XML for z/OS COBOL

Developers
Troy Coleman
CA Technologies
Session Code: E11
Wednesday, 10 November 2010 13:00 14:00
Platform: z/OS

Agenda

Basic XML terminology

XML Benefits and use on z/OS
Enterprise COBOL XML Support
Native XML Parsing and Publishing in COBOL
Native XML Parsing and Publishing in COBOL using DB2
9 for z/OS
Future XML improvements with DB2 10

XML Terminology
XML Extensible Markup Language
Universal means of exchanging data
Expose structure and content of documents
Textual Data Format

Tag Language
<Name>
<FirstName> Troy </FirstName>
<LastName>Coleman</LastName>

</Name>

The eXtensible Markup Language (XML) is seen by some people as the universal format for sharing data between applications.
Any application running on any platform using any database or file system can process data as long as that application understands how to process XML.
The XML document is self describing. That is you know what each element and repeating group of elements are based on tags imbedded within the document. Each element has a begin-tag and end-tag. The structure or order of the data
is inherent through the hierarchical order of tags.

XML Terminology
Well Formed Document

Contains a single root element

Start and End tags matching on all elements
Proper Nesting
Attribute values in quotes

Well Formed:
<Name><First>Troy</First><Last>Coleman</Last></Name>
Note well formed:
<Name><First>Troy<Last>Coleman</First></Last></Name>

The XML document is considered well-formed when each element has a start-tag <> and a corresponding end-tag </>.
The last example is not well formed because the <First> tag overlaps with the <Last> tag.

XML Terminology
Document Validation
DTD Document Type Definition (oldest schema language)
Schema XSD XML Schema Definition
Validation Support
XML System Services in z/OS R10 supports optional XML Validation
COBOL XML PARSE validation added in V4.2
DB2 V9
No DTD support
XML Schema registered in XML Schema Repository
Must use explicit function DSN_XMLVALIDATE

The well formed XML language does not enforce the structure of the document. The DTD language was developed to validate and enforce the document structure.
The DTD was an important step toward the development of defining a family of documents that will be shared between producers and consumers of the data.
The XML Schema is a W3C-recommended language that improves upon the limited capabilities of DTD. The language supports governing the order of elements, Boolean predicates along with data types used to govern the content of
elements, and specialized rules to specify uniqueness and referential integrity (RI) constraints.

XML Terminology
Parser A program that can read an XML document and
provide programmatic access to the document.

SAX Simple API for XML

XML EVENTS generate a Callback to your program for processing.

DOM Document Object Model

The program can traverse the document which is represented in a
tree structure.

The parser processes the markup tags in the XLM document and passes structured information back to an application.
Some parsers process the XML document through the SAX specification and others process using the DOM specification.

XML Benefits and use on z/OS

XML Universal Format for Data
Why z/OS?
Why COBOL?

The next few slides will go into details on the benefits of using XML along with the benefits of using XML specifically on z/OS with COBOL.

XML Universal Format for Data

Data is platform-independent
B2B Industry Standard Schema
Self Documenting Data Structure
Change content and structure does not require program
changes

Benefits:
With XML, applications can more easily read information from a variety of
platforms. The data is platform-independent, so now the sharing of data between
you and your customers can be simplified.
B2B - Businesses are developing DTDs and schemas for their industry. The ability to parse
standardized XML documents gives business products an opportunity to be
exploited in the B2B environment.
Self Doc Data: Applications understand how to read a schema which describes the data in the XML document.
Changes to content and structure is easier in XML. The data is tagged so you can add and remove elements without impacting existing
elements. You will be able to change the data without having to change the application.

Why z/OS?
No platform can compete with enterprise scale workloads

Reliability
Scalability
Security
Availability

Manage the worlds business data and transactions.

Finance Industry
Transportation Industry
Health Industry
Government

Why process XML in COBOL?

Exploit your existing assets/skills
Keeps development controls in one place/style
Modernize existing applications as web services

You have lots of experience mainframe developers who develop in COBOL.

Retain that asset and skills by having them bridge the new XML Web UI with the people and systems that have been running your system.
Using IBMs SOA architecture you can modernize existing COBOL applications to web services.

10

Enterprise COBOL XML Support

Enterprise COBOL for z/OS version 4.2 takes advantage
of the robust z/OS XML System Services
Released August 2009
http://www-01.ibm.com/software/awdtools/cobol/zos/

Adds Validation against SCHEMA (no DTD support)

Performance for non-validating parsing improved
Java 5 and 6 interoperability

Each new release of Enterprise COBOL for z/OS continues to improve on XML features as well as performance processing XML.
The latest release V4.2 takes advantage of the new z/OS XML System Services which provides the ability to valid a document through a Schema.

11

COBOL XML Parse Special Registers

XMLPARSE(COMPAT) compiler option
XML-CODE - XML event set code then pass control to procedure
XML-EVENT Set event name then pass control to procedure
XML-NTEXT Set with document fragment that are returned as national character data
(UNICODE)

XML-TEXT Set with document fragment that are returned as alphanumeric data

New COBOL V4: XMLPARSE(XMLSS) additional registers

XML-NAMESPACE or XML-NNAMESPACE
XML-NAMESPACE-PREFIX or XML-NNAMESPACEPREFIX

To take advantage of the new XML services you will need to use the compiler option XMLPARSE(XMLSS).
A few new special registers are included when you use this option. The XML Namespace and Namespace prefix along with national encoded Namespace and Namespace prefix.

12

COBOL XML PARSE Processing

XML PARSE xml-document
PROCESSING PROCEDURE xml-handler
ON EXCEPTION
display XML document error XML-CODE
NOT ON EXCEPTION
display XML document successfully parsed
END-XML

The XML PARSE statement takes a document string as input xml-document.

It sends this document to the parser and after each XML event callback is made to your program to the xml-handler paragraph.
A set of special registers are set of which one includes the XML structure which caused the event.
You program can then inspect and process the data before control is handed back to the parser.

13

COBOL Parsing Events

<?xml version="1.0"?><msg type="short">Hello, World!</msg>

XML-EVENT

XML-TEXT

START-OF-DOCUMENT
VERSION-INFORMATION

1.0

START-OF-ELEMENT

msg

ATTRIBUTE-NAME

type

ATTRIBUTE-CHARACTERS

short

CONTENT-CHARACTERS

Hello, World!

END-OF-ELEMENT

msg

END-OF-DOCUMENT

This is a simple sample of what the parser will return for each event.

14

COBOL Parsing Events Catagories

XML text nodes

XML element nodes
XML processing instructions
XML comments

The previous example showed specific events. In events are triggered on either a text node, element node, processing instruction or comment.

15

Sample COBOL Programs

First Program shows XMLPARSE
Compiler Options & Identification Division
Working Storage Variables
Procedure Division Parse Statement
XML Handler Paragraph
XML Output
Second program for XMLGENERATE

In this example I ran into an error due to the bracket [ used with the CDATA parameter. The default codepage was causing an error.
I had to specify the compiler option codepage(1047) to correct the problem.

16

COBOL Compiler Options (codepage)

cbl codepage(1047)
IDENTIFICATION DIVISION.
PROGRAM-ID.
XMLPARS1
DATA DIVISION.
WORKING-STORAGE SECTION.
Codepage(1047) needed to support bracket [
<![CDATA[We should add a <picture> element in the future!]]>

In this example I ran into an error due to the bracket [ used with the CDATA parameter. The default codepage was causing an error.
I had to specify the compiler option codepage(1047) to correct the problem.

17

Inline Sample Document

01

xml-document.
05 pic x(39)
05 PIC X(39)
05 PIC X(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)
05 pic x(39)

value
value
value
value
value
value
value
value
value
value
value
value
value
value
value
value
value

'<?xml version="1.0" encoding="ibm-1140.

'?><!-Employee HR Information
-->'.
'<emp><empno>123456</empno><name>
'.
' <firstnme>Troy
</firstnme>
'.
' <lastname>Coleman
</lastname> '.
' </name><workdept>D01</workdept>
'.
' <phoneno>555-123-4459</phoneno>
'.
' <hiredate>03/30/2001</hiredate>
'.
' <job>MANAGER </job>
'.
' <edlevel>16</edlevel><sex>M</sex>
'.
' <birthdate>01/13/1963</birthdate>
'.
' <salary>000078250.00</salary>
'.
' <bonus>000000500.00</bonus>
'.
' <comm>000000000.00</comm>
'.
' <![CDATA[We should add a <picture> ele'.
'ment in the future!]]>
'.
'<?Sign Contract?></emp>
'.

To jump start working with XML I didnt want to take the time to read a file or interface with CICS or IMS.
I decided to create an in working storage document to practice with.
This is a sample of an XML document I used to learn about the different events and parsing features.

18

Sample Program Working Variables

01
01
01
01
01
01
01
01
77

xml-text-len computational
current-element
ws-compensation
ws-dis-salary
ws-salary
ws-bonus
ws-comm
ws-empno
WS-START-OF-DOC
88 WS-START-DOC
88 WS-NOT-START-DOC

comp
comp
comp
comp

pic
pic
pic
pic
pic
pic
pic
pic
pic

999999999.
x(30).
s9(9)v99 value zero.
$$$,$$$,$$9.99.
s9(9)v99 value zero.
s9(9)v99 value zero.
s9(9)v99 value zero.
x(06)
value spaces.
X.
VALUE 'N'.
VALUE 'Y'.

This is a simple set of working storage variables used to process the XML document and do some computations for the final output.

19

Procedure Division parse program

Procedure Division.
mainline section.
DISPLAY 'XML Event
XML Text'.
SET WS-NOT-START-DOC TO TRUE.

XML PARSE xml-document PROCESSING PROCEDURE xml-handler

ON EXCEPTION
display 'XML document error ' XML-CODE
NOT ON EXCEPTION
display 'XML document successfully parsed'
END-XML.
DISPLAY ' '.
DISPLAY '-----+++++***** Using information from XML '*****+++++-----'.
DISPLAY ' '.
DISPLAY ' Employee Number: ' ws-empno.
COMPUTE ws-compensation = ws-salary + ws-bonus + ws-comm.
move ws-compensation to ws-dis-salary.
DISPLAY ' Total Compensation: ' ws-dis-salary.
GOBACK.

To keep things simple Im setting the start of processing at the beginning of the program.
Im then invoking the XML PARSE statement passing the working storage xml-document.
When the xml event is triggered control will be passed back to paragraph xml-handler.
Once XML processing is complete the rest of mainline will be processed.

20

XML-Handler
xml-handler section.
if ws-start-doc then
compute xml-text-len = function length(xml-text)
*
display 'doc length:' xml-text-len
end-if.
evaluate XML-EVENT
* ==> Order XML events most frequent first
when 'START-OF-ELEMENT'
display 'Start element tag: <' XML-TEXT '>'
move XML-TEXT to current-element
when 'CONTENT-CHARACTERS'
display 'Content characters: <' XML-TEXT '>'
* ==> Transform XML content to operational COBOL data
item...
evaluate current-element
when 'salary'

xml-handler will look for the event type then perform some task.
In my example at start of document processing I compute the length of the entire document just as a debugging aid.
Now I display all the different XML Events along with the data that triggered the event. Another learning aid.
I also want to do some processing with some element so Im looking for the element name salary, bonus and Comm.
Continue on the next slide

21

XML-Handler
* ==> Using function NUMVAL-C...
compute ws-salary = function numval-c(XML-TEXT)
when 'bonus'
compute ws-bonus = function numval-c(XML-TEXT)
when 'comm'
compute ws-comm = function numval-c(XML-TEXT)
when 'empno'
move xml-text
to ws-empno
end-evaluate
when 'END-OF-ELEMENT'
display 'End element tag: <' XML-TEXT '>'
move spaces to current-element
when 'START-OF-DOCUMENT'
display 'Start of document processing'
SET WS-START-DOC TO TRUE
when 'END-OF-DOCUMENT'
display 'End of document.'

This is a continuation of the previous slide.

22

XML-Handler
when 'VERSION-INFORMATION'
display 'Version: <' XML-TEXT '>'
when 'ENCODING-DECLARATION'
display 'Encoding: <' XML-TEXT '>'
when 'STANDALONE-DECLARATION'
display 'Standalone: <' XML-TEXT '>'
when 'ATTRIBUTE-NAME'
display 'Attribute name: <' XML-TEXT '>'
when 'ATTRIBUTE-CHARACTERS'
display 'Attribute value characters: <' XML-TEXT '>'
when 'ATTRIBUTE-CHARACTER'
display 'Attribute value character: <' XML-TEXT '>'
when 'START-OF-CDATA-SECTION'
display 'Start of CData: <' XML-TEXT '>'
when 'END-OF-CDATA-SECTION'
display 'End of CData: <' XML-TEXT '>'
when 'CONTENT-CHARACTER'
display 'Content character: <' XML-TEXT '>'

Continue showing the different event types and associated data

23

XML-Handler
when 'PROCESSING-INSTRUCTION-TARGET'
display 'PI target: <' XML-TEXT '>'
when 'PROCESSING-INSTRUCTION-DATA'
display 'PI data: <' XML-TEXT '>'
when 'COMMENT'
display 'Comment: <' XML-TEXT '>'
when 'EXCEPTION'
display 'Exception' XML-CODE
display 'text:' xml-text
display 'len:' xml-text-len
when 'other'
display 'Unexpected XML event:' XML-EVENT '.'
end-evaluate .
End program XMLPARS1.

Continue showing the different event types and associated data

24

Parsed Output
XML Event
XML Text
Start of document processing
Version: <1.0>
Encoding: <ibm-1140>
Comment: < Employee HR Information >
Start element tag: <emp>
Start element tag: <empno>
Content characters: <123456>
End element tag: <empno>
Start element tag: <name>
Content characters: <
>
Start element tag: <firstnme>
Content characters: <Troy
>
End element tag: <firstnme>

*** Cut a bunch of lines from this sample output ****

Start of CData: <>
Content characters: <We should add a <picture> element
in the future!>
End of CData: <>
Content characters: <
>
PI target: <Sign>
PI data: <Contract>
End element tag: <emp>
End of document.
XML document successfully parsed
-----+++++***** Using information from XML *****+++++--Employee Number: 123456
Total Compensation: $78,750.00

Page 1

Page 2

This is page 1 and 2 of the sample output generated by parsing the working storage xml-document

25

Publish XML (Generate)

XML GENERATE xml-document
FROM cobol-data-item
ON EXCEPTION
display XML document error XML-CODE
NOT ON EXCEPTION
display XML document successfully published
END-XML

The XML GENERATE command is used to build an XML document.

The input or source data from the cobol-data-item will be formatted into an XML document and stored in xml-document.

26

COBOL-Data-Item
01 EMP.
10 EMPNO
10 NAME.
15 FIRSTNME
15 LASTNAME
10 WORKDEPT
10 PHONENO
10 HIREDATE
10 JOB
10 EDLEVEL
10 SEX
10 BIRTHDATE
10 SALARY
10 BONUS
10 COMM

PIC X(6)

value 123456.

pic x(12)
value Troy .
pic x(15)
value Coleman .
PIC X(3)
value D01.
PIC X(4)
value 4459.
PIC X(10)
value 03/30/2001.
PIC X(8)
value MANAGER .
PIC S9(4) USAGE COMP value 16.
PIC X(1) value M.
PIC X(10) value 01/13/1963.
PIC 9(7)V9(2) USAGE display value 78250.00.
PIC 9(7)V9(2) USAGE display value 500.00.
PIC 9(7)V9(2) USAGE display value zero.

In my sample program Im using a static working storage area for the source of
input. You of course will read this in from CICS, IMS, DB2, or files.
Remember that the XML tag names are taken from the COBOL variable
names.
Is this case the document will be EMP and will have a tag for EMPNO and
NAME and .. COMM

27

Sample COBOL code to publish XML

MOVE SPACES TO XML-DOC.
XML GENERATE xml-doc
FROM EMP
WITH XML-declaration
ON EXCEPTION
display XML document error XML-CODE
NOT ON EXCEPTION
display XML document successfully generated

END-XML

The procedure division sample code.

First clear out the xml document before calling XML GENERATE.
Now generate an XML document xml-doc from the COBOL field EMP.
The with XML-DECLARATION tells the generator to add the XML version
at the beginning of the doc.

28

XML document generated from EMP data item

<?xml version=1.0 encoding=IBM-1047?>
<EMP>
<EMPNO>123456</EMPNO>
<NAME>
<FIRSTNME>Troy</FIRSTNME>
<LASTNAME>Coleman</LASTNAME>
</NAME>
<WORKDEPT>D01</WORKDEPT>
<PHONENO>4459</PHONENO>
<HIREDATE>03/30/2001</HIREDATE>
<JOB>MANAGER</JOB>
<EDLEVEL>16</EDLEVEL>
<SEX>M</SEX>
<BIRTHDATE>01/13/1963</BIRTHDATE>
<SALARY>78250.00</SALARY>
<BONUS>500.00</BONUS>
<COMM>.00</COMM>
</EMP>

As you can see the first line of text has the XML DECLARATION then the
EMP document.
The names are taken from the COBOL variable names.

29

DB2 Scalar XML Publishing Functions

XMLDOCUMENT (XML Expression)
The XMLDOCUMENT function returns an XML value with a
single document node and zero or more nodes as its
children.
REMEMBER: the first thing you should have in the SQL SELECT is the
function XMLDOCUMENT followed by XML expressions. This will
ensure you have returned a well formed XML document.

Now I would like to talk about publishing XML Documents from DB2.
The first XML function needed as you are building the document is
XMLDOCUMENT.
Otherwise you will not have a well formed document and insert into an XML
column will fail.

30

DB2 Scalar XML Publishing Functions

XMLELEMENT(NAME element name
, xmlnamespace declaration
, xmlattributes-function
, element content expression
OPTION empty on null | null on null )
Can be one or more expressions
<customer xmlns:info="http://troylcoleman.com cid="1">
<info:name>Troy Coleman</info:name>
</customer>

The XMLELEMENT function is used to build the element which includes its
name, optional namespace, attribute, and content.
In this example Ive built the element CUSTOMER which is made up of
namespace, attribute CID and element NAME

31

DB2 Scalar XML Publishing Functions

XMLNAMESPACES( namespace uri as namespace prefix)
This function returns a namespace declaration which
provides a unique context to the element.
XMLNAMESPACES('http://troylcoleman.com' as "info")
Note: The name is case sensitive. If you define the name in
uppercase INFO but then reference in lowercase info,
you will receive SQL error -20275 reason code 2

I used the XMLNAMESPACE function to provide a unique prefix on the

element names.

32

DB2 Scalar XML Publishing Functions

XMLATTRIBUTES (attribute value expression AS attribute-name)
The attribute function provides additional information
about the element. In the DB2 world this is usually an
identifying key value.
In this example the primary key customerid will be
returned as CID.
XMLATTRIBUTES(c.customerid as CID)

The XMLATTRIBUTES function is used when you need to provide what I

think of as the PRIMARY KEY for the element.
This is the identifying key value used for searching.

33

DB2 Scalar XML Publishing Functions

XMLCOMMENT (string expression)
You may find at times you have a need to put comments
into the XML document. The comment function returns a
single comment node from a string expression.
XMLCOMMENT('CID is the Customer ID')
<!--CID is the Customer ID-->

You may find that you need to add a comment in the XML document. The
XMLCOMMENT function will insert that comment for you.

34

More DB2 Scalar XML Publishing Functions

XMLCONCAT
XMLFOREST
XMPARSE
XMLPI
XMLQUERY
XMLSERIALIZE
XMLTEXT

There are many more XML publish functions that can be used for more
complex xml documents.
This is a list of these functions.

35

Publishing XML from DB2 Relational Data

SELECT BIGINT(C.CUSTOMERID)
, XMLDOCUMENT(
XMLELEMENT(NAME "customer"
,XMLNAMESPACES('http://troylcoleman.com'
as "info")
,XMLATTRIBUTES(C.CUSTOMERID AS "cid")
,XMLCOMMENT('CID is the Customer ID')
,XMLELEMENT(NAME "info:name"
,XMLELEMENT(NAME "info:firstname",C.FIRSTNAME)
,XMLELEMENT(NAME "info:lastname",C.LASTNAME)
)
,XMLELEMENT(NAME "info:addr"
,XMLATTRIBUTES('USA' AS "country")
,XMLELEMENT(NAME "info:street",C.STREET_ADDRESS)
,XMLELEMENT(NAME "info:city", C.CITY)
,XMLELEMENT(NAME "info:state",C.STATE)
,XMLELEMENT(NAME "info:zip", C.ZIP) )
,XMLELEMENT(NAME "info:agegroupid", C.AGEGROUPID)
,XMLELEMENT(NAME "info:income_rangeid",C.INCOME_RANGEID)
,XMLELEMENT(NAME "info:marital_status",C.MARITAL_STATUS)
,XMLELEMENT(NAME "info:gender",C.GENDER)
,XMLELEMENT(NAME "info:ethnicgroupid",C.ETHNICGROUPID)
)
)
FROM CUSTOMERS C

This is a sample SQL statement used to publish relational data into an XML
Document.
Notice I start with the XMLDOCUMENT function followed by an
XMLELEMENT function that is made up of a list of other XMLELEMENT
functions.

36

XML Document from DB2 Relational Data

<?xml version="1.0" encoding="IBM1047"?>
<customer xmlns:info="http://troylcoleman.com" cid="1">
<!--CID is the Customer ID-->
<info:name>
<FIRSTNAME>JOHN</FIRSTNAME>
<LASTNAME>DOE</LASTNAME>
</info:name>
<info:addr><STREET_ADDRESS>4356 SARATOGA AVE</STREET_ADDRESS>
<CITY>SAN JOSE</CITY>
<STATE>CALIFORNIA</STATE>
<ZIP>94534-0567</ZIP>
</info:addr>
<AGEGROUPID>1</AGEGROUPID>
<INCOME_RANGEID>1</INCOME_RANGEID>
<MARITAL_STATUS>U</MARITAL_STATUS>
<GENDER>M</GENDER>
<ETHNICGROUPID>1</ETHNICGROUPID>
</customer>

This is the result of the previous SQL statement using the XML publishing
functions.

37

Convert XML Document into Relational Table

XMLTABLE(xmlnamespace-declaration,
row-xquery-expression-constant
PASSING row-xquery-argument
COLUMNS xml-table-regular-column-definition )

You may find that you have some XML data and would like to publish this
data as a relational table to be used in join processing with other tables.
The XMLTABLE function will do this for you.
In my example on the next slide I did not use XMLNAMESPACE.
I did use $CUST as the row-xquery-expression-constant which is used by the
passing row-xquery-argument.
The COLUMNS function looks like what you would see in a CREATE table
statement for each column being defined.

38

XMLTABLE Statement
select cx.cust_id, cx.firstname, cx.lastname, cx.street
, cx.city, cx.state, cx.country
from dsn8910.customer c
, xmltable('$cust/customer' passing info as "cust"
columns
cust_id
integer
path '@cid'
, firstname
varchar(10) path 'name/firstname'
, lastname
varchar(10) path 'name/lastname'
, street
varchar(20) path 'addr/street'
, city
varchar(11) path 'addr/city'
, state
varchar(10) path 'addr/state'
, country
varchar(03) path 'addr/@country'
) as cx
order by cx.lastname asc

In this SQL statement Im taking the XML column INFO found in the
customer table and parsing it with the XMLTABLE function to return what
looks like a table. The table returned is CX. Notice Im doing an order by on
CX.LASTNAME.

39

Insert sample XML Document

INSERT INTO COLTR05.CUSTOMER
SELECT BIGINT(1000)
, XMLDOCUMENT(
XMLELEMENT(NAME "customer"
,XMLNAMESPACES('http://troylcoleman.com' as "info")
,XMLATTRIBUTES(1000 AS "cid")
,XMLCOMMENT('CID is the Customer ID')
,XMLELEMENT(NAME "name"
,XMLELEMENT(NAME "firstname" ,'TROY')
,XMLELEMENT(NAME "lastname" ,'COLEMAN')
)
,XMLELEMENT(NAME "addr"
,XMLATTRIBUTES('USA' AS "country")
,XMLELEMENT(NAME "street" ,'123 Main Street')
,XMLELEMENT(NAME "city" , 'Palatine')
,XMLELEMENT(NAME "state" ,'IL')
,XMLELEMENT(NAME "zip" ,'60067')
)
)
)
FROM SYSIBM.SYSDUMMY1

When Im testing my SQL I usually prototype the statement and verify that it
is a valid XML document.
To do this I publish an XML document using the SELECT publishing
functions and once it looks good I prefix the SELECT statement with an
INSERT statement.
In this case I have two columns. CID and INFO. The CID is set to 1000 and
INFO will contain the published document.

40

XMLTABLE results in relational data

CUST_ID FIRSTNAME
LASTNAME
STREET
CITY
STATE
-----+---------+---------+---------+---------+---------+---------+---------+
1000 TROY
COLEMAN
123 Main Street
Palatine
IL

The result of XMLTABLE is a relational table that can be

joined and processed just like any other DB2 table.

Using the XMLTABLE function against the row I just inserted into INFO
would product the following relational table which can be joined with other
tables.

41

Future XML Improvements with DB2 10

pureXML Enhancements

Schema Validation in Engine

Update of Subparts of XML Document
Multi-Versioning
Binary XML Exchange Format
XML Data Type in Routines
XML Index matching with Date/Timestamp
XML CHECK Utility

Many new enhancements have been delivered in DB2 10 pureXML to improve

performance and functionality.
A few key enhancements are Schema validation, Multi-Versioning, and
Update of subparts of XML document

42

DB2 10 Schema Validation in Engine

INSERT INTO customer CID, INFO
VALUES (1000, XMLVALIDATE(:xml-document,
SYSXSR.CUSTSCHEMA));
Warning: Avoid high overhead of XML validation in
production. Use INSERT trigger in development to verify
and make sure the XML documents are valid.

DB2 now supports validation of the XML document in the engine. In this case
the CUSTSCHEMA was registered before processing this statement.
XML Validation is very CPU intensive. You will want to avoid this in a
production environment.

43

DB2 10 Update of Subparts of XML Document

UPDATE MYCUSTOMER
SET INFO = XMLMODIFY(
'declare default element namespace "http://posample.org";
replace value of node /customerinfo/addr/street
with "42 Rosedale"')
WHERE CID=1000#

Its been a long time requirement to have the ability to update a portion of the
XML document and not take the performance hit with updating the entire
document.
The XMLMODIFY provides the ability to update a portion of the XML
Document.

44

DB2 10 Multi-Versioning
XML Column supporting multiple versions
Universal Table Space
DB2 10 NFM

INSERT Version number assigned

UPDATE Improved Concurrency and Memory Usage
Entire document new version
Partial document new version

Multiple versions of an XML document can coexist in an XML table. The

existence
of multiple versions of an XML document can lead to improved concurrency
through lock avoidance. In addition, multiple versions can save real storage by
avoiding a copy of the old values in the document into memory in some cases.

45

DB2 10 Binary XML Exchange Format

JDBC or SQLJ Application
JDBC 4.0 using java.sql.SQLXML data type

ODBC using SQL_C_BINARYXML data type

LOAD & UNLOAD utility using field specifications
CHAR BLOBF template-name BINARYXML
VARCHAR BLOBF template-name BINARYXML
XML BINARYXML

The Binary data type is not supported at this point in time by COBOL.
The current API format is for JDBC, SQLJ, ODBC, or UNLOAD utility

46

DB2 10 XML Data Type in Routines

Native XML Data Type for Parameters
Native SQL Procedures
SQL Scalar Functions
SQL Table Functions

Native XML Data Type for Variables

Native SQL Procedures
SQL Scalar Functions

Native SQL routines support parameters and variables with the XML data
type.
XML parameters can be used in SQL statements in the same way as variables
of any other data type.
In addition, variables with the XML data type can be passed as parameters to
XQuery expressions in XMLEXISTS, XMLQUERY and XMLTABLE
expressions.

47

DB2 10 XML Index matching Date/Timestamp

Data, Time, and Timestamp data type support

xs:dateTime
xs:time
xs:date
xs:duration
xs:yearMonthDuration
xs:datTimeDuration

Version 10 includes date and time support for XML data types and functions.
This includes matching index support using comparison operators on duration,
date, and time values

48

DB2 10 XML CHECK Utility

Verify all nodes in document are structurally intact
Node ID index is consistent with XML Table Space
Schema Validation - XML Type Modifier
Validate all documents with at least one XML Schema with an
XML type modifier

Version 10 adds functionality to the CHECK DATA utility, so that you can
use this utility to verify the consistency of XML documents that are stored in a
separate XML table space.
The CHECK DATA utility verifies that all nodes in that XML document are
structurally intact and that the node ID index is consistent with the content that
is
in the XML table space.
In addition, this utility verifies that all of the XML documents of an XML
column are valid against at least one XML schema that is
specified in the XML type modifier

49

Troy Coleman
CA Technologies
troy.coleman@ca.com
E11
XML for z/OS COBOL Developers

50