Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
345 views

1 Structured Query Language: 2.1 Create Database

Structured Query Language (SQL) is used to work with relational databases. It has two primary parts: Data Definition Language (DDL) for defining database structure (e.g. CREATE TABLE), and Data Manipulation Language (DML) for manipulating data (e.g. SELECT, INSERT). DDL commands include CREATE DATABASE, CREATE TABLE, DROP TABLE, ALTER TABLE. DML commands include INSERT for adding data, SELECT for querying data, UPDATE for modifying data, and DELETE for removing data. SQL also includes Data Control Language (DCL) for security with GRANT and REVOKE statements. While SQL is standardized, each database vendor adds their own proprietary extensions.

Uploaded by

KAVIARASU.D
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
345 views

1 Structured Query Language: 2.1 Create Database

Structured Query Language (SQL) is used to work with relational databases. It has two primary parts: Data Definition Language (DDL) for defining database structure (e.g. CREATE TABLE), and Data Manipulation Language (DML) for manipulating data (e.g. SELECT, INSERT). DDL commands include CREATE DATABASE, CREATE TABLE, DROP TABLE, ALTER TABLE. DML commands include INSERT for adding data, SELECT for querying data, UPDATE for modifying data, and DELETE for removing data. SQL also includes Data Control Language (DCL) for security with GRANT and REVOKE statements. While SQL is standardized, each database vendor adds their own proprietary extensions.

Uploaded by

KAVIARASU.D
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

1 Structured Query Language

SQL, or Structured Query Language is the most popular declarative language used to work
with Relational Databases. Originally developed at IBM, it has been subsequently standard-
ized by various standards bodies (ANSI, ISO), and extended by various corporations adding
their own features (T-SQL, PL/SQL, etc.).
There are two primary parts to SQL: The DDL and DML (& DCL).

2 DDL - Data Definition Language


DDL is a standard subset of SQL that is used to define tables (database structure), and
other metadata related things. The few basic commands include: CREATE DATABASE, CREATE
TABLE, DROP TABLE, and ALTER TABLE.
There are many other statements, but those are the ones most commonly used.

2.1 CREATE DATABASE


Many database servers allow for the presence of many databases1 . In order to create a
database, a relatively standard command ‘CREATE DATABASE’ is used.
The general format of the command is:
CREATE DATABASE <database-name> ;
The name can be pretty much anything; usually it shouldn’t have spaces (or those spaces
have to be properly escaped). Some databases allow hyphens, and/or underscores in the
name. The name is usually limited in size (some databases limit the name to 8 characters,
others to 32—in other words, it depends on what database you use).

2.2 DROP DATABASE


Just like there is a ‘create database’ there is also a ‘drop database’, which simply removes the
database. Note that it doesn’t ask you for confirmation, and once you remove a database,
it is gone forever2 .
DROP DATABASE <database-name> ;

2.3 CREATE TABLE


Probably the most common DDL statement is ‘CREATE TABLE’. Intuitively enough, it is used
to create tables. The general format is something along the lines of:
CREATE TABLE <table-name> (
...
);
1
Note that some severs, like Oracle, use ‘name spaces’ to separate the database into logical parts used by
various users—so there is less emphasis on individual users having ‘their own’ database.
2
It is very important to make regular backups!

1
The ... is where column definitions go. The general format for a column definition is the
column name followed by column type. For example:
PERSONID INT
Which defines a column name PERSONID, of type INT. Column names have to be comma
separated, ie:
CREATE TABLE PERSON (
PERSONID INT,
LNAME VARCHAR(20),
FNAME VARCHAR(20) NOT NULL,
DOB DATE,
PRIMARY KEY(PERSONID)
);
The above creates a table named person, with person id, last name, first name, and date
of birth. There is also the ‘primary key’ definition. A primary key is a column value that
uniquely identifies a database record. So for example, we can have two ‘person’ records with
the same last name and first name, but with different ids.
Besides for primary key, there are many other flags we can specify for table columns. For
example, in the above example, FNAME is marked as NOT NULL, which means it is not allowed
to have NULL values3 .
Many databases implement various extensions to the basics, and you should read the
documentation to determine what features are present/absent, and how to use them.

2.4 DROP TABLE


Just like there is a ‘create table’ there is also a ‘drop table’, which simply removes the table.
Note that it doesn’t ask you for confirmation, and once you remove a table, it is gone forever.
DROP TABLE <table-name> ;

2.5 ALTER TABLE


There is a command to ‘alter’ tables after you create them. This is usually only useful if the
table already has data, and you don’t want to drop it and recreate it (which is generally much
simpler). Also, most databases have varying restrictions on what ‘alter table’ is allowed to
do. For example, Oracle allows you do add a column, but not remove a column.
The general syntax to add a field is:
ALTER TABLE <table-name>
ADD <field-name> <data-type>
The field declaration is pretty much exactly what it is in the ‘create table’ statement.
The general syntax to drop a field is:
3
Note that just because something is not null, doesn’t mean it cannot have an ‘empty’ value; i.e.: a string
” is empty, but not null.

2
ALTER TABLE <table-name>
DROP <field-name>
Note that very few databases let you drop a field. The drop command is mostly present to
allow for dropping of constraints (such as indexes, etc.) on the table.
The general syntax to modify a field (change its type, etc.) is:
ALTER TABLE <table-name>
MODIFY <field-name> <new-field-declaration>
Note that you can only do this to a certain extent on most databases. Just as with ‘drop’,
this is mostly useful for working with table constraints (changing ‘not null’ to ‘null’, etc.)

3 DML - Data Manipulation Language


This is a standard subset of SQL that is used for data manipulation. Intuitively, we need to
first inset data into the database. Once it’s there, we can retrieve it, modify it, and delete
it. These directly correspond to: INSERT, SELECT, UPDATE, and DELETE statements.

3.1 INSERT Statement


To get data into a database, we need to use the ‘insert’ statement. The general syntax is:
INSERT INTO <table-name> (<column1>,<column2>,<column3>,...)
VALUES (<column-value1>,<column-value2>,<column-value3>);
The column names (i.e.: column1, etc.) must correspond to column values (i.e.: column-value1,
etc.). There is a short-hand for the statement:
INSERT INTO <table-name>
VALUES (<column-value1>,<column-value2>,<column-value3>);
In which the column values must correspond exactly to the order columns appear in the ‘cre-
ate table’ declaration. It must be noted, that this sort of statement should (or rather, must)
be avoided! If someone changes the table, moves columns around in the table declaration,
the code using the shorthand insert statement will fail.
A typical example, of inserting the ‘person’ record we’ve created earlier would be:
INSERT INTO PERSON(PERSONID,LNAME,FNAME,DOB)
VALUES(1,’DOE’,’JOHN’,’1956-11-23’);

3.2 SELECT Statement


Probably the most used statement in all of SQL is the SELECT statement. The select state-
ment has the general format of:
SELECT <column-list>
FROM <table-list>
WHERE <search-condition>

3
The column-list indicates what columns you’re interested in (the ones which you want
to appear in the result), the table-list is the list of tables to be used in the query, and
search-condition specifies what criteria you’re looking for.
An example of a short-hand version to retrieve all ‘person’ records we’ve been using:
SELECT * FROM PERSON;

3.3 The WHERE Clause


The WHERE clause is used in UPDATE, DELETE, and SELECT statements, and has the same
format in all these cases. It has to be evaluated to either true or false. Table 1 lists some of
the common operators.

= equals to
> greater than
< less than
>= greater than or equal to
<= less than or equal to
<> not equal to

Table 1: SQL Operators

There is also IS, which can be used to check for NULL values, for example:
column-name IS NULL
We can also use AND, OR4 and parenthesis to group expressions.
Besides for these operators, we can also call built-in functions (as well as stored procedures
we define ourselves—that is, if the database supports stored procedures).
An example of the operators in use would be: something < 5 OR something is NULL
AND somedate = TO DATE(’01/03/93’,’MM/DD/YY’)5 .

3.4 UPDATE Statement


The update statement is used for changing records. The general syntax is:
UPDATE <table-name>
SET <column1> = <value1>, <column2> = <value2>, ...
WHERE <criteria>
The criteria is what selects the records for update. The ‘set’ portion indicates which columns
should be updated and to what values. An example of the use would be:
UPDATE PERSON
SET FNAME=’Clark’, LNAME=’Kent’
WHERE FNAME=’Superman’;
4
Usually, the AND operator has higher precendence than OR.
5
The function TO DATE is available on Oracle.

4
3.5 DELETE Statement
The ‘delete’ is used to remove elements from the database. The syntax is very similar to
update and select statements:

DELETE FROM <table-name>


WHERE <criteria>

Basically we select which records we want to delete using the where clause. An example use
would be:

DELETE FROM PERSON


WHERE PERSONID=12345;

4 DCL - Data Control Language


This is a standard subset of SQL that is used for security management. Most databases have
their own flavored syntax, but generally, there exist two commands: GRANT, and REVOKE,
these ‘grant’ and ‘remove’ privileges.

4.1 GRANT Statement


The general format is something like this:

GRANT <privilege> ON <object> TO <who> ;

Basically, a privilege can be something like ‘update’ or ‘select’, etc., or it can be ‘all’ when
granting ‘all’ privileges. An ‘object’ can be pretty much anything, but is often a database,
or database table. The ‘who’ is generally a database login/user.
Some databases (like MySQL) will actually create the user if they don’t already exist.
In some cases, you also have the option of specifying the user password, via: ‘identified by’
parameter.

4.2 REVOKE Statement


The revoke is the opposite of GRANT. The general format is:

REVOKE <privilege> ON <object> FROM <who> ;

The individual elements are the same as for GRANT statement.

5 Flavors of SQL
Basically every database implements its own version of SQL. The basic statements (like
the ones described) are almost always there and are almost always exactly the same on all
databases.

5
Extensions come in when you’re trying to do something fancy, like create stored proce-
dures, or work with a specific database environment. Microsoft SQL Server and Sybase use
a variant called T-SQL, for Transact-SQL. To contrast, Oracle uses PL/SQL, for Procedural
Language SQL. Some databases allow developers to write stored procedures in Java (like
Oracle) or something as simple as C language (PostgreSQL). And some databases don’t
even have stored procedures (at this time), like MySQL.

You might also like