Data Server Driver or if you are using some
of the data types not supported by IBM Data Server Driver then go for the
Informix driver. The Data Server Driver for ODBC and CLI requires an Informix
database of Version v11.x or later. If your database server is exclusively Informix,
290
7884ch09.fm
then there is no problem using the Informix driver. You can also use both drivers
simultaneously if such need arises.

To develop an PHP application for Informix, your application development system
should have the following software:
PHP drivers and extensions:
Choose one of the PHP driver or extension supported by Informix. The PHP
Driver for IBM Informix requires the Informix Client SDK and the PHP Driver
for Data Server clients requires the IBM Data Server Driver for ODBC and
CLI. Refer to 2.2, Client setup on page 31 for the details.
Web Server
You can either install Apache Web server or use OpenAdmin Tool (OAT)
which is a PHP-based Web browser administration tool. Installing OAT is an
easier option because it contains everything that is required for PHP
connectivity. At the time of writing this book, OAT version 2.2.7 has:
Apache 2.2.4
PHP 5.2.4
PDO_INFORMIX 1.2.6
In this section we discuss OAT installation and driver configuration. For installing
the latest Apache Web server, refer to the website
http://httpd.apache.org/
9.2.1 Installation of OAT

The OAT product is available for Windows, Linux, and Macintosh OS X (64-bit
only) at the OpenAdmin Tool website at
http://www.openadmintool.org
This will lead to the IBM site that requires you to register and login.
Once you have completed the download, the installation procedure is as follows:
1. Start the installation with one of the following methods:
GUI Mode: Launch the installation executable:
On Windows: Run install.exe.

On Linux: Run install.bin.
MAC OS X: Unzip the install.zip file, and run install.app.
Chapter 9. Working with PHP
291
7884ch09.fm
Console Mode: Enter the following commands:
Windows: install.exe -i console

Linux: install.bin -i console
2. Accept the license agreement to continue.

3. Select the installation directory. The default is:
For Windows: C:\Program Files\OpenAdmin
For Linux: /opt/OpenAdmin
For MAC OS X: /Applications/OpenAdmin/
4. Choose an available port number for the Web server.
5. For Windows users, provide an Apache service name.
6. Specify the host name which is the name of the computer where the database
server is located.
7. In the Security Features panel, to enable password-protection for the
OpenAdmin Tool administration pages, select the check box.
8. In the OAT Administrator login setup panel, enter a user name and password.
9. In the Plug-in page, ensure that the plug-ins that you want to install are
selected. Accept the license agreement.
10.In the Pre- Installation Summary panel, review your selections and proceed
the installation. When the installation is complete, the following message is
displayed
OpenAdmin Tool has been installed successfully. Please visit
http://servername:portnumber/openadmin/ to use the OpenAdmin Tool
Where:
servername is the name of the machine where the Web server is running.
This can be localhost on Windows.
portnumber is the port number that you provided in step 4.
11.Click Done.
For Linux and MAC OS X: The OpenAdmin Tool configuration page opens
in your default Web browser.
For Windows: Your machine needs reboot. You can access OAT from the
Start All Programs OpenAdmin Tool for IDS.
For an insight into OpenAdmin Tool, read the release notes and the README file
which comes bundled with the product.
292
7884ch09.fm
9.2.2 PDO_INFORMIX
We discuss the installation of the PDO_INFORMIX driver in 2.2.3, Setting up
IBM Data Server drivers on page 40. PDO_INFORMIX driver is also installed
automatically when you install OAT.
We use the PDO_INFORMIX driver and the Apache web server that is installed
with OAT in a Windows system for the examples in this chapter.
On successful installation of OAT, the shared library php_pdo_informix.ddl (.so
in Linux) is placed in the PHP extension directory. For example, if your installation
directory is C:\Program FIles\OpenAdmin, the shared library is in the directory
C:\Program Files\OpenAdmin\PHP_5.2.4\ext.
To verify if the PHP extensions are working, in a web browser give the following
URL:
http://localhost:8080/phpinfo.php
This should give a web page similar to the one in Figure 9-1. The picture shown
here is only the initial part of the entire html page.
Figure 9-1 partial picture of output from phpinfo() script
293
7884ch09.fm
9.2.3 PDO_IBM
The installation of PDO_IBM driver is discussed in 2.2.3, Setting up IBM Data
Server drivers on page 40. Depending on your platform copy, add the .dll or .so
to the OpenAdmin extension directory and change the php.ini to include the
extension php_pdo_ibm.dll or php_pdo_ibm.so.
9.2.4 Verifying the PDO setup

You can see which PDO driver has been loaded by the Apache web server by
typing the URL http://localhost:8080/phpinfo.php in your browser. Figure 9-2
shows that the PDO_INFORMIX and PDO_SQLITE driver are available to use.
Figure 9-2 output from phpinfo() function

You can use a simple PHP program to verify the connectivity to the Informix
server using the installed Informix PHP drivers or extensions.
Example 9-1 on page 295 shows a simple PHP script which connects to the
database server using the PDO_INFORMIX driver. In the conn_string variable,
set the proper host name, user name, and password, and the respective locales.
The first keyword in the connection string identify which PDO driver is been used.
The PDO_INFORMIX driver uses informix and the PDO_IBM driver uses ibm.
294
7884ch09.fm
Example 9-1 connect.php

<?php
$informixdir = getenv("INFORMIXDIR");
$uname = "informix"; $password= "123456";
$conn_string = "informix: host=9.12.4.65; service=9088; database=stores_demo;
server=ol_svr_custom; protocol=onsoctcp; CLIENT_LOCALE=EN_US.CP1252 ;
$sql = "SELECT dbinfo('version', 'major') version FROM systables WHERE
tabid=1";
try {
print "<li>informixdir = $informixdir</li>\n";
$conn = new PDO($conn_string, $uname, $password);
$conn->setAttribute(PDO::ATTR_ERRMODE, PDO::ERRMODE_EXCEPTION);
print "<li>Got a connection</li>\n";
$stmt = $conn->query($sql);
if ( ! $stmt ) {
print "Error in execute: stmt->execute()\n";
print "errInfo[0]=>$err[0]\nerrInfo[1]=>$err[1]\nerrInfo[2]=>$err [2]\n";
}
$row = $stmt->fetch();
print "<li>IDS version " . $row['VERSION'] . "</li>\n";
}
catch (Exception $e) {
print "Exception messsage:{$e->getMessage()}\n";
exit(0);}
?>
If you run the script from a browser, you should see an output like below:
informixdir = C:\Program Files\IBM\Informix\
Got a connection
IDS version 11
You can use the same script to test the connectivity using the PDO_IBM driver,
by just changing the connection string. Example 9-2 shows a typical connection
string using the PD_IBM driver.
Example 9-2 PDO_IBM connection string.
$conn_string = "ibm: DRIVER={IBM DB2 ODBC
DRIVER};DATABASE=stores_demo;HOSTNAME=kodiak;PORT=9089;PROTOCOL=TCPIP;";
At this point, you know that your environment is set properly and you are ready to
start using PHP with Informix.
295
7884ch09.fm
9.3 Developing a PHP application

PHP is a powerful server-side scripting language that was invented and designed
for creating dynamic Web applications with non-static content. The PHP code
can be a standalone program as well as an insert inside HTML (Hypertext
Markup Language) or XHTML (Extensible Hypertext Markup Language). The
PHP syntax is based mostly on and similar to C, Java, and Perl. You can use
PHP with an open-source license. You can run the PHP program directly from
command line. The following are reasons for using PHP for our applications:
Easy to use:
PHP is a scripting language included directly in HTML. This means that
getting started is easy. There is no need to compile PHP programs or spend
time learning tools to create PHP. You can simply insert statements and get
quick turnaround as you make changes.
Fully functional:
The PHP language has built-in functions to access your favorite database.
With PHP, your HTML pages can reflect current information from databases.
You can use information of the user viewing your HTML Web page to
customize the page specifically for that user. You can create classes for
object-oriented programming, or use flat file or Lightweight Directory Access
Protocol (LDAP) databases. It also includes a spell checker, XML functions,
image generation functions, and more.
Compatible and quick:
PHP is compatible with all Web browsers, because PHP generates plain
HTML.
Secure:
Although PHP is open source, it is a secure environment. One of its
advantages is that the Web clients can only see the pure HTML code. The
logic of the PHP program is never exposed to the client, therefore, reducing
security exposures.
Open source:
PHP is an open-source programming language. It is easy to get started and
find examples from websites, such as http://www.sourceforge.net

Regardless of which PHP driver is used, PDO_INFORMIX or PDO_IBM, you
must decide what type of connection to use for any database connection and
what user type to use for authentication.
296
7884ch09.fm
Connection type
The connection type behavior is controlled by the Web server. There are two
connection types:
Persistence connection:
For the Persistent connection, the Web server will leave the connection to the
database open after the PHP script finishes its work. The next attempt to
connect to the database with the same parameters will reuse the connection.
Standard connection:
For the Standard connection, the database connection will be closed when
the script finishes its execution.
Here are some considerations for persistent connection:
When using the persistent connection, the same connection will be used for
the next request with the same connection parameters. This could cause
some kind of connection pooling depending on the implementation and how
you decide if the connection will be reused or not. Unintentional connection
pooling is most likely a source of serious problems, even if you use
transactions.
Opening up a persistent connection only makes sense in a Web server
environment if the connection will not be closed in the same script with a
connection close call. Calling the PHP command line processor on the shell
will close the connection to Informix database server at the end of the
execution of the script.
The Web server spawns more than one process to handle the incoming Web
requests, so reusing the existing persistent connection is only valid for a
specific Web server process. So it easily results in the database server having
more connections open than actual running Web server processes.
Restarting the database server without cleaning up the remaining persistent
connections in the Web server environment will produce no errors at
connection time when the connection is reused, but will fail at the first
database statement. This type of error is hard to diagnose.
The user type

The user type is controlled by the database and has to do with authentication at
the database server. You can choose an authentication with user ID and
password in the connection string, or a trusted user connect where the database
server does not apply an authentication. This will be controlled by settings on
operating system resources, such as .rhosts or /etc/hosts.equiv files. Since
you have to specify user names and passwords in the PHP scripts, pay higher
attention to these files in regard to security. Use encapsulation for files containing
297
7884ch09.fm
the passwords and put these files in directories secured with the .htaccess or
httpd.conf mechanism of the Apache Web server.
Informix PDO requires, at connection time, a defined set of parameters:
database name and database server name. We suggest you use a user ID and
password for untrusted user sessions. There are some optional parameters that
can influence the cursor behavior. In this section, we focus on the required
connection parameters.
Example 9-3 gives you a detailed overview about the various parameter settings
in an attempt to connect to the Informix database server. The examples shown
are for PDO_INFORMIX. Choose one of the connection statements which meets
your requirements.
Example 9-3 PDO_INFORMIX connection strings
/*--- standard connect ---*/
$dbc = new PDO("informix:; database=sysmaster;
server=ol_svr_custom;","informix", "123456");
/*--- standard connect trusted user ---*/
$dbc = new PDO("informix:; database=sysmaster; server=ol_svr_custom;");
/*--- persistent connect untrusted user ---*/
server=ol_svr_custom;","informix","123456",array(PDO::ATTR_PERSISTENT=> true));
/*--- persistent connect trusted user ---*/
server=ol_svr_custom;",NULL,NULL,array(PDO::ATTR_PERSISTENT=> true));
PDO_IBM is based on the Data Server Driver CLI driver, this means it uses the
same connectivity information as the ODBC and CLI driver. You can store this
information as well as user credentials in the db2cli.ini file located in the user
profile directory. Example 9-4 shows an example db2cli.ini file.
Example 9-4 db2cli.ini file
[dsc_dsn]
Protocol=TCPIP
Port=9089
Hostname=kodiak
PWD=password
UID=password
298
7884ch09.fm
Having the details inside the db2cli.ini file makes it easier to managed the
information needed for the connection. You can use either of the connection
strings mentioned in the Example 9-5. The second conn_string is using the file
dbscli.ini mentioned in Example 9-4 on page 298.
Example 9-5 PDO_IBM connection strings
$conn_string = "ibm: DRIVER={IBM DB2 ODBC DRIVER};DATABASE=stores_demo;
HOSTNAME=kodiak;PORT=9089;PROTOCOL=TCPIP;";
$conn_string = "ibm: DSN=dsc_dsn";

Informix supports both static and dynamic SQL statements for the client
applications such as PHP programs. In this section we provide an in-depth
discussion about the abilities of handling dynamic and static SQL of PDO
Informix. We focus on the INSERT statement, but also give short examples for
DELETE and UPDATE statements.
Example 9-6 shows various ways of executing a static INSERT with Informix
PDO.
Example 9-6 INSERT, UPDATE and DELETE using PDO_INFORMIX
<?php
$informixdir = getenv("INFORMIXDIR");
$dbc = new PDO("informix:; database=stores_demo;
$sql = "INSERT INTO customer VALUES(0,'Carla','Gomes','All kitchen supplies',
'2440 Cavaleras blvd', '', 'San Jose','CA','93086', '408-777-8075');";
$dbc->exec($sql);
$error=$dbc->errorInfo();
if ( $error["1"]) {
printf(" execute insert failed with %s <br>",$error["1"]);
exit(1);
}
/*--- Using PDO::query for the insert ---*/
$dbc->query($sql);
if ( $error["1"]) {
exit(1);
}
/*--- Using PDO::prepare and PDO::sql::execute for the insert ---*/
299
7884ch09.fm

$stmt=$dbc->prepare($sql);
if ( $error["1"]) {
printf(" prepare insert failed with %s <br>",$error["1"]);
exit(1);
}
$stmt->execute();
if ( $error["1"]) {
exit(1);
}
printf("RowCount: %d <br>",$stmt->rowCount());
$stmt=$dbc->query("SELECT COUNT(*) FROM customer ");
$row=$stmt->fetch(PDO::FETCH_NUM);
printf("<br>Total Rows in Table %d <br>", $row[0]);
?>
In a Web application, most of the time the input data comes from a HTML form
entered by a web user. Example 9-7 shows, based on a very simple HTML form,
how the data flows from the HTML over the PHP script into the database table.
Example 9-7 HTML sample
basic html format
<?php
printf("<form method=\"post\" action=\"<your file name>.php\" >");
printf("fname <input name=\"fname\" /><br />");
printf("lname <input name=\"lname\" /><br />");
printf("company <input name=\"company\" /><br />");
printf("address1 <input name=\"address1\" /><br />");
printf("address2 <input name=\"address2\" /><br />");
printf("city <input name=\"city\" /><br />");
printf("state <input name=\"state\" /><br />");
printf("zipcode <input name=\"zipcode\" /><br />");
printf("phone <input name=\"phone\" /><br />");
printf("<input type=\"submit\" value=\"Confirm\" name=\"Button\"/>");
printf("<input type=\"submit\" value=\"Abort\" name=\"Cancel\"/>");
printf("</form>");
?>
Example 9-8 shows the PHP script which handles the HTML form. The array
$_POST contains all the settings for the buttons and the values for the text fields
from the HTML.
Example 9-8 PHP Script handling the insert
<?php
300
7884ch09.fm
$dbc = new PDO("informix:host=9.14.23.34; database=stores_demo;

server=ol_svr_custom; protocol=onsoctcp ", "informix", "123456");
if (!dbc) {
exit();
}
/* --data from the Input Format
0|Carla|Gomes|Red Socks|217 Milpitas|Alum Rock|San Jose|CA|408-988-9887|
---*/
$statement="INSERT INTO customer VALUES (0,".
"'" . $_POST["fname"] . "'," .
"'" . $_POST["lname"] . "'," .
"'" . $_POST["company"] . "'," .
"'" . $_POST["address1"] . "'," .
"'" . $_POST["address2"] . "'," .
"'" . $_POST["city"] . "'," .
"'" . $_POST["state"] . "'," .
"'" . $_POST["zipcode"] . "'," .
"'" . $_POST["phone"] . "');";
printf ("STATEMENT = %s", $statement);
$dbc->query($statement);
if ( $error["1"]) {
printf(" prepare insert failed with %s \n",$error["1"]);
exit(1);
}
?>
In Example 9-9, we show how to perform an update and a delete using the PDO
Informix extension.
Example 9-9 Update PHP script
<?php
if (!dbc) {
exit();
}
/*--- Static update that changes Redwood city to Fremont ---*/
$statement="UPDATE customer SET city='Fremont' where city='Redwood City'";
$stmt=$dbc->query($statement);
if ( $error["1"]) {
printf(" update failed with %s <br>",$error["1"]);
}
printf("#Records updated: %d <br>",$stmt->rowCount());
/*--- Delete all records beloning to the company Sportstown; ---*/
$dbc->beginTransaction();
301
7884ch09.fm

$statement="DELETE FROM customer WHERE company='Sportstown'";
$stmt=$dbc->prepare($statement);
$stmt->execute();
printf("#Records deleted: %d <br>",$stmt->rowCount());
if ($stmt->rowCount()>10000)
{
$dbc->rollback();
}
else {
$dbc->commit();
}
?>
Dynamic SQL and Informix PDO

Dynamic SQL allows an IBM Informix client side programs to build an SQL
statement at runtime, so that the contents of the statement can be determined by
user input. Informix PDO provides several ways to use dynamic SQL. You have to
use placeholders, such as ? or :parameter, to prepare a statement for
dynamic usage. These placeholders will later be substituted with the actual
values. The prepare once execute multiple times capability of dynamic
statements allows you to specify different values at every execution time. To find
out which statements and clauses allow placeholders, refer to
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.sqls.doc/sq
ls.htm.
This document also is a comprehensive reference of Informix SQL statements.

In Example 9-10, we show different INSERT statements using Informix PDO. To
learn more about using placeholders (?) and bind parameters, refer to PHP
manual at
http://php.net/manual/en/pdostatement.bindparam.php
Example 9-10 Parametrized insert
<?php
$dbc = new PDO("informix:host=9.14.23.34; database=stores_demo; server=ol_svr_custom;
protocol=onsoctcp ", "informix", "123456");
if (!dbc) {
exit();
}
/*--- dynamic inserts with "?" and bindParam ---*/
$stmt=
$dbc->prepare("INSERT INTO customer VALUES (0,?,?,?,?,?,?,?,?,?)");
if ( $error["1"]) {
printf(" prepare insert1 failed with %s \n",$error["1"]);
exit(1);
302
7884ch09.fm

}
$stmt->bindParam(1, $_POST["fname"]);
$stmt->bindParam(2, $_POST["lname"]);
$stmt->bindParam(3, $_POST["company"]);
$stmt->bindParam(4, $_POST["address1"]);
$stmt->bindParam(5, $_POST["address2"]);
$stmt->bindParam(6, $_POST["city"]);
$stmt->bindParam(7, $_POST["state"]);
$stmt->bindParam(8, $_POST["zipcode"]);
$stmt->bindParam(9, $_POST["phone"]);
$stmt->execute();
if ( $error["1"]) {
exit(1);
}
/*--- dynamic inserts with :<> placeholders and bindParam ---*/

$stmt=
$dbc->prepare("INSERT INTO customer VALUES (0,:p1,:p2,:p3,:p4,:p5,:p6,:p7,:p8,:p9)");
$stmt->bindParam(':p1', $_POST["fname"]);
$stmt->bindParam(':p2', $_POST["lname"]);
$stmt->bindParam(':p3', $_POST["company"]);
$stmt->bindParam(':p4', $_POST["address1"]);
$stmt->bindParam(':p5', $_POST["address2"]);
$stmt->bindParam(':p6', $_POST["city"]);
$stmt->bindParam(':p7', $_POST["state"]);
$stmt->bindParam(':p8', $_POST["zipcode"]);
$stmt->bindParam(':p9', $_POST["phone"]);
$stmt->execute();
if ( $error["1"]) {
exit(1);
}
/*--- dynamic inserts with "?" placeholders and bindValue ---*/
$stmt=
$dbc->prepare("INSERT INTO customer VALUES (99,?,?,?,?,?,?,?,?,?)");
$stmt->bindValue(1, $_POST["fname"],PDO::PARAM_STR);
$stmt->bindValue(2, $_POST["lname"],PDO::PARAM_STR);
$stmt->bindValue(3, $_POST["company"],PDO::PARAM_STR);
$stmt->bindValue(4, $_POST["address1"],PDO::PARAM_STR);
$stmt->bindValue(5, $_POST["address2"],PDO::PARAM_STR);
$stmt->bindValue(6, $_POST["city"],PDO::PARAM_STR);
$stmt->bindValue(7, $_POST["state"],PDO::PARAM_STR);
$stmt->bindValue(8, $_POST["zipcode"],PDO::PARAM_STR);
$stmt->bindValue(9, $_POST["phone"],PDO::PARAM_STR);
$stmt->execute();
if ( $error["1"]) {
exit(1);
}
303
7884ch09.fm
/*--- dynamic inserts with :<> placeholders and bindValue ---*/

$stmt=
$dbc->prepare("INSERT INTO customer VALUES
(199,:p1,:p2,:p3,:p4,:p5,:p6,:p7,:p8,:p9)");
$stmt->bindValue(':p1', $_POST["fname"],PDO::PARAM_STR);
$stmt->bindValue(':p2', $_POST["lname"],PDO::PARAM_STR);
$stmt->bindValue(':p3', $_POST["company"],PDO::PARAM_STR);
$stmt->bindValue(':p4', $_POST["address1"],PDO::PARAM_STR);
$stmt->bindValue(':p5', $_POST["address2"],PDO::PARAM_STR);
$stmt->bindValue(':p6', $_POST["city"],PDO::PARAM_STR);
$stmt->bindValue(':p7', $_POST["state"],PDO::PARAM_STR);
$stmt->bindValue(':p8', $_POST["zipcode"],PDO::PARAM_STR);
$stmt->bindValue(':p9', $_POST["phone"],PDO::PARAM_STR);
$stmt->execute();
if ( $error["1"]) {
printf(" prepare4 insert failed with %s \n",$error["1"]);
exit(1);
}
?>l
9.3.3 Handling complex data types

In this section, we discuss how to use complex data types in a PHP program. We
cover row types, collection types such as SET, LIST and MULTISET, as well as
the BLOB and SBLOB data types. We explain how to work with the complex data
types using examples with Informix PDO.
Named row types

Example 9-11 shows the DDL to create a ROW type and the table that uses the
ROW type.
Example 9-11 DDL for creating ROW type and table
/*Create a row type and a table by name customer_rtype as below
CREATE ROW TYPE address_rtype (
street_num int,
street_name char(20),
city
char(20),
state
char(20),
zipcode
char(10)
);
CREATE TABLE customer_rtype (
customer_num serial,
lname
char(15),
fname
char(15),
company
char(20),
304

address
phone
);
7884ch09.fm
address_rtype,
char(18)
The PHP script in Example 9-12 shows SQL statements using ROW type. This
script inserts two rows, then selects the data from the database with and without
a filter on the row type. It then updates one of the rows and delete the rows from
the table.
Example 9-12 Insert complex types sample
<?php
if (!dbc) {
exit();
}
printf(" <br>[ INSERT ] <br><br>");
/*--- Insert a row with a rowtype -- add a row to the newly created table ---*/
$dbc->query(' INSERT INTO customer_rtype VALUES (
0,"Carla","Gomes","All Sports",row(12345,"Broadway","San
Fransisco","CA","12345")::address_rtype, "408-908-8887")');
$dbc->query(' INSERT INTO customer_rtype VALUES (
0,"Smith","jones","Collin sports",row(1222,"Almeda
blvd","Fremont","CA","12345")::address_rtype,"345-908-8887")');
printf(" <br>[ SELECT without filter] <br><br>");
/*--- Select the customer row without any filters ---*/
$stmt1=$dbc->query(" SELECT customer_num,fname,address.city,address.state,phone from
customer_rtype");
$row=$stmt1->fetch(PDO::FETCH_ASSOC);
while($row) {
print_r($row);
printf("<br>");
}
printf(" <br>[ SELECT with row_type filter ] <br><br>");
/*--- Select the row with a row_type filter ( a given address ) ---*/
$stmt1=$dbc->query(' SELECT customer_num , address.state, phone FROM customer_rtype
WHERE row(1222,"Almeda blvd","Fremont","CA","12345")::address_rtype = address');
while($row) {
print_r($row);
printf("<br>");
}
printf(" <br>[ UPDATE ] <br><br>");
/*--- update customer address ---*/
305
7884ch09.fm

$stmt1=$dbc->query(' UPDATE customer_rtype SET
address=row(1234,"Almeda blvd","Fremont","CA","12345")::address_rtype WHERE
row(1222,"Almeda blvd","Fremont","CA","12345")::address_rtype = address');
printf(" <br>[ SELECT ] <br><br>");
/*--- Verification ---*/
$stmt1=$dbc->query(' SELECT customer_num,fname, address::lvarchar FROM customer_rtype
WHERE address = row(1234,"Almeda blvd","Fremont","CA","12345")::address_rtype');
while($row) {
print_r($row);
printf("<br>");
}
printf(" <br>[ DELETE ] <br><br>");
/*--- close the buisness ---*/
$dbc->query(' DELETE FROM address_rowtype ');
$dbc->query(' DELETE FROM customer_rtype ');
?>
Example 9-13 shows the output from the SELECT statements to give you an idea
about how the array with row types looks in PHP.
Example 9-13 Output of ROW type example
[ INSERT ]
[ SELECT without row_type filter ]

Array ( [CUSTOMER_NUM] => 1 [FNAME] => Gomes [CITY] => San Fransisco [STATE] => CA
[PHONE] => 408-908-8887 )
Array ( [CUSTOMER_NUM] => 2 [FNAME] => jones [CITY] => Fremont [STATE] => CA [PHONE] =>
345-908-8887 )
[ SELECT with filter ]
Array ( [CUSTOMER_NUM] => 2 [STATE] => CA [PHONE] => 345-908-8887 )
[ UPDATE ]
[ SELECT ]
Array ( [CUSTOMER_NUM] => 2 [FNAME] => jones [] => ROW(1234 ,'Almeda blvd ','Fremont
','CA ','12345 ') )
[ DELETE ]
306
7884ch09.fm
Collection data types

The collection data types are another set of complex data types provided by the
Informix. Collection data types include SET, LIST, and MULTISET. In this section,
we cover these collection data types by showing an example that inserts into and
selects from a table containing collection data types.
Example 9-14 Using collection data types in PHP
<?php
/*create the tables containing collection data types
CREATE TABLE t_collections
(
seq
serial not null,
l1 list
(integer not null),
s1 set
(integer not null),
m1 multiset(integer not null)
);
*/
server=ol_svr_custom; protocol=onsoctcp ", "informix", "123456",
array(PDO::ATTR_PERSISTENT=> true));
if (!dbc) {
exit();
}
/*--- Insert a row ---*/
printf(" <br>[ Insert ] <br><br>");
$dbc->query(" INSERT INTO t_collections VALUES ( 0,
'LIST{-1,0,-2,3,0,0,32767,249}', 'SET{-1,0,-2,3}', 'MULTISET{-1,0,0,-2,3,0}')
");
$dbc->query(" INSERT INTO t_collections VALUES ( 0,
'LIST{-1,0,-2,3,0,0,55555,249}', 'SET{-11,0,-2,3}',
'MULTISET{-1,0,0,-2,3,0,9,10}') ");
/*--- select the row without any filters ---*/
printf(" <br>[ SELECT without filter] <br><br>");
$stmt1=$dbc->query(" SELECT l1::lvarchar LIST, s1::lvarchar SET, m1::lvarchar
MULTISET FROM t_collections");
while($row) {
print_r($row);
printf("<br>");
}
/*--- select the row with a set filter for the color of the car ---*/
printf("<br> [ SELECT with filter] <br><br>");
$stmt1=$dbc->query(" SELECT l1::lvarchar LIST, s1::lvarchar SET, m1::lvarchar
MULTISET FROM t_collections WHERE 32767 IN l1");
307
7884ch09.fm

while($row) {
print_r($row);
printf("<br>");
}
?>

Example 9-15 The output of collection data type
[ Insert ]
[ SELECT without filter]

Array ( [LIST] => LIST{-1 ,0 ,-2 ,3 ,0 ,0 ,32767 ,249 } [SET] => SET{-1 ,0 ,-2
,3 } [MULTISET] => MULTISET{-1 ,0 ,0 ,-2 ,3 ,0 } )
,3 } [MULTISET] => MULTISET{-1 ,0 ,0 ,-2 ,3 ,0 ,9 ,10 } )
[ SELECT with filter]
,3 } [MULTISET] => MULTISET{-1 ,0 ,0 ,-2 ,3 ,0 } )
BLOB and SBLOB data types

With BLOB and SBLOB data types, Informix IDS can handle large objects as
binary data in a BYTE data type and text objects in a TEXT data type. The BYTE
and TEXT data types, commonly knows as BLOB data types or Simple Large
Objects, provides you the capability to store images and entire documents in the
database. BLOB data types can be stored within all the other data in the table
space, or in a separate specified BLOB space.
BLOB data types can be used in several operational areas. Commonly document
retrieval systems and geographic information systems are based on this data
type. The retrieval of BLOB data with SQL needs the define of keywords, which
are stored together with the BLOB in the data row. The benefit of storing large
data in the database is, beside an easy search via keywords and a combination
of data stored in different rows but belonging together, also the opportunity of
backup and restore and the apply of delete and update for specific data. Storing
the data in a file in the operating system would cause much more maintenance
effort.
308
7884ch09.fm
More advanced than the BLOB data types, Smart Large Objects (commonly
know as SBLOB), provide more flexibility for searching in the data such as
random I/O access to the data, which was impossible with Simple Large
Objects).
There are two types of SBLOBs, BLOB and CLOB. the BLOB data type is used
to stored binary data when the CLOB data type is used only for character data.
There types require an additional set of functions defined in the server that
provides an API for access. They are stored in an SBLOB space (smart BLOB
space) in the database server.
You must create a sbspace to store the BLOB types in the server. You can create
them with following command:
onspaces -c -S sbspace -p <PATH> -o 0 -s 4000
We change the catalog table in stores database to use BLOB and CLOB
instead of BYTE and TEXT using the following command:.
ALTER TABLE catalog MODIFY (cat_descr TEXT, cat_descr CLOB, cat_picture BYTE,
cat_picture BLOB)
In the Example 9-16 we show how to insert and retrieve from data with SBLOB
data types, that is, BLOB and CLOB. The example shows how to insert the large
data from files into a table and select the same data back into files in a local file
system.
Example 9-16 Smart Large Object sample
<<?php
if (!dbc) { exit(); }
/*--- try to insert the BLOB ---*/
$stmt= $dbc->prepare("INSERT INTO catalog VALUES (0,302,'KAR',?,?,'All sports
Goods')");
$file = fopen ("picture.jpg","r");
$image = fread ( $file, 100000) ;
fclose ( $file);
$file = fopen ("README.txt","r");
$text = fread ( $file, 100000) ;
fclose ( $file);
$stmt->bindParam(1, $text );
$stmt->bindParam(2, $image);
$stmt->execute();
if ( $error["1"]) {
printf(" execute insert blobs failed with %s \n",$error["1"]);
309
7884ch09.fm

exit(1);
}
/*--- Get the BLOB from the database back ---*/
$query=$dbc->query("SELECT * FROM catalog where stock_num=302 and
manu_code='KAR' ");
if ( $error["1"]) {
printf(" select blobs failed with %s \n",$error["1"]);
exit(1);
}
$count=1;
$row=$query->fetch(PDO::FETCH_ASSOC);
while ( $row ) {
$file = fopen ("README.$count.txt","w");
$test=fread($row["CAT_DESCR"],100000);
while($test) {
fwrite ( $file, $test) ;
}
fclose ( $file);
$file = fopen ("PICTURE.$count.jpg","w");
$test1=fread($row["CAT_PICTURE"],100000);
while($test1) {
fwrite ( $file, $test1) ;
}
fclose ( $file);
$count++;
}
?>
Using TEXT and BYTE (BLOB data types) is similar to SBLOB types. Users may
try Example 9-16 on page 309 and Example 9-17 using TEXT and BYTE instead
of CLOB and BLOB respectively.
The BLOB data to be inserted can be read from the file without placing it in a
variable first. This simplifies the code as can be seen in Example 9-17.
Example 9-17 Insert BLOB with file
<?php
/*--- try to insert the BLOB ---*/
310
7884ch09.fm
$stmt= $dbc->prepare("INSERT INTO catalog VALUES (0,302,'KAR',?,?,'All sports

Goods')");
$file = fopen ("C:\README.txt","r");
$file1 = fopen ("C:\picture.jpg","r");
$stmt->bindParam(1,$file, PDO::PARAM_LOB);
$stmt->bindParam(2,$file1, PDO::PARAM_LOB);
$stmt->execute();
fclose($file);
fclose($file1);
if ( $error["1"]) {
printf(" execute insert blobs failed with %s \n",$error["1"]);
exit(1);
}
/*---try to get the BLOB from the database back ---*/
$query=$dbc->query("SELECT * FROM catalog where stock_num=302 and
manu_code='KAR' ");
if ( $error["1"]) {
printf(" select blobs failed with %s \n",$error["1"]);
exit(1);
}
$count=1;
while ( $row ) {
$file = fopen ("C:\README.$count.txt","w");
while($test) {
fwrite ( $file, $test) ;
}
fclose ( $file);
$file = fopen ("C:\PICTURE.$count.jpg","w");
while($test1) {
fwrite ( $file, $test1) ;
}
fclose ( $file);
$count++;
}
?>
In Example 9-17 on page 310, the data was selected as a stream in string data
type. Example 9-18 on page 312 shows how to bind variables to a SELECT.
Binding The data is placed in the variables by column.
311
7884ch09.fm

Example 9-18 Select BLOB with bind.
<?php
/*--- try to select the BLOB from the database into bind strings ---*/
$stmt=$dbc->query("SELECT catalog_num, cat_descr, cat_picture FROM catalog
where stock_num=302 and manu_code='KAR'");
$count=1;
$str="";
$str1="";
$id=0;
$stmt->bindColumn(1, $id, PDO::PARAM_INT);
$stmt->bindColumn(2, $str, PDO::PARAM_STR,100000);
$stmt->bindColumn(3, $str1, PDO::PARAM_STR,100000);
while ( $stmt->fetch(PDO::FETCH_BOUND) )
{
$file = fopen ("C:\README.$count.txt","w");
$file1 = fopen ("C:\PICTURE.$count.jpg","w");
fwrite($file,$str);
fwrite($file1,$str1);
fclose ($file);
fclose ($file1);
$count++;
}
?>
Example 9-19 shows how to update existing BLOB fields with Informix PDO
using a prepare statement and parameter for the BLOB columns.
Example 9-19 BLOB update
<?php
/*
try to update the BLOB columns
*/
$stmt= $dbc->prepare("UPDATE catalog SET cat_descr=? , cat_picture=? where
stock_num=302 and manu_code='KAR'");
if ( $error["1"]) {
printf(" prepare update blob columns failed with %s \n",$error["1"]);
exit(1);
}
$descr="This is an PDO descr clob text";
312
7884ch09.fm

$file= fopen ("C:\picture1.jpg","r");
$stmt->bindParam(1, $descr);
$stmt->bindParam(2, $file);
$stmt->execute();
?>
9.3.4 Working with PHP extensions

In addition to the PDO drivers for the PHP PDO extension, there are two PHP
extensions, PHP_INFORMIX and IBM_DB2, that allow you to connect to an
These extensions provide a set of additional PHP functions to work with an
Informix database. In addition to the normal create, read, update, and write
database operations, they also offer extensive access to the database metadata.
Table 9-1 lists a few of the functions included in the Informix PHP extensions as
examples.
Table 9-1 PHP extension functions
Function
Description
ifx_connect()
Open Informix server connection
ifx_fieldproperties()
List of SQL fieldproperties
ifxus_open_slob()
Opens an slob object
ifx_create_blob()
Creates an blob object
ifx_fetch_row()
Get row as an associative array
ifx_query()
Send Informix query
db2_connect()
Returns a connection to a database
db2_client_info()
Returns information describing DB2 database client
db2_primary_keys()
Returns a result set listing primary keys for a table
db2_special_columns()
Returns a result set listing the unique row identifier

columns for a table
db2_bind_param()
Binds a PHP variable to an SQL statement parameter
db2_commit()
Commits a transaction
313
7884ch09.fm
For a complete list of all the functions implemented inthe PHP_INFORMIX and
IBM_DB2 extensions refer to the PHP open source website at
http://www.php.net/manual/en/ref.ifx.php
http://www.php.net/manual/en/ref.ibm-db2.php
PHP_INFORMIX
A PHP script can connect to an Informix database using the ifx_connect()
function provided in the PHP_INFORMIX extension.
Example 9-20 shows a simple PHP script used to test connectivity with the
PHP_INFORMIX extension.
Example 9-20 connect.php
C:\work>type connect.php
<?php
$conn= ifx_connect ($argv[1], $argv[2],$argv[3]);
echo "Connection succeeded.\n";
ifx_close($conn);
?>
C:\work>php connect.php stores_demo informix password
Connection succeeded.
C:\work>
Functions such as ifx_query() and ifx_affected_rows() can be used to run

SQL statements and retrieve the number of rows affected. Example 9-21
demonstrates how to delete a row from the state table using the ifx_query()
function.
Example 9-21 delete.php
C:\work>type delete.php
<?php
$conn= ifx_connect ("stores_demo", "informix", "password");
$result = ifx_query("DELETE FROM state where code='".$argv[1]."'", $conn);
printf("Deleted %d records", ifx_affected_rows($result));
ifx_close($conn);
?>
C:\work>php delete.php UK
Deleted 1 records
C:\work>
The application can select data from an Informix database using the
ifx_prepare() and ifx_fetch_row() functions.
314
7884ch09.fm
Example 9-22 shows how to select the first three rows from the state table using
a prepared statement and the ifx_fetch_row() function.
Example 9-22 select.php
C:\work>cat select.php
<?php
$conn= ifx_connect ("stores_demo", "informix", "password");
$rid = ifx_prepare ("SELECT FIRST 3 code,sname FROM state",$conn, IFX_SCROLL);
if (! ifx_do ($rid)) {
die ("error\n");
}
$row = ifx_fetch_row ($rid, "NEXT");
while (is_array($row)) {
for (reset($row); $fieldname=key($row); next($row)) {
$fieldvalue = $row[$fieldname];
printf ("%s = %s ", $fieldname, $fieldvalue);
}
printf("\n");
$row = ifx_fetch_row($rid, "NEXT");
}
ifx_free_result ($rid);
ifx_close($conn);
?>
C:\work>php select.php
code = AK sname = Alaska
code = HI sname = Hawaii
code = CA sname = California
C:\work>
For more information and examples regarding the PHP_INFORMIX extension,

refer to the PHP open source website at:
http://www.php.net/manual/en/book.ifx.php
IBM_DB2
This PHP extension provides access to IBM Data Servers including IBM Informix
and IBM DB2.
In the same way as the PDO driver, PDO_IBM, The IBM_DB2 extension requires
the IBM CLI driver to communicate with the database server. The IBM CLI driver
is included as part of the IBM Data Server Driver for ODBC and CLI package.
This PHP extension provides functions such as db2_connect(), db2_exec(), and
db2_server_info() that can be used to perform typical tasks against an Informix
database server.
315
7884ch09.fm
Example 9-23 demonstrates how to use the db2_connect() and

db2_server_info() functions to retrieve metadata information from the database.
Example 9-23 connect_ibm.php
C:\work>type "C:\Documents and Settings\Administrator\db2cli.ini"
[dsc_dsn]
Protocol=TCPIP
Port=9089
Hostname=kodiak
C:\work>cat connect_ibm.php
<?php
$conn = db2_connect('dsc_dsn', 'informix','password');
if ($conn) {
$server = db2_server_info( $conn );
printf ("Database name
= %s\n", $server->DBMS_NAME);
printf ("Datbase version
= %s\n", $server->DBMS_VER);
db2_close($conn);
}
else {
echo "Connection failed.";
}
?>
C:\work>php connect_ibm.php
Database name
= IDS/NT64
Datbase version
= 11.50.0000
C:\work>
The IBM_DB2 extension can use the connection details in the db2cli.ini
configuration file. Refer to the 3.2.2, IBM Data Server Driver for ODBC and CLI
on page 66 for more information regarding the db2cli.ini configuration file.
There are several functions that can be used to select data from the database.
Example 9-24 shows how to use the db2_fecth_object() function to retrieve the
first three rows of the state table as a PHP object.
Example 9-24 select_ibm.php
C:\work>cat select_ibm.php
<?php
$conn = db2_connect('dsc_dsn', 'informix','password');
316
7884ch09.fm
if ($conn) {
$stmt = db2_exec($conn, "SELECT FIRST 3 code,sname FROM state");
while ($row = db2_fetch_object($stmt)) {
printf ("%s, %s\n", $row->code,$row->sname);
}
db2_close($conn);
}
else {
echo "Connection failed.";
}
?>
C:\work>php select_ibm.php
AK, Alaska
HI, Hawaii
CA, California
C:\work>
A full description of the IBM_DB2 extension can be found in the Informix

Information Center website at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9/topic/com.ibm.db2.udb.apdv.p
hp.doc/doc/t0023132.htm
9.3.5 Exception handling

Error handling or exception handling is important for all applications, including
web-based applications. Consider an Informix PHP application that shows the
following error message while the user is using the application?
Warning: odbc_connect(): SQL error: [unixODBC][Informix][Informix
ODBCDriver][Informix]User (informix) password 123456 is not able to connect
for the database server, Server is down, SQL state 28000 in SQLConnect in
/usr/local/apache2/htdocs/odbc/error/error.php on line 4
Fail in exception handling in this case creates a security exposure. In the this
section we describe many aspects related to exception handling using Informix
PDO.
PHP 5 has introduced, as a part of the new object-oriented programming
interface, the exception handling mechanism that are already used for other
programming languages. We strongly suggest you consider exceptions for the
usage of Informix PDO. Additionally, you can advance this interface by creating,
throwing, and catching your own exceptions.
317
7884ch09.fm
Informix PDO defines, in comparison with the procedural-oriented interface, two

ways of expressing an error in a database environment:
Exceptions raised internally by the PDO based on an error condition.
Database-generated errors. This type of exception can be captured and
handled by calling the PDO class errorInfo() or errorCode() function.
Example 9-25 shows two exceptions raised by Informix PDO functions. One is a
connection request to the database that failed because the specified database
does not exist. The other is an attempt to start a transaction twice.
Example 9-25 Without exception handling sample
<?php
$dbc = new PDO("informix:; database=stores; server=ol_svr_custom;","informix",
"123456");
?>
Fatal error: Uncaught exception 'PDOException' with message 'SQLSTATE=HY000,
SQLDriverConnect: -329 [Informix][Informix ODBC Driver][Informix]Database not
found or no system permission.' in C:\Program
Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php:2 Stack trace: #0 C:\Program
Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php(2): PDO->__construct('informix:;
data...', 'informix', '123456') #1 {main} thrown in C:\Program
Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php on line 2
<?php
?>
Fatal error: Uncaught exception 'PDOException' with message 'There is already
an active transaction' in C:\Program
Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php:4 Stack trace: #0 C:\Program
Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php(4): PDO->beginTransaction() #1
{main} thrown in C:\Program Files\OpenAdmin\Apache_2.2.4\htdocs\excep1.php on
line 4
If these exceptions are not caught and processed, the application terminates.
Example 9-26 on page 319 shows the usage of the basic exception handler
provided by PHP 5 to cover these errors. The action that should be taken once
an exception is caught depends on where the error originates. For example, if the
error occurs in the connecting to the database phase and the application cannot
continue, the action should be to generate an out of order web page with
contact details. If it is a minor database error, logging the error and retrying the
activity should be an appropriate action.
318
7884ch09.fm

Example 9-26 Simple exception handle code
<?php
try
{
}
catch (PDOException $e )
{
printf("Error: %s \n",$e->getMessage());
}
?>
Output:
Error: There is already an active transaction.
The error information generated during executing the SQL statements in the
database server is different from the exceptions generated by the Informix PDO
extension. The Informix PDO function errorInfo() can be used to capture the
status of the last executed SQL statement in the database. This function returns
an array with three elements, the SQLSTATE, the SQLCODE, and the error
message. The details of the meaning of the codes are described in
ii-11-40709.htm#sii-11-40709.
The function errorCode() is available to retrieve the SQL statement status. This
function returns only the SQLSTATE information. Example 9-27 shows how to
use the errorInfo() and what the output looks like.
Example 9-27 ErrorInfo() sample
<?php
$stmt=$dbc->query('SELECT * FROM nonexistingtable ');
/*
question the error code
output of the error Routines
*/
print_r($error);
if (!$error[1])
$row=$stmt->fetch(PDO::FETCH_ASSOC);
?>
319
7884ch09.fm
Output:
Array
(
[0] => 42S02
[1] => -206
[2] => [Informix][Informix ODBC Driver][Informix]The specified table
(nonexistingtable) is not in the database. (SQLPrepare[-206] at at
ext\pdo_informix\informix_driver.c:118)
)
In addition to using the generic exceptions provided by the PHP 5, you are open
to extend the exception class of your own exceptions. For example, you can
define different severities for SQL errors. Critical database errors are, for
instance, tables which do not exist or the database connections cannot be
established. The application will not be able to continue with these critical errors.
Non-critical errors, such as locking errors, could be handled by a retry.
Example 9-28 demonstrates how to define your own exceptions with PHP 5 and
Informix PDO. This example extends the standard exception class with two new
database exception classes, and depending on the severity, different actions are
taken.
Example 9-28 Custom exceptions
<?php
/*
own Exception classes for minor and major errors
*/
class CriticalDatabaseErrors extends Exception
{
public function __construct($message, $code = 0) {
parent::__construct($message, $code);
}
}
class NonCriticalDatabaseErrors extends Exception
{
public function __construct($message, $code = 0) {
parent::__construct($message, $code);
}
}
try
{
$stmt=$dbc->exec("SET ISOLATION REPEATABLE READ");
$stmt=$dbc->query('SELECT * FROM carr');
if (!$error[1]) {
do
320
7884ch09.fm

{
$row=$stmt->fetch(PDO::FETCH_ASSOC);
if ($error[1]) throw new NonCriticalDatabaseErrors($error[1]);
}
while($row) ;
}
else {
throw new CriticalDatabaseErrors($error[2]);
}
}
catch ( CriticalDatabaseErrors $cde )
{
printf("<pre>CritError!: %s \n</pre>",$cde->getMessage()) ;
}
catch ( NonCriticalDatabaseErrors $ncde )
{
printf("<pre>NonCritError!: %s \n</pre>",$ncde->getMessage()) ;
} catch (PDOExecption $ncde) {
printf("<pre>Error!: %s \n</pre>",$e->getMessage()) ;
exit;
}
?>
Output:
CritError!: [Informix][Informix ODBC Driver][Informix]The specified table (carr) is not
in the database. (SQLPrepare[-206] at ext\pdo_informix\informix_driver.c:131)
In this section we describe a couple of commonly seen errors in setting the
database connectivity:
Missing environment variable setting
Setting the environment variable INFORMIXDIR is required for starting the
Apache. If this variable is not set properly, the database connection will also
fail. Example 9-29 shows the error message when you have improper
INFORMIXDIR set. The message occurs at the first line of any PHP program
which is generally the connection string. If you get this message check your
INFORMIXDIR setting and correct it if its wrongly set.
Example 9-29 Unspecified error
Error!: SQLSTATE=HY000, SQLDriverConnect: -23101
[Informix][Informix ODBC Driver][Informix]Unspecified System Error =
-23101.
Mismatched settings
321
7884ch09.fm
Another important item affecting the connectivity is the setup of the Informix
run-time environment. Informix Connect or Informix Client SDK provides the
connectivity at run time. If the settings between the environment variables and
the sqlhosts do not match, you will see messages similar to the one shown in
Example 9-30.
Example 9-30 Wrong connection information
/usr/local/bin/php pdoconnect.php
Error!: SQLSTATE=HY000, SQLDriverConnect: -25555
[Informix] [Informix ODBC Driver][Informix]Server ol_svr_custom is not
listed as a dbserver name in sqlhosts.
322
7884ch10.fm
10
Chapter 10.
User defined routines

In this chapter we discuss the creation and use of User Defined Routines
(UDRs), a routine that you create, that can be invoked in an SQL statement, by
the database server, or from another UDR. A routine is a collection of program
statements that perform a particular task. By understanding UDRs, you can take
the next step and extend the database server, either a little bit or in steps that
lead to something bigger, such as a bladelet or DataBlade module.
We undertake to do this by looking at three areas:
An Overview of UDR and database extensions
Developing User defined routines
DataBlades and bladelets
Extending the database server requires an understanding of the parts required to
implement the extending. An extension can be as simple as one user defined
routine, or as complex as a new data type with many supporting routines. We first
discuss UDRs and user defined types (UDTs), to provide the scope for what is
needed. Then we provide examples.
323
7884ch10.fm
10.1 An Overview of UDRs and database extensions

IBM Informix database servers have the built-in ability to retrieve, store,
manipulate, and sort a number of standard data types, some of which are unique
to Informix servers. A developer can choose to modify or extend the built-in data
types, and modify or extend how various operations work with the resulting new
data type extensions. You can extend the data types in the following ways:
Extend operations that are used to process built-in data types.

Create complex data types based on built-in data types.
Create user defined data types (UDTs), both distinct and opaque data types.
Create new operations to process extended data types.
Database extensions and extended data types allow the developer to make
customized routines and functions transparent, since you have the capability to
actually build it into the database server. The transparency is visible as:
Better performance: Through optimized routines, faster queries. and reduced
network traffic.
Simpler applications: Modules you write result in streamlined code, and easy
to upgrade applications.
Transaction control: Modules have automatic recovery, backup, and rollback
capability provided by the server.
Scalability: Database extensions automatically scale with the database size
and user count.
In the context of discussing user defined routines and database extensions, we
may also mention DataBlade modules. A DataBlade module is a software
package that extends the functionality of the IBM Informix database server. Each
package includes SQL statements and supporting code written in an external
language or in Stored Procedure Language. A DataBlade module enables the
same level of support for new data types as the database server provides for
built-in data types. DataBlade modules can also use SQL queries or the
DataBlade API to access data types and routines in other DataBlade modules.
Note: Extended Parallel Server and Standard Engine support stored
procedures but not UDRs.
10.1.1 Considerations for user defined routines

A user define routine can be written using the IBM Informix Stored Procedure
Language (SPL) or with an external language. A routine written with SPL is
simple to use and easy to implement. An SPL based UDR has flow-control
324
7884ch10.fm
extensions (conditional clauses and while loops), and works with SQL
statements. Once the routine has been created and is ready to use, the database
server parses and optimizes it, and stores it in system catalog tables, ready to
execute.
The system catalog tables are used to keep track of the information the database
server uses to manage the database server. UDR related information is stored in
a small group of catalog tables that can tell us a lot about the UDR layouts. The
catalog tables that are common to all UDRs are:
sysprocedures table: This table is used to track the name, owner, and to
indicate whether the UDR is a user-defined function or a user-defined
procedure (functions return values. procedures do not).
sysprocauth table: This table tracks who can execute the procedure.
sysprocbody table: This table contains the actual code for the SPL routines.
UDRs can also be written with an external language. The body of an
external-language routine allows language-based operations such as flow control
and looping, while also using special Informix library calls to access the database
server. The database server also stores information for external-language
routines in three system catalog tables:
sysprocedures table: The information kept in this table is same as the SPL
user define routines.
sysroutinelangs table: This table tracks the language information.
syslangauth table: This table tracks the users of the server who can use the
particular external language.
You need to use an appropriate compiler to parse and compile an
external-language routine into an executable format. We discuss this with
specific examples of the language and API methods that are available for writing
these extensions. The external languages that can be used are:
C: To execute SQL statements in C UDRs, you must use the DataBlade API,
and you cannot use ESQL/C. To write routines in C, you need a C compiler.
Additional information on writing UDRs in C, is found in the IBM Informix
DataBlade API Programmers Guide, Version 11.50, SC23-9429-03, and IBM
Informix DataBlade API Function Reference, Version 11.50, SC23-9428-01.
Java: To write Java routines, you must use IBM Informix database server with
J/Foundation and with the Java Development Kit (JDK) to compile your Java
routines. Additional information about how to write Java UDRs, is found in
J/Foundation Developer's Guide, Version 10.0, G251-2291-00.
Chapter 10. User defined routines
325
7884ch10.fm
10.1.2 About user-defined routines

Table 10-1 lists the UDR types that are supported by SPL, C, and Java in IBM
Table 10-1 Supported UDR tasks by languages
UDR task
SPL routines
C routines
Java routines
Cast function
Yes
Yes
Yes
Cost function
No
Yes
No
End-user routine
Yes
Yes
Yes
Iterator function
No
Yes
Yes
Negator function
Yes
Yes
Yes
Opaque data type support
No
Yes
Yes
Parallelizable UDR
No
Yes
Yes
Statistics function
No
Yes
Yes
Selectivity function
No
Yes
No
User-defined aggregate
Yes
Yes
Some
Operator function
Yes
Yes
Yes
Operator-Class function
Yes
Yes
Yes
The following are the definitions of the terminologies in this table:

Cast: An routine to convert from one data type to another. Built-in data types
have automatic casts between most data types. For User Defined Data Types
(UDTs), casting have to be defined from scratch.
Cost function: A routine that informs the optimizer of cost factors for execution
of a particular UDR.
End-user function: A routine used to encapsulate multiple SQL statements
into one function.
Iterator function: A routine that is designed to return to its calling SQL
statement several times, each time returning a value.
Negator function: A routine used for a NOT condition that involves a Boolean
UDR.
Opaque data type support: When you create a new data type, you must
provide several basic support functions for your UDT. The following functions
are required:
326
7884ch10.fm
Text input and output routines

Binary send and receive routines
Text import and export routines
Binary import and export routines
Parallel UDR: A routine that can run in parallel within parallel queries.
Statistics function: A routine to create distribution statistics for a UDT.
Selectivity function: A routine to determine the percentage of rows for which a
Boolean UDR is expected to return true.
User-defined aggregate: An SQL invoked routine that takes values selected
and returns information about those rows (a summarizing method).
Operator function: A routine used within expressions with a symbol, such as
+,-,<,>,=. Built-in data type operators cannot be extended. All UDTs require
some operators in order to function within an SQL context.
Operator-Class function: A set of operators that the server will associate with
how to build an access method (that is, an index), how to arrange values in
the access method, how to select values based on operator function, and
ways to allow the query optimizer to consider using the access method to
return results for a query.
For more details on any of these functions, and for details on functionality not
discussed here, see the IBM Informix User-Defined Routines and Data Types
Developer's Guide, Version 11.50, SC23-9438-00.
Invoking a UDR
You can invoke a UDR implicitly or explicitly. Implicit invocation would be the
result of an operator function, an implicit cast, or some type of query processing.
For this handbook, we mainly discuss explicit invocation. You can use either
EXECUTE PROCEDURE or EXECUTE FUNCTION statements to run a UDR.
When the database server executes an SQL statement that contains a UDR, it
loads the UDR executable code as a shared-object into memory. It determines
which shared-object file to load from the externalname column of the row in the
sysprocedures system catalog table that describes the UDR. The sysprocedures
entry is created when you register the UDR, as a result of the CREATE
FUNCTION or CREATE PROCEDURE statement.
In more general terms, when you invoke a UDR, the database server parses the
statement into syntactic parts, call the system catalog to resolve the routine
parts, generate a query plan, and then execute the query. If the query involves
more than one database, each database requires that all the UDRs and UDTs
must be registered in the participating databases.
327
7884ch10.fm
10.1.3 Considerations for extending data types

A significant aspect of UDR is the support for an extended data type. IBM
Informix database servers have several built-in data types. Why do we need
more? An extended data type is a new data type, which has different core
properties, new functionality, and new operator methods that go beyond a basic
data type. As an example, consider a datetime data type. Imagine, as a
programmer if you had to add a day and a half to a specific datetime event.
Intuitively, we recognize that the value day and a half is an interval of time, but
the database serve would not recognize this construct as an interval. We would
need to manually convert it to a usable interval, then add it to a datetime value.
Using a user defined routine, an unconventional interval name such as this could
be passed through a function, which automatically translates it to an normal
interval, applies the addition operation, and returns a datetime value with the
result. To do this, we would be applying a functional behavior change, which
would extend the data type. The result of processes like these can simplify a
repeating data task and make the development task easier.
A quick review of the data type hierarchy is useful. Figure 10-1 shows a summary
of these data types.
Figure 10-1 Hierarchy of data types in Informix Servers
The Extended data types break out into user defined data types and complex
data types. We are especially interested in user defined types. These can be
described as follows:
Distinct: Internally stored the same as a source data type, but it is overlaid
with different casts and functions defined beyond the basic source type. The
server sees distinct types as different from the source type. It is necessary to
tell the server:
Source data type information and how the internal structure is defined.
The functions of how this data type interacts with its internal structure.
328
7884ch10.fm
The operations that are valid with this distinct data type.
Any secondary access methods on how to handle this type.
Cast functions to move data in and out of the distinct type are automatic.
Opaque: A fundamental, user defined data type. It cannot be broken into
smaller pieces, though it can serve as the building block for other data types.
The internal structure of the opaque data type is invisible to the server. When
you define and use an opaque type, the developer must provide all of the
following:
How the internal structure is defined.
The functions that enable routines to interact with its internal structure.
The operations that are valid with this distinct data type.
Any secondary access methods on how to handle this type.
Cast functions to move data in and out of the distinct type need to be
provided.
10.2 Developing user defined routines

To develop a user define routines, you should plan it then write the routine. We
follow the directions given in the UDR and Data Types Developers Guide for
planning our routines:
Use a sensible name.
Avoid modal arguments (an argument in the function determines how the
function will work).
Always declare routine parameter data types.
Declare the type that is returned when the routine is returning a value
The source for an external routine resides in a separate text file. To prepare UDR
source code:
You will compile the C-language UDR and store the executable version in a
shared-object (.so or .o on UNIX) file.
Compile the Java-language UDR and store the executable version in a .jar
file.
You must install shared object files and .jar files on all database servers that
need to run the UDRs, including database servers involved in Enterprise
Replication (ER) and High-Availability Data Replication (HDR). The shared object
files and .jar files need to be installed under the same absolute path name.
329
7884ch10.fm
Tip: Although not required, use the DataBlade Developers Kit (DBDK) to help
write UDRs is advantageous. DBDK books and software can help enforce
standards that facilitate migration between different versions of the database
server. Because external-language routines are external to the database, be
aware that you must compile the UDR source code and store it where the
database server can access it when the routine is invoked.
For information about C UDRs, refer to the IBM Informix DataBlade API
Programmers Guide, Version 11.50, SC23-9429-03 and the IBM Informix
DataBlade API Function Reference, Version 11.50, SC23-9428-01. For
information about Java UDRs, refer to the J/Foundation Developer's Guide,
Version 10.0, G251-2291-00.
10.2.1 User defined routines in SQL

You can use Informix Stored Procedure Language (SPL) statements to write
routines and store these SPL routines in the database. SPL extends SQL and
helps to reduce SQL coding, by minimizing the visible code in large SQL
operations. It has the advantage that it is run as a server side routine, the
executable code stays inside the engine, and it is optimized only as needed. The
end result is lower application startup costs and better performance. As an SQL
extension, it can do flow control, such as looping and branching. SPL routines
can also execute routines written in C or other external languages, and other
UDR routines can execute SPL routines.
We provide examples of different UDRs written in SQL in the following sections.
As we do so, a couple of rules for SQL UDRs to keep in mind are:
If you use any parameters, they have to be declared as built-in or user-defined
data types.
An SPL routine does not have access to the user state of its execution
sequence. If the routine is going to be called more than once and you want to
retain information about previous executions in the transaction, you will want
to use an SPL routine that states WITH RESUME as a part of the RETURN
statement for multiple executions of the same SPL routine within the same
routine sequence.
When an SPL routine is executed, the parameters (also known as the
dependency list) will be checked. If it is determined that an item in the
dependency list needs reoptimization, optimization occurs at this point. If an
item needed in the execution of the SQL statement is missing (for example, a
column or table has been dropped), an error is returned.
330
7884ch10.fm
UDRs can be overloaded. This means that a function can have more than one
way to operate, depending on the list of data types that are provided as
parameters. There is a precedence hierarchy to decide how the parameter list
is to be executed. This may be important if you have more than one UDR with
the same name, but a different parameter list. See the IBM Informix
User-Defined Routines and Data Types Developer's Guide, Version 11.50,
SC23-9438-00 for more details.
Routine examples in SQL

Using a stored procedure method for a UDR is simple, as long as you can
recognize the incoming and outgoing parameters properly. Using a stored
procedure method for a UDR is simple, as long as you can recognize the
incoming and outgoing parameters properly. The incoming parameters, which
are provided in the function definition, must have their data types defined when
the parameter is first created and declared. The parameter values, which are to
be returned from the function, also should have declared data types. By following
this general rule, you can avoid many of the initial problems you can get with SPL
UDRs.
For the routines with SPL, we use the Informix stores sample database, defined
as stores@demo_on.
One variable in, one result out

For this example, we want a function that provides a count of all the orders
received in a numerical month (N) from the orders table of the stores database.
This example demonstrates what happens when selecting an aggregating value,
once.
Example 10-1 shows the function code and ways to invoke the function to get
output.
Example 10-1 CREATE FUNCTION new_orders (month_num INT)
CREATE FUNCTION new_orders ( month_num INT )
RETURNING INT ;
DEFINE nrows INT;
SELECT COUNT(order_date) INTO nrows FROM orders
WHERE month(order_date)=month_num;
RETURN nrows;
END FUNCTION;
# -- Execute it as a function:
EXECUTE FUNCTION new_orders(5);
(expression)
7
331
7884ch10.fm
# -- Execute it as a select statement.

# -- Note that we have to force first 1 as a criteria so we can assure we are
# -- only getting one value. Otherwise we would get an error.
SELECT FIRST 1 new_orders(5) FROM orders;
(expression)
7
One variable in, several results out

In Example 10-2, we perform three SELECT statements, and each does an
aggregation, returning one value from each select. In this case we return three
values and label each one in the returned result. This shows how to aggregate
three separately selected value returns in one call, and still not use a cursor.
Example 10-2 Selecting three and returning three
RETURNING INT as TotalOrders,INT as ShippedOrders,INT as Backorders;
DEFINE nrows INT;
DEFINE mrows INT;
DEFINE b_rows INT;
SELECT COUNT(ship_date) INTO mrows FROM orders
WHERE month(ship_date)=month_num;
SELECT COUNT(order_date) INTO nrows FROM orders
WHERE MONTH(order_date)=month_num;
SELECT COUNT(order_date) INTO b_rows FROM orders
WHERE MONTH(order_date)=month_num AND ship_date IS NULL;
RETURN nrows,mrows,b_rows;
END FUNCTION;
# -- Since we are returning three values with labels, calling the execute
# -- function is the most appropriate way to make our SQL call.
totalorders shippedorders
7
backorders
1
Using the WITH RESUME clause to return a cursor result

In Example 10-3 on page 333, we collect better details for the new orders. To
return a detailed listing of orders made, we use WITH RESUME in the function.
In this case, we assume (and expect) more than one row is returned, so the
WITH RESUME clause is needed.
332
7884ch10.fm

Example 10-3 A WITH RESUME ROUTINE that uses a cursor
RETURNING INT as Num, INT as ord_num;
DEFINE ord_num INT;
DEFINE Num INT;
LET Num=1;
FOREACH cursor1 FOR
SELECT order_num INTO ord_num FROM orders
WHERE month(order_date)=month_num
RETURN Num, ord_num WITH RESUME;
LET Num=Num+1;
END FOREACH;
END FUNCTION;
------num
ord_num----1
2
3
4
5
6
7
1001
1002
1003
1004
1005
1006
1007
Multi-table select using a cursor

In Example 10-4, we perform a multi-table join with a summary expression return
and a group by. We ask for all the orders placed in a specific month, the name of
the person placing the order, and the total amount for each order.
Example 10-4 A multi-table select using a cursor
RETURNING INT as Ord_Num, char(15) as ord_fname,
char(15) as ord_lname, money(8,2) as Amt;
DEFINE ord_num INT;
DEFINE ord_fname char(15);
DEFINE ord_lname char(15);
DEFINE Amt money(8,2);
FOREACH cursor1 FOR
SELECT o.order_num, c.fname,c.lname,sum(i.quantity*i.total_price)
INTO ord_num, ord_fname, ord_lname, Amt
FROM orders o, customer c, items i
WHERE month(order_date)=month_num
AND o.customer_num=c.customer_num
AND i.order_num=o.order_num
group by o.order_num,c.fname,c.lname
333
7884ch10.fm

RETURN ord_num, ord_fname, ord_lname, Amt WITH RESUME;
END FOREACH;
END FUNCTION;
EXECUTE FUNCTION new_orders(5)
----ord_num ord_fname
1001
1003
1006
1004
1005
1002
1007
Anthony
Anthony
Margaret
George
Jean
Ludwig
Arnold
ord_lname
Higgins
Higgins
Lawson
Watson
Parmelee
Pauli
Sipes
amt
$250.00
$1355.00
$1904.00
$1416.00
$4864.00
$1200.00
$1696.00
10.2.2 User defined routines in Java

It may be helpful to point out that we must distinguish between JDBC and Java
Virtual Machine (JVM) applications. JDBC can be used to write standalone
applications. If you want to connect with other databases that support Java, you
will typically write standalone Java (JDBC) applications since these applications
require specific driver methods to communicate with other database servers.
When you write a Java UDR, you must use the IBM Informix JDBC Driver, which
is based on the JDBC 2.0 API. The generated code is processed by a Java
Virtual Machine that runs as a process inside the IBM Informix Server. The
generated code (jar file) is stored in an sbspace, and may also refer to a jar file in
a storage location outside of the server
Java allows you to create user-defined routines, cast support functions,
aggregates, and opaque data type support routines. However, Java routines
cannot handle row or collection data types.
For ordinary UDR with IBM Informix database servers, you can use the Java
Developers Kit (version 1.5 at this writing). A Java Virtual Machine comes
pre-installed with the IBM Informix database server (with J/Foundation). To
confirm it is present, simply make sure you have an existing directory path to
INFORMIXDIR/extend/kraratoa.
Java based UDR, once developed, is placed into a Java archive (jar) file. The jar
file is stored inside an sbspace, or it may have additional supporting files on the
file system. If the jar file is very large, or perhaps proprietary, you can leave it on
the file system. Smaller jar files, or something you may want to update frequently,
should be stored in the sbspace.
334
7884ch10.fm
Configuration
Make sure you use the JVM system that came with your server engine, and
make sure the onconfig and environment variables are set to accurate working
paths. The environment settings for our testing setup had the following variables
in the environment:
JRE_HOME=/usr/lib/jvm/java/jre
JAVA_BINDIR=/usr/lib/jvm/java/bin
JAVA_HOME=/usr/lib/jvm/java
SDK_HOME=/usr/lib/jvm/java
JJDK_HOME=/usr/lib/jvm/java
JAVA_ROOT=/usr/lib/jvm/java
The JAVA_ROOT environment variable is mostly determined by the developer. It is

the path to the file system directory where you are developing your jar files:
CLASSPATH=location_for_your_java_classes:.:.:.
The INFORMIXDIR/etc/ONCONFIG file also has a small group of parameters that

must have verified settings. The following are standard:
JVPJAVAHOME
JVPHOME
JVPPROPFILE
JVPJAVALIB
JVPJAVAVM
$INFORMIXDIR/extend/krakatoa/jre
$INFORMIXDIR/extend/krakatoa
$INFORMIXDIR/extend/krakatoa/.jvpprops
/bin
jvm
The JVPJAVALIB and JVPJAVAVM onconfig parameters are important for

development. This one specifies where Java problems are written to (make sure
the path exists):
JVPLOGFILE
$INFORMIXDIR/jvp.log
If you are going to use external jar files, you must add them to JVPCLASSPATH:
JVPCLASSPATH $INFORMIXDIR/extend/krakatoa/krakatoa.jar:$INFORMIXDIR/extend/krakatoa/jdbc.jar
Note: Jar paths added to JVPCLASSPATH are only visible after being added
to the ONCONFIG file, and the database engine must be restarted.
Routine examples in Java

In the examples which follow, we create our Java routines in /work/, which
defines the UNIX directory we use for our working CLASSPATH. You must
change the directory path in some of the code expressions if you use a different
directory. Once your code is compiled, the source code is not used for execution,
but you should keep it in a safe place in case you want to improve it later.
335
7884ch10.fm
A function extension
This example illustrates a simple way to extend functionality. This routine
provides an SQL function to multiply a value times ten. The developing
procedure is as follows:
1. Create a simple class file, named Times.java in the working JVPCLASSPATH
directory. Example 10-5 shows the source code.
Example 10-5 A java function to multiply a value times ten
/*Times.java */
public class Times {
public static int TimesTen(int x) {
return x * 10;
}
}
2. At a command prompt, compile the file using the following command:

javac Times.java
3. Compress it into a jar file:

jar cvf Times.jar Times.class
As this runs, it shows output similar to the following:

added manifest
adding: Times.class(in = 245) (out= 187)(deflated 23%)
4. Register the routine:

Start dbaccess to connect to the database and run the following SQL to
register and add the jar file into our sbspace:
execute procedure sqlj.install_jar ("file://work/Times.jar" , "Times_jar");
Routine executed.
5. With the jar file in storage and accessible, from dbaccess, we create a function
that calls our routine in the jar file. See Example 10-6.
Example 10-6 Calling routine
create function times_ten(value int) returning int
with (class = "jvp")
external name "Times_jar:Times.TimesTen"
language JAVA;
6. The procedure is ready to run. We test it in dbaccess:

EXECUTE function times_ten(13)
(expression)
130
336
7884ch10.fm
For an alternate method, try:

SELECT ship_charge, times_ten(ship_charge) X10 FROM orders;
A Java user-defined routine with two input parameters

Here we demonstrate how to create a Java routine with multiple input
parameters. Java language and IBM Informix servers express decimals in
different ways. In this example, we create a new sales tax function Salestax that
includes a new Java function and the existing Times_ten function created in A
function extension on page 336. We need to include an external standard Java
library that has a math class. You might find that multiplying decimal values in
Java is different than what you may be used to. In addition, Java uses a different
naming convention than Informix data types. If you get the class wrong or the
library is incorrect, things simply do not work and you will get errors.
We use the following procedure to create the salestax function:
1. Create a Java class called Tax.java, using an editor such as vi.
Example 10-7 shows the source code.
Example 10-7 Tax.java
/*Tax.java*/
import java.math.BigDecimal;
public class Tax {
public static BigDecimal salestax(BigDecimal x,BigDecimal xtax )
{
BigDecimal ratePlusOne = xtax.add(BigDecimal.valueOf(1));
BigDecimal afterTax = x.multiply(ratePlusOne);
afterTax = afterTax.setScale(2, BigDecimal.ROUND_HALF_UP);
return (afterTax);
}
}
2. Compile the file. At a command prompt, run the following command:

javac Tax.java
3. Compress it and add it to our existing jar file Times.jar.

jar cvf Times.jar Tax.class
As this runs it shows a line similar to the following:

added manifest
adding: Tax.class(in = 546) (out= 318)(deflated 41%)
4. Add the updated jar file to sbspace.

For this example we want to update an existing function that already exists in
our sbspace jar. Since we can not replace an existing function in a sbspace
jar directly, in order to update an sbspace jar function, we must replace the
337
7884ch10.fm
entire jar. To do this, we must remove (drop) the sbspace jar file, then replace
it with an updated version from our working path.
A jar file in an sbspace must be empty, in order to drop the jar. Drop all the
UDRs in the jar to empty the jar. Otherwise, you will receive the error
message Invalid jar removal. All dependent UDRs not dropped.
Our Times_jar file exists in the database. To drop the function, we start
dbaccess, connect to the database, then run the following command:
DROP FUNCTION times_ten;
Now, we can remove the jar:

EXECUTE PROCEDURE sqlj.remove_jar ("Times_jar");
5. Add the updated jar file into sbspace. Run the following SQL,:
EXECUTE PROCEDURE sqlj.install_jar ("file://work/Times.jar" , "Times_jar");
6. The updated jar is now in sbspace storage. Recreate the dropped Times_ten
function:
create function times_ten(value int) returning int
external name "Times_jar:Times.TimesTen"
language JAVA;
7. Add the new function to call on the same jar file:

create function salestax(value decimal,xtax decimal)
returning decimal(8,2)
external name "Times_jar:Tax.salestax"
language JAVA;
8. Now we can test the expanded function from dbaccess:

EXECUTE FUNCTION salestax(250.00,.065);
(expression)
266.25
As an alternative test using SQL, try:

SELECT o.order_num,
salestax (sum(i.quantity*i.total_price), .065) Amt_w_Tax
FROM orders o, items i WHERE month(order_date)=5
AND i.order_num=o.order_num group by o.order_num
Creating a routine that uses external Java APIs

In this example, we demonstrate how to create a Java UDR that requires support
from one or more Java APIs which lie outside the database server. The import
references in the code indicate that an external Java API exists to help support
the UDR. The references on the import list remain outside the database server.
338
7884ch10.fm
The Java Virtual Machine runs using the jar calls in the sbspace storage location
and sends calls to the external Java functions based on the import reference in
the jar file.
Note that the import reference does not have a full directory path. Any Java API
or jar file that is not residing in an sbspace is external to the instance. In order to
help the server instance find the external jar files, you must supply all Jar API
path locations in the onconfig file.
Note: The onconfig file must be updated so the JVM knows the directory path
for any supporting APIs. The full path location for the supporting Java file
specified on an import list must be included in JVPCLASSPATH.
For our example, JVPCLASSPATH is set to:
/usr3/11.50/extend/krakatoa/krakatoa_g.jar:/usr3/11.50/extend/krakatoa/jdbc_g.j
ar:/usr3/11.50/extend/krakatoa/jre/lib/rt.jar:/work/mailapi.jar:/work/activatio
n.jar:/work/smtp.jar
For our example to work, the engine needs the Java mail API classes. The files
required are mailapi.jar, activation.jar, and smtp.jar from the Java website
at
http://java.sun.com/products/javamail/downloads/index.html
We use these steps to create the sendmail routine:

1. Install the downloaded jar files and add jar files with full path to
JVPCLASSPATH.
In this example, the paths are /work/mailapi.jar, /work/activation.jar,
and /work/smtp.jar.
2. Create the file MailClient.java as shown in Example 10-8. You have to
update the italicized references in the example with your own information.
Example 10-8 A SQL based sendmail() UDR
---- MailClient.java --import javax.mail.*;
import javax.mail.internet.*;
import java.io.*;
import java.util.Properties;
public class MailClient
{
public static void send(String to, String text)
{
try
{
MailClient client = new MailClient();
339
7884ch10.fm

Properties props = System.getProperties();
props.put("mail.smtp.host", "smtp.server.com");
Session session = Session.getDefaultInstance(props, null);
Message message = new MimeMessage(session);
message.setFrom(new InternetAddress("Name_showing@fromfield"));
message.addRecipient(Message.RecipientType.TO, new InternetAddress(to));
message.setSubject("Message from the database");
message.setText(text);
Transport.send(message);
}
catch(Exception e)
{
e.printStackTrace(System.out);
}
}
}
---- MailClient.java ---
3. Compile and compress your file:

javac MailClient.java
jar cvf MailClient.jar MailClient.class
4. Install the jar into the sbspace:

EXECUTE PROCEDURE sqlj.install_jar ("file://work/MailClient.jar" ,
"MailClient_jar");
Make sure JVPCLASSPATH is set as indicated in step 1.

5. Create the sendmail procedure from dbaccess:
CREATE PROCEDURE sendemail(to LVARCHAR, message LVARCHAR)
WITH (class = "jvp")
EXTERNAL NAME "MailClient_jar:MailClient.send"
LANGUAGE JAVA;
6. Test the procedure:

EXECUTE PROCEDURE sendemail('dba@mybiz.com','Error deleting from table');
Troubleshooting tips
Sometimes you may have trouble getting a Java UDR up and running. There are
three points of exposure for errors:
At the time of a Java compile.
If you get an error here, the issue relates to a Java language problem, most
likely resulting from syntax or a Java method. Consult a Java Programming
Language Guide for assistance.
340
7884ch10.fm
At the jar install point or later.

Check the jvp.log at the path in the onconfig file specified by the JVPLOG
parameter. You should see no errors at the time of install, and no errors at
runtime. When you start the server instance the JVM starts, and jar files
should be loaded into the process memory as needed. The MSGPATH file
(online.log) will often reveal the success or failure of JVM and jar loading. If
the files cannot load, they cannot run.
UDR runtiime errors
If there are errors, study the Java error messages returned, and determine
the cause for whatever did not resolve.
Troubleshooting can be a trial and error approach. Once you have an
understanding of how the provided UDR examples work, do them over again,
and break something in the Java code. Go through the example with slightly
broken syntax and review the results. For example, In the BigDecimal code
example, change BigDecimal to Float or Double, and work through the example
again.
10.2.3 User defined routines in C

C allows you to create user-defined routines, cast support functions, aggregates,
and opaque data type support routines. C can also handle row and collection
data types. Working with C UDR adds an extra layer of difficulty since these
routines have to be compiled using a compiler, specific to the machine and
operating system where you have products installed. Compiled C is not stored in
sbspace, but rather, it is stored outside the server.
In addition to compiler difficulties, it is a good idea to develop and test UDRs on a
development server and not in a production environment. C UDR code runs as a
database server process that works closely with internal structures. The routine
should not do anything that would negatively affect the database server. A poorly
designed C UDR is likely to crash the server. If you plan to work with C UDRs,
there are three good reference books:
IBM Informix User-Defined Routines and Data Types Developer's Guide,
Version 11.50, SC23-9438-00
IBM Informix DataBlade API Programmers Guide, Version 11.50,
SC23-9429-03
IBM Informix DataBlade API Function Reference, Version 11.50,
SC23-9428-01
341
7884ch10.fm
If you are interested in using C++ Datablade modules, the IBM Informix
Developer Zone that provides the latest recommendations on C++ programming
options is at this website:
http://www.ibm.com/software/data/developer/informix
Routine examples in C
In this section, we undertake a few examples written in C. With Java we have
portability that extend across platforms, but some of our functionality is limited. C
is flexible since it can extend the database by way of data types, new functions,
and new operators. For each database across an enterprise, the extensibility that
you provide through C must be specifically compiled with each operating system.
For the examples in this section, we use a Solaris machine.
A simple function using C

For our first example, we start with a function similar to what we did in our Java
example, a multiplying function. For each of our C examples, we must include the
DataBlade API files (referenced in the example as mi.h and milib.h). The
DataBlade APIs provide the interface to allow C language calls to interface with
IBM Informix database SQL calls. To prevent a conflict with our times_ten Java
UDR function, we name this example times_five. We use the following steps to
create this routine:
1. Create the text file that holds the C code. Example 10-9 shows the source file
times.c.
Example 10-9 A C UDR which multiplies times 5
#include <stdlib.h>
#include "mi.h"
#include "milib.h"
mi_integer* TimesFive(mi_integer value);
mi_integer* TimesFive(mi_integer value)
{
return (mi_integer *) (value * 5);
}
2. Compile and link the C code.

The compile command (cc or gcc) depends on the operating system. The first
way to identify your compiler is to examine the text of man cc (on UNIX), or
check the command for your operating system compiler in the
INFORMIXDIR/release/en_us/0333/ids_machine_notes_vers.txt file.
At a minimum, your compile line for UDR preparation usually includes:
-DMI_SERVBUILD -KPIC -I$INFORMIXDIR/incl/public -I/$INFORMIXDIR/incl
342
7884ch10.fm
Where
-DMI_SERVBUILD is the flag to indicate that this is a server-oriented C UDR
application which uses the DataBlade API (required).
-KPIC is the flag to indicate that the symbol table is dynamic (for UNIX and
Linux only).
-I$INFORMIXDIR/incl/public -I/$INFORMIXDIR/incl are the location of
the mi_ libs.
The compile command for our code sample (on Solaris) is:
cc -DMI_SERVBUILD -KPIC -I$INFORMIXDIR/incl/public -I/$INFORMIXDIR/incl -o
times.o -c times.c
Run the link command to put the compiled object file into our Blade library file:
ld -dy -G -Bsymbolic -o times.bld times.o
3. Set the bld file permissions as user informix:

chmod 555 ./times.bld
4. In dbaccess, after connecting to the database, create the times_five

function:
CREATE FUNCTION times_five(value int) RETURNING int
WITH (handlesnulls)
EXTERNAL NAME "/work/times.bld(TimesFive)"
LANGUAGE C;
Adjust the external work path as needed.

5. Test the routine:
execute function times_five(20);
(expression)
100
In Java UDR, the jar file serves as a library (collection-repository) for all of the
compiled routines. In C, a collection of compiled routines is stored in a shared
library (.so or .o) file. On Windows, a shared object file has a .dll extension
(dynamic link library).
Creating a C routine using large object column

In our next example we explore the use of working with a large object column
reference. Without a text search DataBlade, searching a character large object
(CLOB) file for a particular value can be a laborious SQL task. This example
shows how to access a CLOB file, copy the CLOB contents into an LVARCHAR,
and search for a specific text item, while using a simple function. We create a C
file that handles two parameters. The first parameter is a CLOB column
343
7884ch10.fm
reference, the second parameter is the LVARCHAR text value for our search. It
returns a count for the number of successful finds.
We use the following procedure to implement our search routine:
1. Create the CLOB column search routine source file un.c as shown in
Example 10-10.
Example 10-10 A UDR for searching a CLOB
#include
#include
#include
#include
#include
#include
<ifxgls.h>
<mi.h>
<milib.h>
<fcntl.h>
<stdio.h>
<stdlib.h>
typedef unsigned char byte;

mi_integer contains(MI_LO_HANDLE* loptr, mi_lvarchar* pattern, MI_FPARAM* fp);
mi_integer contains(MI_LO_HANDLE* loptr, mi_lvarchar* value, MI_FPARAM* fp)
{
mi_unsigned_integer crc = (mi_fp_argisnull(fp, 1) == MI_TRUE) ? 0 :
(mi_unsigned_integer)1;
MI_CONNECTION *conn;
MI_LO_SPEC *lo_spec = NULL;
MI_LO_FD lo_fd;
MI_LO_STAT *lo_stat = NULL;
char
*buff = NULL;
char
*pattern = NULL;
mi_integer buffsize = 4096;
mi_integer found=0;
mi_integer result=0;
pattern = mi_lvarchar_to_string(value);
if ((conn = mi_open(NULL, NULL, NULL)) == NULL) return (mi_integer) -1;
if ((buff = (char*)mi_alloc(buffsize)) == NULL) return (mi_integer) -2;
if ((lo_fd = mi_lo_open(conn, loptr, MI_LO_RDONLY)) == MI_ERROR)
return (mi_integer) -3;
do {
if ((result = mi_lo_read(conn, lo_fd, buff, buffsize)) == MI_ERROR)
break;
if (result == 0)
break;
if (strstr(buff,pattern)!=NULL)
{
found=1;
break;
}
if (result < buffsize)
break;
} while(1);
mi_lo_close(conn,lo_fd);
344
7884ch10.fm

if (buff)
mi_free(buff);
return (mi_integer)found;
}
2. Compile and link the C routine. We use the following Solaris compile line:
cc -DMI_SERVBUILD -KPIC -I$INFORMIXDIR/incl/public
-I$INFORMIXDIR/incl/public -I$INFORMIXDIR/incl/esql -I/$INFORMIXDIR/incl -o
un.o -c un.c
Here is the link line:

ld -dy -G -Bsymbolic -o un.bld un.o
3. Set the bld file permissions as user informix:

chmod 555 ./un.bld
4. Create the function in dbaccess.

create function contains(clob,lvarchar)
returns integer
external name '/work/un.bld(contains)'
language C;
5. Set up for testing. Create a table and populate it with our c file.
CREATE TABLE tclob (c1 INT, c2 CLOB);
INSERT INTO tclob VALUES (1,filetoclob('un.c','server'));
6. Test the routine:

SELECT
SELECT
SELECT
SELECT
c2 FROM tclob;
CONTAINS(c2,'pattern') FROM tclob;
CONTAINS(c2,'nopattern') FROM tclob;
c1 FROM tclob WHERE CONTAINS(c2,'buff')=1;
Troubleshooting tips
In order to track down the cause of problems with C UDRs, The most affective
approach is to use a debugger. To debug your UDR, use a debugger that can
attach to the active server process and access the symbol tables of the
dynamically loaded shared object files. On UNIX and Linux, the debugger and
dbx utilities meet these criteria. To start a debugger, enter the following command
at the shell prompt, in which pid is the process identifier of the CPU or virtual
processor:
debugger - pid
This command starts the debugger on the server virtual-processor process

without starting a new instance of the virtual processor. For more information
about available debugger commands, see the debugger manual page, and learn
345
7884ch10.fm
more at the Information Center web page that describes Debugging a UDR
located at:
dapip.doc/sii111026637.htm
10.3 DataBlades and bladelets

Your initial collection of UDRs will increase over time. It may not be long until you
have UDRs, special functions, and stored procedure UDRs that reduce work
complexity and provide great functionality. If you choose not to write your own
DataBlade, you can still choose from a nice selection of BataBlade modules. Any
DataBlade module you use will provide functions that can dramatically increase
the usefulness of business data, or perhaps generate income for you -- as a
software developer that develops and sells licensed DataBlades. IBM Informix
has several books on the subject of planning, designing, and implementing
DataBlades. With a taste of the programming examples presented earlier in this
chapter, you may be ready to take the next step and program a DataBlade in C or
Java. From a technical perspective, the IBM Informix DataBlade Developers Kit
User's Guide, Version 4.20, G229-6366-01 is specifically oriented to the
development of DataBlade modules. It provides detailed help for programming
DataBlade modules in Java and C.
If you would like to work with a DataBlade before deciding on whether you have a
need to write your own, in the sections that follow we will take a tour through the
DataBlades available with IBM Informix Server.
10.3.1 Configuration
Configuration for a Datablade module only requires three steps:
Prepare the database serve.r
Install the DataBlade.
Register the DataBlade.
Conceptually, the process is the same, regardless of the operating system and
hardware. Small differences exist in interface or command lines, which we will
point out.
Database preparation
Database preparation involves setting up environment variables, and making
sure the database you are going to use with a DataBlade is set to a logged
database in advance.
346
7884ch10.fm
The environment settings required are

On UNIX and Linux: LD_LIBRARY_PATH, INFORMIXSERVER, and ONCONFIG
On Windows: INFORMIXSERVER and ONCONFIG
When you get to the step for BladeManager usage, you also require an
environment setting for LD_LIBRARY_PATH.
Though a logged database is not required for every DataBlades, a logged
database can help avoid concurrency problems and thus is recommended. Also,
while it is the case that not all DataBlade modules require a logged database,
some of them do require logged sbspace. Keep this in mind as you install and
configure the DataBlade you select.
DataBlade installation
Any DataBlades installed with IBM Informix Servers are installed in separate
subdirectories under INFORMIXDIR/extend directory. Several subdirectories for
DataBlades are established when the IBM Informix Server is installed. There
may be some variation to the list, based on your exact version and operating
system. Example 10-11 shows the subdirectory listing for 11.50.UC6 on Linux
with J/Foundation.
Example 10-11 A sample INFORMIXDIR/extend subdirectory
opt/IBM/informix/extend:> ls
binaryudt.1.0
bts.2.00
ifxbuiltins.1.1
ifxmngr
LLD.1.20.UC2
ifxrltree.2.00
mqblade.2.0
krakatoa
Node.2.0
spatial.8.21.UC3
web.4.13.UC4
wfs.1.00.UC1
If you do not see the DataBlade directory reference for the one you want, you
must acquire it by way of a download or CD.
Installation on UNIX is simply a matter of uncompressing the new DataBlade
module into a temporary directory. Once the files are expanded, run the install
script, ./install. The install script creates a new module directory under the
INFORMIXDIR/extend directory. Some file expansions may result in more than
one new module. If there is more than one module, you must run the ./install
for each one.
Installation on Windows requires that you go to the install location and run
setup.exe. Select Typical installation option. The other dialog verification is the
INFORMIXDIR location. With those items confirmed, the software is installed.
When the installation is complete, the module directory is installed under
INFORMIXDIR\extend directory.
347
7884ch10.fm
Datablade registration
With a DataBlade directory in place, the active database server is not aware of
the DataBlade directories until the software is registered in the database.
Registration is the process of executing the SQL statements that create the
DataBlade module database objects and identify the DataBlade module shared
object file or dynamic link library to the database server.
With the release of 11.50.XC4, IBM Informix began providing two distinct
methods for DataBlade registration:
BladeManager: This method has been around since the beginning of Informix
datablades. The BladeManager is an interface that automates the registration
process, by performing a series of SQL steps in the database engine.
sysbldprepare(): The is an Informix function for DataBlade registration. At its
simplest, you can run the command inside dbaccess, with your target
database open, to install a DataBlade. It has a few usage restrictions which
you can read about in the IBM Informix DataBlade Module Installation and
Registration Guide, Version 4.20, G229-6368-03. Here is an example of how
the bts DataBlade is registered with this interface:
EXECUTE FUNCTION sysbldprepare('bts.*','create');
The following are steps to register a DataBlade module using BladeManger:

1. Start BladeManager:
On UNIX, or at the MS-DOS prompt, the BladeManager is started with the
blademgr command.
To start BladeManager on Windows, select Start Programs
Informix program group name BladeManager or double-click the
BladeManager icon in the Informix program group.
If the BladeManager fails to start, it is either not installed or you do not have
the environment variables set, as noted in the previous section.
2. Set confirmations.
If you would like an auto confirmation after each step, turn the prompt on:
set confirm on
Commands run when you press the carriage return.

3. Connect to an Informix instance:
show servers
set server demo_on < Use you own server name for demo_on>
If you want to connect as a different user, try:

set user <username>
348
7884ch10.fm
At the password prompt, enter your password. Validation does not occur until
connection, on the next step.
To connect to a database, run one of the following commands:
list stores
register module_name database_name
unregister module_name database_name
The module_name represents the name of the DataBlade module directory.

These names typically follow the form of the DataBlade module name
followed by the version number.
When BladeManager registers a DataBlade module, it executes a series of SQL
CREATE statements to register each database object in the module. You must
have resource permissions on the database to register the DataBlade. In
addition, if your server has implemented the ONCONFIG EXTEND role, you
must be granted the EXTEND role by user informix.
If the registration of a module fails, BladeManager returns the database to its
prior state. To see the SQL statements that failed, look at the corresponding log
file and check the procedure in Appendix A. Troubleshooting Registration
Problems of the IBM Informix DataBlade Module Installation and Registration
Guide, Version 4.20, G229-6368-03, for possible solutions.
Occasionally, DataBlade modules have more than one interface. If there are
additional modules, there are also dependencies. You have to make sure that
each of the interfaces are registered correctly in order for the DataBlade to work.
BladeManager automatically checks for dependencies and registers any
dependencies it might need. If the BladeManager cannot do the registration, it
will prompt you to do so manually.
Important: BladeManager does not verify the integrity of DataBlade modules
that have additional interfaces, nor does it not check for the presence of
required database objects.
Datablade modules written in the Java language can only be registered in IBM
Informix Servers with J/Foundation database servers.
10.3.2 IBM Informix provided DataBlades

Table 10-2 on page 350 the DataBlades that are available with IBM Informix
database servers.
349
7884ch10.fm

Table 10-2 Informix provided DataBlades
IBM Informix
DataBlade
module name
350
Description
Special notes
Large Object
Locator
A foundation Datablade module for large

objects management that can be used by
other modules that create or store
large-object data.
Available on
standard install
(LLD.1.20.UC2)
MQ DataBlade
Allows IBM Informix database

applications to communicate with other
MQSeries applications with MQ
messaging.
Available on
standard install
(mqblade.2.0)
Binary DataBlade
This module includes binary data types to

store binary-encoded strings that can be
indexed for quick retrieval.
Available on
standard install.
(binaryudt.1.0)
Basic Text Search
Permits text search of words and phrases

in an unstructured document repository
stored in a column of a table.
Available on
standard install.
(bts.2.00)
Node DataBlade
This module is for the hierarchical data

type, to represent hierarchical data within
a relational database.
Available on
standard install.
(Node.2.0)
Web Feature
Service
This module is an add-on to allow Open

Geospatial Consortium (OGC) web
feature service as a presentation layer for
the Spatial and Geodetic DataBlade
modules
Available on
standard install.
(wfs.1.00.UC1)
J/Foundation
krakatoa
A library of classes and interfaces that

allow programmers to create and execute
Java user-defined routines that access
Informix database servers
Available as a part
of Informix Server
with J/Foundation
ifxbuilt-ins
This is not really a DataBlade, but it sets

up definitions and functions for the
standard data types offered in the informix
server.
Available on
standard install
(ifxbuiltins.1.1)
ifxmngr.2.00
This is the API that supports the

BladeManager.
Available on
standard UNIX
install
7884ch10.fm
IBM Informix
DataBlade
module name
Description
Special notes
ifxrltree.2.00
This is a foundational, multidimensional,

index called Region tree (R-tree) (also
known as Range Tree). This blade is
needed for both spatial and time related
data management.
Available on
standard install
Image Foundation
This module is a foundation DataBlade

which provides a base on which new or
specialized image types and image
processing technologies can be added or
changed. Because the foundation is open,
secure, and scalable, it provides a clear
path toward reusing and repurposing
valuable image assets.
No charge
download
Excalibur Text
Search
This module enables provides extensive

text-searching capabilities; It supports
full-text indexing, including extensive
fuzzy-search logic for indexing scanned
text. Can search document types
including: ASCII, Word, Excel, HTML,
PowerPoint, WordPerfect, and PDF.
Includes an adaptive pattern recognition
process (APRP) and capabilities such as
multiple stop-word lists, proximity
searching and synonym lists.
License fee applies
Geodetic
This blade supports global space- and

time based queries. It is designed to treat
earth as a globe rather than a flat plane.
Supports client-side Geographic
Information Systems (GIS) software.
License fee applies
Spatial
This blade transforms locations and

traditional 2-d map data into useful
information. Uses SQL-based spatial data
types and functions that can be used
directly through standard SQL queries or
with client-side Geographic Information
Systems (GIS) software.
No charge
download
351
7884ch10.fm
IBM Informix
DataBlade
module name
Description
Special notes
TimeSeries
This module supports data for managing

time-series and temporal data. A time
series is any set of data that is accessed
in sequence by time and can be
processed and analyzed in a
chronological order.
License fee
applies.
Video Foundation
This blade allows you to incorporate video

servers, external control devices,
compression codes, and cataloging tools
to manage video content and metadata or
information about the content. Allows
metadata elements in the database, while
allowing video content to be maintained
on disk, video tape, video server, or other
external storage devices.
License fee
applies.
Web
This module supports most Web server

APIs and has a web client application to
build and run SQL queries to work with
your database. Enables customized Web
applications. Allows you to track
persistent session variables between
AppPages.
License fee applies
10.3.3 Developing a bladelet routine

We can define a bladelet as a small, unofficial DataBlade module. It is meant to
be useful (and complete with source code), from the time you set it up, but it
becomes your own application (with no support or warranty). If you have tried out
our UDR development examples, you have a bladelet.
As you may have observed from having to drop and recreate the jar API in the
server in our earlier example, you can understand that if you have a large number
of UDRs and have gone to the trouble of creating user defined data types, the
whole package of tasks required to set up, change or update a DataBlade object
inside the server may not be a convenient task.
On a large scale, if you have API dependencies, dozens of UDRs, and other
DataBlade related objects, you will want to move them and install them as a
package. IBM Informix provides a Windows-based interface for this, which has
the ability to do the package preparation work for you. The package preparation
352
7884ch10.fm
interface, called BladeSmith, allows you to populate a properties definition

dictionary.
When the properties are all defined and you proceed, BladeSmith creates,
assembles, and arranges a directory structure with a complete tree layout of all
the components that are required and handled in a DataBlade registration
process. The resulting directory tree layout and assembly pieces provide a
prepared package that is ready to ship to an operating system of your choice.
Likewise, any language source code that is output for the preparation task is
parcelled out for the appropriate component nodes also.
The dispensation for the BlabdeSmith file components and directory structure is
laid out based on the type of component node, described in Table 10-3.
Table 10-3 Bladesmith file package creation layout
Component node
What is generated
Source
All source code in the coding languages you use for your
DataBlade module objects
Client
Client code (activeX or Java)
Server
Server code in the coding language you specified for BladeSmith
Individual language
Source code for the represented language (C, Java, or SPL)
For further information on creating datablade objects using BladeSmith

generating files, visit the IBM Information Center website under the Datablade
Developers Kit Users Guide Section at
dbdk.doc/sii-smith-27272.htm
For further information or examples of existing bladelets and downloads

developed by users, check out the article Downloadable Bladelets and Demos at
http://www.ibm.com/developerworks/data/zones/informix/library/samples/db_downlo
ads.html
There is also a downloads page featuring bladelets at the International Informix

Users Group site, called ORDBMS - Object-Relational Database Extensibility,
DataBlades at
http://www.iiug.org/software/index_ORDBMS.html
353
7884ch10.fm
354
7884ch11.fm
11
Chapter 11.
Working with Ruby on Rails

This chapter contains information about how to develop applications against an
IBM Informix database using the Ruby programming language and the Rails web
development framework.
In this chapter, we discuss the following topics:
Introduction to Ruby and Ruby on Rails

Installing Ruby on Rails
Options available to use an IBM Informix database with Ruby
How to perform basic database operations
Use of specific Informix data types
Using Rails with an Informix database.
355
7884ch11.fm
11.1 A brief overview of Ruby on Rails

Ruby is a open source programming scripting language with a focus on simplicity
and productivity. Ruby is similar to other scripting languages like Perl or Python
with the difference that Ruby is an object-oriented language. Ruby is both
technology and platform independent. You can find implementations of the Ruby
runtime on C, Java, and even .NET. This makes Ruby a really useful option for
any scripting needs.
Refer to the Ruby website for more information about the Ruby language
http://www.ruby-lang.org/en/about/
Rails is an open source Ruby framework for developing database-backed web

applications. Rails explodes the object-orientated core design of Ruby helping
developers to build websites and applications with minimum coding efforts.
Rails is based on two key principles:
Convention over Configuration (CoC) where developers need to focus only on
the exceptions to the conventions. Every other aspect of the application, from
design to implementation, is automatically done by the defined conventions.
The conventions define rules such as use a plural for the table names. For
example, if the application uses an entity call Book, the table that stores this
entity must be called Books, if it stores details about a Person, the table must
be called People.
Dont Repeat Yourself (DRY). Information is located in one place only.
Database object definitions, documentation, configuration scripts all are kept
in one global place and consulted when information about them is required.
Refer to the Ruby on Rails website for more information at
http://rubyonrails.org/documentation
11.1.1 Architecture of Ruby on Rails

Ruby on Rails is build with the Model-View-Controller (MVC) architecture that is
typically used in web based GUI programming. This architecture has three main
concepts:
Model: This is the business logic of the system. This also encompasses the
persistence layer because it interacts with a database back end.
View: This is the GUI interface of the Model to the end user. There can be
many Views for one Model.
356
7884ch11.fm
Controller: This is the action taken by the user using the View. It takes inputs
from the end user through the View and executes the business logic
encapsulated in the Model.
The Rails framework provides a set of utilities and components designed to
facilitate the development of web applications:
Rake:
Rake is a build tool bundled with the Ruby programming language. It is the
equivalent to the Make on UNIX.
WEBrick:
WEBrick is the web server bundled along with Ruby on Rails.
Active Record:
ActiveRecord is the object-relational mapper of Rails and provides for
persistence. It presents the database table as a class, which in Rails is called
model.
Action Controller
Action Controller is the component that manages the controllers in a Rails
application. It also processes and dispatches incoming requests.
Action View
Action View manages the views in a Rails application.
11.1.2 Ruby driver and Rails adapter

IBM Informix supports database access for client applications written in the Ruby
programming language and web application development with the Rail
framework.
Ruby driver
To use an database with Ruby, the application requires a Ruby driver. This driver
provides the layer that connect the Ruby runtime with the database server.
There are two drivers which allows Ruby to connect to an IBM Informix database:
Ruby Informix driver:
The Ruby Informix driver is an Open Source project supported by the Open
Source community. It allows Ruby to connect to any IBM Informix database
server. The Ruby driver is developed using the IBM Informix ESQL/C
language that provides full support for all the Informix database features and
data types. Because it uses the Informix Client SDK libraries,
Ruby Driver for IBM Data Servers driver (IBM_DB):
Chapter 11. Working with Ruby on Rails
357
7884ch11.fm
The Data Server Ruby driver, or IBM_DB, is provided, supported, and

developed by IBM as an Open Source project. The Ruby driver is bundled
together with the Rails Adapter in the Rails Adapter/Driver for IBM Data
Servers package.
Rails adapter
A Rails adapter is a Ruby script that allows you to use a specific Ruby driver
within the Rails framework. It provides the required Ruby objects, for example the
ActiveRecord object, that enables the full use of the Ruby driver inside the Rails
framework.
IBM Informix supports two Rails adapters:
Informix_adapter.rb, used in conjunction with the Ruby Informix driver.
Requires the IBM Informix Client SDK libraries for the communication with the
ibm_db_adapter.rb, used with the Ruby Driver for IBM Data Servers. Requires
the IBM Data Server Driver for ODBC and CLI package.
Both adapters are available from the Ruby repository as Ruby gems. Ruby Gems
are self-contained packages containing all the libraries, source files, and scripts
needed for the Ruby component.

This section describes how to set up and configure both Ruby drivers and Rails
adapters for use with an IBM Informix database.
11.2.1 Ruby Informix driver

The Ruby Informix driver is available for download at the SourceForge website at:
http://rubyforge.org/projects/ruby-informix
You can use the Ruby utility Gem to download the Ruby Informix driver
automatically from the Ruby repository and install it in the Ruby environment.
Ruby Gems are self-contained packages with the libraries, source code, and
scripts for Ruby libraries.
Run the following command from your Ruby session to install the Ruby Informix
driver:
gem install ruby-informix
358
7884ch11.fm
Because the driver shared library, informixc.so, is built during the installation
process, the environment should contain the correct settings for compiling
ESQL/C applications. Refer to chapter Chapter 4, Working with ESQL/C on
page 121 for ESQL/C the setting details.
You can find more information about the install process in the README file
inside the Gem directory. See Example 11-1 for the list of files included with the
Ruby Informix driver.
Example 11-1 Gem directory
Directory of C:\work\Ruby187\lib\ruby\gems\1.8\gems\ruby-informix-0.7.3
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
19:34
<DIR>
.
19:34
<DIR>
..
19:19
8,383 Changelog
19:19
1,470 COPYRIGHT
19:29
<DIR>
ext
19:35
<DIR>
lib
19:19
4,500 README
19:19
<DIR>
test
3 File(s)
14,353 bytes
5 Dir(s) 76,168,769,536 bytes free
Configuration
The Ruby Informix uses the same connectivity information as other Informix
Client SDK components. It uses the environment variable INFORMIXDIR to locate
the libraries and resources such as error message files or configuration files.
By default, the Ruby driver connects to the database server specified in the
INFORMIXSERVER environment variable. Same as the other Informix Client SDK
components, the information regarding the INFORMIXSERVER value is stored on the
sqlhosts file or the Windows registry. For more information, refer to 2.2.2, Setting
up IBM Informix Client Software Development Kit on page 33.
The shared library search path variables, for example, LD_LIBRARY_PATH or
SHLIB_PATH, must contain the $INFORMIX/libl and $INFORMIX/lib/esql
directories, otherwise the Ruby driver may fail to load the ESQL/C libraries that it
requires for work.
Data types
The Ruby Informix driver provides the data types to be used against an IBM
Informix database. The driver provides specific types such as
Informix::IntervalYTM or Informix::Slob to handle specific Informix types.
Table 11-1 on page 360 shows the data type mapping between the Ruby Informix
driver and the Informix database.
359
7884ch11.fm

Table 11-1 Ruby Informix data type mapping
Informix data type
Ruby data type
SMALLINT, INT, INT8, FLOAT, SERIAL

and SERIAL8
Numeric
CHAR, NCHAR, VARCHAR,

NVARCHAR
String
DATE
Date
DATETIME
TIME
INTERVAL
Informix::IntervalYTM, Informix::IntervalDTS
DECIMAL, MONEY
BigDecimal
BOOL
TrueClass, FalseClass
BYTE, TEXT
StringIO, String
CLOB, BLOB
Informix::Slob
Ruby includes an interactive shell called irb that can be used to run simple Ruby
statements. irb is located in the bin directory of the Ruby installation.
The driver name used inside the Ruby scripts to reference the Ruby Informix
driver is informix.
To test if the Ruby driver can connect to a database, you must load the Ruby
driver and then create an connection using the Ruby Informix object.
Example 11-2 demonstrates how to load the Ruby Informix driver and connect to
an IBM Informix database. The fist command, require 'informix', tells Ruby
runtime to load the Ruby Informix driver. After that we create a Ruby Informix
connection object and print the database version information.
Example 11-2 Testing connection with Ruby informix
C:\work>irb
irb(main):001:0> require 'informix'
=> true
irb(main):002:0> db=Informix.connect('stores_demo','informix','password')
=> #<Informix::Database:0x8176110>
irb(main):003:0> puts db.version
IBM Informix Dynamic Server Version 11.50.FC6
=> nil
irb(main):004:0>
360
7884ch11.fm
11.2.2 Data Server Ruby driver

The Data Server Ruby driver uses the Data Server CLI driver to connect the
Informix database. It uses the DRDA protocol so the version of the IBM Informix
database server must be 11.10 or 11.50.
The driver is included as part of the IBM Data Server Driver package and it is
also available to download directly from the RubyForge website at:
http://rubyforge.org/projects/rubyibm/
The complete package can be installed using the Ruby Gem utility by running the
following command from a Ruby session:
gem install ibm_db
For information about the build process and setup process for the Data Server
Driver, consult the README file in the driver directory.
Example 11-3 shows the Data Server Ruby driver directory.
Example 11-3 ibm_db gem directory
Directory of C:\work\Ruby187\lib\ruby\gems\1.8\gems\ibm_db-0.10.0-x86-mswin32
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
03/07/2010
17:30
<DIR>
.
17:30
<DIR>
..
17:30
6,063 CHANGES
17:30
<DIR>
ext
17:30
1,656 init.rb
18:06
<DIR>
lib
17:30
1,088 LICENSE
17:30
299 MANIFEST
17:30
13,402 README
17:30
<DIR>
test
5 File(s)
22,508 bytes
5 Dir(s) 76,167,290,880 bytes free
You can find additional information regarding the Ruby driver for IBM Data
Servers at
http://publib.boulder.ibm.com/infocenter/db2luw/v9r5/topic/com.ibm.db2.luw.apdv
.ruby.doc/doc/t0052765.html
On some platforms, such as Windows, the shared library for the Ruby driver is
already included in the package, therefore, there is no need for a C build
environment.
361
7884ch11.fm
Example 11-4 shows the Ruby driver directory from the Data Server Client on
Windows after the installation.
Example 11-4 Windows Data Server Ruby directory
C:\work>dir "C:\Program Files\IBM\IBM DATA SERVER DRIVER\ruby"
Directory of C:\Program Files\IBM\IBM DATA SERVER DRIVER\ruby
17/05/2010
17/05/2010
30/05/2009
02:16
<DIR>
.
02:16
<DIR>
..
11:09
198,144 ibm_db-0.10.0-mswin32.gem
1 File(s)
198,144 bytes
2 Dir(s) 76,167,372,800 bytes free
C:\work>
Configuration
The Data Server Ruby driver uses the Data Server ODBC/CLI driver for the
connection to the database.
The configuration details are the same as with the ODCB/CLI driver. These
details are usually keep in the db2profile.ini file. Refer to the 2.2.3, Setting up
IBM Data Server drivers on page 40 chapter for detail information about the
Data Server Driver configuration.
Data Types
There are no specific Ruby objects to use Informix data types. The Data Server
Ruby driver supports the same Informix data types as the Data Server Driver for
CLI.
We use a irb session to test the connectivity against an IBM Informix database
server.
The reference name of the Data Server Ruby driver is ibm_db, you must load this
driver before creating the Ruby connection object.
Example 11-5 shows how to load the Ruby driver and how to connect to the
Informix server.
Example 11-5 Testing connection with data server ruby driver
C:\work>irb
362
7884ch11.fm
irb(main):001:0> require 'mswin32/ibm_db'

=> true
irb(main):002:0> db=IBM_DB.connect 'dsc_dsn','informix','password'
=> #<IBM_DB::Connection:0x81fb6d0>
irb(main):003:0>
First we load the Ruby driver with require 'mswin32/ibm_db', then open the
connection using the IBM_DB::connect method.
Note: Because we are using the Windows version of the Ruby driver we must
prefix the driver name with the mswin32 directory
You can also made a dsn-less connection by specifying all the required
parameters in the connection string:
IBM_DB.connect 'DRIVER={IBM DB2 ODBC DRIVER};DATABASE=stores_demo;
HOSTNAME=kodiak;PORT=9089;PROTOCOL=TCPIP;UID=informix;PWD=password;', '', ''
11.2.3 Rails adapters

The Ruby informix Rails Adapter provides the Ruby ActiveRecord object that
makes the use of the Ruby driver to work on the Rails framework possible.
The Rails adapter for the Ruby Informix driver is a free download available at
http://rubyforge.org/projects/rails-informix/
It can be automatically installed as a Ruby Gem using the gem utility.

Example 11-6 shows how to install the Ruby Informix adapter.
Example 11-6 Installing Ruby Informix adapter
C:\work>gem install activerecord-informix-adapter -v 1.1.1
Successfully installed activerecord-informix-adapter-1.1.1
1 gem installed
Installing ri documentation for activerecord-informix-adapter-1.1.1...
Installing RDoc documentation for activerecord-informix-adapter-1.1.1...
C:\work>
The configuration of the adapter depends on the version of Rails installed in the
Ruby environment. On versions older than 2.x the adapter script file
informix_adapter.rb has to be copied into the connection_adapters directory.
Example 11-7 on page 364 shows the connection_adapter directory with both
Rails adapters installed:
363
7884ch11.fm

Example 11-7 connection _adapter directory
Directory of
C:\work\Ruby187\lib\ruby\gems\1.8\gems\activerecord-1.15.6\lib\active_record\co
nnection_adapters
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
3/07/2010
18:02
<DIR>
.
18:02
<DIR>
..
17:01
<DIR>
abstract
17:01
5,235 abstract_adapter.rb
17:01
6,957 db2_adapter.rb
17:01
27,749 firebird_adapter.rb
17:01
30,751 frontbase_adapter.rb
17:30
70,854 ibm_db_adapter.rb
20:06
9,916 informix_adapter.rb
17:01
13,774 mysql_adapter.rb
17:01
11,531 openbase_adapter.rb
17:01
25,897 oracle_adapter.rb
17:01
21,513 postgresql_adapter.rb
17:01
13,162 sqlite_adapter.rb
17:01
22,087 sqlserver_adapter.rb
17:01
22,622 sybase_adapter.rb
13 File(s)
282,048 bytes
3 Dir(s) 76,166,725,632 bytes free
When using versions of Rails older than 2.x, it is also required to include the
reference name of the Ruby driver in the RAILS_CONNECTION_ADAPTER
parameter in the ActiveRecord ruby script. See Example 11-8 for an example.
Example 11-8 active_record file
C:\work\Ruby187\lib\ruby\gems\1.8\gems\activerecord-1.15.6\lib>grep informix
active_record.rb
RAILS_CONNECTION_ADAPTERS = %w( mysql postgresql sqlite firebird sqlserver
db2 oracle sybase openbase frontbase informix ibm_db )
C:\work\Ruby187\lib\ruby\gems\1.8\gems\activerecord-1.15.6\lib>
After creating a Rails project, the project configuration file that contains the
database information, database.yml, must be updated with the connection
details of the IBM Informix database. This file is located in the <project>/config
directory and has three sections, development, test and production. These
sections point to databases on the respective environments. The database
connectivity properties include:
adapter: The Ruby driver used. You do not have to give the complete version.
database: Database to be connected to.
username and password: To connect to the Informix server.
364
7884ch11.fm
server: The machine that Informix server is running on.

port: The DRDA port on the Informix server.
Example 11-9 shows a typical database.yml file with the details for an IBM
Informix database. The Rails application is named stores7 and we are
connecting to an Informix server called demo_on.
We use the Ruby Informix driver reference name for the adapter parameter
informix. The server parameter identifies the IBM Informix server to connect to.
The information about the Informix server such as host and port is retrieved from
the sqlhosts file or Windows registry.
Example 11-9 database.yml for the Ruby informix Adapter
C:\work\stores7\config>type database.yml
development:
adapter: informix
database: stores7
pool: 5
timeout: 5000
server: demo_on
username: informix
password: password
# Warning: The database defined as 'test' will be erased and
# re-generated from your development database when you run 'rake'.
# Do not set this db to the same as development or production.
test:
adapter: mysql
database: stores7
username: root
password:
host: localhost
production:
adapter: informix
database: stores7
pool: 5
timeout: 5000
server: demo_on
username: informix
password: password
C:\work\stores7\config>
365
7884ch11.fm
Rails Adapter for IBM Data Server

The Rails Adapter for the IBM Data Server Ruby driver is bundled with the Ruby
driver. It is installed at the same time as the Ruby driver is installed.
In the same way as with the Ruby Informix Rails adapter, if the Rails Adapter for
IBM Data Server is installed on a version of Rails older than 2.x, the adapter
script (informix_adapter.rb) must be copied in the connection_adapter
directory, and the RAILS_CONNECTION_ADAPTER parameter in the
active_record.rb Ruby script has to be updated.
The database configuration file for a Rails project, database.yml, has to be
updated with the details for the IBM Informix server.
Example 11-10 shows a database.yml used in a Rails project that connects to an
IBM Informix database.
Example 11-10 database.yml for the Rails adapter for IBM Data Servers
C:\work\stores7\config>type database.yml
development:
adapter: ibm_db
database: stores7
pool: 5
timeout: 5000
host: kodiak
port: 9089
username: informix
password: password
test:
adapter: mysql
database: stores7
username: root
password:
host: localhost
C:\work\stores7\config>
The adapter parameter is set to the reference name of the Data Server Ruby
driver ibm_db. It also contains the host name and the port parameter with the
details about the DRDA Informix instance. These details are the same as used in
the Data Server driver for ODBC/CLI.
366
7884ch11.fm
For a list of all the parameters refer to Info Center documentation at

ent.ruby.doc/doc/t0052780.html
11.3 Database operations

This section provides examples of using both Ruby drivers to perform basic
operations against an IBM Informix database server. We also demonstrate how
to use the Rails adapter to create basic web applications.
Using Ruby Informix driver

In this section we discuss the basic database operation using the Ruby Informix
driver.
Connection to the database

Example 11-11 shows how to connect to an Informix database. The database
name, user name, and password are passed as the parameters in the command
line.
Example 11-11 ifx_connect.rb output
C:\work>type ifx_connect.rb
# load the informix driver
require 'informix'
# Connect to the database
db = Informix.connect(ARGV[0],ARGV[1],ARGV[2])
# print database information
print "Connected to #{db.version}"
db.close
C:\work>ruby ifx_connect.rb stores_demo informix password
Connected to IBM Informix Dynamic Server Version 11.50.FC6
C:\work>
Getting information about a table

Example 11-12 demonstrate how to use the Ruby connection object to retrieve
metadata information about the state table. We use the columns() method of the
Informix connection object to retrieve a Ruby array with all the metadata
information.
Example 11-12 ifx_metadata.rb output
C:\work>cat ifx_metadata.rb
367
7884ch11.fm

require 'informix'
db = Informix.connect(ARGV[0])
print "Connected to #{db.version}\n"
db.columns(ARGV[1]).each {
|name| name.each {|elem| print " #{elem[0]} #{elem[1]}\n"}
}
db.close
C:\work>ruby ifx_metadata.rb stores_demo state
precision 0
type 0
scale 0
length 2
xid 0
nullable true
name code
stype CHAR
...
Executing a simple SQL statement

Using the Ruby driver you can run an SQL statement directly from the Ruby
connection object or through a prepared statement object.
Example 11-13 demonstrates how to run a simple SQL statement using a
prepared statement Ruby object. The execute() method returns the number of
affected rows.
Example 11-13 ifx_execute.rb output
C:\work>cat ifx_execute.rb
require 'informix'
db = Informix.connect(ARGV[0])
# create a prepare object with the SQL passed
stmt = db.prepare(ARGV[1])
# Execute the prepared statement
rc=stmt.execute()
print "Result=#{rc}"
db.close
368
7884ch11.fm
C:\work>ruby ifx_execute.rb stores_demo "DELETE from STATE where CODE='AR'"

Result=1
C:\work>
If the SQL statement returns only one row, the execute() method can be used to
retrieve that value. See Example 11-14 for a example.
Example 11-14 ifx_execute.rb output
C:\work>ruby ifx_execute.rb stores_demo "SELECT sname FROM state WHERE
code='CA'"
Result=snameCalifornia
C:\work>
Using parameters
Example 11-15 shows how to prepare and execute a parametrized INSERT
statement using the Ruby driver.
Example 11-15 ifx_parameters output
C:\work>cat ifx_insert.rb
require 'informix'
db = Informix.connect('stores_demo')
stmt = db.prepare('INSERT INTO state(code,sname) VALUES (?,?)')
# Execute the statement using parameters
rc=stmt.execute(ARGV[0],ARGV[1])
db.close
C:\work>ruby ifx_insert.rb AR Arizona
Result=1
C:\work>
369
7884ch11.fm
Selecting data from the database

When selecting multiple records from the database the application must create a
cursor object to fetch the selected rows. Example 11-16 illustrates how to return
information using an Informix cursor.
Example 11-16 ifx_cursor.rb output
C:\work>cat ifx_cursor.rb
require 'informix'
db = Informix.connect('stores_demo')
# display database information
# Create a Cursor object
cur = db.cursor(ARGV[0])
# Open the cursor and fetch some rows
cur.open.each {|rows|
puts rows*" = "
}.close
# Close connection
db.close
C:\work>ruby ifx_cursor.rb "SELECT FIRST 3 sname FROM state"
Alaska
Hawaii
California
C:\work>
Using the Ruby Informix IfxSlob class

The Slob class is the Ruby interface for handling Smart Large Objects. It
provides methods for every action applicable with a smart large object. By using
the Informix::Slob class, it is possible to perform the same operations against a
smart large object as with other programming languages and drivers. The Slob
methods such as Seek() or Lock() allow random I/O to individuals part of the
large object that can only be achieved using this Ruby driver.
Example 11-17 shows how to select a BLOB from a database table. It retrieves
the catalog_advert CLOB column from the catalog table and prints out the
content of the large object and the size of the large object.
Example 11-17 ifx_blob.rb output
C:\work>cat ifx_blob.rb
require 'informix'
370
7884ch11.fm

db = Informix.connect("stores_demo")
# Creates an Informix Slob object
slob = Informix::Slob
# Opens a cursor to retrieve
cur = db.cursor("SELECT catalog_num,advert_descr FROM catalog WHERE
catalog_num=?")
cur.open(ARGV[0]).each {|rows|
slob = rows[1].open
print "Number
= #{rows[0]}\n"
# Reads the blob data as a String
print "Clob data= #{rows[1].read(rows[1].size)}\n"
print "Clob Size= #{rows[1].size}\n"
# Close the Slob object
slob.close
}
# Close database connection
db.close
C:\work>
C:\work>ruby ifx_blob.rb 10001
Number
= 10001
Clob data= Brown leather. Specify first baseman's or infield/outfield style.
Clob Size= 98
Number
= 10027
Clob data=
Double or triple crankset with choice of chainrings or chunky bacon. For double
crankset...
Clob Size= 154
Number
= 10031
Clob data= No buckle so no plastic touches your chin. Meets both ANSI and
Snell...
Clob Size= 123
C:\work>
For examples and a full description of all the method implemented by the
Informix::Slob class, refer to the Ruby Informix driver documentation at
http://ruby-informix.rubyforge.org/doc/classes/Informix/Slob.html
Using the Ruby Informix INTERVAL

The Ruby Informix driver provides a specific class to deal with the Informix
INTERVAL data type.
371
7884ch11.fm
Example 11-18 shows how to define and use the Informix::Interval class. In
this example we create an Interval Year to Month with one year and one month
as the value. The code performs a simple arithmetic operation adding the Interval
to the current date.
Example 11-18 ifx_interval.rb output
# Creates an Informix Interval object
minterval = Informix::Interval.year_to_month(1, 1)
print "Interval
\t=#{minterval}\n"
today = Date.today
print "Current date \t=#{today}\n"
print"Interval+today\t=#{minterval + today}\n"
C:\work>ruby ifx_interval.rb
Interval
=1-01
Current date
=2010-07-04
Interval+today =2011-08-04
C:\work>
You can fine the full documentation about all the methods supported by the Ruby
Informix driver at
http://ruby-informix.rubyforge.org/doc/
Using Data Server Ruby driver

In this section we discuss the basic database operation using the Data Server
Ruby driver.
Connecting to the database

The syntax for opening a connection with an Informix database server using the
Data Server Ruby driver differs from the syntax used by the Ruby Informix driver.
Example 11-19 shows a simple Ruby script that creates a connection object and
opens the connection. In the example code we also retrieve information about
the Informix server using the IBM_DB::server_info class.
Example 11-19 dsc_connect.rb
C:\work>cat dsc_connect.rb
require 'mswin32/ibm_db'
db = IBM_DB.connect(ARGV[0],ARGV[1],ARGV[2])
info = IBM_DB.server_info(db)
372
7884ch11.fm

print "Connected to #{info.DBMS_NAME} #{info.DBMS_VER}"
C:\work>ruby dsc_connect.rb testdsc informix password

Connected to IDS/NT64 11.50.0000
C:\work>
The Data Server Ruby driver is based on calls to the CLI driver, this means it
takes the connection details from the db2cli.ini configuration file. Example 11-20
shows the contents of the db2cli.ini used for this test.
C:\work>type "c:\Documents and Settings\Administrator\db2cli.ini"
[dsc_dsn]
Protocol=TCPIP
Port=9089
Hostname=kodiak
PWD=password
UID=informix
C:\work>
Executing an SQL statement

Example 11-21 demonstrate how to execute an SQL statement using the
prepare() and execute() methods of the IBM_DB driver. Both methods,
prepare() and execute(), require the connection and statement objects as a
parameters.
Example 11-21 dsc_execute.rb output
C:\work>cat dsc_execute.rb
db = IBM_DB.connect(ARGV[0],'','')
print "Connected to #{info.DBMS_NAME} #{info.DBMS_VER}\n"
stmt = IBM_DB.prepare(db,ARGV[1])
rc=IBM_DB.execute(stmt)
373
7884ch11.fm
C:\work>ruby dsc_execute.rb dsc_dsn "DELETE from STATE WHERE code='NW'"

Result=true
C:\work>
Parametrized SQL statement

When using parameters for an SQL statement, the application can use the
execute() method and provide the parameters as second argument for the
method call.
Example 11-22 insert a new record into the state table with the values passed to
the script through the command line.
Example 11-22 dsc_param.rb output
C:\work>cat dsc_param.rb
db = IBM_DB.connect('dsc_dsn','','')
sql = 'INSERT INTO state(code, sname) VALUES (?,?)'
# create a prepare object
stmt = IBM_DB.prepare(db,sql)
# display Statement and parameters
print "SQL=#{sql}\n"
rc=IBM_DB.execute(stmt,[ARGV[0],ARGV[1]])
C:\work>
C:\work>ruby dsc_param.rb "NW" "NewState"
SQL=INSERT INTO state(code, sname) VALUES (?,?)
Result=true
C:\work>
Selecting data
The Data Server Ruby driver has several methods that allow retrieving data from
the database, such as fetch_array(), fetch_assoc(), and fetch_row().
374
7884ch11.fm
Example 11-23 shows a simple Ruby script that returns the first two columns of
an SQL SELECT statement passed through the command line. The script uses
the fecth_array() method to retrieve the rows as an array object.
Example 11-23 Select data using fetch_array()
C:\work>cat dsc_fetch.rb
db = IBM_DB.connect(ARGV[0],'','')
stmt = IBM_DB.prepare(db,ARGV[1])
IBM_DB.execute(stmt)
while row = IBM_DB.fetch_array(stmt)
puts "#{row[0]}:#{row[1]}"
end
C:\work>ruby dsc_fetch.rb dsc_dsn "SELECT FIRST 3 code,sname FROM state"

AK:Alaska
HI:Hawaii
CA:California
C:\work>
Using Smart Large Objects

The Data Server Ruby driver handles smart large objects as normal data types. It
does not support all the smart features that are normally available with other
drivers. However, it simplify the code needed to deal with these data types.
Example 11-24 retrieves a CLOB column from the catalog table and displays the
contents.
Example 11-24 dsc_blob.rb
C:\work>cat dsc_blob.rb
db = IBM_DB.connect('dsc_dsn','','')
375
7884ch11.fm

sql = "SELECT catalog_num,advert_descr FROM catalog WHERE catalog_num=?"
stmt = IBM_DB.prepare(db,sql)
IBM_DB.execute(stmt,ARGV)
while row = IBM_DB.fetch_array(stmt)
puts "#{row[0]}:#{row[1]}"
end
C:\work>ruby dsc_blob.rb 10001
10001:Brown leather. Specify first baseman's or infield/outfield style.
Specify right- or left-handed.
C:\work>
You can find additional documentation about the methods for the Ruby for IBM
Data Server at
http://rubyibm.rubyforge.org/docs/driver/2.0.0/doc/
or in the IBM DB2 Information Center website at

ent.ruby.doc/doc/c0052760.html
11.4 Using the Rails adapter with Ruby Informix

In this section we demonstrate how to create a basic web application using the
Rails framework with the Ruby Informix Adapter.
One of the key concepts of Rails is Convention over Configuration. This means
that you must follow the convention rules when designing your database so the
Rails framework can automatically generate code to handle typical operations
with the database.
Some of the convention rules required by the Rails framework are:
Table name must be a plural name for the entity it contains. For example, if
the table contains information about books, it should be called Books.
The table must contains an unique primary key column and should be called
ID.
376
7884ch11.fm
For both Ruby drivers, an Informix SQL SEQUENCE must be created for
each of the tables used. The name of the sequence must be tablename_seq.
You can find additional information about Rails conventions on the RailsGuides
website at:
http://guides.rubyonrails.org/
Creating database objects

The following examples use the orders and items tables from the stores7
database. Due to the convention rules used on Rails, we must modify the
schema of the tables to follow the Rails conventions.
Example 11-25 shows the SQL script we use to change the name of the two
tables, create views to incorporate the ID column, and create the SQL sequence
required for object reference.
Example 11-25 setup.sql script
-- Orders table
RENAME TABLE orders TO order;
CREATE VIEW orders(
id,
order_date,
customer_num,
ship_instruct,
backlog,
po_num,
ship_date,
ship_weigh,
ship_charge,
paid_date) AS
SELECT * FROM order;
CREATE SEQUENCE orders_seq;
-- Items table
RENAME TABLE items TO item;
CREATE VIEW items (
id,
order_num,
stock_num,
manu_code,
quantity,
total_price) AS
SELECT * FROM item;
CREATE SEQUENCE items_seq;
377
7884ch11.fm
Creating the Rails application

You must create a Rails application using Rails commands before adding any
objects or definitions.
Example 11-26 shows the output of the rail stores command
Example 11-26 Output of the rail stores command
C:\work>rails
create
create
create
create
create
create
create
create
create
create
create
create
create
create
create
...
stores
app/controllers
app/helpers
app/models
app/views/layouts
config/environments
components
db
doc
lib
lib/tasks
log
public/images
public/javascripts
public/stylesheets
For more information refer to the Rails documentation at

Modifying the database configuration file

The database.yml must include the connection details of the database.
Example 11-27 shows the database configuration file used by our application.
Example 11-27 database.yml file
C:\work\stores>type config\database.yml
development:
adapter: informix
database: stores7
pool: 5
timeout: 5000
server: demo_on
username: informix
password: password
378
7884ch11.fm

test:
adapter: mysql
database: stores7
username: root
password:
host: localhost
production:
adapter: informix
database: stores7
pool: 5
timeout: 5000
server: demo_on
username: informix
password: password
C:\work\stores>
Creating the Rails model and controllers

To make Rails aware of the database tables, you create model and controller in
your application. We use the model and controller Ruby script to create model
and controller for the two tables used in our application.
Example 11-28 shows the batch script used to generate the model and controller
for the Items and Orders table.
Example 11-28 objects.cmd
ruby
ruby
ruby
ruby
script\generate
script\generate
script\generate
script\generate
model Order
controller Order
model Item
controller Item
Example 11-29 shows the output of each of the commands in the objects.cmd
batch script.
Example 11-29 Output of objects.cmd
C:\work\stores>ruby script\generate model Order
exists app/models/
exists test/unit/
exists test/fixtures/
create app/models/order.rb
create test/unit/order_test.rb
create test/fixtures/orders.yml
exists db/migrate
379
7884ch11.fm

create db/migrate/004_create_orders.rb
C:\work\stores>ruby script\generate controller Order
exists app/controllers/
exists app/helpers/
create app/views/order
exists test/functional/
create app/controllers/order_controller.rb
create test/functional/order_controller_test.rb
create app/helpers/order_helper.rb
C:\work\stores>ruby script\generate model Item

exists app/models/
exists test/unit/
exists test/fixtures/
create app/models/item.rb
create test/unit/item_test.rb
create test/fixtures/items.yml
exists db/migrate
create db/migrate/005_create_items.rb
C:\work\stores>ruby script\generate controller Item
exists app/helpers/
create app/views/item
create app/controllers/item_controller.rb
create test/functional/item_controller_test.rb
create app/helpers/item_helper.rb
C:\work\stores>
After the objects are created, you must modify the controller Ruby file for each
object to build the object scaffold.
We use Rails 1.2.6 on our examples. Rails 2.x does not support dynamic
scaffolding. This means it can not retrieve the information for the table column
dynamically. In this case, the scaffold for the objects must be created manually
while creating the model object. We add the instruction scaffold: object_name
to each of the files for creating scaffold.
Example 11-30 shows the Ruby script file for the Item and Order controllers.
Example 11-30 Controller script
C:\work\stores>cat app/controllers/order_controller.rb
class OrderController < ApplicationController
380
7884ch11.fm

scaffold :Order
end
C:\work\stores>cat app/controllers/item_controller.rb
class ItemController < ApplicationController
scaffold :Item
end
C:\work\stores>
For more information about the changes in Rails 2.x refer to Rails documentation
at
Starting the Rails web server

To start the Rails web server use the rails scripts/server command.
Example 11-31 demonstrates how to start the server.
Example 11-31 Rails web server
C:\work\stores>ruby script/server
=> Booting WEBrick...
=> Rails application started on http://0.0.0.0:3000
=> Ctrl-C to shutdown server; call with --help for options
[2010-07-05 10:40:00] INFO WEBrick 1.3.1
[2010-07-05 10:40:00] INFO ruby 1.8.7 (2010-01-10) [i386-mingw32]
[2010-07-05 10:40:00] INFO WEBrick::HTTPServer#start: pid=1092 port=3000
Demonstrating Website application

At this point, Rails should have constructed the application for you and you can
browse and change the information for your table.
We can open a Web browser and navigate to the local web server to browse the
the Order and Item tables:
http://127.0.0.1:3000/item
http://127.0.0.1:3000/order
Figure 11-1 on page 382 shows the Item table listing web page. The web page
contains links to perform all the typical operations associated with a database
table (select, insert, update, and delete).
381
7884ch11.fm
Figure 11-1 Item listing web page
Figure 11-2 shows the New Item website with all the fields from the Item table
ready to be used to insert a new record into the table.
Figure 11-2 Item New web page
Figure 11-3 on page 383 shows the Order listing page.
382
7884ch11.fm
Figure 11-3 List Order webpage
The Web applications development with Ruby on Rails was designed to be an

effortless task. With just four commands and an little configuration, we have
created a website that can handle the common table operations for a database
application.
11.5 Using the Rails adapter with IBM_DB

In this section we demonstrate how to create a basic web application using the
Rails framework with the IBM_DB Adapter. The sample program is a simple
telephone directory application for user to list, add, update, and delete phone
entries.
Rails can be used to generate the Data Definition Language (DDL) for the
database objects, we show how to create tables with Ruby on Rails.
Creating the Rails application

Use the rails command to create the Rails application. We create our
application sample in the directory C:\RailsProjects.
Example 11-32 shows how to create a rails application and the rails command
output of our application.
Example 11-32 Creating a Rails application
C:\RailsProjects>rails sample
create
383
7884ch11.fm

create
create
create
create
create
create
...
app/controllers
app/helpers
app/models
app/views/layouts
config/environments
config/initializers
Modifying the database configuration file

Update the database configuration file database.yml with the database
connectivity details.
Example 11-33 shows the configuration file for our development database. The
name of the adapter, ibm_db, correspond to the Ruby Adapter for IBM Data
Servers.
Example 11-33 database.yml file
development:
adapter: ibm_db
database: ruby
username: informix
password: Ifmx4you
host: kefka.lenexa.ibm.com
port: 9089
# Warning: The database defined as "test" will be erased and
# re-generated from your development database when you run "rake".
test:
production:
Creating model, control, and view

Ruby on Rails worked on a Model-Control-View architecture. You can create
model, control, and view components in stages. Here we show a quick way of
using the scaffold command to have the Rails create the complete application
for us including all the directories and necessary files.
The telephone table of our application use strings for first name, last name, and
phone number. This table layout forms the model for our application. We specify
this model right on the scaffold command as shown in Example 11-34 on
page 385.
We invoke the command from the root directory of our project with the ruby
script\generate script. The syntax for the scaffold options is:
384
7884ch11.fm
scaffold <model name> <column name: attributes> .. <column name: attributes>
The output shows that Rails creates the necessary models, views and
controllers. This command also generate the script to create the tables
necessary to associate with the model.
Example 11-34 Creating a scaffold
C:\RailsProjects\sample>ruby script\generate scaffold phonedir
first_name:string last_name:string phone:string
exists app/models/
exists app/helpers/
create app/views/phonedirs
exists app/views/layouts/
exists test/unit/
create test/unit/helpers/
exists public/stylesheets/
create app/views/phonedirs/index.html.erb
...
dependency model
exists
app/models/
exists
test/unit/
exists
test/fixtures/
create
app/models/phonedir.rb
create
test/unit/phonedir_test.rb
create
test/fixtures/phonedirs.yml
create
db/migrate
create
db/migrate/20100704234009_create_phonedirs.rb
Migrating the model

Ruby on Rails calls the creation of the database objects backing the model as
migration. The scaffold command created a migrate script for creating the table
attached to the model.
Example 11-35 shows the migration file
db/migrate/20100704234009_create_phonedirs.rb created by the scaffold
command. This script has both create and drop sections that means you can
rollback any migration. We created a model by name phonedir. Ruby on Rails
then created the table phonedirs which is a plural form of the model name. This
is the Ruby on Rails convention.
Example 11-35 the phonedir migration file
class CreatePhonedirs < ActiveRecord::Migration
def self.up
create_table :phonedirs do |t|
385
7884ch11.fm

t.string :first_name
t.string :last_name
t.string :phone
t.timestamps
end
end
def self.down
drop_table :phonedirs
end
end
Use the rake utility to migrate the file and create table in the database. rake a
simple ruby build script with capabilities similar to make. It can be used to
generate the database schema using a migration file.
Example 11-36 shows how to run the rake script to create the model in the
database.
Example 11-36 Creating model
C:\RailsProjects\sample>rake db:migrate
(in C:/RailsProjects/sample)
== CreatePhonedirs: migrating ================================================
-- create_table(:phonedirs)
-> 0.0781s
== CreatePhonedirs: migrated (0.0781s) =======================================
In Example 11-37, we use the Informix dbschema utility to export the table
schema to check the tables created Ruby on Rails migration. We started with an
empty database and the output shows that two tables were created:
schema_migrations: Ruby on Rails uses this table to keep track of the various
version of the table. This allows you to roll back to the previous version.
phonedirs: In this application table, Ruby added a few columns that we did not
specify:
id: This serial column is for the primary key required by Ruby on Rails.
created_at and updated_at: Ruby on Rails uses these optionally.
Example 11-37 dbschema on database ruby output
% dbschema -d ruby
DBSCHEMA Schema Utility
grant dba to "informix";
...
386
INFORMIX-SQL Version 11.50.FC7
7884ch11.fm

create table "informix".schema_migrations
(
version varchar(255) not null
);
...
create table "informix".phonedirs

(
id serial not null ,
first_name varchar(255),
last_name varchar(255),
phone varchar(255),
created_at datetime year to fraction(5),
updated_at datetime year to fraction(5),
primary key (id)
);
...
create unique index "informix".unique_schema_migrations on "informix"
.schema_migrations (version) using btree ;
Starting the Rails web server

Example 11-38 shows how to start the Rails WEBrick server from the root
directory. The http port number is 3000.
Example 11-38 Starting the WEBrick server
C:\RailsProjects\sample>ruby script\server
=> Booting WEBrick
=> Rails 2.3.8 application starting on http://0.0.0.0:3000
=> Call with -d to detach
=> Ctrl-C to shutdown server
[2010-07-04 17:25:22] INFO WEBrick 1.3.1
[2010-07-04 17:25:22] INFO ruby 1.9.1 (2010-01-10) [i386-mingw32]
[2010-07-04 17:25:22] INFO WEBrick::HTTPServer#start: pid=292 port=3000
Checking the application from website

You can start the web application using the local web server address:
http://localhost:3000/phonedirs
Figure 11-4 on page 388 shows the initial screen of our application brought up
with http://localhost:3000/phonedirs.
387
7884ch11.fm
Figure 11-4 Initial screen of the application
We add one phone entry to our directory as shown in Figure 11-5.
Figure 11-5 Creating a new phone listing
Figure 11-6 phone entry has been added
Figure 11-6 Listing phonedir after the new entry
For more information regarding the development with Ruby and Ruby on Rails
refer to the Ruby on Rails website at:
388
7884ch12.fm
12
Chapter 12.
Informix 4GL Web Services

In this chapter we introduce the new Web Services feature of IBM Informix 4GL.
This chapter provides an overview and configuring and building Web Services
using Informix 4GL.
In this chapter, we discuss the following:
Informix 4GL
Web Services and Service Oriented Architecture (SOA)
Environment setup when developing Web Services with I4GL
Using I4GL to publish and consume a Web Service
How to diagnose typical problems
389
7884ch12.fm
12.1 Basic concepts

In this section we provide an introduction of the products and technologies
mention in this chapter.
12.1.1 IBM Informix 4GL

Informix 4GL is a programming language developed by IBM for interacting with
Informix database servers. It provides a rich environment for easy development
of relational database applications. Informix 4GL provides all the component
needed to develop character based applications using an Informix database, for
example, project management, reports, debugger, and so on.
Informix 4GL supports compilers that can convert the I4GL applications to C
language or to generate platform-independent pseudo code that can be executed
using a I4GL runner.
For more information about Informix 4GL, refer to the 4GL Reference Manual at
http://publib.boulder.ibm.com/infocenter/ifxhelp/v0/index.jsp?topic=/com.ibm.to
ols.doc/4gl.html.
12.1.2 Service Oriented Architecture and Web Services

Service Oriented Architecture or SOA is an architectural style that provides
methods for systems development and integration, allowing applications
developed with different technologies or programing languages to exchange data
with one another.
This exchange of data is accomplished through the use of web services. Web
services are saleable functions that can be accessed independent of platforms
and programming languages. These functions take a set of inputs and return a
set of outputs to accomplish a specific task.
Refer to the IBM Service Oriented Architecture portal for more information at
http://www.ibm.com/software/solutions/soa/
12.1.3 Web Services development

A Web Service can be created with any web-aware technology. The most
common language used for Web Services development is Java. however, Web
Services can be created with other language and technologies such as C or
.NET
390
7884ch12.fm
IBM provides several options for SOA development including:

IBM SOA Sandbox
IBM Rational Application Developer
eKit: Enterprise Architect for SOA
The development of a Web Service that requires the use of an IBM Informix
database can be achieved using the following languages:
Java, using any of the JDBC driver for an Informix database
Any .NET language, using any of the Informix .NET providers available
IBM Informix 4GL
12.1.4 Informix 4GL and Web Services

Starting form version 7.50 of IBM Informix, 4GL developers can manage and
create Web Services using the 4GL language.
One of the key benefits of using Informix 4GL is the easy interaction with the
Informix database server. The use of SQL statements to access the database
objects does not require any specific code as with other programming languages
because SQL is embedded in the I4GL language.
The ability of create Web Services directly with I4GL allows the reuse of legacy
code. Existing solutions developed with I4GL can be converted to Web solutions
without much effort.
With Informix 4GL you can publish existing I4GL functions as web services as
well as consume exiting Web Services from any I4GL application.
12.1.5 Components
Informix 4GL uses the Axis2 web service wrapper API to implement the interface
required to communicate between the web server and the Informix 4GL libraries.
The components used for a typical I4GL Web Service solution are:
Apache AXIS2C server

Web Service
Informix 4GL
Informix Database Server
There are only two possible operations to perform with a Web Service, Create or
Consume:
Create is the process of creating the Web Service and publish it to made it
available to consumers.
Chapter 12. Informix 4GL Web Services
391
7884ch12.fm
The term used in Informix 4GL for this task is Publish.

Consume is the process of using the Web Service, providing input
parameters and retrieving the result as output parameters for the function.
The term used in Informix 4GL for this task is Subscribe.
Axis2 C functions are used for both operations as the wrapper code between the
Web Service and Informix 4GL.

In this section, we discuss the setup and configuration needed to develop Web
Services with Informix 4GL.
12.2.1 Prerequisites and supported platforms

The following is the requisites for using the Web Services with Informix 4GL.
Apache Axis2/C version 1.5.1 (bundled with Informix 4GL)

Apache Axis2/Java version 1.3.1 (bundled with Informix 4GL)
IBM Informix database server version 10 or later
Java Runtime Environment (JRE) 1.5 or later
Perl 5.8.8
IBM Informix 4GL 7.50 is supported in the following platforms:
HP-IA 11.23 and 11.31

AIX 5.3 and 6.1
Solaris 5.9 or 5.10
Red Hat Enterprise Linux 4 and 5
SUSE Linux Enterprise Server 10
Note: The Web Service feature was added to version 7.50.xC1 of Informix
4GL but it was only available for Linux platforms. Since version 7.50.xC3, all
the above platforms are supported.
12.2.2 Environment
The utilities for using Web Service with Informix 4GL are installed in the same
directory as Informix 4GL.
We do not cover how to install and set up Informix 4GL in this book, refer to the
Informix 4GL Installation Guide at
392
7884ch12.fm
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.4gl_install
.doc/fgl_ing_010.htm
To use any of the I4GL Web Services tools, the variables listed in Table 12-1
must be set in the development environment. These variables define the location
of those 4GL, Java, and Axis2 resources that you need for application
development and deployment.
Table 12-1 Environment variables
Variable
Description
AXIS2C_HOME
Specifies the Axis2 installation directory
CLASSPATH
Path to the required Java classes
DBPATH
Path for support files. Must be $INFORMIXDIR/etc
INFORMIXDIR
Directory where the 4GL files are installed
INFORMIXSERVER
Default database server
INFORMIXSQLHOSTS
Optional: Specifies the location of the sqlhosts file,

which contains database connectivity information
JAVA_HOME
Must be set to point to JRE 1.5 or later
Load Library Path

for example: LD_LIBRARY_PATH
Specifies which directories to search for client or

shared IBM Informix general libraries
PATH
Specifies which directories to search for

executable programs. Must include
$INFORMIXDIR/bin and $JAVA_HOME/bin
SOA_ERR_LOG
Optional: Specifies the directory where the log file

(w4glerr.log) is created, defaults to /tmp
PROGRAM_DESIGN_DBS
Optional: Database used for storing Web service

definitions, defaults to syspgm4gl
Example 12-1 shows a typical shell script to set up these variables.

Example 12-1 setup.ksh
AXIS2C_HOME=$INFORMIXDIR/AXIS2C
AXJDIR=$INFORMIXDIR/AXIS2C/AXIS2JARS
CLASSPATH=$AXJDIR/wsdl4j-1.6.2.jar:$AXJDIR/backport-util-concurrent-2.2.jar:$AX
JDIR/XmlSchema.jar:$AXJDIR/XmlSchema-1.3.1.jar:$AXJDIR/xbean-2.2.0.jar:$AXJDIR/
axiom-dom-1.3.1.jar:$AXJDIR/axiom-impl-1.3.1.jar:$AXJDIR/axiom-api-1.3.1.jar:$A
XJDIR/neethi-1.3.1.jar:$AXJDIR/axis.jar:$AXJDIR/commons-logging.jar:$AXJDIR/wsd
l2ws.jar:$AXJDIR/commons-discovery.jar:$AXJDIR/jaxrpc.jar:$AXJDIR/saaj.jar:$AXJ
393
7884ch12.fm

DIR/wsdl4j.jar:$AXJDIR/axis2-java2wsdl-1.3.1.jar:$AXJDIR/axis2-codegen-1.3.1.ja
r:$AXJDIR/axis2-kernel-1.3.1.jar
DBPATH=$INFORMIXDIR/etc
LD_LIBRARY_PATH=$INFORMIXDIR/AXIS2C/lib:$LD_LIBRARY_PATH
export AXIS2C_HOME AXJDIR CLASSPATH DBPATH LD_LIBRARY_PATH
Some of the I4GL Web Services utilities such as w4gl keep design and
configuration information in a database on the Informix database server. The
default name for this database is syspgm4gl. The database is created the first
time the w4gl tool is invoked. You can specify your own database name using the
environment variable PROGRAM_DESIGN_DBS.
12.3 I4GL Web Service tools

This section describes the tools available within Informix 4GL to publish, deploy,
package, and subscribe web services.
12.3.1 Web Service Compiler w4glc

The I4GL Web Service Compiler w4glc is a script based on Perl. w4glc performs
all the required task to use Web Services within 4GL, from the creating process
to the deploying and packaging.
The w4glc script is used by other I4GL utilities in a non-interactive way. Any error
or failure generated during the execution is written to the SOA_ERR_LOG log file.
The syntax for executing the w4glc tool is:
w4glc {-option} <configuration-file>
The w4glc utility does not keep any information in the database, it only uses
configuration files. The configuration file is a text file that specifies the details
about the I4GL function to be created as a Web service. It contains information
such as the location of the 4GL source files, input and output parameters, and
database connectivity information.
See Table 12-2 for a description of all the parameters available with the w4glc
tool.
Table 12-2 w4glc parameter list
394
Option
Description
check
Reads the configuration file and performs basic checks, such

as ensuring that the identified source files exist.
7884ch12.fm
Option
Description
compile
Compile generated intermediate code
deploy
Deploy web service on the AXIS2C server
force
Overwrite existing service with identical name
generate
Generate intermediate code for publish/subscribe
help
Provides basic help information
keep
Retain intermediate source files for troubleshooting
package
Bundling a web service for production deployment
silent
Generates code without on-screen display
version
Prints the version number
Example 12-2 shows the output for the generate option.

Example 12-2 w4glc generate output
informix@irk:/work$ w4glc -generate ./ws_zipcode_irk.4cf
Begin environment check ...
Environment check is completed.
Generating code. Please wait ...
Generating Wrapper code ....
The wrapper file is /tmp/w4gl_informix/zipcode_details_wrap.c
Generating WSDL ....
Generating headers ....
Generating skeletal code ....
Code generation completed.
Removing /tmp/w4gl_informix/zipcode_details_wrap.c ...
Removing /tmp/w4gl_informix/zipcode_details.wsdl ...
Removing /tmp/w4gl_informix/axis2_skel_ws_zipcode.h ...
Removing /tmp/w4gl_informix/axis2_svc_skel_ws_zipcode.c ...
Removing /tmp/w4gl_informix/services.xml ...
Removing /tmp/w4gl_informix/axis2_skel_ws_zipcode.c ...
informix@irk:/work$
For additional information about the I4GL Web Service Compiler refer to
http://publib.boulder.ibm.com/infocenter/idshelp/v115/topic/com.ibm.4gl_admin.d
oc/fgl_wsg_500.htm
395
7884ch12.fm
12.3.2 The w4gl utility

The character-based Interface w4gl utility is the main tool when using Web
Services with Informix 4GL. It allows you to perform the same task as the
command line tool, w4glc, however, it also manages the data within the program
design database, syspgm4gl.
The design database contains information about the Web Services environment
including:
Information for the host machine, such as name or temporary directories

Location for the Axis2 web server
Details for the Informix servers to which the Web Services connects
Definition for Web Services, such as function names or parameters types
The Web Service definition specified through the tool is saved in database tables
and hence is available for future reuse and modification thus reducing
development effort.
The w4gl tool uses the same I4GL character-based interface as other 4GL tools
to accomplish the creation and consumption of Web Services.
Example 12-3 shows the main menu of the w4gl tool.
Example 12-3 w4gl main menu
+------------------------------------------------------------------------------+
|W4GL:
Publish Subscribe Host name App server Exit
|
|Create and Deploy web services from I4GL functions.
|
|--------------------------------[ irkpgm4gl ]-------------------[Help: CTRL-W]|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
396
7884ch12.fm
The main options of the w4gl are

Publish: This option is used to create a new web service.
Subscribe: This option is used to consume a web service.
Host name: This option is used to managed the host information stored in the
design database.
Application server: This option is used to managed the Axis2 information
stored in the design database.
For a details description of all the w4gl menu options, refer to the 4GL Web
Services Administration Guide at
4gl_admin.doc/fgl_wsg_006.htm
12.3.3 Web Service Description Language Parser (wsdl_parser)

Web Service Description Language (WSDL) is an XML-based language for
describing network services. You can use WSDL to describe all the functions and
parameters for a Web Service
When you subscribe a Web service, the wsdl_parser tool parses the WSDL file
to retrieve all the information that is required to define a Web service.
The syntax for invoking the wsdl_parser tool is:
wsdl_parser sid wsdl_path ws_func i4gl_func target_dir target_file
Table 12-3 describes each parameter. All of the parameters are required for the
tool to work correctly.
Table 12-3 wsdl_parse parameters list
Parameter
Description
sid
An integer that uniquely identifies the subscriber.
wsdl_path
Location of the file that describes the complete description of the

Web service, can be an online or a local copy of the WSDL file
ws_func
Function to consume within the designated Web service.
i4gl_func
Name for the wrapper function to be used by the I4GL program.
target_dir
The path where files are stored while the Web service is being
consumed.
target_file
The file name containing the generated subscriber client code
397
7884ch12.fm
Note: The SID parameter is used only when the wsdl_parser is invoked from
the w4gl tool. The parameter is ignored if wsdl_parse is used from the
command line.
Example 12-4 shows the wsdl_parser tool been used to generate the
configuration file for a ws_zipcode Web Service. We pass the WSDL definition
directly from a web server as wsdl_path parameter
http://irk:9876/axis/services/ws_zipcode?wsdl
Example 12-4 wsdl_parser output
informix@irk:/work$ wsdl_parser 0 http://irk:9876/axis/services/ws_zipcode?wsdl
zipcode_details zipcode4gl `pwd`/publish zipcode4gl.c
informix@irk:/work$ ls publish
local.wsdl zipcode4gl.c_zipcode4gl.4cf
informix@irk:/work$
12.3.4 I4GL Web Services process

Figure 12-1 illustrates how the I4GL Web Services tools work together.
Figure 12-1 I4GL Web Service tools
12.4 Developing a Web Service with I4GL

In this section we demonstrate the steps required to create a Web Service using
IBM Informix 4GL. Publish and subscribe are two operations to perform with a
Web Service:
Publish
398
7884ch12.fm
The tasks required to create a new Web Service with the w4gl tool are:
a. Add the Host name and an Axis2 server information.
b. Add the Web service definition details.
c. Generate the Web service configuration file for publishing.
d. Deploy a Web service registering the service in the Axis2 web server.
e. Package a Web service as a single file (tar file) ready for the production
server.
Subscribe
The tasks needed to consume a Web Service are with the w4gl tool are:
a. Add details for the Web service to consume.
b. Compile the Web Services wrapper code which generates the
configuration file for subscription.
We publish a simple I4GL function as a Web Service and show the function can
be used from other languages such as Java.
12.4.1 Example I4GL function

We use a basic I4GL function state_name() for the Web Service. This function
connects to the database server and retrieves the name for a specific state code.
Example 12-5 shows the 4GL code state_name() saved in the state_name.4gl
file. This function queries the state table from the stores_demo database. The
state_name() function takes one input parameter state code of type CHAR(2)
and returns the state name that has CHAR(15) data type.
Example 12-5 state_name.4gl
FUNCTION state_name(code)
DEFINE state_rec RECORD
code CHAR(2),
sname CHAR(15)
END RECORD,
code CHAR(2),
sel_stmt CHAR(100);
LET sel_stmt= "SELECT code, sname FROM state WHERE code = ?";
PREPARE st_id FROM sel_stmt;
DECLARE cur_id CURSOR FOR st_id;
OPEN cur_id USING code;
FETCH cur_id INTO state_rec.*;
399
7884ch12.fm

CLOSE cur_id;
FREE cur_id;
FREE st_id;
RETURN state_rec.sname
END FUNCTION
To make sure the 4GL code is correct, we compile the function with the I4GL
compiler, c4gl as shown in Example 12-6.
Example 12-6 Compile 4GL function
informix@irk:/work$ c4gl -c state_name.4gl
informix@irk:/work$
12.4.2 Host and application details

Before adding any of the details for the Web Service we must provide details
about the machine where the Web Service will run and the Axis2 Server used.
This tasks are done using the Host name and Application server menus from the
w4gl tool.
Example 12-7 shows the HOST INFORMATION form with a host name irk and a
temporary directory /tmp/w4gl_informix.
Example 12-7 Host name menu
+------------------------------------------------------------------------------+
|HOST NAME:
Query Next Previous Add Modify Remove Exit
|
|See the existing host name details.
|
|[1 of 1]------------------------[ irkpgm4gl ]-------------------[Help: CTRL-W]|
|
|
|
HOST INFORMATION
|
|
|
| Machine ID
[
1]
|
| Host Name
[irk
]
|
| Temporary Directory [/tmp/w4gl_informix
]
|
|
[
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
400
7884ch12.fm
+------------------------------------------------------------------------------+
There are three fields for the host information:

Machine ID: This is an automatic identifier number.
Host Name: This is the machine name where the Axis2 server is installed.
Temporary Directory: Directory used by the w4glc tool for creating the
temporary files.
After defining the host information we add the information for the Axis2
application server using the App server menu.
Example 12-8 shows the APP SERVER form.
Example 12-8 Application menu
+------------------------------------------------------------------------------+
|APP SERVER: Query Next Previous Add Modify Remove Exit
|
|Go to the next app server
|
|
|
|
APP SERVER INFORMATION
|
| Server ID
[
1]
|
| Server Name [axis
] |
| Host Name
[irk
] |
| Port Number [ 9876]
|
|
ENVIRONMENT VARIABLES
|
| INFORMIXDIR
[/usr3/4gl750
] |
| INFORMIXSQLHOSTS [/usr3/sqlhosts
] |
| CLIENT_LOCALE
[en_us.utf8
]
|
| DBDATE
[Y4MD]
|
| Notes
[
] |
|
[
] |
|
[
] |
|
[
] |
|
|
|
|
|
|
+------------------------------------------------------------------------------+
An application server is identified by the following fields:
Server ID: This is an automatic identifier number.

Server Name: This is the name of the Axis2 server.
Host Name: This is the machine name where the Axis2 server runs.
Port Number: This is the port used for incoming connections to the web serve.
The App serv option is also used to stored specific environment information for
the application server. This information is required because the Axis2 server
401
7884ch12.fm
binary loads the I4GL libraries that might need additional resource files located in
the $INFORMIXDIR directory.
12.4.3 Definition of the Web Service

In this section we demonstrate how to define a Web Service from the w4gl tool.
The task to perform are:
Add the IBM Informix database the Web Service would use.
Add the definition for the Web Service such as the name of the service or the
name of the 4GL function.
Add specific details for the Web Service such as input and output variables.
Add the location of the 4GL source code file.
Add Informix database details

This task is done using the Database menu from the w4gl tool.
Example 12-9 shows the DATABASE INFORMATION form with the details of the
database used in our example.
Example 12-9 Database option
+------------------------------------------------------------------------------+
|DATABASE: Query Next Previous Add Modify Remove Exit
|
|Go to the previous database record.
|
|
|
|
DATABASE INFORMATION
|
|
|
| Database ID
[
1]
|
| Database Name [stores7
] |
| Database Server [irk1150
] |
| IDS Version
[11.50
]
|
| DB_LOCALE
[en_US.819
]
|
| Notes
[
] |
|
[
] |
|
[
] |
|
[
] |
|
|
|
|
|
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
The fields used in this menu option are:
402
7884ch12.fm
Database ID: This is an automatic number that identifies a database

definition.
Database name: The database name to which the Web Server connects.
IDS Version: This is Informix database server version number.
DB_LOCALE: This is the locale of the database. The default DB_LOCALE is
en_US.819.
Add details about the Web service

The definition of the Web Service is done using Add option in the Web Service
menu.
Example 12-10 Shows this option.
Example 12-10 Add of Web Service menu
+------------------------------------------------------------------------------+
|WEB SERVICE:
Query Next Previous Add Modify Remove Install ...
|
|Specify a new service record.
|
|
|
|
|
|
|
The first step is to add the Web Service name and the function name using the
Detail menu option. Example 12-11 shows this input form with the details of our
Web Service.
Example 12-11 Web Service details
+------------------------------------------------------------------------------+
|ADD:
Detail Variable File Exit
|
|Specify the web service parameters.
|
|
|
| Webservice ID
[
4]
|
| Webservice Name [ws_statename
] |
| Function Name
[state_name
] |
| Notes
[Returns the state name for a given code
] |
|
[
] |
|
[
] |
|
[
] |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
403
7884ch12.fm

|
|
|
|
+------------------------------------------------------------------------------+
The input details on this form describes a Web Service:

Web Service ID: This is an automatic generated identifier for the Web
Service.
Web service Name: This is the name of the Web Service.
Function Name: This is the name of the 4gl function.
Input and output variables

The next step on the definition of a Web Service is to add the input and output
variables used by the Web Service. This is done using the Variable menu option.
Example 12-12 shows this menu option. Our function takes an CHAR(2) input
parameter and returns a CHAR(15) value.
Example 12-12 Web Service Variable option
+------------------------------------------------------------------------------+
|VARIABLE: Input Output Exit
|
|Exit the Variable menu.
|
|
|
| [ 1] Input parameter - Variable name
Data type
|
|
|
| [ 1][code
][CHAR(2)
] |
| [ ][
][
] |
| [ ][
][
] |
| [ ][
][
] |
| [ ][
][
] |
|
|
| [ 1] Output parameter - Variable name Data type
|
|
|
| [ 1][sname
][CHAR(15)
] |
| [ ][
][
] |
| [ ][
][
] |
| [ ][
][
] |
| [ ][
][
] |
|
|
|
|
+------------------------------------------------------------------------------+
Source file information

This is the last step required to define a Web Service. We stored the location of
the I4GL source file that contains our function using the File menu option.
404
7884ch12.fm
Example 12-13 shows the fields used by the File option.

Example 12-13 Web Service File option
+------------------------------------------------------------------------------+
|ADD:
Detail Variable File Exit
|
|Exit the web services Add menu.
|
|
|
| Service Name
[ws_state_name
] |
| Function Name
[state_name
] |
|
|
| File Number
[ 1]
|
| Directory
[/work
] |
| File Name
[state_name.4gl
] |
|
|
| File Number
[ ]
|
| Directory
[
] |
| File Name
[
] |
|
|
| File Number
[ ]
|
| Directory
[
] |
| File Name
[
] |
|
|
|
|
|
|
+------------------------------------------------------------------------------+
A I4GL function may require more than one I4Gl file, you must supply all the
required file names and their directories. We supply the 4gl file state_name.4gl
containing the state_name() function.
12.4.4 Generate the configuration file

With the details of the Web Service stored in the design database we can
proceed to generate the configuration file (.4cf) for a Publish operation. The
Generate option is inside the Install menu option.
Example 12-14 shows the Generate form. You must fill out every field in the form.
If you have stored the definition for all those items in the design database, you
can retrieve them using the Ctrl+B key shortcut.
Example 12-14 Generate form
+------------------------------------------------------------------------------+
|INSTALL: Generate Deploy Package Exit
|
|Generate the configuration file for a web service.
|
|
|
|
+-----------------------------------------------------+ |
405
7884ch12.fm

|
|DATABASES:
Query Previous query Exit
| |
| Service ID [
4]
M|Make a new selection
| |
|
|-----------------------------------------------------| |
| Service Name
[ws_s|
DATABASE INFORMATION
| |
| Host Name
[irk |
| |
| Temp Directory [/tmp|ID
Database Name
| |
|
[
|[
1] [stores7
]| |
| App-Server Name [axis|[
] [
]| |
| Port Number
[ 98|[
] [
]| |
| Database Name [
|[
] [
]| |
| Database Server [
|[
] [
]| |
|
|[
] [
]| |
|
|
| |
|
|Arrow key - Press Esc to Accept or press Ctrl+C to Ca| |
|
+-----------------------------------------------------+ |
|
|
+------------------------------------------------------------------------------+
After all the information is entered, use the Generate option to create the
configuration file. Example 12-15 shows the output of the Generate task.
Example 12-15 Generate option
+------------------------------------------------------------------------------+
|
|Generate the configuration file for a web service.
|
|
|
|
GENERATE CONFIGURATION
|
|
|
| Service ID [
4]
Machine ID [
1]
Server ID [
1] Database ID [
1] |
|
|
| Service Name
[ws_state_name
] |
| Host Name
[irk
] |
| Temp Directory [/tmp/w4gl_informix
] |
|
[
] |
| App-Server Name [axis
] |
| Port Number
[ 9876]
|
| Database Name [stores7
] |
| Database Server [irk1150
] |
|
|
|
|
|
|
|
|
| Generated configuration file ws_state_name_irk.4cf
|
+------------------------------------------------------------------------------+
The Generate operation creates the configuration file required for publishing a
Web Service. The name of the file is constructed using the Web Service name,
the Host name, and .4cf extension. In our example the name of the file is
ws_state_name_irk.4cf
406
7884ch12.fm
Example 12-16 shows the configuration file for the ws_state_name Web Service.
Example 12-16 ws_state_name_irk.4cf file
[SERVICE]
TYPE = publisher
INFORMIXDIR = /usr3/4gl750uc3
DATABASE = stores7
CLIENT_LOCALE = en_us.utf8
DB_LOCALE = en_US.819
INFORMIXSERVER = irk1150
HOSTNAME = irk
PORTNO =
9876
I4GLVERSION = 7.50.xC3
WSHOME = /usr3/4gl750uc3/AXIS2C
WSVERSION = axis
TMPDIR = /tmp/w4gl_informix
SERVICENAME = ws_state_name
[FUNCTION]
NAME = state_name
[INPUT]
[VARIABLE] NAME = code TYPE = CHAR(2) [END-VARIABLE]
[END-INPUT]
[OUTPUT]
[VARIABLE] NAME = sname TYPE = CHAR(15) [END-VARIABLE]
[END-OUTPUT]
[END-FUNCTION]
[DIRECTORY]
NAME = /work
FILE = state_name.4gl,
[END-DIRECTORY]
[END-SERVICE]
A configuration file (4cf) can be used with the Web 4GL compiler (w4glc) tool to
perform tasks, such as generate and compile, directly from the command line.
Example 12-17 shows how to perform all the steps that the w4gl deploy options
does using the w4glc utility.
Example 12-17 w4glc use
informix@irk:/work$ w4glc -generate -compile -deploy ws_state_name_irk.4cf
The wrapper file is /tmp/w4gl_informix/state_name_wrap.c
407
7884ch12.fm

Generating shared object for service ws_state_name ....
Compiling code. Please wait...
Executing: c4gl --shared -o /tmp/w4gl_informix/libws_state_name.so
-I/usr3/4gl750/AXIS2C/include/axis2-1.5.0 -L/usr3/4gl750/AXIS2C/lib
-laxis2_engine -laxutil -laxis2_axiom /tmp/w4gl_informix/state_name_wrap.c
/tmp/w4gl_informix/axis2_skel_ws_state_name.c
/tmp/w4gl_informix/axis2_svc_skel_ws_state_name.c /work/state_name.4gl
/usr3/4gl750/lib/tools/w4glutil.a /usr3/4gl750/lib/tools/lib4gl.a
Compilation done.
Deploying service ws_state_name ...
Deploying the service. Please wait ...
Copying /tmp/w4gl_informix/libws_state_name.so ...
Copying /tmp/w4gl_informix/services.xml ...
Copying /tmp/w4gl_informix/state_name.wsdl ...
Service name: ws_state_name
Generating shared object for service ws_state_name ....
Executing: c4gl --shared -o /tmp/w4gl_informix/libws_state_name.so
-laxis2_engine -laxutil -laxis2_axiom /tmp/w4gl_informix/state_name_wrap.c
/tmp/w4gl_informix/axis2_skel_ws_state_name.c
/tmp/w4gl_informix/axis2_svc_skel_ws_state_name.c /work/state_name.4gl
/usr3/4gl750/lib/tools/w4glutil.a /usr3/4gl750/lib/tools/lib4gl.a
Compilation done.
Deploying service ws_state_name ...
Deploying the service. Please wait ...
Copying /tmp/w4gl_informix/libws_state_name.so ...
Copying /tmp/w4gl_informix/services.xml ...
Copying /tmp/w4gl_informix/state_name.wsdl ...
Service name: ws_state_name
Deployed at : /usr3/4gl750/AXIS2C/services/ws_state_name
Removing
Removing
Removing
Removing
Removing
Removing
Removing
/tmp/w4gl_informix/state_name_wrap.c ...
/tmp/w4gl_informix/state_name.wsdl ...
/tmp/w4gl_informix/axis2_skel_ws_state_name.h ...
/tmp/w4gl_informix/axis2_svc_skel_ws_state_name.c ...
/tmp/w4gl_informix/services.xml ...
/tmp/w4gl_informix/axis2_skel_ws_state_name.c ...
/tmp/w4gl_informix/libws_state_name.so ...
The output of the command line tool shows all the steps done while creating and
compiling the wrapper function. When errors appear during the deployment
408
7884ch12.fm
process, running each individual step with the w4glc may help to diagnose the
reason for the problem.
12.4.5 Deployment of the Web Service

This process generates the auxiliary code required to link the Axis2 application
server with the IBM Informix 4GL function. During the deployment, the C code for
the Web Service is created and compiled automatically using the 4Gl libraries.
The result of this process is a shared library ready to be use in the application
server.
This process also creates a Web Services Description Language (WSDL) file for
the Web Service and copy this file, together with the Web Service shared library,
into the application server directory.
The Deploy option is part of the Install menu. Example 12-18 shows the
deployment of the ws_state_name Web Service.
Example 12-18 Deploy form
+------------------------------------------------------------------------------+
|
|Deploy the web service.
|
|
|
|
CONFIGURATION TO DEPLOY
|
|
|
| File Name [ws_state_name_irk.4cf
]
|
|
[
]
|
|
[
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Deployed ws_state_name.
|
+------------------------------------------------------------------------------+
After a successful deployment, the Web Service would be ready to be consume

by any application. The files are automatically copied into the Axis2 services
directory.
409
7884ch12.fm
Example 12-19 on page 410 shows the Web Service files on the application
server.
Example 12-19 Application Server services directory
informix@irk:/work$ ls $INFORMIXDIR/AX*/services/ws_state_name
libws_state_name.so services.xml state_name.wsdl
informix@irk:/work$
12.4.6 Packaging of the Web Service

Packaging a Web Service is the process of creating a compress file with all the
components required by the Web Service. The compressed file, in tar format, can
be used on other Axis2 servers.
Example 12-20 shows the Package menu option.
Example 12-20 Package option.
+------------------------------------------------------------------------------+
|
|Package the web service.
|
|
|
|
SERVICE PACKAGING
|
|
|
| File Name [ws_state_name_irk.4cf
] |
|
[
] |
|
[
] |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Packaging successful. Check $TMPDIR directory (in config file) for .tar file |
+------------------------------------------------------------------------------+
The compressed file is left in the $TMPDIR directory. Example 12-21 shows the
contents of the packaged file for the ws_state_name Web Service.
Example 12-21 Packaged tar file
informix@irk:/work$ tar tvf /tmp/w4gl_informix/ws_state_name.tar
drwxr-xr-x informix/informix 2010-07-05 21:01 ws_state_name/
410
7884ch12.fm
-rw-r--r-- informix/informix 2010-07-05 21:01 ws_state_name/services.xml

-rw-r--r-- informix/informix 2010-07-05 21:01 ws_state_name/state_name.wsdl
-rwxr-xr-x informix/informix 2010-07-05 21:01 ws_state_name/libws_state_name.so
informix@irk:/work$
12.4.7 Starting the Axis2 application server

An Axis2 application server is included with the Informix 4GL package. It is
located in the AXIS2C directory under $INFORMIXDIR.
The Axis2 application server must be started to consume a Web Service. It is not
required during the development or deploy process.
Example 12-22 demonstrates how to start the Axis2 application server.
Example 12-22 Starting Axis2
informix@irk:/usr3/4gl750/AXIS2C/bin$ ./axis2_http_server -p 9876
Started Simple Axis2 HTTP Server ...
For more information about the Axis2 application server, refer to the Axis2/C
Manual at
http://ws.apache.org/axis2/c/docs/axis2c_manual.html
12.4.8 Consuming the I4GL Web Service

The main feature of a Web Service is that it can be used by any application in any
platform. A Web Service is not tied to the technology or programming language
used to develop the service. Any application that supports the Simple Object
Access Protocol or SOAP, can make use of the functions implemented inside a
Web Service. SOAP is a simple XML-based protocol used to exchange
information over a HTTP link.
In this section we demonstrate how to use the ws_state_name Web Service using
basic Java application.
Getting a list of available Web Services

You can get a list of available Web Services in the Axis2 application server by
opening a web browser and connecting to the Axis2 server.
http://hostname:port/axis/services
Figure 12-2 on page 412 shows the Web Services available in our example.
411
7884ch12.fm
Figure 12-2 Axis2 Web Services
You can use the ?wsdl keyword to retrieve the WSDL file as follows:
http://hostname:port/axis/services/ws_state_name?wsdl
Figure 12-3 shows the WSDL information for the ws_state_name service.
Figure 12-3 WSDL information for the ws_state_name service
412
7884ch12.fm
Java application
The main advantages of Web Services is that the application that uses the
services does not need to know anything about how the service is implemented.
The information needed for consuming a Web Service is defined in the WSDL
file, name of the operations it supports and parameters it requires.
Example 12-23 demonstrates how the I4GL Web Service ws_state_name can be
used from a simple Java application.
Example 12-23 state_name.java
informix@irk:/work$ cat state_name.java
import org.apache.axis.client.Call;
import org.apache.axis.client.Service;
import javax.xml.namespace.QName;
public class state_name {
public static void main(String [] args) {
try {
String endpoint = "http://irk:9876/axis/services/ws_state_name";
String qname = "http://www.ibm.com/ws_state_name";
Service service = new Service();
Call
call
= (Call) service.createCall();
call.setTargetEndpointAddress( new java.net.URL(endpoint) );
call.setOperationName(new QName(qname, "state_name"));
String ret = (String) call.invoke( new Object[] { args[0] } );
System.out.println("Sent 'CA', got '" + ret + "'");
} catch (Exception e) {
System.err.println(e.toString());
}
}
}
informix@irk:/work$ javac state_name.java

informix@irk:/work$ java state_name CA
Sent 'CA', got 'California
'
informix@irk:/work$ java state_name AR
Sent 'CA', got 'Arkansas
'
informix@irk:/work$
413
7884ch12.fm
12.5 Consuming a Web Service with I4GL

In this section we demonstrate all the steps required to consume a Web Service
using IBM Informix 4GL.
12.5.1 Web service to consume

You can use the w4gl tool to insert details about the specific Web Service you
want to consume. The menu option Subscribe allows you to manage definitions
for Web Services in the design database.
Example 12-24 shows the Subscribe form with the details referring to the
ws_state_name Web Service.
Example 12-24 Subscribe form
+------------------------------------------------------------------------------+
|SUBSCRIBE:
Query Next Previous Add Modify Remove Compile Exit
|
|Specify a new web service definition.
|
|
|
| Subscription ID
[
3]
|
| WSDL Path
[http://irk:9876/axis/services/ws_state_name?wsdl
] |
|
[
] |
| Webservice Function[state_name
] |
| I4GL Function
[statename
] |
| Target Directory [/work/publish
] |
| Target File Name [statename
] |
| Notes
[Wrapper for ws_state_name Web Service
] |
|
[
] |
|
[
] |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
The details for this form are:

Subscription ID: This is an automatic generated identifier of the Web Service.
WSDL Path: This is the complete path of the WSDL file for the service. This
can be a local file or a URL.
Web service Function: This is the function name provided by the Web
Service.
414
7884ch12.fm
I4GL Function: This is the name of the wrapper C function. This is the function
our I4GL would use to invoke the Web Service operation.
Target Directory: This is the directory where the wrapper and configuration
files are created.
Target File Name: This is the name of the C source file with the I4GL function.
The configuration file for a publish operation is automatically generated by the
w4gl tool, but it can also be created using the wsdl_parser as shown in
Example 12-25.
Example 12-25 Generating configuration file
informix@irk:/work$ wsdl_parser 0
http://irk:9876/axis/services/ws_state_name?wsdl state_name wsstatecode
`pwd`/publish statecode.c
informix@irk:/work$ ls publish
local.wsdl statecode.c_wsstatecode.4cf
informix@irk:/work$ cat publish/statecode.c_wsstatecode.4cf
[SERVICE]
TYPE = subscriber
I4GLVERSION = 7.50.xC4
WSHOME = 0
TARGET_DIR = /work/publish
I4GL_FUNCTION = wsstatecode
TARGET_FILE = statecode.c
[WSDL_INFO]
WSDL_PATH = /work/publish/local.wsdl
WSDL_NAME_SPACE = http://www.ibm.com/state_name
[FUNCTION]
SERVICENAME = ws_state_name
NAME = state_name
[INPUT]
[VARIABLE] NAME = code TYPE = char(2) [END-VARIABLE]
[END-INPUT]
[OUTPUT]
[VARIABLE] NAME = sname TYPE = char(15) [END-VARIABLE]
[END-OUTPUT]
[END-FUNCTION]
[END-WSDL_INFO]
[END-SERVICE]
12.5.2 Compiling the wrapper code

The next step to consume a Web Service is to generate and compile the wrapper
code. This is achieved using the Compile menu option of the w4gl tool.
415
7884ch12.fm
Example 12-26 shows the w4gl screen after compiling the wrapper code for the
ws_state_name Web Service.
Example 12-26 Compile option.
+------------------------------------------------------------------------------+
|SUBSCRIBE:
Query Next Previous Add Modify Remove Compile Exit
|
|Compile the subscriber client code.
|
|
|
| Subscription ID
[
3]
|
| WSDL Path
[http://irk:9876/axis/services/ws_state_name?wsdl
] |
|
[
] |
| Webservice Function[state_name
] |
| I4GL Function
[statename
] |
| Target Directory [/work/publish
] |
| Target File Name [statename
] |
| Notes
[Wrapper for ws_state_name Web Service
] |
| +---------------------------------------------------------------------------+|
| |
||
| |Subscriber code has been compiled successfully.
||
| +---------------------------------------------------------------------------+|
|
|
|
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
After this process is complete, the C files for the wrapper function, including the
object file containing the function code, are left in the Target Directory specified in
the Web Service definition.
Example 12-27 shows the files for the ws_state_name Web Service. These files
can be used from an Informix 4GL program to consume the Web Service.
Example 12-27 Web Service files
informix@irk:/work/publish$ ls
axis2_stub_ws_state_name_statename.h
statename.c
informix@irk:/work/publish$
statename.c_statename.4cf
statename.o
12.5.3 Using the Web Service from an I4GL application

To consume a Web Service from I4GL we have to link the object file for the
wrapper function in our 4GL code, or use the C source file generated during the
subscribe compile process.
416
7884ch12.fm
Example 12-28 shows a simple 4GL program which uses the wrapper function
for the ws_state_name Web Service. The wrapper function can be called like any
other C function in I4GL.
Example 12-28 i4gl subscriber code. wsstate.4gl
informix@irk:/work/publish$ cat wsstate.4gl
MAIN
DEFINE sname CHAR(15)
WHENEVER ERROR STOP
CALL statename("CA") RETURNING sname
DISPLAY "State name: ",sname
END MAIN
informix@irk:/work/publish$ c4gl wsstate.4gl statename.c -o wsstate
$INFORMIXDIR/lib/tools/w4glutil.a -I$AXIS2C_HOME/include/axis2-1.5.0
-L$AXIS2C_HOME/lib -laxis2_engine
informix@irk:/work/publish$ ./wsstate
State name: California
informix@irk:/work/publish$
For more information about Informix 4GL, refer to the 4GL Reference Manual at
http://publib.boulder.ibm.com/infocenter/ifxhelp/v0/index.jsp?topic=/com.ibm.to
ols.doc/4gl.html.
12.6 Troubleshooting
In this section we discuss typical problems seen when developing a Web Service
using the Informix 4GL Web Services tools and how to obtain diagnostic
information through the use of tracing or log files.
12.6.1 Typical problems

Here we discuss a few frequently seen problems.
Connection
The connection errors normally appears when there is something wrong in the
configuration details for the development environment or the Web Service.
Any Web Service developed in I4GL requires the use of the Informix 4GL
communication libraries for the database connection. I4GL tools like w4gl keep
the details about the environment in the Design database, therefore, it is critical
417
7884ch12.fm
that the communication with the IBM Informix database was configure correctly
before starting any project.
Environment variables such as INFORMIXDIR and INFORMIXSQLHOSTS must contain
valid details for the Informix database server to use during development.
Other things to check if there are problems related to connection include:
Informix database server details. Informix Web Services tools and Web
Services running on the Axis2 server uses the database information defined
in the Design database.
Connection information about the Axis2 server such as host name or port
number is stored as part of an Application Server definition. Check if these
values are valid.
Consuming Web Services with an 4GL application requires a TCP connection
from the I4GL program to the application server running the Web Service.
Make sure that the location of the Web Service is correct and the Application
Server can be reached from the I4GL process.
Compilation errors
A common reason for having this type of errors is due to an incorrect setup or
incorrect use of the Informix 4GL Web Services utilities.
Environment variables such as PATH and LD_LIBRARY_PATH (or the suitable
variable for the platform) may cause compilation and runtime errors while
developing a Web Service with Informix 4GL. When a compile problem appears
while using one of the I4GL Web Service tools, the information about the error is
written into the W4GL log file. The default name and location of this file is
/tmp/w4glerr.log.
Example 12-29 shows an error message generated by the w4gl tool during the
Deploy process.
Example 12-29 w4gl error.
The file "state_name.err" has been written.
-CDCAK0012: The web service not deployed. Check error log '/tmp/w4glerr.log'
The Deploy menu option performs several operations automatically including the
generation of the wrapper code, compiling, and moving the Web Services files
into the Axis2 server. To know which specific task is failing, use the Web compiler
tool (w4glc) to manually perform each individual task.
By default, the w4glc script deletes the temporary files used to perform a task, so
use the -keep flag to avoid files deletion.
418
7884ch12.fm
Example 12-30 shows how to run the Generate process using the w4glc tool.
Example 12-30 w4glc generate keeping the files
informix@irk:/work$ w4glc -generate -keep ws_state_name_irk.4cf
With the wrapper files in the temporary directory, you can execute the Compile
process from the command line using w4glc and examine the output for any
errors. Example 12-31 shows the output of the Compile process when an
compilation error occurs.
Example 12-31 Compilation output
informix@irk:/work$ w4glc -compile -keep ws_state_name_irk.4cf
Generating shared object for service ws_state_name ...
Executing: c4gl -keep --shared -o /tmp/w4gl_informix/libws_state_name.so
-laxis2_engine -laxutil -laxis2_axiom...
The compilation was not successful. Errors found: 1.
The file "state_name.err" has been written.
Error executing: c4gl -keep --shared -o /tmp/w4gl_informix/libws_state_name.so
-I/usr3/4gl750/AXIS2C/include/axis2-1.5.0 ...
/usr3/4gl750/lib/tools/w4glutil.a /usr3/4gl750/lib/tools/lib4gl.a at
/usr3/4gl750/lib/globals.pm line 583.
Error executing: c4gl -keep --shared -o /tmp/w4gl_informix/libws_state_name.so
-laxis2_engine -laxutil ...
informix@irk:/work$
Similar to any I4GL program, if the error was inside the Informix 4GL code, a .err
file is created containing the error message. Example 12-32 shows the content of
the state_name.err file.
Example 12-32 state_name.err
informix@irk:/work$ cat state_name.err
FUNCTION state_name(code)
419
7884ch12.fm

DEFINE state_rec RECORD
code CHAR(2),
sname CHAR(15)
END RECORD,
code CHAR(2),
sel_stmt CHAR(100);
LET sel_stmt= "SELECT code, sname FROM state WHERE code = ?";
PEPARE st_id FROM sel_stmt;
|________^
|
|
A grammatical error has been found on line 11, character 10.
| The construct is not understandable in its context.
| See error number -4373.
DECLARE cur_id CURSOR FOR st_id;
OPEN cur_id USING code;
FETCH cur_id INTO state_rec.*;
...
Keeping the wrapper C files in the temporary directory might be useful when the
compilation error appears inside the Axis2 functions.
Example 12-33 shows the temporary files created with the Generate option for
the ws_state_name Web Service.
Example 12-33 temporary files
informix@irk:/work$ ls /tmp/w4gl_informix/
axis2_skel_ws_state_name.c
services.xml
axis2_skel_ws_state_name.h
state_name_wrap.c
axis2_svc_skel_ws_state_name.c state_name.wsdl
informix@irk:/work$
tmpXMLReqFile
ws_state_name.tar
While compiling any Web Service, during creation or consume, make sure all the
required libraries and include files are correctly passed to the compiler. These
requirement may vary from platform to platform, so, always check the release
notes.
The release and documentation files for Informix 4GL are located in the
release/en_us/0333 directory inside INFORMIXDIR. These files contain
additional information relevant to the version of I4GL installed such as known
issues or compiler requirements which may help diagnosing the problem.
420
7884ch12.fm
Consuming the Web Service

This types of errors may be caused by an invalid Web Service definition, items
such as function name or parameters type are defined in the configuration file
and used to create the WSDL file the Application Server would use to define the
Web Service.
You can retrieve the WSDL information for a Web Service using the location of
the service and the ?wsdl suffix from any Web browser.
Figure 12-4 shows the WSDL information for the ws_state_name Web Service.
Figure 12-4 WSDL for ws_state_name service.
Make sure the operations, parameters, and data types are correct for the I4GL
code the function uses.
Testing a Web Service

Informix I4GL does not provide any specific tool for testing Web Services. While
you can create a simple 4GL code to consume the Web Service, depending on
the complexity of the Web Service it could be useful to perform a more complete
testing.
You can use open source tools such as soapUI to test your Web Service before
publish it on the production environment. For more information about soapUI
refer to the soapUI at
http://www.soapui.org/
421
7884ch12.fm
Figure 12-5 on page 422 shows the soapUI program testing the ws_state_name
Web Service.
Figure 12-5 soapUI test
12.6.2 Tracing
There are two types of tracing a developer can use when diagnosing an Informix
4GL Web Service problem:
Application server trace, can be used to diagnose problems between a client
application and the Axis2 application server.
Database trace, used when diagnostic problems specific to operations with
the database server such as SQL errors or incorrect data being returned.
Application server trace

The Axis2c application server provides a method to trace all the exchanged
messages between the application server and a client consuming a web service.
You can set this trace using the -l log_level option when the Axis2 server is
started. To launch the Axis2 server with trace enable, run the following command:
422
7884ch12.fm
axis2_http_server -p 9876 -l 6 -f /tmp/w4gl_informix/axis_trace.log
The option -l is used to specify the logging level for the application server. The
maximum value is 6 that enables full tracing.
You can specify the location of the trace file using the -f log_file option.
Example 12-34 shows the contents of a typical trace file.
Example 12-34 application_trace sample
[debug]
[debug]
[debug]
[debug]
[debug]
[debug]
[debug]
[debug]
...
phase_resolver.c(139) Service name is : ws_zipcode

phase_resolver.c(1123) Operation name is : zipcode_details
phase_holder.c(139) Add handler AddressingOutHandler to phase
phase_resolver.c(222) svc name is:ws_state_name
phase_resolver.c(139) Service name is : ws_state_name
phase_resolver.c(1123) Operation name is : state_name
For more information about the logging options with the Axis2c application
server, refer to the Apache Axis2/C Manual at
http://ws.apache.org/axis2/c/docs/axis2c_manual.html#simple_axis_server
Database trace
IBM Informix 4GL uses the SQLI protocol to exchange data with the Informix
server. This means that any Web Service developed using Informix 4GL also
uses the SQLI protocol for any communication with the database server.
The environment variable SQLIDEBUG can be used to collect all the messages
between the Application server and the Informix database server.
To enable this trace, create the SQLIDEBUG environment variable before
starting the Axis2 Application server.
Example 12-35. demonstrates how to set SQLIDEBUG and run sqliprint to
un-encode the SQLI file.
Example 12-35 sqlidebug client side
informix@irk:/usr3/4gl750/AXIS2C/bin$ export SQLIDEBUG=2:/tmp/sqlitrace
informix@irk:/usr3/4gl750/AXIS2C/bin$ ./axis2_http_server -p 9876
Started Simple Axis2 HTTP Server ...
...
...
informix@irk:/work$ ls /tmp/sqlitrace*
423
7884ch12.fm

/tmp/sqlitrace_17008_0_8c819d0
informix@irk:/works$ sqliprint -o tracefile.txt /tmp/sqlitrace_17008_0_8c819d0
informix@irk:/works$
Note: The sqliprint tool is included in the Informix Client SDK package.
The SQLIDEBUG trace can also be used at the server side. For more
information, refer to 3.3.6, Troubleshooting on page 113.
424
7884ch13.fm
13
Chapter 13.
Application development
considerations
In this chapter we examine some of the considerations a developer may need to
address in a multi-user environment. A single user workstation which connects to
an exclusive-use database does not have to worry about the issue of two or more
independent uses of the same data at the same time. However, in a multi-user
environment, concurrency is a challenge.
IBM Informix database servers are designed to provide features to help in
handling concurrency and sorting facilities. The application developer should
design applications to take advantage of these built-in features, rather than
attempt to implement their own facilities in the application. In this chapter we
examine the factors that cause concurrency problems, and focus on ways to
keep the scope and duration of locks to a minimum. To do this, we consider
isolation levels, sharing data, and data contention issues that can occur when
two or more attempts are made to access or change the same row of data.
In the last two sections, we focus on configuration parameters that effect the
application development, and how to monitor issues when the applicaiton
developers work with the database administrator to tune the engine and the
application to work effectively.
425
7884ch13.fm
13.1 Concurrency and locking

Concurrency involves two or more independent uses of the same data at the
same time. In a database system with many users, each user needs to be able to
access and modify data. Unless the developer and database system impose
controls, there can be negative consequences. Programs might access old data
that is in the process of being changed by another user, and changes may seem
to disappear even though it seems like the change was performed successfully.
To take advantage of database server controls, tables in a multiple user
environment should be logging tables. At a minimum, if you must use a
nonlogging table within a transaction, either set Repeatable Read isolation level
or lock the table in exclusive mode.
To avoid concurrency problems, the database server imposes a system of locks.
A lock is a claim, or reservation, that a program can place on a piece of data. The
database server assures that no other program can modify it, as long as the lock
is in place. When another application requests the data, the database server
either makes the program wait or turns it back with an error.
The application developer can control the effect of lock access using a
combination of SQL statements and with the buffering mode selected for the
database. The most used SQL statements are SET LOCK MODE, SET
ISOLATION, and SET TRANSACTION. These are discussed in more detail later.
For now, we need a better understanding of the types of locks that we can
encounter.
IBM Informix offers several server edition. IBM Informix Extended Parallel Server,
Informix Online, and Standard Engine will have different syntax for concurrency
related commands. The command syntax for the discussion in this chapter
follows is found in the IBM Informix Guide to SQL: Syntax, v11.50,
SC27-3611-00.
13.1.1 Types of locks

A lock is implemented as a variable associated with a data item. It can be
explicitly placed by an application, or, more frequently, it is implicitly handled by
the database management system. The lock is used to mark a data item as
reserved; the type of lock designation determines what actions are permitted by
users in regard to the data item.
In an IBM Informix instance, there are several types of locks:
Shared lock
426
7884ch13.fm
A shared lock reserves its object for reading only. It prevents the object from
changing while the lock remains. More than one program can place a shared
lock on the same object. More than one object can read the record while it is
locked in shared mode. In the lock list output visible from onstat -k, a thread
with a shared lock is be designated with an S. If an object is currently locked
in exclusive mode and the user thread wants to acquire a shared lock, the
designation is IS (intent-shared).
When a session first connects to a database, IBM Informix Servers place a
shared lock on the database, to prevent another session from acquiring an
exclusive lock on the same database. SELECT queries place a shared lock at
the the table level, since it is faster for the engine to find a table level lock than
it is to search through potentially thousands of row locks.
Intent-exclusive lock
Intent locks are automatically set by Informix. If a row in a table is updated, an
exclusive lock is placed on the row and an intent-exclusive lock is placed on
the table. This assures that no other session can place a shared or exclusive
lock on the table as long as an individual row is locked exclusively. In the lock
list output visible from onstat -k, a thread with an intent exclusive lock is
designated with an IX. A related designation that is sometimes visible is
SIX. This designation indicates the object is currently shared, with
Intent-exclusive when the chance arrives.
Exclusive lock
An exclusive lock reserves its object for the use of a single application. This
lock type is used when the application needs to change the object. You
cannot place an exclusive lock where any other kind of lock exists. After you
place an exclusive lock, you cannot place another lock on the same object. In
the lock list output visible from onstat -k, a thread with an exclusive lock is
designated with an X.
Update lock
A promotable (or update) lock establishes an intent to update. You can only
place it where no other updatable or exclusive lock exists. You can place an
updatable lock on records that already have shared locks. When the
application is about to change the locked object, you can promote the update
lock to an exclusive lock, but only if no other locks, including shared locks, are
on the record at the time the lock would change from update to exclusive. If a
shared lock was on the record when the update lock was set, you must drop
the shared lock before the update lock can be promoted to an exclusive lock.
In the lock list output visible from onstat -k, a thread with an update lock is
designated with an U.
Chapter 13. Application development considerations
427
7884ch13.fm
13.1.2 Lock duration

The length of time a lock remains in effect is known as lock duration. The
duration of a lock is determined by the application, the closing of a database, and
the type of transaction method used by the application and database
If the database does not use transactions (no transaction log exists and you do
not use a COMMIT WORK statement), an explicit table lock remains until it is
removed by the execution of the UNLOCK TABLE statement.
If database transactions are in use, the end-of-lock-duration event occurs when
the transaction ends or a COMMIT WORK is issued in the application. The
ending transaction causes a release of all table, row, page, and index locks that
were on hand during the transaction.
13.1.3 Lock granularity

With IBM Informix Servers, the developer can apply locks to databases, tables,
disk pages, data rows, or index-key values. At a database level, an exclusive lock
is simple to enforce for a developer, but it has a big impact on users. No one gets
access until the database lock processing is completed. Concurrency drops to
zero, and performance is maximized for the exclusive use of a single user
running an application.
At the other end of the granularity scope for locks, a transaction can exclusively
lock a row, and have no impact on other users in their work efforts (if they are not
trying to access the same row, and the lock mode is row). Concurrency is
maximized, but performance will tend to slide as the number of users for a
database or table increases.
Table locks
It is the responsibility of both database administrator and developer to select and
enable the best level of lock granularity for users and for their database system.
At the database level, administrative activities such as imports and exports are
usually the task of the database administrator. For such activity, the database
administrator would use a command such as
DATABASE database_name EXCLUSIVE;
Another task, principally for the database administrator, is to enable a change for
a table or an index structure. The task of enabling work on an entire table can be
done with a command similar to the following:
LOCK TABLE table_name IN EXCLUSIVE MODE;
or
428
7884ch13.fm
LOCK TABLE table_name in SHARE MODE;
When the task is finished (end of statement or transaction reached), the table is
implicitly unlocked. It could be unlocked explicitly with:
UNLOCK TABLE table_name;
Note: A table lock on a table can decrease update concurrency radically. Only
one update transaction can access that table at any given time, and that
update transaction locks out all other transactions. However, multiple
read-only transactions can simultaneously access the table. This behavior is
useful in a data warehouse environment where the data is loaded and then
queried by multiple users.
Page locks
Lock mode PAGE is the default for Informix tables, and it is considered the
optimal level in lock efficiency when rows are being accessed and modified in
physical order. If your tables are large in row count and small in row size, a page
level lock can be severely limiting, since it will lock a large number of rows on a
page, and discourage user access.
In this case, it would be more efficient to change the default lock mode. For all
new tables, this can be done by way of the Informix onconfig file, using the
parameter DEF_TABLE_LOCKMODE. For example:
DEF_TABLE_LOCKMODE ROW;
Note:
Use a small (or default) page size if your application contains small sized
rows. Increasing the page size for an application that randomly accesses
small rows can decrease performance. In addition, a page lock on a larger
page will lock more rows, which is likely to reduce concurrency in some
situations.
Tables that use page locks cannot support the USELASTCOMMITTED
concurrency feature.
Row and key locks

Row and key locks generally provide the best overall performance when you are
updating a relatively small number of rows, because they increase concurrency.
However, the database server incurs some overhead in obtaining a lock. For an
operation that changes a large number of rows, obtaining one lock per row may
not be cost effective. For operations that consistently change a large number of
rows, page locks may be a better option.
429
7884ch13.fm
If a table is not created with row locking and you want row or key locks, you must
alter the table. Here is an example to show how to create a table with row locking
on:
CREATE TABLE table_namer(field1 serial,field2 char(20)...)
LOCK MODE ROW;
The ALTER TABLE statement can also change the lock mode. An example for
this command syntax is:
ALTER TABLE table_name LOCK MODE (ROW);
When the lock mode is ROW and you insert or update a row, the database server
creates a row lock. In some cases, you place a row lock by simply reading the
row with a SELECT statement.
When the lock mode is ROW and you insert, update, or delete a key (performed
automatically when you insert, update, or delete a row), the database server also
creates a lock on the key in the index.
Key-value locks
When a user deletes a row within a transaction, the row cannot be locked
because it becomes a non-existent row. However, the database server must
somehow record that a row existed until the end of the transaction. The database
server uses key-value locking to lock the deleted row. Key locks are used
identically to row locks. When the table uses row locking, key locks are
implemented as locks on imaginary rows.
When the table uses page locking, a key lock is placed on the entire index page
that contains the key or that would contain the key if it existed. A page lock on an
index page can decrease concurrency more substantially than a page lock on a
data page. Index pages are dense and hold a large number of keys. By locking
an index page, you make a potentially large number of keys unavailable to other
users until you release the lock.
Note: To determine the current lock mode for a table:
On UNIX, try oncheck -pt dbname:tablename | grep Locking
On Windows, examine the output of oncheck -pt dbname:tablename
Database locks
The act of opening a database places a shared lock on the database name. The
statements which open a database are CONNECT, DATABASE, or CREATE
DATABASE. As long as a database is open, the shared lock on the database will
prevent any other program from dropping the database or putting an exclusive
lock on it.
430
7884ch13.fm
Locking a database for exclusive use are not needed very often, since this would
prevent other users and programs from accessing it for the duration of the lock.
The usual reason for a database lock would be the need for a major structure
change across several tables, when implementing a series of related indexes on
several tables, or when you have an application that needs uninterrupted access
to the database for a period of time. To lock a database in exclusive mode, the
syntax is DATABASE database_name EXCLUSIVE;
Tasks that run from the sysadmin database (dbscheduler tasks, such as Auto
Update Statistics), occasionally may prevent exclusive database access. In this
case, you must temporarily disable the dbscheduler. In the sysadmin database,
stop the scheduler API with:
execute function task(scheduler shutdown)
and restart the scheduler API with:

execute function task(scheduler start)
Smart-Large-Object locks
Smart large objects are quite different in the way they work from the rest of the
structures and processes in IBM Informix databases. The locking and locking
granularity for smart large objects is also different. The database server uses one
of the following granularity levels for locking smart large objects:
The sbspace chunk header partition
The smart large object
A byte range of the smart large object
The default locking granularity for a smart BLOB is at the level of the smart large
object. When you update a smart large object, the database server locks the
smart large object that is being updated. Concurrently, there is an update lock
placed on the sbspace chunk header partition while the object is being updated.
Byte locks
Byte locks, also known as byte-range locks, are used to lock a specific byte range
of a smart large object. Byte-range locking is advantageous because it allows
multiple users to update the same smart large object simultaneously, as long as
they are updating different parts of it. Also, users can read a part of a smart large
object while another user is updating or reading a different part of the same
smart large object.
How the database server manages byte-range locks

The database server manages byte-range locks in the lock table in a similar
fashion to other locks placed on rows, pages, and tables. However, the lock table
must store the byte range as well. If a user places additional locks on a byte
431
7884ch13.fm
range, any new byte locks in the range that are contiguous are consolidated into
one lock range.
Likewise, if a user unlocks a portion of the bytes included within a byte-range
lock, the database server will split into multiple byte-range locks.
To enable use of byte-range locks

By default, the database server places a lock on the entire smart large object. At
the time of sbspace creation, there is an option to use byte-range locking.
When the DBA sets the default locking mode for the sbspace to byte-range
locking, the database server locks only the necessary bytes when it updates any
smart large objects stored in the sbspace. To set byte-range locking for the
sbspace that stores the smart large object, the database administrator must use
the onspaces utility. The following example sets byte-range locking for a new
sbspace:
onspaces -c -S sblob -g 2 -p /ifmx/sblob1 -o 0 -s 1000 -Df LOCK_MODE=RANGE
When byte-range locking is set for the individual smart large object, the database
server implicitly locks only the necessary bytes when it selects or updates the
smart large object. The application developer can set byte-range locking for the
smart large object when it is opened, using one of the following methods:
Set the MI_LO_LOCKRANGE flag in the mi_lo_open() DataBlade API function.
Set the LO_LOCKRANGE flag in the ifx_lo_open() ESQL/C function.
To lock a byte range explicitly, use one of the following functions:
mi_lo_lock()
ifx_lo_lock()
These functions lock the range of bytes that is specified for the smart large
object. If the developer specifies an exclusive lock with either function, UPDATE
statements do not place locks on the smart large object if they update the locked
bytes.
The database server releases exclusive byte-range locks placed with
mi_lo_lock() or ifx_lo_lock() at the end of the transaction. The database
server releases shared byte-range locks placed with mi_lo_lock() or
ifx_lo_lock() based on the same rules as locks placed with SELECT
statements, depending upon the isolation level. The application can also release
shared byte-range locks with mi_lo_unlock() or ifx_lo_unlock().
For more information about these DataBlade API functions, see the IBM Informix:
DataBlade API Programmer's Guide at
432
7884ch13.fm
http://publib.boulder.ibm.com/infocenter/idshelp/v10/index.jsp?topic=/com.ibm.d
apip.doc/dapip.htm
You can fine additional details about the ESQL/C functions in the IBM Informix:
ESQL/C Programmer's Manual, located at
http://publib.boulder.ibm.com/infocenter/idshelp/v10/index.jsp?topic=/com.ibm.e
sqlc.doc/esqlc.htm
13.2 Locking issues and performance

Row and key locks generally provide the best performance whenever a database
system only needs to update a small number of rows at a time. However, the
database incurs additional overhead in obtaining a lock as the granularity gets
smaller. The developer should help decide the granularity based on the character
of the trnasactions.The following are guidelines for your decision making:
For low row count transactions, use row level locking.
For large processes that update an entire table, set locking to page or lock the
table in exclusive mode before processing.
For massive updates to many tables or the whole database, lock the database
in exclusive mode before processing.
The type of isolation can affect overall performance because it affects
concurrency. Before you execute a SELECT statement, you can set the isolation
level with one of these options:
The SET ISOLATION statement (an extension to ANSI SQL-92 standard)
Can be executed more than once in a transaction.
Can change the enduring isolation level for a session
Has an additional isolation level - Cursor Stability.
SET TRANSACTION (ANSI/ISO-compliant)
Can only be executed once in a transaction.
Dirty Read isolation (or ANSI Read Uncommitted) level does not place any
locks on any rows fetched during a SELECT statement. Dirty Read isolation is
appropriate for static tables that are used for queries. Use Dirty Read isolation
with care if update activity occurs at the same time. With Dirty Read, the
reader can read a row that has not been committed to the database and might
be eliminated or changed during a rollback.
433
7884ch13.fm
13.2.1 Deadlocks
A deadlock is a situation in which a pair of programs blocks the progress of each
other. Each program has a lock on some object that the other program wants to
access. In a database with local database queries only, a deadlock arises only
when all programs concerned set their lock modes to wait for locks.
An Informix database server detects deadlocks immediately when they only
involve data at a single network server. Before a lock is granted, the database
server examines the lock list for each user. If a user holds a lock on the resource
that the requestor wants to lock, the database server traverses the lock wait list
for the user to see if the user is waiting for any locks that the requestor holds. If
so, the requestor receives a deadlock error. It prevents the deadlock from
occurring by returning an error code (error -143 ISAM error: deadlock detected)
to the second program to request a lock. The error code is the one the program
receives if it sets its lock mode to not wait for locks. If your program receives an
error code related to locks even after it sets lock mode to wait, you know the
cause is an impending deadlock.
The best way to avoid deadlocks is to set lock mode to wait for a specific number
of seconds (N), where N is the number of seconds it is reasonable for your
applications to wait on a lock. Example syntax:
SET LOCK MODE TO WAIT 30;
Deadlock errors can be unavoidable when applications update the same rows
frequently. However, certain applications might always be in contention with each
other. Examine applications that are producing a large number of deadlocks and
try to run them at different times. To monitor the number of deadlocks, monitor
the deadlks field in the output of onstat -p.
Distributed transactions
Deadlock handling is more challenging when you are using a distributed query,
which involves more than one database server. In this case, the server cannot
monitor the locks of the other database, so deadlocks cannot be detected before
they occur. Occasionally, there may be a need for each database to wait on a
lock from another local transaction. When the wait occurs, the entire distributed
transaction may need to wait until the action is cleared up. If both servers get into
a wait-for-the-other-server to complete mode, you have a multi-server deadlock
which neither server can easily get out of. In order to avoid this as much as
possible, set the onconfig parameter DEADLOCK_TIMEOUT. This parameter
tells the server how much time to wait before returning an error code (error -143
ISAM error: deadlock detected). For more information on using this configuration
parameter, see your IBM Informix Dynamic Server Administrator's Guide,
v11.50, SC27-3606-00.
434
7884ch13.fm
Distributed transactions, also known as global transactions, can have additional

problems beyond concurrency and locking. In a distributed transaction, we can
have queries that include multiple, separate databases spanning multiple host
machines across a network. With the added task of reaching out to different
servers, we are confronted with delays caused by network traffic, connectivity
issues, additional user authentication, and user activity on the other database
server that can slow or temporarily prevent quick access to the remote database
server. In the context of a query that is doing an update, insert, select, or delete,
this new group of interfering factors can occasionally present problems.
Developers should include coding conventions to accommodate distributed (XA)
transactions and include code to avoid certain problems. We have included an
ESQL/C routine (xa_tool.ec) in Appendix B, Accommodating distributed
transactions on page 455, designed to help provide a start for handling XA
problems.
13.3 Isolation levels

The number and duration of locks placed on data during a SELECT statement
depends on the level of isolation that the user sets. The type of isolation can
affect overall performance because it affects concurrency.
You can set the isolation level with the SET ISOLATION or the ANSI SET
TRANSACTION statement before you execute the SELECT statement. The
main differences between the two statements are that SET ISOLATION has an
additional isolation level, Cursor Stability, and SET TRANSACTION cannot be
executed more than once in a transaction as SET ISOLATION can.
Dirty Read Isolation

Dirty Read isolation (or ANSI Read Uncommitted) does not place any locks on
any rows fetched during a SELECT statement. Dirty Read isolation is appropriate
for static tables that are used for queries. It offers the best performance of all
isolation levels, because the database server does not check or place any locks
for queries.
If update activity can occur at the same time as a dirty read, there is a chance
that a rollback may need to occur during the dirty read. The reader could read a
row that has not been committed (this is known as a phantom read) to the
database and the data would be lost during a subsequent rollback. Because of
potential problems with uncommitted data that is rolled back, use Dirty Read
isolation with care.
435
7884ch13.fm
Databases that do not have logging turned on (and hence do not allow
transactions) use Dirty Read as a default isolation level. In fact, Dirty Read is the
only isolation level allowed for databases that do not have logging turned on.
Committed Read isolation

Committed Read isolation (or ANSI Read Committed) removes the problem of
phantom reads. A reader with this isolation level checks for locks before it returns
a row. By checking for locks, the reader cannot return any uncommitted rows.
The database server does not actually place any locks for rows read during
Committed Read. It simply checks for any existing rows in the internal lock table.
Committed Read is the default isolation level for databases with logging, and it is
an appropriate isolation level for most activities.
Cursor Stability isolation

A reader with Cursor Stability isolation acquires a shared lock on the row that is
currently fetched. This action assures that no other user can update the row until
the user fetches a new row.
If you do not use a cursor to fetch data, Cursor Stability isolation behaves in the
same way as Committed Read. No locks are actually placed.
Repeatable Read isolation

Repeatable Read isolation (also known as ANSI Serializable and ANSI
Repeatable Read) is the strictest isolation level. With Repeatable Read, the
database server locks all rows examined (not just fetched) for the duration of the
transaction.
Repeatable Read is useful during any processing in which multiple rows are
examined, but nothing will change during the transaction. The original application
holds a read lock on each account that it examines until the end of the
transaction, so the attempt by the second application to change the first account
fails (or waits, depending upon SET LOCK MODE).
Note: With Repeatable Read, since even examined rows are locked, if the
database server reads the table sequentially, a large number of rows
unrelated to the query result can be locked.
Use Repeatable Read isolation for tables when the database server can use an
index to access a table. If an index exists and the optimizer chooses a sequential
scan instead, you can use directives to force use of the index. However, forcing a
change in the query path might negatively affect query performance.
436
7884ch13.fm
13.4 Configuration options

We have brought configuration parameters into our discussion several times. In
this section we examine IBM Informix server configuration options that directly
affect application development. We point out what the developers should know,
and how these parameters affect the application and transactions. For more
information on any of the onconfig parameters or environment variables relating
to configuration, see the IBM Informix Dynamic Server Administrator's
Reference, v11.50, SC27-3607-00.
13.4.1 Server identification

When the Informix server needs to resolve a database instance name, the
oninit process uses the contents of a file known as the sqlhosts file. Each line
in the sqlhosts file provides a database server naming reference, called the
dbservername, along with a protocol, a host machine reference, and a listening
protocol service reference. On Windows, this information is stored in the registry,
and is updated and accessed using the setnet32.exe utility. Multiple
dbservername references allow for an evenly distributed connection distribution
over a variety of protocols
In addition to the sqlhosts information, the dbservername crosses reference to
the onconfig file, which holds the configuration information describing how the
database server is to operate. When a client application connects to a database
server, it will specifiy a dbservername, which helps the server to know which
protocol will be used for the connectivity.
The dbservername used in the application connection must be consistent with a
name specified in sqlhosts and a dbservername defined in the onconfig file.
In the onconfig file there is a default (main instance) dbservername, and one or
more alternate dbservernames available. The DBA specifies the default name to
be used with the onconfig parameter DBSERVERNAME. For alternative
application processing, the DBA can allow for or specify alternative protocols,
referenced by an a different dbservername. The alternate dbservername or
names are specified by the onconfig parameter DBSERVERALIASES. An
application developer should check with the DBA to determine whether to use the
DBSERVERNAME or one of the DBSERVERALIASES for connecting with his
applcation. Example SQL statements include CONNECT, DATABASE, CREATE
TABLE, and ALTER TABLE, which can specify the database server instance.
437
7884ch13.fm
13.4.2 Storage space identifiers

When tables are created, the database creator has syntax options available to
define where tables and indexes are physically to be placed in a disk structure.
Typically the database administrator defines the disk structure layout. The disk
layout is assigned by way of dbspaces, and each dbspace is further broken down
into allocations of chunks. The dbspace is the parent structure unit, used to hold
storage information for tables and indexes, with low level definitions to describe
where the actual data pointers for row data resides, as well as information about
indexes, fragments, and table scope. The dbspace area that tracks pointer
locations for data is known as the table space.
When a database administrator defines the table and index schema, he can
specify a dbspace location to hold the table and/or the index. There are other
space specifications for default storage if none is specified at creation time.
These default locations define where to store binary large objects (BLOBs) and
character large objects (CLOBs), and temporary storage locations.
The default location for the location of temp tables and sorting space, if not
predefined, is the dbspace where a table is accessed, or the root dbspace
(rootdbs). If a default onconfig definition for the location of temp dbspace is
defined, we have more control over the location used
DBSPACETEMP
The DBSPACETEMP parameter specifies a list of dbspaces that the database
server uses to globally manage the storage of temporary tables. When a
temporary space is available, it improves performance by enabling the database
server to spread out I/O for temporary tables across multiple disks efficiently. The
database server also uses temporary dbspaces during backups to store the
before-images of data that are overwritten while the backup is occurring. More
than one dbspace can be specified for this parameter. Simply list them as comma
separated values.
DBSPACETEMP can contain dbspaces with a non-default page size, but all of
the dbspaces in the DBSPACETEMP list must have the same page size.
When using a logged database, file activity in temporary dbspaces is not logged.
If the developer writes a query that requires tempspace, it is useful to include the
phrase WITH NO LOG, so that you can force the query to use the designated
temporary dbspace.
SBSPACENAME
The SBSPACENAME parameter specifies the name of the default sbspace. If
your database tables include smart-large-object columns that do not explicitly
438
7884ch13.fm
specify a storage space, that data is stored in the sbspace that SBSPACENAME
specifies. The default sbspace is also used by the built-in encryption and
decryption functions to store BLOB or CLOB values.
An sbspace is a specialized type of dbspace, used for storing smartblobs, binary
large object dbspaces. (Binary data or some type of multidimensional data is
common to any datablade which uses R-tree indexing).
When a create table statement includes a column that defines a CLOB or BLOB
object, the column information will be stored in an sbspace. If the PUT clause is
not specified in the create table statement when the defined CLOB or BLOB is
created, the default location where the column data will be stored is the sbspace
designated by SBSPACENAME.
If you are using IBM Informix with J/Foundation, you must provide a smart large
object where the database server can store the Java archive (JAR) files. These
JAR files contain the Java user-defined routines (UDRs). If you use Java UDRs,
you will want to create additional, separate sbspaces for storing smart large
objects.
SBSPACETEMP
The SBSPACETEMP parameter specifies the name of the default temporary
sbspace for storing temporary smart large objects without metadata or user-data
logging. If you store temporary smart large objects in a standard sbspace, the
metadata is logged. For more information on using temporary smart large
objects, see the IBM Informix DataBlade API Programmer's Guide, V11.50,
SC23-9429-03.
SYSSBSPACENAME
The SYSSBSPACENAME parameter specifies the name of the sbspace in which
the database server stores statistics that the UPDATE STATISTICS statement
collects for certain user-defined data types. Normally, the database server places
statistics in the sysdistrib system catalog table.
Note: For information on writing user-defined statistics, see the performance
chapter in IIBM Informix User-Defined Routines and Data Types Developer's
Guide, V11.5, SC23-9438-00.
For information on providing statistics data for a smart BLOB column, see the
IBM Informix DataBlade API Programmer's Guide, V11.50, SC23-9429-03.
439
7884ch13.fm
13.4.3 Limiters and limits

In addition to the storage location identifiers mentioned in 13.4.2, Storage space
identifiers on page 438, a number of onconfig parameters impose resource
limits. Limits help the engine and machine to utilize appropriate resources rather
than consume all the resources on the machine and end up in a crash. When the
resources are no longer needed, they are released to be used by other
processes. These configuration limits can be adjusted in cooperation with the
database administrator, to accommodate for application concerns that may need
them.
MULTIPROCESSOR
The MULTIPROCESSOR parameter is used to determine the type of processing
method the engine will use. If it is set to 0, locking is done in a way that is suitable
for a single- processor computer, and processor affinity is ignored.
If the nearby onconfig parameter SINGLECPU_VP is non-zero (On),
MULTIPROCESSOR and user defined VPCLASSes (of any kind) will not work.
VPCLASS
The VPCLASS parameter allows you to designate a class of virtual processors
(VPs), create a user-defined VP, and specify the following information for it:
The number of virtual processors that the database server should start
initially.
The maximum number of virtual processors allowed for this class.
The assignment of virtual processors to CPUs if processor affinity is available.
The disabling of priority aging by the operating system if the operating system
implements priority aging.
You can have multiple VPCLASS parameter definitions. Use one VPLCLASS
reference for each class of virtual processor, one per line. Basic guidelines for
defining a VPLCASS are in the IBM Informix Dynamic Server Administrator's
Reference, V11.50, SC27-3607-00.
For information on creating a user-defined virtual process, see the IBM Informix
User-Defined Routines and Data Types Developer's Guide, V11.5,
SC23-9438-00 or the J/Foundation Developer's Guide, V11.5, SC23-9434-00.
NETTYPE
The NETTYPE parameter provides tuning options for the protocols that
dbservername entries define in the sqlhosts file or registry. Each dbservername
entry in the sqlhosts file or registry is defined in relation to either the
440
7884ch13.fm
DBSERVERNAME parameter or the DBSERVERALIASES parameter in the

ONCONFIG file. The NETTYPE configuration parameter describes a network
connection as follows:
The protocol (or type of connection).
The number of poll threads assigned to manage the connection.
The expected number of concurrent connections.
The class of virtual processor that will run the poll threads.
You can specify a NETTYPE parameter for each protocol that you want the
database server to use.
There is a special NETTYPE setting available if you want to enable the
database server to use multiplexed connections on UNIX. To do this, you
designate a NETTYPE with the value sqlmux, as in the following example.
NETTYPE sqlmux
LOCKS
The LOCKS parameter specifies the initial size of the lock table. Every SQL
session that connects to a database, and accesses tables and rows, generates a
lock. If the lock is non-exclusive, it is shared; if something needs to change, it is
exclusive. If there is an intent to change an object, it is an intent-exclusive lock.
The lock table holds an entry for each type of lock entry, and each of these lock
entries will have a small allocation in resident memory.
When the number of locks exceeds the lock table value, on 32-bit servers, the
database server will increase the size of the lock table by doubling the lock table
value, up to 99 times, or to the maximum value for the server allowance
(whichever comes first). On 32 bit servers, the maximum limit is 8,000,000 locks.
On 64 bit servers, the lock table limit is based on the maximum starting locks
value (500,000)+ 99 allowed increments of 1,000,000 locks for a total of
599,000,000 locks
The amount of memory storage per lock will range from 100 to 200 bytes,
depending on the byte-word size and the platform.
Note: All lock table increments are kept in virtual memory. If the server engine
has limited shared memory, locks can become a memory resource drain.
441
7884ch13.fm
STACKSIZE
The STACKSIZE parameter specifies the stack size for database server user
threads. Setting a value for STACKSIZE that is too large wastes virtual memory
space and can cause swap-space problems.
For 32-bit platforms, the default STACKSIZE value of 32 kilobytes is sufficient for
non-recursive database activity. For 64-bit platforms, the recommended
STACKSIZE value is 64 kilobytes.
In recursive SQL routines, the server checks for the possibility of stack-size
overflow and automatically expands the stack.
User-defined routines should increase the stack size for a routine as needed,
using the stack modifier in the CREATE FUNCTION statement
USELASTCOMMITTED
This parameter specifies the isolation level whenever the LAST COMMITTED
feature of the COMMITTED READ isolation level is implicitly in effect. The LAST
COMMITTED feature can reduce the risk of locking conflicts between concurrent
transactions on tables that have exclusive row locks.
Important: USELASTCOMMITTED only works with tables that have been
created or altered to have ROW as their locking granularity.
In order for USELASTCOMMITTED to work as expected, and given that a table
has been created or altered to have ROW as their locking granularity, the
following considerations also apply:
If SET TRANSACTION is enabled with READ COMMITTED or READ
UNCOMMITTED, USELASTCOMMITTED will work.
Tables created without any explicit lock mode setting will use the default
setting in DEF_TABLE_LOCKMODE.
DEF_TABLE_LOCKMODE
The DEF_TABLE_LOCKMODE parameter sets the lock mode for every newly
created table for all sessions connected to a logging or nonlogging database, and
has no effect on existing tables in the database.
The specified value can be ROW, which sets the lock mode to ROW for every
new table connected to a database. If the specified value is PAGE (default), any
exclusive locked table remains inaccessable.
There is also an environment variable on the client side,
DEF_TABLE_LOCKMODE, which can be set with a lock mode value (PAGE or
442
7884ch13.fm
ROW). The environment variable has rules of precedence involved if the onconfig
value is also set. For more information, see the IBM Informix Dynamic Server
Administrator's Reference, V11.50, SC27-3607-00.
DEADLOCK_TIMEOUT
The DEADLOCK_TIMEOUT parameter specifies the maximum number of
seconds that a database server thread can wait to acquire a lock. Use this
parameter only for distributed queries that involve a remote database server.
Note: This parameter is applies only to distributed queries. There is a
separate, automated mechanism for deadlocks internal to a local database.
OPTCOMPIND
The OPTCOMPIND parameter determines the method used by the engine to
process a query. All queries will be optimized to run based on the guideline
recommended by this parameter, unless the developer forces a different
guideline using a directive.
When OPTCOMPIND is set to one of the three values which follow, a given query
is optimized using the guideline assigned:
0: When appropriate indexes exist for tables in the query, the optimizer
chooses index scans, without consideration of the cost, over table scans.
1: As long as the isolation is not Repeatable Read, the optimizer will use cost
based decisions. If Repeatable Read is in use, it behaves as though
OPTCOMPIND =0. Setting 1 is recommended for optimal performance.
2: The optimizer uses cost to determine an execution path regardless of
isolation level. Index scans are not given preference; the optimizer decision is
purely based on cost. Setting 2 is the default if the parameter is not set.
Warning: The same OPTCOMPIND value should be used in development
and in the production environment. Performance and query behavior may
have variations if OPTCOMIND is different in development and production
environments.
DIRECTIVES
The DIRECTIVES parameter enables or disables the use of SQL directives. SQL
directives allow you to specify behavior for the query optimizer in developing
query plans for SELECT, UPDATE, and DELETE statements.
Set DIRECTIVES to 1, which is the default value, to enable the database server
to process directives.
443
7884ch13.fm
Set DIRECTIVES to 0 to disable the database server from processing directives.

Note: Client programs have the option to set the IFX_DIRECTIVES
environment variable to ON or OFF to enable or disable processing of
directives by the database server. The setting of the IFX_DIRECTIVES
environment variable overrides the setting of the DIRECTIVES configuration
parameter. If you do not set the IFX_DIRECTIVES environment variable, all
sessions for a client inherit the database server configuration for processing
SQL directives.
MAX_PDQPRIORITY
The MAX_PDQPRIORITY parameter limits the PDQ resources that the database
server can allocate to a given DSS query. MAX_PDQPRIORITY is a factor that is
used to scale the value of PDQ priority set by users. For example, if the database
administrator sets MAX_PDQPRIORITY to 80, and a user sets the
PDQPRIORITY environment variable to 50 and issues a query, the database
server silently processes the query with a PDQ priority of 40. The database
administrator can use the onmode utility to change the value of
MAX_PDQPRIORITY while the database server is online.
In Informix database server, PDQ resources include memory, CPU, disk I/O, and
scan threads. MAX_PDQPRIORITY lets the database administrator run decision
support concurrently with OLTP, without a deterioration of OLTP performance.
However, if MAX_PDQPRIORITY is too low, the performance of decisionsupport queries can degrade.
Table 13-1 lists the MAX_PDQPRIORITY value that you can set.
Table 13-1 MAX_PDQPRIORITY value
Value
Database Server Action
Turns off PDQ. DSS queries use no parallelism.
Fetches data from fragmented tables in parallel (parallel scans) but

uses no other form of parallelism.
100
Uses all available resources for processing queries in parallel.
An integer
between 1-100
Sets the percentage of the user-requested PDQ resources to be

allocated to the query.
13.4.4 Java configuration parameters

The configuration parameters listed Table 13-2 on page 445 allow you to use
J/Foundation, which incorporates an embedded Java virtual machine on the
444
7884ch13.fm
database server. For more information on these parameters, see the

J/Foundation Developer's Guide, V11.5, SC23-9434-00.
Table 13-2 Java configuration parameters
Parameter name
Description
AFCRASH 0x10
When the 0x10 bit is on for AFCRASH, all the messages that the
Java Virtual Machine generates are logged into the JVM_vpid
file, where vpid is the process ID of the Java virtual processor.
This file is stored in the directory where the JVPLOG file is stored.
JVPDEBUG
When set to 1, tracing messages are written to the JVPLOG file.
JVPHOME
Directory where the classes of the IBM Informix JDBC Driver are
installed.
JVPLOGFILE
Absolute pathname for your Java VP log files.
JVPPROPFILE
Absolute pathname for the Java VP properties file.
JVPJAVAVM
Libraries to use for the Java Virtual Machine (JVM).
JVPJAVAHOME
Directory where the Java Runtime Environment (JRE) for the

database server is installed.
JVMTHREAD
Thread package (green or native) to use for the JVM.
JVPVJAVALIB
Path from JVPJAVAHOME to the location of the Java VM

libraries.
JVPCLASSPATH
Initial Java class path setting.
VPCLASS jvp=n
Number of Java virtual processors that the database server

should start.
For additional onconfig parameters and how they relate to development, see
Appendix A, Onconfig parameters on page 453.
13.5 Working with your database administrator

There is an assumption throughout this handbook that our main audience is the
developer, with a distinctive separate role from the database administrator (DBA).
The database server administrator has the task of allocating and handling
resources effectively in the server side of a client -server environment. When a
performance issue occurs, or a transaction process is not working as it should, it
becomes necessary to work with your database administrator to find a solution.
Present the problem to your database administrator. He or she may have a good
445
7884ch13.fm
idea about a particular configuration area that may relate to the problem you are
seeing. There are a few areas we can discuss here, which may help target the
common problems that developers experience.
Configuration parameters that affect logging

Checkpoints, logging, and page cleaning are necessary to maintain database
consistency. A direct trade-off exists between the frequency of checkpoints or the
size of the logical logs and the time that it takes to recover the database in the
event of a failure. A database administrator has to consider ways to reduce the
overhead for these activities based on the acceptable delay during recovery.
Sometimes these decisions are influenced by the behavior of the application.
Here we discuss the configuration parameters that effect logging.
LOGBUFF and PHYSBUFF

The LOGBUFF and PHYSBUFF configuration parameters effect logging I/O
activity because they specify the respective sizes of the logical-log and
physical-log buffers that are in shared memory. The size of these buffers
determines how quickly they fill and therefore how often they need to be flushed
to disk. If checkpoints tend to have long duration, you can seek further
information related to buffer tuning and checkpoint tuning information in the IBM
Informix Dynamic Server Performance Guide, V11.50, SC27-3618-00.
LOGFILES
The LOGFILES parameter specifies the number of logical-log files. This
parameter only becomes important with respect to the long transaction high
water mark (LTXHWM and LTXEHWM), if we are having trouble with transactions
that need to be rolled back. See the sections on DYNAMIC_LOGS, LTXHWM
and LTXEHWM, later in this chapter for more details.
LOGSIZE
Choose a log size based on how much logging activity occurs and the amount of
risk in case of catastrophic failure. If you cannot afford to lose more than an
hour's worth of data, create many small log files that each hold an hour's worth of
transactions. If your system is stable with high logging activity, choose larger logs
to improve performance.
Note: A backup process can hinder transaction processing if data is located
on the same disk as the logical-log files. It is better to keep logical logs on
separate disks from data, if possible. If extra disks are not an option for
separate logical log space, however, you can wait for periods of low user
activity before you back up the logical-log files.
446
7884ch13.fm
You need to adjust the size of the logical log when your transactions include
simple large objects or smart large objects, as the following sections describe.
Estimating logical-log size when logging simple large objects

To obtain better overall performance for applications that perform frequent
updates of TEXT or BYTE data in blobspaces, reduce the size of the logical log.
Blobpages cannot be reused until the logical log to which they are allocated is
backed up. When TEXT or BYTE data activity is high, the performance impact of
more frequent checkpoints is balanced by the higher availability of free
blobpages.
When you use volatile blobpages in blobspaces, smaller logs can improve
access to simple large objects that must be reused. Simple large objects cannot
be reused until the log in which they are allocated is flushed to disk. In this case,
you can justify the cost in performance because those smaller log files are
backed up more frequently.
Estimating logical-log size when logging smart large objects

If you plan to log smart-large-object user data, you must ensure that the log size
is considerably larger than the amount of data being written. Smart-large-object
metadata is always logged even if the smart large objects are not logged.
Use the following guidelines when you log smart large objects:
If you are appending data to a smart large object, the increased logging
activity is roughly equal to the amount of data written to the smart large
object.
If you are updating a smart large object (overwriting data), the increased
logging activity is roughly twice the amount of data written to the smart large
object. The database server logs both the before-image and after-image of a
smart large object for update transactions. When updating the smart large
objects, the database server logs only the updated parts of the before and
after image.
Metadata updates affect logging less. Even though metadata is always
logged, the number of bytes logged is usually much smaller than the smart
large objects.
DYNAMIC_LOGS
The default value for the DYNAMIC_LOGS configuration parameter is 2, which
means that the database server automatically allocates a new logical log file after
the current log file when it detects that the next log file contains an open
transaction. The database server automatically checks if the log after the current
log still contains an open transaction at the following times:
447
7884ch13.fm
Immediately after it switches to a new log file while writing log records (not
while reading and applying log records)
At the beginning of the transaction cleanup phase which occurs as the last
phase of logical recovery
Logical recovery happens at the end of fast recovery and at the end of a cold
restore or roll forward. For more information on the phases of fast recovery, see
IBM Informix Dynamic Server Administrator's Guide, Version 11.50,
SC27-3606-00.
If you set DYNAMIC_LOGS to 0, the database server still checks whether the
next active log contains an open transaction when it switches log files. If it finds
an open transaction in the next log to be active, it issues the following warning:
WARNING: The oldest logical log file (%d) contains records from an open
transaction (0x%p), but the Dynamic Log Files feature is turned off.
LTXHWM and LTXEHWM

The LTXHWM parameter indicates how full the logical log is when the database
server starts to check for a possible long transaction and to roll it back.
LTXEHWM indicates the point at which the database server suspends new
transaction activity to locate and roll back a long transaction. With the dynamic
log file feature, long transaction high watermarks are no longer as critical
because the database server does not run out of log space unless you use up
the physical disk space available on the system. Under normal conditions, you
should keep the default values for LTXHWM and LTXEHWM. If a rollback is ever
needed, it can indicate a serious problem within an application.
After Version 9.4, LTXHWM and LTXEHWM are not in the onconfig.std file and
default to the following values, depending on the value of the DYNAMIC_LOGS
configuration parameter:
With DYNAMIC_LOGS set to 1 or 2, the long transaction high watermark
default values are 80 for LTXHWM and 90 for LTXEHWM. Because the
database server does not run out of logs, other users can still access the log
during the rollback of a long transaction.
With DYNAMIC_LOGS to 0, the default values are 50 for LTXHWM and 60 for
LTXEHWM.
You might want to change these default values for one of the following reasons:
To allow other transactions to continue update activity (which requires access
to the log) during the rollback of a long transaction. In this case, you increase
the value of LTXEHWM to raise the point at which the long transaction
rollback has exclusive access to the log.
448
7884ch13.fm
To perform scheduled transactions of unknown length, such as large loads

that are logged. In this case, you increase the value of LTXHWM so that the
transaction has a chance to complete before reaching the high watermark.
13.5.1 Parameters for negotiation

While the database administrator should make final decisions for server
parameters, there are several parameters for which you need agreement to
achieve the most effective results.
VPCLASS
If you decide to use user-defined routines (UDRs), a UDR can use the
processing power in the CPU class that users currently have at their disposal, or
you can define the UDR with its own processor class when you create the
function. You define a new class of virtual processors to isolate UDR execution
from other transactions that execute on the CPU virtual processors. This is
typically used when you write UDRs to support user-defined data types.The
class name that you specify in the VPCLASS parameter must match the name
specified in the CLASS modifier of the CREATE FUNCTION statement.
Note: Since a user defined CPU class is treated as a virtual processor, the
onconfig parameter SINGLE_CPU_VP must be off (0) and
MULTIPROCESSOR must be on (1).
Guidelines for setting VPCLASS
For uniprocessor computers, it is recommended that you use one CPU virtual
processor.
VPCLASS cpu,num=1.
For multiprocessor systems with four or more CPUs that are primarily used as
database servers, it is recommended that you set the VPCLASS num option
to one less than the total number of processors. For example, if you have four
CPUs, use the following specification:
VPCLASS cpu,num=3
Process priority aging

On some operating systems, priority aging occurs when the operating system
lowers the priority of long-running processes as they accumulate processing
time. You should always disable the priority aging parameter of VPCLASS
because it can cause the performance of the database server processes to
decline over time. For further information, check the machine notes for your
database server.
449
7884ch13.fm
Processor affinity
Processor affinity distributes the computation impact of CPU virtual processors
and other processes. On computers that are dedicated to the database server,
assigning CPU virtual processors to all but one of the CPUs achieves maximum
CPU utilization. On computers that support both database server and client
applications, you can bind applications to certain CPUs through the operating
system. By doing so, you effectively reserve the remaining CPUs for use by
database server CPU virtual processors, which you bind to the remaining CPUs
with the VPCLASS configuration parameter. Set the aff option of the VPCLASS
parameter to the numbers of the CPUs on which to bind CPU virtual processors.
For example, on an 8 CPU machine, the following VPCLASS setting assigns
CPU virtual processors to processors 4 to 7, and CPUs 0,1,2,and 3 would be
most available to the application:
VPCLASS cpu,num=4,aff=4-7
For additional methods that can be used with the VPCLASS, see your IIBM
Informix Dynamic Server Administrator's Guide, V11.50, SC27-3606-00.
13.5.2 Monitoring isolation levels

As a general discovery tool, the onstat -g sql command can be used for
determining the isolation levels and lock modes for SQL statements actively
running against the server, their associated databases and any SQL errors that
might have occurred. The onstat -g sql command can also be run with an
optional session ID. The output contains the most recent and the current SQL
statements being run by the session.
13.5.3 Monitoring locks

The onstat -k command displays the current database locks held within the
system. The type of lock is determined by the logging mode, isolation levels, and
application design. Example 13-1shows the output of a sample onstat -k
command.
Example 13-1 onstak -k output
IBM Informix Dynamic Server Version 11.50.UC6 -- On-Line -- Up 03:55:17 -15360 Kbytes
Locks
address wtlist
owner lklist type tblsnum rowid
key#/bsiz
a095f78 0
a4d9e68 0 HDR+S
100002 203
0
1 active, 2000 total, 2048 hash buckets, 0 lock table overflows
450
7884ch13.fm
The information in this output explains lock aspects. The type column explains
the lock types associated with each open lock. The abbreviations in the Type
column defines lock types as follows:
HDR
B
S
I
X
U
IX
IS
SIX
Header
Byte lock
Shared
Intent
Exclusive
Update
Intent exclusive
Intent-shared
Shared, Intent exclusive
The lock level is determined by the value of the tblsnum (tblspace ID) and the row
ID as show in Table 13-3 and Table 13-4.
Table 13-3 Determining lock level from Row ID
If RowID
Lock level is
Equals 0
Table
Ends in 00
Page
<6 digits & non zero
Row
Table 13-4 Determining lock level from tblsnum

If tblsnum ID
Lock level is
Equals 100002
Database
<10000
ER Pseudo lock
13.5.4 Monitoring user threads

The onstat -u command output reveals all of the current user threads running
on a server, though the output can be cryptic. In a highly active system it is
normal to see user threads waiting on various conditions. These wait conditions
change rapidly as resources are engaged and released. To identify a problem
when a particular session is having a problem, monitor the primary user threads
by watching the flags over a few minutes. If the user thread remains and the
thread and wait conditions are unchanging, it provides a sign that the time has
come to focus in on a problem user thread. Watch the flag values, then narrow
down the problem by zeroing in the details of the problem with additional onstat
commands. The steps to do this are:
451
7884ch13.fm
1. Look at flag position 5: Primary threads (P) are the only ones of interest
(Ignore other user threads).
2. Look at flag position 1 (Wait conditions): B =buffer, C =checkpoint, G=write of
the Logical Log buffer, L = lock, S = mutex, T =transaction, X=transaction
cleanup, Y=condition. (Watch for persistent L,S,T, or Y).
3. Look at flag position 3 (thread activity): A=DBSpace backup, B= Begin work,
P=Preparing/prepared, X=XA prepare, C=Commiting/committed,
R=Aborting/aborted, H=Heuristic aborting/aborted. (Watch for P, X,C,R,H).
For more information on onstat commands and the meanings of the columns
check out the onstat portal, located at
adref.doc/ids_adr_0608.htm
452
7884ax01.fm
Appendix A.
Onconfig parameters
13.4, Configuration options on page 437 discusses several parameters that
have a direct impact on developer methods and understandings. The list
provided was limited to the direct-impact related parameters. It is likely that
application developers will not have much use for the remaining onconfig
parameters most of the time. Normally, this remaining configuration information is
only a concern for the database administrators. Occasionally, there will come a
time when performance of an application may come into question. Sometimes
slow performance is caused by the application, and is under control of the
developer. At other times it may be an onconfig setting in the IBM Informix server
that requires adjustment. The parameters listed in Table A-1 on page 454 may
help the developer and the database administrators to narrow the target area for
further investigation. It is not an exhaustive list; parameters that do not apply or
which have already been discussed in this book are not mentioned. For further
information on any of these parameters, see IBM Informix Dynamic Server
Administrator's Reference, v11.50, SC27-3607-00.
From the developers point of view, it makes sense to categorize the configuration
parameters from a functional perspective, rather than a strict listing by name.
Toward this objective, the parameter names are listed by function groupings to
help narrow down the parameter list that can be considered for an area of
specific interest. The functional groupings do not reflect specific naming
conventions used in the onconfig file.
453
7884ax01.fm

Table A-1 Onconfig parameters
454
Related function area
Onconfig parameters
Storage Space
ROOTNAME, ROOTOFFSET, ROOTSIZE, MIRROR,

MIRROROFFSET, LOGFILES, LOGSIZE,
DBSPACETEMP, SBSPACETEMP, SBSPACENAME,
SYSSBSPACENAME, CLEANERS, STAGEBLOB,
TBLTBLFIRST, TBLTBLNEXT, DATASKIP
Path locations
ROOTPATH, JVPJAVALIB, JVPJAVAVM, MSGPATH,

PLOG_OVERFLOW_PATH
Buffer Movement
BUFFERPOOL,PHYSBUFF, PHYSFILE
Memory Cache
DD_HASHSIZE, DD_HASHMAX, DS_HASHSIZE,

DS_POOLSIZE, PC_HASHSIZE, PC_POOLSIZE,
STMT_CACHE, STMT_CACHE_HITS,
STMT_CACHE_SIZE, STMT_CACHE_NOLIMIT,
STMT_CACHE_NUMPOOL, PLCY_POOLSIZE,
PLCY_HASHSIZE, USRC_POOLSIZE,
USRC_HASHSIZE
Memory control
RESIDENT, SHMVIRTSIZE, SHMADD, EXTSHMADD,

SHMTOTAL, SHMVIRT_ALLOCSEG,
SHMNOACCESS,VP_MEMORY_CACHE_KB,
ONLIDX_MAXMEM
Virtual processor and

connection
NETTYPE, LISTEN_TIMEOUT,
MAX_INCOMPLETE_CONNECTIONS, FASTPOLL,
MULTIPROCESSOR, VPCLASS, SINGLE_CPU_VP,
AUTO_AIOVPS, DIRECT_IO
SQL Control
LOCKS, OPTCOMPIND, DIRECTIVES,

EXT_DIRECTIVES, OPT_GOAL, IFX_FOLDVIEW,
RA_PAGES, RA_THRESHOLD
Transaction control
DEF_TABLE_LOCKMODE, BLOCKTIMEOUT,
TXTIMEOUT, HETERO_COMMIT,
DEADLOCK_TIMEOUT, DYNAMIC_LOGS, LOGBUFF,
LTXHWM, LTXEHWM, TEMPTAB_NOLOG,
AUTO_REPREPARE
Decision support
DS_MAX_QUERIES, DS_TOTAL_MEMORY,
DS_MAX_SCANS, DS_NONPDQ_QUERY_MEM
7884ax02.fm
Appendix B.
Accommodating distributed
transactions
This appendix discusses the use of distributed transactions with an IBM Informix
database server.
455
7884ax02.fm
B.1 Distributed transactions

A transaction is a series of actions performed as a single logical unit of work in
which either all of the actions are performed or none of them are.
A distributed transaction is a transaction that runs in multiple processes, usually
on several machines, and normally involves actions against two or more
databases. Each participant of a distributed transaction must agree to commit
the changes before the distributed transaction can be committed.
There are three core components on a distributed transaction:
Application program (AP)
Application program implements the desired business function. It specifies a
sequence of operations that involve resources such as databases. An
application program participates in one or more units of work, and might
decide to commit or roll back each unit of work independently.
Resource manager (RM)
Resource manager manages access to shared resources such as databases.
The resource manager provides the services to manage the data involved in
the distributed transaction.
Transaction manager (TM)
Transaction manager manages global transactions and coordinates the
decision to commit them or roll them back ensuring their atomicity. The
transaction manager also coordinates recovery activities of the resource
managers when necessary, such as after a component failure.
The XA standards, set forth by the Open Group's X/Open Distributed Transaction
Processing (DTP) model, define the interfaces between the transaction manager,
the application program, and the resource manager in a DTP environment.
These interfaces are implemented in the Informix TP/XA Interface Library.
B.2 TP/XA Transaction Manager XA Interface Library

TP/XA is a library of functions that allows the IBM Informix database server act
as a resource manager in the X/Open DTP environment.The TP/XA library
facilitates communication between a third-party transaction manager and the
database server.
TP/XA is supplied with IBM Informix ESQL/C included on the IBM Informix Client
SDK package.
456
7884ax02.fm
In addition to the TP/XA library, a header file xa.h, is supplied containing the
definition for the functions and common structures such as XID or xa_switch_t
used by the transaction manager and resource manager.
Table 13-5 describes the functions used to work with XA transactions.
Table 13-5 TP/XA Macro definitions
Function
Description
xa_open()
Initializes the resource manager (database server) for a XA

transaction
xa_close()
Close a currently open resource manager
xa_start()
Starts a XA transaction
xa_end()
Dissociates from a XA transaction
xa_rollback()
Tells the resource manager to roll back a XA transaction
xa_prepare()
Asks the resource manager to prepare to commit a XA transaction
xa_commit()
Tells the resource manager to commit a XA transaction
xa_recover()
Obtains a list of XIDs that are currently in a prepared or

heuristically completed state
xa_forget()
Tells the resource manager to forget a heuristically completed

transaction
xa_complete()
Test Completion of asynchronous XA Request. This function is

provided only for compliance with the X/Open XA Specification
B.3 XA_TOOL ESQL/C sample

Example B-1is a basic ESQL/C application which demonstrates how to use
some of the TP/XA functions such as xa_prepare() and xa_rollback() to
perform operations with a distributed transaction.
Example B-1 xa_tool.ec
/*************************************************************
*
*
WARNING:
USE OF THIS PROGRAM MAY HAVE UNDESIRABLE AFFECTS
*
TO THE DATABASE SERVER, INCLUDING DATA CORRUPTION!!
*
*
TITLE:
xa_tool.ec
*
*
Compile with:
Appendix B. Accommodating distributed transactions
457
7884ax02.fm

*
on UNIX systems:
*
esql -o xa_tool xa_tool.ec -lifxa
*
on Windows systems
*
esql xa_tool.ec
*
* This program will connect to a global transaction branch
* and let the user manipulate the transaction.
*
* Usage:
*
xa_tool <fID> <gtl> <bql> <hex data>
*
* The fID, gtl, bql, and hex data should be provided from the
* transaction desired to be manipulated.
The information is
* found by execution of the onstat -G command.
Example output
* of the onstat -G command:
*
*
Global Transaction Identifiers
*
address
flags
fID
gtl
bql
data
*
cb2a964
0x8442a
0
2
4
4D4E4F000000
*
1 active, 128 total
*
* If transaction cb2a964 is the transaction that is desired
* to be manipulated, then this program should be executed
* with the following command:
*
*
xa_tool 0 2 4 4D4E4F000000
*
*
*************************************************************
*/
#include
#include
#include
#include
<stdio.h>
<stdlib.h>
<ctype.h>
"xa.h"
extern struct xa_switch_t infx_xa_switch;

#define xa_open(info, rmid, flags)
\
((*infx_xa_switch.xa_open_entry)(info,rmid,flags))
#define xa_start(gtrid,rmid, flags) \
((*infx_xa_switch.xa_start_entry)(gtrid,rmid,flags))
#define xa_rollback(gtrid,rmid, flags) \
((*infx_xa_switch.xa_rollback_entry)(gtrid,rmid,flags))
#define xa_commit(gtrid,rmid, flags)
\
((*infx_xa_switch.xa_commit_entry)(gtrid,rmid,flags))
#define xa_rollback(gtrid,rmid, flags)
\
((*infx_xa_switch.xa_rollback_entry)(gtrid,rmid,flags))
#define xa_end(gtrid,rmid, flags)
\
((*infx_xa_switch.xa_end_entry)(gtrid,rmid,flags))
#define xa_prepare(gtrid,rmid, flags)\
((*infx_xa_switch.xa_prepare_entry)(gtrid,rmid,flags))
#define xa_close(info,rmid, flags)\
458
7884ax02.fm
((*infx_xa_switch.xa_close_entry)(info,rmid,flags))
#define xa_forget(gtrid,rmid, flags)\
((*infx_xa_switch.xa_forget_entry)(gtrid,rmid,flags))
#define xa_recover(gtrid, count, rmid, flags)\
((*infx_xa_switch.xa_recover_entry)(gtrid,count,rmid,flags))
/* Doesn't matter what database is opened... */
#define OPEN_DATABASE "sysmaster"
/* Initialize the xid */

void
setup_myxid (XID *xid, int x_fID, int x_gtl, int x_bql, char* x_data)
{
int ii, c;
xid->formatID = x_fID;
xid->gtrid_length = x_gtl;
xid->bqual_length = x_bql;
for(ii=0; ii<x_gtl+x_bql; ii++)
if (sscanf( &x_data[ii*2], "%02x", &c) != 1)
{
printf("Invalid GTRID data!\n");
exit(1);
}
else
xid->data[ii] = (char) c;
}
void xa_tool(XID *xid)

{
int choice = 0;
int cc;
while( choice != 'Q')
{
printf("\tP -- XA_PREPARE\n");
printf("\tC -- XA_COMMIT\n");
printf("\tR -- XA_ROLLBACK\n");
printf("\tF -- XA_FORGET\n");
printf("\tQ -- terminate program\n");
printf("Enter Choice: ");
choice = getchar();
choice = toupper(choice);
printf("%c\n\n", choice);
switch(choice)
{
case 'P':
printf("Executing XA_PREAPRE\n");
if ((cc = xa_prepare(xid, 0, TMNOFLAGS)) != XA_OK)
{
459
7884ax02.fm

printf("XA_PREPARE failed with %d\n", cc);
}
else
printf("... XA_PREPARE finished\n\n");
break;
case 'C':
printf("Executing XA_COMMIT\n");
if ((cc = xa_commit(xid, 0, TMNOFLAGS)) != XA_OK)
{
printf("XA_COMMIT failed with %d\n", cc);
}
else
printf("... XA_COMMIT finished\n\n");
break;
case 'R':
printf("Executing XA_ROLLBACK\n");
if ((cc = xa_rollback(xid, 0, TMNOFLAGS)) != XA_OK)
{
printf("XA_ROLLBACK failed with %d\n", cc);
}
else
printf("... XA_ROLLBACK finished\n\n");
break;
case 'F':
printf("Executing XA_FORGET\n");
if ((cc = xa_forget(xid, 0, TMNOFLAGS)) != XA_OK)
{
printf("XA_FORGET failed with %d\n", cc);
}
else
printf("... XA_FORGET finished\n\n");
break;
case 'Q':
break;
default:
printf("%c is not a valid option!\n\n", choice);
choice = 0;
}
if (choice != 'Q') {
choice = getchar();
choice = 0;
}
}
}
int main(int arc, char *argv[])
{
$int cc;
XID xid;
int xid_fID, xid_gtl, xid_bql;
char xid_data[300];
if (arc != 5)
460
7884ax02.fm
{
printf("Error:
printf("Usage:
exit(2);
Incorrect number of parameters.\n");

%s <fID> <gtl> <bql> <data>\n", argv[0]);
}
printf("\n\n");
xid_fID = atoi(argv[1]);
xid_gtl = atoi(argv[2]);
xid_bql = atoi(argv[3]);
strcpy(xid_data, argv[4]);
/* setup the XID which is used to identify the transaction */
setup_myxid(&xid, xid_fID, xid_gtl, xid_bql, xid_data);
/* establish connection to the database */
printf("Calling xa_open ... \n");
if ((cc = xa_open(OPEN_DATABASE, 0, TMNOFLAGS)) != XA_OK)
{
printf("xa_open failed with %d, sqlcode = %d\n", cc, SQLCODE);
exit(1);
}
else
printf("... xa_open finished\n\n");
/** This is for xa_tool...
xa_tool(&xid);
**/
/* close the connection */

printf("Calling xa_close...\n");
if ((cc = xa_close("", 0, TMNOFLAGS)) != XA_OK)
{
printf("xa_close failed with %d\n", cc);
exit(1);
}
else
printf("... xa_close finished\n\n");
}
You also can use the xa_tool example to complete unresolved transactions left
in the database server. Local transactions are automatically roll back by the
Informix database server if the connection between the client application and the
database server was lost. This happens because the database server is in direct
control of the transaction.
When using distributed transactions the database server can not automatically
resolved a transaction, the Transaction Manager is the only process in charge of
committing or aborting the global transaction. This means that if there is an
461
7884ax02.fm
communication error or a failure within the resource manager, there is a chance

of leaving unresolved transactions in the database server.
These transactions can be resolved using the xa_commit(), xa_rollback(), and
xa_forget() functions. Example B-2 shows how to compile and use the
xa_tool.ec program to rollback an XA transaction left in the Informix database
server.
Example B-2 xa_tool.ec output
D:\Infx\ids1150>onstat -G
IBM Informix Dynamic Server Version 11.50.FC6
Global Transaction Identifiers
address
flags isol
83246d88
-L--G COMMIT
1 active, 128 total
timeout fID
0
0
-- On-Line -- Up 05:04:17
gtl
2
bql
4
data
4D4E4F000000
D:\Infx\ids1150>
C:\work>esql -nologo xa_tool.ec
IBM Informix CSDK Version 3.50, IBM Informix-ESQL Version 3.50.TC7
xa_tool.c ...
C:\work>xa_tool 0 2 4 4D4E4F000000
Calling xa_open ...
... xa_open finished
P -- XA_PREPARE
C -- XA_COMMIT
R -- XA_ROLLBACK
F -- XA_FORGET
Q -- terminate program
Enter Choice: R
R
Executing XA_ROLLBACK
... XA_ROLLBACK finished
P -- XA_PREPARE
C -- XA_COMMIT
R -- XA_ROLLBACK
F -- XA_FORGET
Q -- terminate program
Enter Choice: q
Q
Calling xa_close...
... xa_close finished
C:\work>
462
7884ax02.fm
For more information related to the TP/XA library refer to the TP/XA Transaction
Manager XA Interface Library User Manual at:
http://publibfp.boulder.ibm.com/epubs/pdf/5193.pdf
463
7884ax02.fm
464
7884bibl.fm
Related publications
The publications listed in this section are considered particularly suitable for a
more detailed discussion of the topics covered in this book.
IBM Redbooks
For information about ordering these publications, see How to get Redbooks on
page 466. Note that some of the documents referenced here may be available in
softcopy only.
Embedding Informix Dynamic Server: An Introduction, SG24-7666
Other publications
These publications are also relevant as further information sources:
IBM Informix Dynamic Server Administrators Guide, v11.50, SC23-7748
IBM Informix Guide to SQL: Reference, v11.50, SC23-7750
Embedded SQLJ Users Guide, Version 2.90, G251-2278
IBM Informix Storage Manager Administrators Guide, v2.2, G229-6388
IBM Informix Security Guide, v11.50, SC23-7754
IBM Informix GLS Users Guide, G229-6373
IBM Informix Dynamic Server Administrators Reference, SC23-7749
IBM Informix Guide to SQL: Syntax, v11.50, SC23-7751
IBM Informix Guide to SQL: Tutorial, G229-6427
IBM Informix Dynamic Server Performance Guide, v11.50, SC23-7757
IBM Informix High-Performance Loader Users Guide, v11.50, SC23-9433
IBM Informix Migration Guide, SC23-7758
IBM Informix JDBC Driver Programmers Guide, SC23-9421
IBM Informix ESQL/C Programmers Manual, v3.50, SC23-9420
IBM Data Server Provider for .NET Programmers Guide, SC23-9848
465
7884bibl.fm
IBM Informix Dynamic Server Installation Guide for UNIX, Linux, and Mac OS
X, GC23-7752
IBM Informix DB-Access Users Guide, SC23-9430
IBM Informix ODBC Driver Programmers Manual, SC23-9423
IBM Informix Dynamic Server Installation Guide for Windows, GC23-7753
Guide to Informix MaxConnect, Version 1.1, G251-0577
IBM Informix Backup and Restore Guide, v11.50, SC23-7756
IBM Informix Dynamic Server Enterprise Replication Guide, v11.50,
SC23-7755
Expand transaction capabilities with savepoints in Informix Dynamic Server
by Uday B. Kale in IBM developerWorks, 26 March 2009:
http://www.ibm.com/developerworks/data/library/techarticle/dm-0903idssavepo
ints/index.html
Online resources
These websites are also relevant as further information sources:
IBM Informix Dynamic Server v11.50 Information Center:
http://publib.boulder.ibm.com/infocenter/idshelp/v115/index.jsp
How to get Redbooks

You can search for, view, or download Redbooks, Redpapers, Technotes, draft
publications and Additional materials, as well as order hardcopy Redbooks
publications, at this website:
ibm.com/redbooks
Help from IBM

IBM Support and downloads
ibm.com/support
IBM Global Services
ibm.com/services
466
7884bibl.fm
Related publications
467
7884bibl.fm
468
7884IX.fm
Index
Symbols
.4cf extension 406
.ec 124
.ecp 124
.NET connection string attributes
Client_Locale 253
Connection Lifetime 253
DB_LOCALE 253
DELIMIDENT 253
Enlist 253
Fetch Buffer Size 254
Host 253
Max Pool Size 253
OPTOFC 254
Packet Size 254
Pooling 254
PWD 254
Skip Parsing 254
UserDefinedTypeFormat 254
XCL 253
Numerics
4GL 13
A
access method 327
ACE Report Writer 7
ActiveRecord 357
ad hoc queries 6
adapter script 363
ADO interface
IAccessor 215
IColumnsInfo 215
ICommand 215
IDBCreateCommand 215
IDBCreateSession 215
IDBDataSourceAdmin 215
IDBProperties 215
IErrorLookup 215
IGetDataSource 215
IRowsetIdentity 215
ISessionProperties 215
ITransaction 215
ADO object 212
adOpenDynamic cursor 219
adOpenForwardOnly cursor 219
adOpenKeyset cursor 219
adOpenStatic cursor 219
advanced access control feature 21
annotation 205
ANSI serializable 436
antlr-2.7.6.jar 187
application programming interface 13
application-programming interface 32
attribute 184
authorized user 4
AutoCommit mode 166
AXIS2C_HOME 393
B
backup 324
backup server 20
begin work 135
BeginTransaction method 267
BeginTransaction() 258
bigint data type 77
bigserial data type 77
binary import routine 327
binary large object 438
binary receive routine 327
binary send routine 327
bind parameter 302
bladelet 352
BLOB 77, 438
BOOLEAN 77
Boolean UDR 326
buffered logging 18
built-in data type 324
bulk load 7
byte 77
byte column 6
byte-range lock 431
byte-range locking 432
469
7884IX.fm
C
call level interface 14
CallableStatement object 167
cast function 326
cast support function 341
certified file 177
character large object 343, 438
checkpoint 21
chunk 438
CLASSPATH environment variable 153
classpath environment variable 47, 188
CLI 14
CLI parameters
database 69
hostname 69
port 69
protocol 69
client connectivity 6
client locale 58
Client SDK 14
client_locale environment variable 145
CLOB 77, 438
collection 9
collection data type 140, 341
collection variable 141
column-level encryption 21
commit work 135
committed read 436
compiler 126
complex data type 328
component object model 210
concentrator 4
concurrency 21, 426
concurrent query 18
concurrent session 4
connection management 186
connection object 166
connection pooling 151
connection string 251
connection string attribute 216
connection_adapters directory 363
connectionstring property 258
constraint 184
continuity 20
controller 357
convention over configuration 376
cost factor 326
cost function 326
cursor behavior 58
470

cursor stability 435436
customer object 184
customized routine 324
D
data access layer 184
Data accessibility 138
data consistency 7
data consumer 12
data definition language 192
data logging 138
data management 6
data manipulation language 192
data mart 6
data provider 12, 248
data query language 184
data source name 57
data type extension 324
data warehouse 6
database driver 41
database extension 324
database instance name 437
database locale 58
database lock 430
database operation 72
datasource 151
datetime 77
db_locale 59
db_locale environment variable 145
db2cli.ini 69
db2jcc.jar 47, 188
db2jcc4.jar 47
db2sdriver.cfg configuration file 255
DbDataReader method 263
dbschema utility 386
dbspace 438
dbspacetemp parameter 438
DDL 192
deadlock 434
deadlock_timeout parameter 443
def_table_lockmode parameter 429, 442
delete 72
dependency list 330
directives parameter 443
dirty read isolation 433
discarded-logging 188
display environment variable 35
distinct 9, 77
7884IX.fm

distinct data type 324, 328
distributed transaction 18, 90, 435, 456
distributed transactions 150
distribution statistic 327
dmg format 24
DML 192
dom4j-1.6.1.jar 187
DRDA protocol 151
DRDADEBUG trace 182
driver manager 9
driver package 41
drsoctcp 28
dyld_library_path environment variable 127
dynamic link library 343
dynamic sql 132
dynamic_logs configuration parameter 447
E
editions 2
element type 140
encryption 177
end-user routine 326
environment variable
ld_library_path 127, 347, 359
lib 127
libpath 127
shlib_path 127, 359
esql command 13
exception 199
exclusive lock 427
exclusive mode 426427
execute into command 130
executeQuery() 163
executeUpdate() 163
export routine routine 327
F
failover 20
fast data loading 6
fetch buffer size 58
filetoblob 139
finderr 33
finderr utility 13
flat file 7
flow-control extension 324
functional test file 14
G
gem utility 363
generator element 194
get diagnostics statement 144
get() method 199
GetIfxMonthSpan 263
global transaction 456
global transactions 435
H
header file 14, 457
Hibernate dialect 190
Hibernate service 189
Hibernate session instance 197
hibernate.cfg.xml 189
hibernate.property 189
hibernate3.jar 187
hibernate-jpa-2.0-api-1.0.0.Final.jar 187
HibernateUtil class 196
high availability clustering 3
high availabiliy replication 17
host name 57
host variable 141
HQL criteria 202
HQL restriction 202
HTTP link 411
I
i4gl_func 397
IBM.Data.DB2 249
IBM.Data.Informix 249
ICommand interface 220
idssecuritylabel 77
ifx_lo_lock() 432
ifx_lo_open 78
ifx_lo_t 137
ifx_lo_write 78
IfxBlob GetIfxBlob() 258, 263
IfxClob GetIfxClob() 258, 263
IfxCommand 262
IfxCommand class 259
IfxCommand CreateCommand() 258
IfxCommand method 265
IfxConnection class 258
IfxConnection object 267
IfxDataReader 262
IfxDataReader ExecuteReader() 260
IfxDateTime class 276
Index
471
7884IX.fm
IfxDateTime GetIfxDateTime() 263
IfxDateTime public method 277
IfxDecimal GetIfxDecimal 263
IfxDecimal object 280
IfxDecimal public method 281
ifxdotnettrace variable 286
IfxError 265
IfxException class 265
ifxjdbc.jar 188
IfxMonthSpan 263
ifxoledbc COM class ID 210
ifxoledbctrace environment variable 243
IfxParameter 266
IfxParameter CreateParameter() 260
IfxParameterCollection object 265
IfxTimeSpan GetIfxTimeSpan 263
IfxTransaction 258
IfxTransaction object 267
iifxoledbc.dll 210
ILogin utility 33
Implict invocation 327
incoming parameter 331
Informix classes 11
Informix OLE DB Provider 34
informixc.so 359
informixdir environment variable 19, 55, 123
informixserver 347
informixserver environment variable 129, 359
informix-sqli 159
inheritance 184
insert 72
installation directory 19
installation ini file 24
installIDSDriver command 41
int8 77
intent locks 427
intent-exclusive lock 427
intent-shared 427
interval day 77
interval day to hour 77
interval day to minute 77
interval day to second 77
interval hour 77
interval hour to minute 77
interval hour to second 77
interval minute 77
interval minute to second 77
interval month 77
interval second 77
472

interval year 77
interval year to month 77
Isolation Level 58, 66
isolation level 18, 426, 435, 450
ISupportErrorInfo interface 229
Iterator function 326
J
jar file 334
Java annotation 185
Java archive file 334
Java class file 188
Java compiler 188
Java interface 11
Java library 187188
Java Naming and Directory 161
java property 189
Java runtime 161
Java Virtual Machine 34, 188
java_bindir environment variable 335
java_home environment variable 335
java_root environment variable 335
java.io.Serializable 11
java.lang.object 12
java.sql.ParameterMetaData 11
javassist-3.9.0.GA.jar 187
javax.sql extensions 158
javax.sql.ConnectionEventListener 11
javax.sql.ConnectionPoolDataSource 11
javax.sql.DataSource 11
javax.sql.PooledConnection 11
javax.sql.XADataSource 11
JDBC API interfaces 151
JDBC class 151
JDBC method 151
JDBC standard 7
jjdk_home environment variable 335
JNDI 161
jre_home environment variable 335
jta-1.1.jar 187
jvp.log 341
K
key 429
Key lock 430
keystore file 178
key-value locking 430
7884IX.fm
L
label-based access control 21
library file 122
licensing metric 4
limited use socket 45
list 140
load() 199
load() method 199
loc_t 137
lock 426
lock duration 428
lock list 434
lock mode page 429
locks parameter 441
logbuff 446
logfiles parameter 446
logging mode 450
logical device 7
logical volume manager 7
logical-log buffer 18, 446
logical-log record 18
long identifier 9
loopback connection 7
LPAR 4
ltxehwm parameter 448
ltxhwm parameter 448
LU socket 5
lvarchar 78
M
make file 14
max_pdqpriority parameter 444
MaxBytes 270
memory addressability 2
messages files 14
method 184
method properties
position 270
ReferenceCount 270
mi_lo_lock() 432
migration 385
model 356357
module 14
money 78
msgpath file 341
multimedia capability 7
multiplexed connection 1617
multiplexer 4
multiprocessor parameter 440

multiset 140
mutexes 16
mutually exclusive 16
N
native-protocol 49
nchar 78
negator function 326
nettype parameter 440
network connection 16
nvarchar 78
O
OAT 7, 15
object ExecuteScalar() 260
object file 122
object processing 185
object-orientated programming 184
object-relational mapper 357
object-relational mapping technology 184
ODBC data source 56
ODBC driver 55
ODBC driver configuration files
odbc 62
odbcinst.ini 62
sqlhosts 62
ODBC libraries
libifcli.a or libcli.a 62
libifcli.so or iclis09b.so 62
libifdrm.so or idmrs09a.so 62
libthcli.a 62
libthcli.so or iclit09b.so 62
ODBC/CLI driver 41
OLE DB consumer 210
OLE DB provider 210
OLE DB providers
DBTYPE_DBDATE 218
DBTYPE_DBTIME 218
DBTYPE_DBTIMESTAMP 218
olsoctcp 28
onconfig 335, 347
onconfig parameter
dbservername 437
jvpclasspath 335
jvphome 335
jvpjavahome 335
jvpjavalib 335
Index
473
7884IX.fm
jvpjavavm 335
jvplog 341
jvplogfile 335
jvppropfile 335
oninit 16
opaque 9
opaque data type 324, 326, 329
open administration tool 7
operator 327
operator function 326
operator-class function 326
optcompind parameter 17, 443
OUT parameter 167
outgoing parameter 331
outgoing parameters 331

primary key 194
priority aging 449
process memory 341
processor affinity 450
processor value unit 3
processor-based pricing 3
program_design_dbs 394
public method 258
publish 392
pure-Java driver 49
pure-Java drivers 150
Q
query optimization 134
query plan 327
query-intensive analytical application 6
P
packaging file 14
page level lock 429
page locking 430
parallel subquery 18
parallelism 18
parallelizable UDR 326
parsing 134
partition 4
password-based authentication 21
path environment variable 123
path variable 55
persistence 357
persistence data 184
persistence framework 184
persistence object 192
persistent connectivity layer 4
phantom read 435
physbuff configuration parameter 446
physical log image 18
physical-log buffer 446
pipe connection 16
placeholder 85, 302
plain old java objects 192
pluggable authentication modules 21
POJO 192
polymorphism 184
precedence heirarchy 331
prepare statement 18
prepared statement 169
PreparedStatement object 163
preprocessor 9
474
R
rails command 383
rake 357
recovery 324
Redbooks Web site 466
Contact us xiii
regsvr32 tool 240
repeatable read 426, 436
resource manager 456
result set 167
role 21
rollback 19, 324, 435
rollback method 267
rollback work statement 135
row and collection functions
ifx_rc_count 103
ifx_rc_create 103
ifx_rc_delete 103
ifx_rc_describe 103
ifx_rc_fetch 103
ifx_rc_free 103
ifx_rc_insert 103
ifx_rc_isnull 103
ifx_rc_setnull 103
ifx_rc_typespec 103
ifx_rc_update 103
rowset 223
R-tree 351
Ruby/Informix Adapter 376
runtime 8
runtime library 32
7884IX.fm
S
SBS 92
sbspacename parameter 438
sbspacetemp parameter 439
scaffold command 384
scalability 5, 324
scalable data warehousing 6
schema authorizations 21
sdk_home environment variable 335
secure sockets layer 17
secured connection method 17
security mechanism 21
select 72
selectivity function 326
serial8 78
server definition 28
server instance 16
server name 57
server_locale environment variable 145
service oriented architecture 390
Session.createCiteria() 199
Session.createQuery() 199
Session.delete() method 203
Session.get() 199
Session.load() 199
Session.save() 199
SessionFactory class 196
set 140
set isolation 426
set lock mode 426
set transaction 426
setnet32 utility 27, 39
shared library 63
shared lock 427
shared memory 16, 18
simple logging facade for Java 188
simple object access protocol 411
simple-logging 188
single install 4
single password CSM 17
single-user product version 7
slf4j-api-1.5.8.jar 187
SLFJ 188
SLFJ logger 188
slob method 370
smart blob space 92, 139
smart large object 9
smart large object functions
ifx_lo_alter 93
ifx_lo_close 93
ifx_lo_col_info 93
ifx_lo_create 93
ifx_lo_def_create_spec 93
ifx_lo_open 93
ifx_lo_read 94
ifx_lo_readwithseek 94
ifx_lo_seek 94
ifx_lo_specget_estbytes 94
ifx_lo_specget_extsz 94
ifx_lo_specget_flags 94
ifx_lo_specget_maxbytes 94
ifx_lo_specget_sbspace 94
ifx_lo_specset_estbytes 94
ifx_lo_specset_extsz 94
ifx_lo_specset_flags 94
ifx_lo_specset_maxbytes 94
ifx_lo_specset_sbspace 94
ifx_lo_stat 94
ifx_lo_stat_atime 94
ifx_lo_stat_cspec 94
ifx_lo_stat_ctime 94
ifx_lo_stat_refcnt 94
ifx_lo_stat_size 94
ifx_lo_tell 94
ifx_lo_truncate 94
ifx_lo_write 94
ifx_lo_writewithseek 95
soa_err_log log file 394
SOAP 411
socket connection 16
software deployment 22
source file 122
source files 14
SQL communications area 144
SQL dialect 190
SQL optimization 190
sql_autocommit_off 87
sql_bigint 7778
sql_bit 77
sql_c_binary 76
sql_ifmx_udt_blob 77
sql_ifmx_udt_clob 77
sql_infx_attr_lo_automatic 101
sql_infx_bigint 77
sql_infx_udt_fixed 76, 78
sql_infx_udt_varying 78
sql_integer 78
sql_interval_day 77
Index
475
7884IX.fm
sql_interval_hour 77
sql_interval_month 77
sql_interval_second 77
sql_interval_year 77
sql_longbinary 101
sql_longvarbinary 77, 93
sql_longvarchar 78, 93, 101
sql_timestamp 77
sqlca 144
sqlexec 27
SQLExecDirect() 79
SQLExecDirect() function 79
SQLGetDiagRec() 81
sqlhosts configuration file 25
sqlhosts file 19
sqlidebug trace 116, 146
sqliprt tool 116
sqlj.zip 47
sqlj4.zip 47
SQLSetConnectAttr() 87
SQLSTATE 144
stacksize parameter 442
standard data type 324
startup cost 330
Statement object 162
static SQL statement 132
statistics function 326
stored procedure 128
structured query language 150
subscribe 392
sysdistrib system catalog table 439
syslangauth table 325
syspgm4gl 396
sysprocauth table 325
sysprocbody table 325
sysprocedures table 325
sysroutinelangs table 325
syssbspacename parameter 439
system catalog table 325
T
table space 438
tar file 24
target_dir 397
target_file 397
temporary space 438
temporary table 438
text export routine 327
476

text import routine 327
Text input routine 327
text output routine 327
third-party vendor 9
thread 427
TimeSpan GetTimeSpan() 263
transaction 456
transaction control 324
transaction management 186
transient 203
transient object 203
translation library 58
translation option 58
Type 4 150
type constructor 140
U
UDRManager 12
UDRMetaData 12
UDTManager 12
UDTMetaData 12
uncommitted data 435
unit of work 456
unresolved transaction 461
updatable lock 427
update 72
uselastcommitted 429, 442
user defined data type 328
user thread 451
user-defined aggregate 326
V
validity checking 134
version numbering convention 2
virtual processor 16
virtual server 4
virtual-processor process 345
VMB character 58
void Prepare() 260
vpclass parameter 440, 449
W
w4glc 407
w4glc parameters
check 394
compile 395
deploy 395
7884IX.fm
force 395
generate 395
help 395
keep 395
package 395
silent 395
version 395
web 381
web 4GL compiler 407
web server 357
web service description language 397
web service function 414
WEBrick 357
workload 3
workload balancing 20
ws_func 397
WSDL Path 414
wsdl_path 397
X
xa.h 457
XML configuration file 184
XML formatted file 189
XML mapping file 191
XML-based protocol 411
Index
477
7884IX.fm
478
7884spine.fm
479
To determine the spine width of a book, you divide the paper PPI into the number of pages in the book. An example is a 250 page book using Plainfield opaque 50#
smooth which has a PPI of 526. Divided 250 by 526 which equals a spine width of .4752". In this case, you would use the .5 spine. Now select the Spine width for
the book and hide the others: Special>Conditional Text>Show/Hide>SpineSize(-->Hide:)>Set . Move the changed Conditional text settings to all files in your
book by opening the book file with the spine.fm still open and File>Import>Formats the Conditional Text Settings (ONLY!) to the book files.
IBM Informix Developers

Handbook
(1.0 spine)
0.875<->1.498
460 <-> 788 pages
7884spine.fm
480
To determine the spine width of a book, you divide the paper PPI into the number of pages in the book. An example is a 250 page book using Plainfield opaque 50#
smooth which has a PPI of 526. Divided 250 by 526 which equals a spine width of .4752". In this case, you would use the .5 spine. Now select the Spine width for
the book and hide the others: Special>Conditional Text>Show/Hide>SpineSize(-->Hide:)>Set . Move the changed Conditional text settings to all files in your
book by opening the book file with the spine.fm still open and File>Import>Formats the Conditional Text Settings (ONLY!) to the book files.
Back cover
IBM Informix Developers

Handbook
Learn application
development with
supported APIs,
drivers, and
interfaces
Understand Informix
supported
programming
environments
Practical examples
IBM Informix is a low-administration, easy-to-use, and

embeddable database that is idea for application
development. It supports a wide range of development
platforms, such as Java, .NET, PHP, and Web services,
enabling developers to build database applications in their
language of choice. Informix is designed to handle RDBMS
data and XML out of the box, and can easily be extended to
handle new data sets.
This IBM Redbooks publication provides fundamentals of
Informix application development. It covers the Informix
Clients installation and configuration for application
development environments. It discusses the skills and
techniques for building Informix applications with Java,
ESQL/C, OLE DB, .NET, PHP, Ruby on Rails, DataBlade, and
Hibernate. Code examples are used to demonstrate how to
develop an Informix application with various drivers, APIs,
and interfaces. This book also provides application
development trouble-shooting and considerations for
performance.
.
INTERNATIONAL
TECHNICAL
SUPPORT
ORGANIZATION
BUILDING TECHNICAL
INFORMATION BASED ON
PRACTICAL EXPERIENCE
IBM Redbooks are developed by
the IBM International Technical
Support Organization. Experts
from IBM, Customers and
Partners from around the world
create timely technical
information based on realistic
scenarios. Specific
recommendations are provided
to help you implement IT
solutions more effectively in
your environment.
For more information:

ibm.com/redbooks
SG24-7884-00
ISBN

Informix Handbook

Uploaded by

Copyright:

Available Formats

Informix Handbook

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Informix Handbook

Uploaded by

Copyright:

Available Formats

Front cover

Draft Document for Review August 23, 2010 10:53 am

Draft Document for Review August 23, 2010 10:53 am

International Technical Support Organization

Draft Document for Review August 23, 2010 10:53 am

First Edition (October 2010)

Copyright International Business Machines Corporation 2010. All rights reserved.

Draft Document for Review August 23, 2010 10:53 am

Copyright IBM Corp. 2010. All rights reserved.

Draft Document for Review August 23, 2010 10:53 am

3.2.1 IBM Informix ODBC Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

6.1.2 Hibernate concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

Draft Document for Review August 23, 2010 10:53 am

9.2 Setup and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

12.1.3 Web Services development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

Draft Document for Review August 23, 2010 10:53 am

13.5.2 Monitoring isolation levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

Copyright IBM Corp. 2010. All rights reserved.

Draft Document for Review August 23, 2010 10:53 am

Distributed Relational Database

The following terms are trademarks of other companies:

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

The team who wrote this book

Copyright IBM Corp. 2010. All rights reserved.

Draft Document for Review August 23, 2010 10:53 am

Javier Sagrera is a software engineer on the Common Client

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

Become a published author

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

Draft Document for Review August 23, 2010 10:53 am

Introduction to IBM Informix

Copyright IBM Corp. 2010. All rights reserved.

Draft Document for Review August 23, 2010 10:53 am

1.1 Server options

1.1.1 Informix servers

Informix Developer Edition

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am

Informix Innovator-C Edition

Processor Value Unit

Calculated using the number of processor cores in

Chapter 1. Introduction to IBM Informix

Draft Document for Review August 23, 2010 10:53 am

A single named user or specific individual

A single logical connection from a client device to

Limited Use Socket

Available only on Informix Growth Edition, this

An Install is an installed copy of an informix

Informix Choice Edition

IBM Informix Developers Handbook

Draft Document for Review August 23, 2010 10:53 am