SQL Server Questions

What is Microsoft SQL Server?
Microsoft SQL Server is a relational database management system (RDBMS) developed by

Microsoft. It runs on Windows systems and uses Transact-SQL as the query language.
Microsoft SQL Server release history:

 1993 - SQL Server 4.21 for Windows NT
 1995 - SQL Server 6.0, codenamed SQL95
 1996 - SQL Server 6.5, codenamed Hydra
 1999 - SQL Server 7.0, codenamed Sphinx
 1999 - SQL Server 7.0 OLAP, codenamed Plato
 2000 - SQL Server 2000 32-bit, codenamed Shiloh (version 8.0)
 2003 - SQL Server 2000 64-bit, codenamed Liberty
 2005 - SQL Server 2005, codenamed Yukon (version 9.0)
 2005 - SQL Server 2005 Express Edition, restricted free version
 How to download Microsoft SQL Server 2005 Express Edition?

 Microsoft SQL Server 2005 Express Edition is the free version of the Microsoft SQL Server 2005.
If you are interested to try SQL Server 2005, you should follow this tutorial to download Microsoft
SQL Server 2005 Express Edition:
 1. Go to the Microsoft SQL Server 2005 Express Edition download page.
 2. Go to the Files in This Download section, and click Download button next to the
"SQLEXPR.EXE - 53.5 MB" file. The File Download box shows up.
 3. Save the download file to C:\temp directory. When the download is done, you should get the
following file:
 Name: SQLEXPR.EXE
 Description: Microsoft SQL 2005 Server Express Edition
 Location: C:\temp
 Size: 56,105,688 bytes
 Version: 9.0.1399.6
 What are the requirements to install SQL Server 2005 Express Edition?
 The following system requirements cover the SQL Server 2005 Express Edition:
 Processor
 32-bit Processor of 600-megahertz (MHz) or faster

 Operating System
 Windows XP with Service Pack 2 or later
 Windows 2000 Professional with SP4
 Windows 2000 Server with Service Pack 4 or later
 Windows Server 2003 Standard, or Enterprise SP1
 Windows Server 2003 Web Edition SP1
 Windows Small Business Server 2003 with SP1
 Vista Home Basic and above

 Framework
 .NET Framework 2.0

 Memory
 512 megabytes (MB) or more recommended

 Hard Disk
 Approximately 425 MB of available hard-disk space
 Why I am getting "The Microsoft .Net Framework 2.0 in not installed" message?
 When you try to install SQL Server 2005 Express Edition, you may get a Microsoft SQL Server
2005 Setup error box with this message: "The Microsoft .Net Framework 2.0 in not installed.
Please install before running setup."
 You are getting this error, because .NET Framework 2.0 is not installed on your system yet. Read
the next tutorial to download and install .NET Framework 2.0.
 How to download and install Microsoft .NET Framework Version 2.0?
 .NET Framework Version 2.0 is required by many Microsoft applications like SQL Server 2005. If
you want download and install .NET Framework Version 2.0, you should follow this turorial:
 1. Go to the Microsoft .NET Framework Version 2.0 Redistributable Package (x86) page.
 2. Click the Download button. Save the download file dotnetfx.exe to c:\temp.
 3. Look at and compare the downloaded file properties with:
 Name: dotnetfx.exe
 Size: 23,510,720 bytes
 Version: 2.0.50727.42
 4. Close all IE (Internet Explorer) windows.
 5. Double click the downloaded file: to c:\temp\dotnetfx.exe. The Microsoft .NET Framework 2.0
Setup windows shows up. Follow the intructions to finish the installtion.
 What is mscorsvw.exe - Process - Microsoft .NET Framework NGEN?
 Process mscorsvw.exe is installed as a system service as part of the .NET Framework 2.0. You
can disable it, if you are not using any applications that require .NET Framework 2.0.
 mscorsvw.exe process and program file info:
 CPU usage: 00%
 Memory usage: 2,704K
 Launching method: System Service
 Directory: C:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\
 File name: mscorsvw.exe
 Description: .NET Runtime Optimization Service
 Size: 66,240 bytes
 Date: Friday, September 23, 2005, 7:28:56 AM
 Version: 2.0.50727.42
 Company name: Microsoft
 System essential: No
 Virus/Spyware/Adware: No
 How to install SQL Server 2005 Express Edition?
 Once you have downloaded SQL Server 2005 Express Edition, you should follow this tutorial to
install it on your system:
 1. Double click SQLEXPR.EXE. The setup window shows up.
 2. Click Next to let the setup program to unpack all files from the downloaded file.
 3. When unpack is down, the setup program will check all required programs on your system.
 4. Then the setup program will start the installation process.
 5. On the Authentication Mode window, click the radio button for Mixed Mode (Windows
Authentication and SQL Server Authentication). And enter "FYIcenter" in the "Specify the sa
logon password below:" fields.
 6. Continue to finish the installation process.
 7. When installation is done, you will see a new program menu entry as: Start > Programs >
Microsoft SQL Server 2005 > Configuration Tools.
 How do you know if SQL Server is running on your local system?
 After installing SQL Server 2006 Express Edition, it will be running on your local system quietly as
a background process.
 If you want to see this process is running, run Windows Task Manager. You should see a process
called sqlservr.exe running in the process list:
 sqlservr.exe 00 1,316 K
 If you select sqlservr.exe and click the "End Process" button, SQL Server will be stopped.
 If you can not find sqlservr.exe in the process list, you know that your SQL Server is running.
 What is sqlservr.exe - Process - SQL Server (SQLEX?PRESS)
 Process sqlservr.exe is the Microsoft SQL Server system service installed as part of the Microsoft
SQL Server 2005 Express Dditon.
 mscorsvw.exe process and program file info:
 CPU usage: 00%
 Memory usage: 1,316K
 Launching method: System Service
 Directory: C:\Program Files\Microsoft SQL Server\MSSQL.1
 \MSSQL\Binn
 File name: sqlservr.exe
 Description: SQL Server Windows NT
 Size: 28,768,528 bytes
 Date: Friday, October 14, 2005, 3:51:46 AM
 Version: 2005.90.1399.0
 Company name: Microsoft
 System essential: No
 Virus/Spyware/Adware: No
 How to download and install Microsoft SQL Server Management Studio Express?
 Microsoft SQL Server Management Studio Express (SSMSE) is a free, easy-to-use graphical
management tool for managing SQL Server 2005 Express Edition and SQL Server 2005 Express
Edition with Advanced Services. If you want to download and install it to your system, follow this
tutorial:
 1. Go to SQL Server Management Studio Express home page.
 2. Go to the "Files in This Download" section.
 3. Click the Download button next to "SQLServer2005_SSMSEE.msi - 38.5 MB" And save it to
c:\temp directory.
 4. Look at and compare the downloaded file properties with:
 Name: SQLServer2005_SSMSEE.msi
 Size: 40,364,032 bytes
 5. Double click to install. The setup window shows up. Follow the instructions to finish the
installation process.
 6. When installation is done, you will see a new program menu entry as: Start > Programs >
Microsoft SQL Server 2005 > SQL Server Management Studio Express
 How to connect SQL Server Management Studio Express to SQL Server 2005 Express?
 Once you have SQL Server 2005 Express installed and running on your local machine, you are
ready to connect SQL Server Management Studio Express to the server:
 Click Start > Programs > Microsoft SQL Server 2005 > SQL Server Management Studio Express
to launch SQL Server Management Studio Express.
 The "Connect to Server" box shows up. The Server Name field has a default value of
"LOCALHOST\SQLEXPRESS". So don't change it. Select "SQL Server Authentication" as the
Authentication. Enter enter "sa" as the Login, and "FYIcenter" as the Password.
.
 Click the Connect button, you should see the SQL Server Management Studio Express window
comes up.
 How to run Queries with SQL Server Management Studio Express?
 1. Launch and connect SQL Server Management Studio Express to the local SQL Server 2005
Express.
 2. Click on the "New Query" button below the menu line. Enter the following SQL statement in the
query window:
 SELECT 'Welcome to FYIcenter.com SQL Server!'
 3. Click the Execute button in the toolbar area. You should get the following in the result window:
 Welcome to FYIcenter.com SQL Server!
 See the following picture to know where the query and result windows are:
.
 How to download and install SQL Server 2005 Books Online?
 1. Go to the SQL Server 2005 Books Online download page.
 2. Click the download button, the File Download box shows up. Save the download file to c:\temp.
 3. Double click on the downloaded file: c:\temp\SqlServer2K5_BOL_Feb2007.msi. The
installation setup window shows up. Follow the instructions to finish the installation.
 4. When the installation is done. You will see a new entry in the Start menu: Start > Programs >
Microsoft SQL Server 2005 > Documentation and Tutorials
 How to run SQL Server 2005 Books Online on your local system?
 SQL Server 2005 Books Online can be accessed by a Web browser over the Internet. But you
can also download it and read it on your local system. If you have downloaded and installed SQL
Server 2005 Books Online package, you follow this tutorial to run it:
 1. Click Start > Programs > Microsoft SQL Server 2005 > Documentation and Tutorials >
Tutorials > SQL Server Tutorials. The SQL Server 2005 Books Online window shows up.
 2. Click the plus sign (+) next to "SQL Server 2005 Tutorials in the Contents window".
 3. Click the plus sign (+) next to "SQL Server Tools Tutorials".
 4. Click "Lesson 1: Basic Navigation in SQL Server Management Studio". The book content
shows up for you to read.
.
 How to use Transact-SQL statements to access the database engine?
 Transact-SQL statements can be used to access the database engine directly. Here are some
good tutorials provided by the SQL Server 2005 Books Online. See the SQL Server 2005
Tutorials > Database Engine Tutorials > Writing Transact-SQL Statements Tutorial section in the
SQL Server 2005 Books Online document.
 This tutorial is intended for users who are new to writing SQL statements. It will help new users
get started by reviewing some basic statements for creating tables and inserting data. This
tutorial uses Transact-SQL, the Microsoft implementation of the SQL standard. This tutorial is
intended as a brief introduction to the Transact-SQL language and not as a replacement for a
Transact-SQL class. The statements in this tutorial are intentionally simple, and are not meant to
represent the complexity found in a typical production database.
 How to create new databases with "CREATE DATABASE" statements?
 This is the first tutorial of a quick lesson on creating database objects with Transact-SQL
statements. This lesson shows you how to create a database, create a table in the database, and
then access and change the data in the table. Because this lesson is an introduction to using
Transact-SQL, it does not use or describe the many options that are available for these
statements. This tutorial assumes that you are running SQL Server Management Studio Express.
 Like many Transact-SQL statements, the CREATE DATABASE statement has a required
parameter: the name of the database. CREATE DATABASE also has many optional parameters,
such as the disk location where you want to put the database files. When you execute CREATE
DATABASE without the optional parameters, SQL Server uses default values for many of these
parameters. This tutorial uses very few of the optional syntax parameters.
 To create a database - In a Query Editor window, type but do not execute the following code:
 CREATE DATABASE TestData
 GO
 Use the pointer to select the words CREATE DATABASE, and then press F1. The CREATE
DATABASE topic in SQL Server 2005 Books Online should open. You can use this technique to
find the complete syntax for CREATE DATABASE and for the other statements that are used in
this tutorial.
 In Query Editor, press F5 to execute the statement and create a database named TestData.
When you create a database, SQL Server makes a copy of the model database, and renames
the copy to the database name. This operation should only take several seconds, unless you
specify a large initial size of the database as an optional parameter.
 How to create new table with "CREATE TABLE" statements?
 This is the second tutorial of a quick lesson on creating database objects with Transact-SQL
 To create a table, you must provide a name for the table, and the names and data types of each
column in the table. It is also a good practice to indicate whether null values are allowed in each
column.
 Most tables have a primary key, made up of one or more columns of the table. A primary key is
always unique. The Database Engine will enforce the restriction that any primary key value
cannot be repeated in the table.
 For a list of data types and links for a description of each, see Data Types (Transact-SQL).
 Note: The Database Engine can be installed as case sensitive or non-case sensitive. If the
Database Engine is installed as case sensitive, object names must always have the same case.
For example, a table named OrderData is a different table from a table named ORDERDATA. If
the Database Engine is installed as non-case sensitive, those two table names are considered to
be the same table, and that name can only be used one time.
 Before you create the table in this tutorial, execute the USE command to change the database
context to the TestData database. Otherwise, you will create the table in the database you were
connected to earlier. That was probably your default database. Unless your default database has
been changed, the default database is the master database. You should not create objects in the
master database.
 Switch the Query Editor connection to the TestData database - In a Query Editor window,
type and execute the following code to change your connection to the TestData database.
 USE TestData
 GO
 To create a table - In a Query Editor window, type and execute the following code to create a
simple table named Products. The columns in the table are named ProductID, ProductName,
Price, and ProductDescription. The ProductID column is the primary key of the table. int,
varchar(25), money, and text are all data types. Only the Price and ProductionDescription
columns can have no data when a row is inserted or changed. This statement contains an
optional element (dbo.) called a schema. The schema is the database object that owns the table.
If you are an administrator, dbo is the default schema. dbo stands for database owner.
 CREATE TABLE dbo.Products
 (ProductID int PRIMARY KEY NOT NULL,
 ProductName varchar(25) NOT NULL,
 Price money NULL,
 ProductDescription text NULL)
 GO
 How to insert and update data into a table with "INSERT" and "UPDATE" statements?
 This is the third tutorial of a quick lesson on creating database objects with Transact-SQL
 Now that you have created the Products table, you are ready to insert data into the table by using
the INSERT statement. After the data is inserted, you will change the content of a row by using
an UPDATE statement. You will use the WHERE clause of the UPDATE statement to restrict the
update to a single row. The four statements will enter the following data.
ProductID ProductName Price ProductDescription
1 Clamp 12.48 Workbench clamp
50 Screwdriver 3.17 Flat head
75 Tire Bar Tool for changing tires
3000 3mm Bracket .52
 The basic syntax is: INSERT, table name, column list, VALUES, and then a list of the values to
be inserted. The two hyphens in front of a line indicate that the line is a comment and the text will
be ignored by the compiler. In this case, the comment describes a permissible variation of the
syntax.
 To insert data into a table - Execute the following statement to insert a row into the Products
table that was created in the previous task. This is the basic syntax.
 -- Standard syntax
 INSERT dbo.Products (ProductID, ProductName, Price, ProductDescription)
 VALUES (1, 'Clamp', 12.48, 'Workbench clamp')
 GO
 The following statement shows how you can change the order in which the parameters are
provided by switching the placement of the ProductID and ProductName in both the field list (in
parentheses) and in the values list.
 -- Changing the order of the columns
 INSERT dbo.Products (ProductName, ProductID, Price, ProductDescription)
 VALUES ('Screwdriver', 50, 3.17, 'Flat head')
 GO
 The following statement demonstrates that the names of the columns are optional, as long as the
values are listed in the correct order. This syntax is common but is not recommended because it
might be harder for others to understand your code. NULL is specified for the Price column
because the price for this product is not yet known.
 -- Skipping the column list, but keeping the values in order
 INSERT dbo.Products
 VALUES (75, 'Tire Bar', NULL, 'Tool for changing tires.')
 GO
 The schema name is optional as long as you are accessing and changing a table in your default
schema. Because the ProductDescription column allows null values and no value is being
provided, the ProductDescription column name and value can be dropped from the statement
completely.
 -- Dropping the optional dbo and dropping the ProductDescription column
 INSERT Products (ProductID, ProductName, Price)
 VALUES (3000, '3mm Bracket', .52)
 GO
 To update the products table - Type and execute the following UPDATE statement to change
the ProductName of the second product from Screwdriver, to Flat Head Screwdriver.
 UPDATE dbo.Products
 SET ProductName = 'Flat Head Screwdriver'
 WHERE ProductID = 50
 GO
 How to read data in a table with "SELECT" statements?
 This is the fourth tutorial of a quick lesson on creating database objects with Transact-SQL
 Use the SELECT statement to read the data in a table. The SELECT statement is one of the most
important Transact-SQL statements, and there are many variations in the syntax. For this tutorial,
you will work with five simple versions.
 To read the data in a table - Type and execute the following statements to read the data in the
Products table.
 -- The basic syntax for reading data from a single table
 SELECT ProductID, ProductName, Price, ProductDescription
 FROM dbo.Products
 GO
 You can use an asterisk to select all the columns in the table. This is often used in ad hoc
queries. You should provide the column list in you permanent code so that the statement will
return the predicted columns, even if a new column is added to the table later.
 -- Returns all columns in the table
 -- Does not use the optional schema, dbo
 SELECT * FROM Products
 GO
 You can omit columns that you do not want to return. The columns will be returned in the order
that they are listed.
 -- Returns only two of the columns from the table
 SELECT ProductName, Price
 GO
 Use a WHERE clause to limit the rows that are returned to the user.
 -- Returns only two of the records in the table
 SELECT ProductID, ProductName, Price, ProductDescription
 WHERE ProductID < 60
 GO
 You can work with the values in the columns as they are returned. The following example
performs a mathematical operation on the Price column. Columns that have been changed in this
way will not have a name unless you provide one by using the AS keyword.
 -- Returns ProductName and the Price including a 7% tax
 -- Provides the name CustomerPays for the calculated column
 SELECT ProductName, Price * 1.07 AS CustomerPays
 GO
 How to create a login to access the database engine using "CREATE LOGIN" statements?
 This is the first tutorial of a quick lesson on creating login and configure users for databases with
Transact-SQL statements. Granting a user access to a database involves three steps. First, you
create a login. The login lets the user connect to the SQL Server Database Engine. Then you
configure the login as a user in the specified database. And finally, you grant that user permission
to database objects. This lesson shows you these three steps, and shows you how to create a
view and a stored procedure as the object. This tutorial assumes that you are running SQL
Server Management Studio Express.
 To access the Database Engine, users require a login. The login can represent the user's identity
as a Windows account or as a member of a Windows group, or the login can be a SQL Server
login that exists only in SQL Server. Whenever possible you should use Windows Authentication.
 By default, administrators on your computer have full access to SQL Server. For this lesson, we
want to have a less privileged user; therefore, you will create a new local Windows Authentication
account on your computer. To do this, you must be an administrator on your computer. Then you
will grant that new user access to SQL Server. The following instructions are for Windows XP
Professional.
 To create a new Windows account - Click Start, click Run, in the Open box, type %SystemRoot
%\system32\compmgmt.msc /s, and then click OK to open the Computer Management program.
 Under System Tools, expand Local Users and Groups, right-click Users, and then click New
User.
 In the User name box type Mary.
 In the Password and Confirm password box, type a strong password, and then click Create to
create a new local Windows user.
 To create a login - In a Query Editor window of SQL Server Management Studio, type and
execute the following code replacing computer_name with the name of your computer. FROM
WINDOWS indicates that Windows will authenticate the user. The optional
DEFAULT_DATABASE argument connects Mary to the TestData database, unless her
connection string indicates another database. This statement introduces the semicolon as an
optional termination for a Transact-SQL statement.
 CREATE LOGIN [computer_name\Mary]
 FROM WINDOWS
 WITH DEFAULT_DATABASE = [TestData];
 GO
 This authorizes a user name Mary, authenticated by your computer, to access this instance of
SQL Server. If there is more than one instance of SQL Server 2005 on the computer, you must
create the login on each instance that Mary must access.
 How to create a user to access a database using "CREATE USER" statements?
 This is the second tutorial of a quick lesson on creating login and configure users for databases
with Transact-SQL statements. Granting a user access to a database involves three steps. First,
you create a login. The login lets the user connect to the SQL Server Database Engine. Then you
 Mary now has access to this instance of SQL Server 2005, but does not have permission to
access the databases. She does not even have access to her default database TestData until
you authorize her as a database user.
 To grant Mary access, switch to the TestData database, and then use the CREATE USER
statement to map her login to a user named Mary.
 To create a user in a database - Type and execute the following statements (replacing
computer_name with the name of your computer) to grant Mary access to the TestData database.

 USE [TestData];
 GO

 CREATE USER [Mary] FOR LOGIN [computer_name\Mary];
 GO
 Now, Mary has access to both SQL Server 2005 and the TestData database
 How to create a view and a stored procedure using "CREATE VIEW/PROCEDURE"
statements?
 This is the third tutorial of a quick lesson on creating login and configure users for databases with
Transact-SQL statements. Granting a user access to a database involves three steps. First, you
create a login. The login lets the user connect to the SQL Server Database Engine. Then you
 Now that Mary can access the TestData database, you may want to create some database
objects, such as a view and a stored procedure, and then grant Mary access to them. A view is a
stored SELECT statement, and a stored procedure is one or more Transact-SQL statements that
execute as a batch.
 Views are queried like tables and do not accept parameters. Stored procedures are more
complex than views. Stored procedures can have both input and output parameters and can
contain statements to control the flow of the code, such as IF and WHILE statements. It is good
programming practice to use stored procedures for all repetitive actions in the database.
 For this example, you will use CREATE VIEW to create a view that selects only two of the
columns in the Products table. Then, you will use CREATE PROCEDURE to create a stored
procedure that accepts a price parameter and returns only those products that cost less than the
specified parameter value.
 To create a view - Execute the following statement to create a very simple view that executes a
select statement, and returns the names and prices of our products to the user.

 CREATE VIEW vw_Names
 AS
 SELECT ProductName, Price FROM Products;
 GO
 Test the view - Views are treated just like tables. Use a SELECT statement to access a view.

 SELECT * FROM vw_Names;
 GO
 To create a stored procedure - The following statement creates a stored procedure name
pr_Names, accepts an input parameter named @VarPrice of data type money. The stored
procedure prints the statement Products less than concatenated with the input parameter that is
changed from the money data type into a varchar(10) character data type. Then, the procedure
executes a SELECT statement on the view, passing the input parameter as part of the WHERE
clause. This returns all products that cost less than the input parameter value.

 CREATE PROCEDURE pr_Names @VarPrice money
 AS
 BEGIN
 -- The print statement returns text to the user
 PRINT 'Products less than ' + CAST(@VarPrice AS varchar(10));
 -- A second statement starts here
 SELECT ProductName, Price FROM vw_Names
 WHERE Price < @varPrice;
 END
 GO
 Test the stored procedure - To test the stored procedure, type and execute the following
statement. The procedure should return the names of the two products entered into the Products
table in Lesson 1 with a price that is less than 10.00.

 EXECUTE pr_Names 10.00;
 GO
 How to grant a permission using "GRANT EXECUTE" statements?
 This is the fourth tutorial of a quick lesson on creating login and configure users for databases
with Transact-SQL statements. Granting a user access to a database involves three steps. First,
you create a login. The login lets the user connect to the SQL Server Database Engine. Then you
 As an administrator, you can execute the SELECT from the Products table and the vw_Names
view, and execute the pr_Names procedure; however, Mary cannot. To grant Mary the necessary
permissions, use the GRANT statement.
 Procedure Title - Execute the following statement to give Mary the EXECUTE permission for the
pr_Names stored procedure.

 GRANT EXECUTE ON pr_Names TO Mary;
 GO
 In this scenario, Mary can only access the Products table by using the stored procedure. If you
want Mary to be able to execute a SELECT statement against the view, then you must also
execute GRANT SELECT ON vw_Names TO Mary. To remove access to database objects, use
the REVOKE statement.
 Note: If the table, the view, and the stored procedure are not owned by the same schema,
granting permissions becomes more complex. For information about how to configure
permissions on objects with different owners, see Ownership Chains.
 About GRANT You must have EXECUTE permission to execute a stored procedure. You must
have SELECT, INSERT, UPDATE, and DELETE permissions to access and change data. The
GRANT statement is also used for other permissions, such as permission to create tables.
 How to delete database objects with "DROP" statements?
 To remove all database objects created by previous tutorials, you could just delete the database.
However, in this tutorial, you will go through the steps to reverse every action you took doing the
tutorial.
 Removing permissions and objects - Before you delete objects, make sure you are in the
correct database:
 USE TestData;
 GO
 Use the REVOKE statement to remove execute permission for Mary on the stored procedure:
 REVOKE EXECUTE ON pr_Names FROM Mary;
 GO
 Use the DROP statement to remove permission for Mary to access the TestData database:
 DROP USER Mary;
 GO
 Use the DROP statement to remove permission for Mary to access this instance of SQL Server
2005:
 DROP LOGIN [\Mary];
 GO
 Use the DROP statement to remove the store procedure pr_Names:
 DROP PROC pr_Names;
 GO
 Use the DROP statement to remove the view vw_Names:
 DROP View vw_Names;
 GO
 Use the DELETE statement to remove all rows from the Products table:
 DELETE FROM Products;
 GO
 Use the DROP statement to remove the Products table:
 DROP Table Products;
 GO
 You cannot remove the TestData database while you are in the database; therefore, first switch
context to another database, and then use the DROP statement to remove the TestData
database:
 USE MASTER;
 GO
 DROP DATABASE TestData;
 GO
What is a database?
A database is a logical container that contains a set of related database objects:

 Tables - Storages of structured data.
 Views - Queries to present data from tables.
 Indexes - Sorting indexes to speed up searches.
 Stored Procedures - Predefined SQL program units.
 Users - Identifications used for data access control.
 Other objects.
 What is the simplest way to create a new database?

 The simplest way to create a new database is to use the "CREATE DATABASE" statement with
this syntax:
 CREATE DATABASE database_name
 For example, run this statement:
 CREATE DATABASE FyiCenterData
 GO
 A new database called "FyiCenterData" should be created in the SQL server. Of course,
FyiCenterData should be empty at this moment - no tables. But it should have some other data
objects automatically created by the server.
 How to set the current database?
 Once you are connected to the SQL Server, you should select a database to work with and set it
as the current database using the "USE" statement with this syntax:
 USE database_name
 The following tutorial example shows you how to set "FyiCenterData" as the current database,
and create a table in "FyiCenterData":
 USE FyiCenterData
 GO
 Changed database context to 'FyiCenterData'.

 CREATE TABLE Links (Name NVARCHAR(32))
 GO

 SELECT name, type_desc, create_date FROM sys.tables
 GO
 name type_desc create_date
 Links USER_TABLE 2007-05-19 23:05:43.700
 How to delete a database?
 If you created a database incorrectly, or you have a database that is not needed any more, you
can delete it with the "DROP DATABASE" statement with this syntax:
 DROP DATABASE database_name
 For example, execute this statement:
 DROP DATABASE FyiCenterData
 GO
 The database "FyiCenterData" created in the previous tutorial should be deleted from the SQL
server.
 Warning, if you delete a database, all tables and their data in that database will be deleted.
 Why I am getting this error when dropping a database?
 If you are trying to drop a database that is in use, you will get an error message like this: 'Cannot
drop database "FyiCenterData" because it is currently in use.'
 Before dropping a database, you must stop all client sessions using this database. If your own
client session is using this database, you should set a different database as the current database
as shown in this tutorial example:
 GO

 GO

 GO
 Msg 3702, Level 16, State 4, Server LOCALHOST\SQLEXPRESS
 Cannot drop database "FyiCenterData" because it is
 currently in use.

 USE master
 GO

 GO
 How to get a list all databases on the SQL server?
 If you don't remember database names you have created, you can get a list of all databases on
the server by query the "sys.databases" view as shown in this tutorial example:
 GO

 SELECT name, database_id, create_date FROM sys.databases
 GO
 name database_id create_date
 master 1 2003-04-08 09:13:36.390
 tempdb 2 2007-05-19 13:42:42.200
 model 3 2003-04-08 09:13:36.390
 msdb 4 2005-10-14 01:54:05.240
 FyiCenterData 5 2007-05-19 20:04:39.310
 As you can see, the newly created database is listed at the end of query result
Where is my database stored on the hard disk?

If a database is created with simple CREATE DATABASE statement, the server will create two
database files on the hard disk to store data and configuration information about that data bases:
 database_name.mdf - SQL Server Database Primary Data File
 database_name_log.ldf - SQL Server Database Transaction Log File
To find out the location of database files, you can query the "sys.database_files" view as shown in
this tutorial example:
USE FyiCenterData
GO
SELECT type_desc, physical_name, size

FROM sys.database_files
GO
type_desc physical_name size
ROWS c:\Program Files\Microsoft SQL Server

\MSSQL.1\MSSQL\DATA\FyiCenterData.mdf 152
LOG c:\Program Files\Microsoft SQL Server

\MSSQL.1\MSSQL\DATA\FyiCenterData_log.LDF 63
Go verify these two files with Windows Explorer.
How to create database with physical files specified?
If you don't like the default behavior of the CREATE DATABASE statement, you can specify the
physical database files with a longer statement:
CREATE DATABASE database_name
ON (NAME = logical_data_name,
FILENAME = physical_data_name,
SIZE = x, MAXSIZE = y, FILEGROWTH = z)
LOG ON (NAME = logical_log_name,
FILENAME = physical_log_name,
SIZE = x, MAXSIZE = y, FILEGROWTH = z)
For example, the following statement will create a database with database files located in the C:\temp
directory:
USE master
GO
DROP DATABASE FyiCenterData

GO
CREATE DATABASE FyiCenterData

ON (NAME = FyiCenterData,
FILENAME = 'C:\temp\FyiCenterData.mdf',
SIZE = 10MB, MAXSIZE = 50MB, FILEGROWTH = 5MB)
LOG ON (NAME = FyiCenterLog,
FILENAME = 'C:\temp\FyiCenterLog.ldf',
SIZE = 1MB, MAXSIZE = 5MB, FILEGROWTH = 1MB)
GO
SELECT type_desc, name, physical_name, size

FROM sys.database_files
GO
type_desc name physical_name size
ROWS FyiCenterData C:\temp\FyiCenterData.mdf 1280
LOG FyiCenterLog C:\temp\FyiCenterLog.ldf 128
How to rename databases?
If don't like the name of a database, you can change it by using the "ALTER DATABASE" statement
with the following syntax:
ALTER DATABASE database_name
MODIFY NAME = new_database_name
The tutorial example below shows you how change the database name from "FyiCenterData" to
"FyiCenterComData":
ALTER DATABASE FyiCenterData
MODIFY NAME = FyiCenterComData
GO
The database name 'FyiCenterComData' has been set
Why I am getting this error when renaming a database?
If you are trying to rename a database that is in use, you will get an error message like this: "The
database could not be exclusively locked to perform the operation."
Before renaming a database, you must stop all client sessions using this database. Otherwise, you
will get an error as shown in this tutorial example:
1. Launch one instance of SQL Server Management Studio and run:

USE FyiCenterComData
GO
2. Keep the first instance running and launch another instance of SQL Server Management Studio:
ALTER DATABASE FyiCenterComData
MODIFY NAME = FyiCenterData
GO
Msg 5030, Level 16, State 2, Server LOCALHOST\SQLEXPRESS
The database could not be exclusively locked to perform
the operation.
Obviously, the first instance is blocking the "ALTER DATABASE" statement.

What are database states?
A database is always in one specific state. For example, these states include ONLINE, OFFLINE, or
SUSPECT. To verify the current state of a database, select the state_desc column in the
sys.databases catalog view. The following table defines the database states.
 ONLINE - Database is available for access. The primary filegroup is online, although the undo
phase of recovery may not have been completed.
 OFFLINE - Database is unavailable. A database becomes offline by explicit user action and
remains offline until additional user action is taken. For example, the database may be taken
offline in order to move a file to a new disk. The database is then brought back online after the
move has been completed.
 RESTORING - One or more files of the primary filegroup are being restored, or one or more
secondary files are being restored offline. The database is unavailable.
 RECOVERING - Database is being recovered. The recovering process is a transient state; the
database will automatically become online if the recovery succeeds. If the recovery fails, the
database will become suspect. The database is unavailable.
 RECOVERY PENDING - SQL Server has encountered a resource-related error during recovery.
The database is not damaged, but files may be missing or system resource limitations may be
preventing it from starting. The database is unavailable. Additional action by the user is required
to resolve the error and let the recovery process be completed.
 SUSPECT - At least the primary filegroup is suspect and may be damaged. The database cannot
be recovered during startup of SQL Server. The database is unavailable. Additional action by the
user is required to resolve the problem.
 EMERGENCY - User has changed the database and set the status to EMERGENCY. The
database is in single-user mode and may be repaired or restored. The database is marked
READ_ONLY, logging is disabled, and access is limited to members of the sysadmin fixed server
role. EMERGENCY is primarily used for troubleshooting purposes. For example, a database
marked as suspect can be set to the EMERGENCY state. This could permit the system
administrator read-only access to the database.
 How to set a database state to OFFLINE?

 If you want to move database physical files, you should take the database offline by using the
"ALTER DATABASE" statement with the following syntax:
 ALTER DATABASE database_name SET OFFLINE
 The following tutorial example will bring "FyiCenterComData" offline:
 ALTER DATABASE FyiCenterComData SET OFFLINE
 GO

 SELECT name, state_desc from sys.databases
 GO
 name state_desc
 master ONLINE
 tempdb ONLINE
 model ONLINE
 msdb ONLINE
 FyiCenterComData OFFLINE

 USE FyiCenterComData
 GO
 Msg 942, Level 14, State 4, Line 1
 Database 'FyiCenterComData' cannot be opened because
 it is offline.
 How to move database physical files?
 If you want to move database physical files to a new location, you can use the "ALTER
DATABASE" statements to bring the database offline, and link it to the files at the new location.
The following tutorial gives you a good example:
 ALTER DATABASE FyiCenterComData SET ONLINE
 GO

 GO

 CREATE TABLE Links (Name NVARCHAR(32))
 GO

 ALTER DATABASE FyiCenterComData SET OFFLINE
 GO
 Now it is safe to move the database physical files to a new location:
 1. Run Windows Explorer
 2. Create a new directory: c:\temp\data
 3. Drag and drop c:\temp\FyiCenterData.mdf to c:\temp\data
 3. Drag and drop c:\temp\FyiCenterLog.mdf to c:\temp\data
 Go back to the SQL client program and run:
 ALTER DATABASE FyiCenterComData
 MODIFY FILE (NAME = FyiCenterData,
 FILENAME = 'C:\temp\data\FyiCenterData.mdf')
 GO
 The file "FyiCenterData" has been modified in the system
 catalog. The new path will be used the next time the
 database is started.

 ALTER DATABASE FyiCenterComData
 MODIFY FILE (NAME = FyiCenterLog,
 FILENAME = 'C:\temp\data\FyiCenterLog.ldf')
 GO
 The file "FyiCenterLog" has been modified in the system
 catalog. The new path will be used the next time the
 database is started.

 ALTER DATABASE FyiCenterComData SET ONLINE
 GO

 GO

 SELECT type_desc, name, physical_name, size
 FROM sys.database_files
 GO
 type_desc name physical_name size
 ROWS FyiCenterData C:\temp\data\FyiCenterData.mdf 1280
 LOG FyiCenterLog C:\temp\data\FyiCenterLog.ldf 128

 SELECT name, type_desc FROM sys.tables
 GO
 name type_desc
 Links USER_TABLE
 As you can see the physical files are moved properly, and the "Links" table is still there
How to set database to be READ_ONLY?
Databases in SQL Server have two update options:

 READ_WRITE - Data objects are allowed to be queried and modified. This is the default.
 READ_ONLY - Data objects are allowed to be queried, but not allowed to be modified.
You can use the "ALTER DATABASE" to change database update options as shown in the tutorial
below:
GO
INSERT Links (Name) VALUES ('dba.FYIcenter.com')

GO
(1 rows affected)
ALTER DATABASE FyiCentercomData SET READ_ONLY

GO
INSERT Links (Name) VALUES ('dev.FYIcenter.com')

GO
Msg 3906, Level 16, State 1, Server SQLEXPRESS, Line 1
Failed to update database "FyiCenterComData" because
the database is read-only.
SELECT * FROM Links

GO
Name
dba.FYIcenter.com
ALTER DATABASE FyiCentercomData SET READ_WRITE

GO
INSERT Links (Name) VALUES ('dev.FYIcenter.com')

GO
(1 rows affected)
As you can see from the output, inserting data into a table is not allowed if the database is in
READ_ONLY mode.
How to set database to be SINGLE_USER?
Databases in SQL Server have three user access options:

 MULTI_USER - All users that have the appropriate permissions to connect to the database are
allowed. This is the default.
 SINGLE_USER - One user at a time is allowed to connect to the database. All other user
connections are broken.
 RESTRICTED_USER - Only members of the db_owner fixed database role and dbcreator and
sysadmin fixed server roles are allowed to connect to the database, but it does not limit their
number.
You can use the "ALTER DATABASE" to change database user access options as shown in the
tutorial below:
GO
ALTER DATABASE FyiCentercomData SET SINGLE_USER

GO
Now connect to server with another client session and try:

GO
Msg 924, Level 14, State 1, Line 1
Database 'FyiCenterComData' is already open and can only
have one user at a time.
Go back to the first session and re-set the database to MULTI_USER:

ALTER DATABASE FyiCenterComData SET MULTI_USER
GO
What are system databases?
System databases are created by the SQL Server itself during the installation process. System
databases are used by the SQL server to help manage other user databases and client execution
sessions. SQL Server 2005 Express Edition uses 4 system databases:
 master - The brain of a SQL server - Stores server configuration, runtime information, and
database metadata.
 model - An empty database model - Used to clone new databases.
 msdb - The background job scheduler - Used for background jobs and related tasks.
 tempdb - The temporary database - Used by the server as a scratch pad.
What is a table?
A table in database is a data object used to store data. Tables have the following features:
 Data is stored in a table with a structure of rows and columns.
 Columns must be pre-defined with names, types and constrains.
 A table object may have other associated data objects like, constrains, triggers, indexes, and
statistics.
For example, a table called Address may have columns defined to store different elements of an
address like, street number, city, country, postal code, etc.
What are DDL (Data Definition Language) statements for tables?
DDL (Data Definition Language) statements are statements to create and manage data objects in the
database. The are three primary DDL statements to create and manage tables:
 CREATE TABLE - Creating a new table.
 ALTER TABLE - Altering the definition of an existing table.
 DROP TABLE - Dropping an existing table.
 How to create new tables with "CREATE TABLE" statements?

 If you want to create a new table, you can use the "CREATE TABLE" statement. The following
tutorial script shows you how to create a table called "tip":
 CREATE TABLE tip (id INTEGER PRIMARY KEY,
 subject VARCHAR(80) NOT NULL,
 description VARCHAR(256) NOT NULL,
 create_date DATETIME NULL)
 GO
 This scripts creates a testing table called "tip" with 4 columns in the current database.
 How To Get a List of All Tables with "sys.tables" View?
 If you want to see the table you have just created, you can use the "sys.tables" system view to
get a list of all tables in the current database. The tutorial script gives you a good example:
 GO
 tip USER_TABLE 2007-05-19 23:05:43.700
 The output shows that there is only one table in the current database.
 How To Get a List of Columns using the "sys.columns" View?
 If you have an existing table, but you don't remember what are the columns defined in the table,
you can use the "sys.columns" system view to get a list of all columns of all tables in the current
database.
 In order to a list of columns of a single table, you need to join sys.columns and sys.tables as
shown in the tutorial example below:
 SELECT * FROM sys.columns c, sys.tables t
 WHERE c.object_id = t.object_id
 AND t.name = 'tip'
 GO
 object_id name column_id user_type_id max_length
 2073058421 id 1 56 4
 2073058421 subject 2 167 80
 2073058421 description 3 167 256
 2073058421 create_date 4 61 8
 You can see the column names easily from the sys.columns view. But you can only see the
column type IDs. This requires another join to get the column type names. You may try the
"sp_columns" stored procedure to get a better list of columns shown in the next tutorial.
 How To Get a List of Columns using the "sp_columns" Stored Procedure?
 If you have an existing table, but you don't remember what are the columns defined in the table,
you can use the "sp_columns" stored procedure to get a list of all columns of the specified table.
The following tutorial script shows you a good example:
 sp_columns tip
 GO
 TABLE_OWNER TABLE_NAME COLUMN_TABLE TYPE_NAME ...
 dbo tip id int ...
 dbo tip subject varchar ...
 dbo tip description varchar ...
 dbo tip create_date datetime ...
 The "sp_columns" stored procedure returns a long list of properties for each column of the
specified table. Take a look at each of them.
 How To Get a List of Columns using the "sp_help" Stored Procedure?
 Another way to get a list of columns from a table is to use the "sp_help" stored procedure.
"sp_help" returns more than just a list of columns. It returns: the table information, the column
information, the identity column, the row GUID column, the primary key, indexes, and constraints.
It you run "sp_help tip" in SQL Server Management Studio, you will see the result as shown in
this picture:
 How To Generate CREATE TABLE Script on an Existing Table?

 If you want to know how an existing table was created, you can use SQL Server Management
Studio to automatically generate a "CREATE TABLE" script The following tutorial shows you how
to do this:
 1. Run SQL Server Management Studio and connect to SQL server.
 2. On the Object Explorer window, follow the object tree: Databases > FyiCenterData > Tables >
dbo.tip.
 3. Click right mouse button on dbo.tip. The context menu shows up.
 4. Select "Script Table as" > "CREATE to" > "New Query Editor Window". The following script will
be displayed:
 USE [FyiCenterData]
 GO
 /****** Object: Table [dbo].[tip]
 Script Date: 05/19/2007 21:34:19 ******/
 SET ANSI_NULLS ON
 GO
 SET QUOTED_IDENTIFIER ON
 GO
 SET ANSI_PADDING ON
 GO
 CREATE TABLE [dbo].[tip](
 [id] [int] NOT NULL,
 [subject] [varchar](80)
 COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
 [description] [varchar](256)
 COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
 [create_date] [datetime] NULL,
 PRIMARY KEY CLUSTERED (
 [id] ASC
 )WITH (PAD_INDEX = OFF,
 IGNORE_DUP_KEY = OFF) ON [PRIMARY]
 ) ON [PRIMARY]
 GO
 SET ANSI_PADDING OFF
 How to create new tables with "SELECT ... INTO" statements?
 Let's say you have a table with many data rows, now you want to create a backup copy of this
table of all rows or a subset of them, you can use the "SELECT ... INTO" statement. The tutorial
script below gives you a good example:
 INSERT INTO tip VALUES (1, 'Learn SQL',
 'Visit dev.fyicenter.com','2006-07-01')
 GO

 SELECT * INTO tipBackup FROM tip
 GO
 (1 rows affected)

 SELECT * FROM tipBackup
 GO
 id subject description create_date
 1 Learn SQL Visit dev.fyicenter.com 2006-07-01

 sp_columns tipBackup
 GO
 dbo tipBackup id int ...
 dbo tipBackup subject varchar ...
 dbo tipBackup description varchar ...
 dbo tipBackup create_date datetime ...
 As you can see, the "SELECT ... INTO" statement created a table called "tipBackup" using the
same column definitions as the "tip" table and copied all data rows into "tipBackup".
 How To Add a New Column to an Existing Table with "ALTER TABLE ... ADD"?
 If you have an existing table with existing data rows, and want to add a new column to that table,
you can use the "ALTER TABLE ... ADD" statement. The tutorial script below shows you a good
example:
 ALTER TABLE tip ADD author VARCHAR(40)
 GO

 sp_columns tip
 GO
 dbo tip id int ...
 dbo tip subject varchar ...
 dbo tip description varchar ...
 dbo tip create_date datetime ...
 dbo tip author datetime ...

 SELECT * FROM tip
 GO
 id subject description create_date author
 1 Learn SQL Visit dev.fyicenter.com 2006-07-01 NULL
 This SQL script added a new column called "author" to the "tip" table. NULL values were added
to this column on all existing data rows.
 How To Delete an Existing Column in a Table with "ALTER TABLE ... DROP COLUMN"?
 If you have an existing column in a table and you do not need that column any more, you can
delete it with "ALTER TABLE ... DROP COLUMN" statement. Here is a tutorial script to delete an
existing column:
 ALTER TABLE tip DROP COLUMN create_date
 GO

 SELECT * FROM tip
 GO
 id subject description author
 1 Learn SQL Visit dev.fyicenter.com NULL
 As you can see, the column "create_date" is gone.
 How to rename an existing column with the "sp_rename" stored procedure?
 If you have an existing column in a table and you want to change the column name, you can use
the "sp_rename ... 'COLUMN'" stored procedure. "sp_rename" allows you to change names of
COLUMN, DATABASE, INDEX, USERDATATYPE, and OBJECT. The tutorial example below
shows you how to rename a column:
 USE master
 GO

 sp_rename 'tip.subject', 'title', 'COLUMN'
 GO
 Msg 15248, Level 11, State 1, Procedure sp_rename, Line 213
 Either the parameter @objname is ambiguous or the claimed
 @objtype (COLUMN) is wrong.

 GO

 sp_rename 'tip.subject', 'title', 'COLUMN'
 GO
 Caution: Changing any part of an object name could break
 scripts and stored procedures.

 SELECT id, title, description, author FROM tip
 GO
 id title description author
 1 Learn SQL Visit dev.fyicenter.com NULL
 You are getting the first error because 'FyiCenterData' is not the current database.
 How to rename an existing column with SQL Server Management Studio?
 If you are using SQL Server Management Studio, you can rename almost any data objects
through the Object Explorer window. The tutorial example below shows you how to rename a
column:
 2. On the Object Explorer window, follow the object tree: Databases > FyiCenterData > Tables >
dbo.tip > Columns > title.
 3. Click right mouse button on "title". The context menu shows up.
 4. Select "Rename", type "subject" over "title", and press Enter key. The column name will be
changed.
 How to change the data type of an existing column with "ALTER TABLE" statements?
 Sometimes, you may need to change the data type of an existing column. For example, you want
increase the string length of a column. You can use the "ALTER TABLE ... ALTER COLUMN"
statements in the following syntax:
 ALTER TABLE table_name ALTER COLUMN column_name new_type
 Here is a good example of change column data types:
 -- Can not make a string column shorter
 ALTER TABLE tip ALTER COLUMN subject VARCHAR(10)
 GO
 String or binary data would be truncated.
 The statement has been terminated.

 -- Can make a string column longer
 ALTER TABLE tip ALTER COLUMN subject VARCHAR(100)
 GO
 Command(s) completed successfully.

 -- Can not change string to numeric
 ALTER TABLE tip ALTER COLUMN subject NUMBER
 GO
 Error converting data type varchar to numeric.
 The statement has been terminated.
 As you can see, the new date type must be compatible with the old data type in order for the
"ALTER TABLE ... ALTER COLUMN" statement to work
 How to rename an existing table with the "sp_rename" stored procedure?
 If you have an existing table and you want to change the table name, you can use the
"sp_rename ... 'OBJECT'" stored procedure. "sp_rename" allows you to change names of
COLUMN, DATABASE, INDEX, USERDATATYPE, and OBJECT (including tables). The tutorial
example below shows you how to rename a table:
 sp_rename 'tip', 'faq', 'OBJECT'
 GO
 Caution: Changing any part of an object name could break
 scripts and stored procedures.

 GO
 faq USER_TABLE 2007-05-19 23:05:43.700
 tipBackup USER_TABLE 2007-05-19 23:25:23.357
 You can also rename a table with on the Object Explorer window of SQL Server Management
Studio. See tutorials on rename table columns.
 How To Drop an Existing Table with "DROP TABLE" Statements?
 If you want to delete an existing table and its data rows, you can use the "DROP TABLE"
statement as shown in the tutorial script below:
 GO
 id subject description create_date
 1 Learn SQL Visit dev.fyicenter.com 2006-07-01

 DROP TABLE tipBackup
 GO

 GO
 Invalid object name 'tipBackup'.
 Be careful, when you use the "DROP TABLE" statement. All data rows are gone too.
What Are DML (Data Manipulation Language) Statements?
DML (Data Manipulation Language) statements are statements to change data values in database
tables. The are 3 primary DML statements:
 INSERT - Inserting new rows into database tables. For example "INSERT INTO fyi_links
VALUES (101, 'dev.fyicenter.com', NULL, 0, '2006-04-30')" inserts a single new row in the
fyi_links table.
 UPDATE - Updating existing rows in database tables .
 DELETE - Deleting existing rows from database tables.
(Continued on next topic...)
How To Create a Testing Table with Test Data?
If you want to practice DML statements, like INSERT, UPDATE and DELETE statements, you should
create a testing table. The tutorial exercise shows you a good example:
CREATE TABLE fyi_links (id INTEGER PRIMARY KEY,
url VARCHAR(80) NOT NULL,
notes VARCHAR(1024),
counts INT,
created DATETIME NOT NULL DEFAULT(getdate()))
GO
SELECT c.column_id as seq, c.name, x.name as type,

c.max_length, c.is_nullable
FROM sys.columns c, sys.tables t, sys.systypes x
WHERE c.object_id = t.object_id
AND c.system_type_id = x.xtype
AND t.name = 'fyi_links'
ORDER BY c.column_id
GO
seq name type max_length is_nullable
1 id int 4 0
2 url varchar 80 0
3 notes varchar 1024 1
4 counts int 4 1
5 created datetime 8 0
You should keep this table to practice other tutorial exercises presented in this collection.
How To Insert a New Row into a Table with "INSERT INTO" Statements?
To insert a new row into a table, you can use the INSERT INTO statement with values specified for all
columns as in the following syntax:
INSERT INTO table_name VALUES (list_of_values_of_all columns)
Note that the list of values of all columns must be specified in the same order as how columns are
defined in the CREATE TABLE statement. The following tutorial example inserts a row into "fyi_links":
INSERT INTO fyi_links VALUES (101,
'dev.fyicenter.com',
NULL,
0,
'2006-04-30')
GO
(1 row(s) affected)
SELECT * FROM fyi_links

GO
id url notes counts created
101 dev.fyicenter.com NULL 0 2006-04-30
The values are stored in the new record nicely.
How To Use Column Default Values in INSERT Statements?
If a column is defined with a default value in a table, you can use the key word DEFAULT in the
INSERT statement to take the default value for that column. The following tutorial exercise gives a
good example:
'dba.fyicenter.com',
NULL,
0,
DEFAULT)
GO
(1 row(s) affected)

GO
102 dba.fyicenter.com NULL 0 2007-05-19
The default value, getdate(), is used for "created" column, which gives the current date
How to provide column names in INSERT Statements?
If you don't want to specify values for columns that have default values, or you want to specify values
to columns in an order different than how they are defined, you can provide a column list in the
INSERT statement. If a column is omitted in the column, SQL Server applies 3 rules:
 If default value is defined for the column, that default value will be used.
 If no default value is defined for the column and NULL is allowed, NULL will be used.
 If no default value is defined for the column and NULL is not allowed, SQL Server will reject the
insert statement with an error.
The following tutorial exercise gives you some good examples:

INSERT INTO fyi_links (url, id)
VALUES ('sqa.fyicenter.com', 103)
GO
(1 row(s) affected)
INSERT INTO fyi_links (id) VALUES (110)

GO
Cannot insert the value NULL into column 'url',
table 'FyiCenterData.dbo.fyi_links'; column does
not allow nulls. INSERT fails.
The statement has been terminated.

GO
103 dba.fyicenter.com NULL NULL 2007-05-19
The first INSERT statement shows that: the order of the columns is reversed; the default value is
taken for the un-specified column "created"; the NULL value is taken for the un-specified column
"counts", since is has no default value defined and null is allowed.
The second INSERT statement shows the error you get for the un-specified column "url", because it
has no default value, and null is not allowed.
What Happens If You Insert a Duplicate Key for the Primary Key Column?
If your table has a primary key column, and you are trying to insert a new row with duplicate key value
on the primary key column, you will get an error. The reason is simple - Primary key column does not
allow duplicate values. The following tutorial exercise gives you a good example:
'sql.fyicenter.com',
NULL,
0,
'2006-04-30')
GO
Violation of PRIMARY KEY constraint
'PK__fyi_links__03317E3D'. Cannot insert duplicate
key in object 'dbo.fyi_links'.
You are getting this error, because value "101" has already been used by an existing row.
How To Insert Multiple Rows with One INSERT Statement?
If you want to insert multiple rows with a single INSERT statement, you can use a subquery instead of
the VALUES clause. Rows returned from the subquery will be inserted the target table. The following
tutorial exercise gives you a good example:
INSERT INTO fyi_links SELECT id+500, REVERSE(url),
notes, counts, created FROM fyi_links
GO
(3 row(s) affected)

GO
103 sqa.fyicenter.com NULL NULL 2007-05-19
601 moc.retneciyf.ved NULL 0 2006-04-30
602 moc.retneciyf.abd NULL 0 2007-05-19
603 moc.retneciyf.aqs NULL NULL 2007-05-19
As you can see, "INSERT INTO ... SELECT ..." is powerful statement. you can use it build up data in
tables quickly.
How To Update Values in a Table with UPDATE Statements?
If you want to update some values in one row or multiple rows in a table, you can use the UPDATE
statement. The tutorial script below shows a good example:
SELECT * FROM fyi_links WHERE id = 101
GO
UPDATE fyi_links SET counts = 999, notes = 'Good.'

WHERE id = 101;
GO
(1 row(s) affected)

GO
101 dev.fyicenter.com Good. 999 2006-04-30
As you can see, the SET clause takes column and value pairs to provide new values, while the
WHERE clause defines which row to apply the update
How To Update Multiple Rows with One UPDATE Statement?
If the WHERE clause in an UPDATE statement matches multiple rows, the SET clause will be applied
to all matched rows. This rule allows you to update values on multiple rows in a single UPDATE
statement. Here is a good example:
SELECT * FROM fyi_links WHERE id >= 500
GO
601 moc.retneciyf.ved NULL 0 2006-04-30
603 moc.retneciyf.aqs NULL NULL 2007-05-19

UPDATE fyi_links SET counts = 9, notes = 'Wrong'
WHERE id >= 500
(3 row(s) affected)

GO
601 moc.retneciyf.ved Wrong 9 2006-04-30
602 moc.retneciyf.abd Wrong 9 2007-05-19
603 moc.retneciyf.aqs Wrong 9 2007-05-19
The UPDATE statement updated 3 rows with the same new values.
How to use old values to define new values in UPDATE statements?

If a row matches the WHERE clause in a UPDATE statement, existing values in this row can be used
in expressions to provide new values in the SET clause. Existing values are represented by column
names in the expressions. The tutorial exercise below shows you a good example:
GO
UPDATE fyi_links SET id = id+200, counts = id*2

WHERE id >= 500
GO
(3 row(s) affected)

GO
This statement increased values in the id column by 200. It also updated the counts column with the
newly increased id value.
Is the Order of Columns in the SET Clause Important?
The answer is NO. The order of columns in the SET clause of the UPDATE statement is NOT
important. You probably already noticed from the previous tutorial. There is a BIG DIFFERENCE
among SQL Server, MySQL and Oracle on update multiple columns with previous values:
 SQL Server provides you the existing values from the database on columns names used in new
value expressions. So the order of columns in the SET clause is NOT important
 MySQL provides you the updated values on columns names used in new value expressions. So
the order of columns in the SET clause is important.
 Oracle provides you the existing values from the database on columns names used in new value
expressions. So the order of columns in the SET clause is NOT important
Here is a good tutorial exercise:

-- Check the old values
SELECT * FROM fyi_links WHERE url = 'dev.fyicenter.com'
GO
-- Update "id" before "counts"

UPDATE fyi_links SET id = id+200, counts = id*2
WHERE url = 'dev.fyicenter.com'
GO
(1 row(s) affected)
-- Check the new values

SELECT * FROM fyi_links WHERE url = 'dev.fyicenter.com'
GO
-- Reset to old values

UPDATE fyi_links SET id = 101, counts = 999
WHERE url = 'dev.fyicenter.com'
(1 row(s) affected)
Notice that the "id" in the "counts" new value expression is taking the old value of the "id" column, not
the updated value, even the "id" column is updated before the "counts" column.
Now try this on a MySQL server, you will get different result.
How To Use Values from Other Tables in UPDATE Statements?
If you want to update values in one table with values from another table, you can use a subquery as
an expression in the SET clause. The subquery should return only one row for each row in the update
table that matches the WHERE clause. The tutorial exercise below shows you a good example:
-- Create another table
CREATE TABLE fyi_rates (id INTEGER,
comment VARCHAR(16))
Go
-- Insert some rows in the new table

INSERT INTO fyi_rates VALUES (101, 'The best')
Go
INSERT INTO fyi_rates VALUES (102, 'Well done')
GO
INSERT INTO fyi_rates VALUES (103, 'Thumbs up')
Go
-- Update fyi_links with values from fyi_rates

UPDATE fyi_links SET notes = (
SELECT comment FROM fyi_rates
WHERE fyi_rates.id = fyi_links.id
)
WHERE id > 0 AND id < 110
GO
(3 row(s) affected)
-- View the updated values

WHERE id > 0 AND id < 110
GO
101 dev.fyicenter.com The best 999 2006-04-30
102 dba.fyicenter.com Well done 0 2007-05-19
103 sqa.fyicenter.com Thumbs up NULL 2007-05-19
Note that if column names are confusing between the inner table and the outer table, you need to
prefix column names with table names, like "fyi_rates.id = fyi_links.id".
What Happens If the UPDATE Subquery Returns No Rows?
If you use a subquery to assign new values in the SET clause in an UPDATE statement, and the
subquery returns no rows for an outer row, SQL Server will provide a NULL value to the SET clause.
The tutorial exercise below shows you a good example:
-- insert a new row
INSERT INTO fyi_links (id, url, notes)
VALUES (0, 'www.fyicenter.com', 'Number one')
GO
(1 row(s) affected)
-- view old values

GO
0 www.fyicenter.com Number one NULL 2007-05-23
-- make sure there is no matching row in fyi_rates

SELECT * FROM fyi_rates WHERE id = 0
GO
0 rows
-- update a subquery returning no rows

)
WHERE id = 0
(1 row(s) affected)
-- view new values

GO
0 www.fyicenter.com NULL NULL 2007-05-23
Column "notes" gets updated with NULL if there is no return rows in the subquery
What Happens If the UPDATE Subquery Returns Multiple Rows?
If a subquery is used in a UPDATE statement, it must return exactly one row for each row in the
update table that matches the WHERE clause. If it returns multiple rows, SQL Server will give you an
error message. To test this out, you can try the following tutorial exercise:
-- insert two rows to fyi_rates
INSERT INTO fyi_rates VALUES (0, 'Number 1')
GO
INSERT INTO fyi_rates VALUES (0, 'Number 2')
GO
-- make sure there are 2 match rows

SELECT * FROM fyi_rates WHERE id = 0
GO
id comment
0 Number 1
0 Number 2
-- update with subquery that returns 2 rows

)
WHERE id = 0
GO
Subquery returned more than 1 value. This is not permitted
when the subquery follows =, !=, <, <= , >, >= or when
the subquery is used as an expression.
It is clear that we are using subquery as an expression, and it must return 0 or 1 row. Otherwise, we
will get an error.
How To Delete an Existing Row with DELETE Statements?
If you want to delete an existing row from a table, you can use the DELETE statement with a WHERE
clause to identify that row. Here is good sample of DELETE statements:
-- insert a row for this test
VALUES ('www.myspace.com', 301)
GO
(1 row(s) affected)
-- view the inserted row

SELECT id, url, notes, counts FROM fyi_links
WHERE id = 301
GO
id url notes counts
301 www.myspace.com NULL NULL
-- delete one row

DELETE FROM fyi_links WHERE id = 301
GO
(1 row(s) affected)
-- try to view the deleted row

WHERE id = 301
no rows
Row with id of 301 is truly deleted.
How To Delete Multiple Rows with One DELETE Statement?
You can delete multiple rows from a table in the same way as deleting a single row, except that the
WHERE clause will match multiple rows. The tutorial exercise below deletes 3 rows from the fyi_links
table:
-- view rows to be deleted
WHERE id > 300
GO
id url notes counts
801 moc.retneciyf.ved Wrong 1202
802 moc.retneciyf.abd Wrong 1204
803 moc.retneciyf.aqs Wrong 1206
-- delete multiple rows

DELETE FROM fyi_links WHERE id > 300
GO
(3 row(s) affected)
-- try to view the deleted row
WHERE id > 300
GO
no rows
How To Delete All Rows with DELETE Statements?
If you want to delete all rows from a table, you have two options:
 Use the DELETE statement with no WHERE clause.
 Use the TRUNCATE TABLE statement.
Here is an example of deleting all rows with a DELETE statement:

SELECT COUNT(*) FROM fyi_links
GO
4
DELETE FROM fyi_links

GO
(4 row(s) affected)

GO
0
How To Delete All Rows with TRUNCATE TABLE Statement?
If you want to delete all rows from a table, you have two options:
 Use the DELETE statement with no WHERE clause.
 Use the TRUNCATE TABLE statement.
The TRUNCATE statement is more efficient the DELETE statement. The tutorial exercise shows you
a good example of TRUNCATE statement:
SELECT COUNT(*) FROM fyi_rates
GO
5
TRUNCATE TABLE fyi_rates

GO
SELECT COUNT(*) FROM fyi_rates

GO
0
How To Join Two Tables in a Single Query?

Two tables can be joined together in a query in 4 ways:
 Inner Join: Returns only rows from both tables that satisfy the join condition.
 Left Outer Join: Returns rows from both tables that satisfy the join condition, and the rest of rows
from the first (left) table.
 Right Outer Join: Returns rows from both tables that satisfy the join condition, and the rest of
rows from the second (right) table.
 Full Outer Join: Returns rows from both tables that satisfy the join condition, the rest of rows from
the first (left) table, and the rest of rows from the second (right) table.
 How To Write a Query with an Inner Join?

 If you want to query from two tables with an inner join, you can use the INNER JOIN ... ON clause
in the FROM clause. The tutorial exercise below creates another testing table and returns output
with an inner join from two tables: fyi_links and fyi.rates. The join condition is that the id in the
fyi_links table equals to the id in the fyi_rates table:
 CREATE TABLE fyi_rates (id INTEGER,
 comment VARCHAR(16))
 GO

 INSERT INTO fyi_rates VALUES (101, 'The best')
 GO
 INSERT INTO fyi_rates VALUES (102, 'Well done')
 GO
 INSERT INTO fyi_rates VALUES (103, 'Thumbs up')
 GO
 INSERT INTO fyi_rates VALUES (204, 'Number 1')
 GO
 INSERT INTO fyi_rates VALUES (205, 'Not bad')
 GO
 INSERT INTO fyi_rates VALUES (206, 'Good job')
 GO
 INSERT INTO fyi_rates VALUES (207, 'Nice tool')
 GO

 SELECT fyi_links.id, fyi_links.url,
 fyi_rates.comment FROM fyi_links
 INNER JOIN fyi_rates ON fyi_links.id = fyi_rates.id
 GO
 id url comment
 101 dev.fyicenter.com The best
 102 dba.fyicenter.com Well done
 103 sqa.fyicenter.com Thumbs up
 How To Define and Use Table Alias Names?
 When column names need to be prefixed with table names, you can define table alias name and
use them to prefix column names. To define an alias for a table name, just enter the alias name
right after the original table name in the FROM clause as shown in the following select statement:
 SELECT l.id, l.url, r.comment FROM fyi_links l
 INNER JOIN fyi_rates r ON l.id = r.id
 GO
 id url comment
 How To Write a Query with a Left Outer Join?
 If you want to query from two tables with a left outer join, you can use the LEFT OUTER JOIN ...
ON clause in the FROM clause. The following query returns output with a left outer join from two
tables: fyi_links and fyi_rates. The join condition is that the id in the fyi_links table equals to the id
in the fyi_rates table:
 LEFT OUTER JOIN fyi_rates r ON l.id = r.id
 GO
 id url comment
 104 www.mysql.com NULL
 105 www.oracle.com NULL
 106 www.php.net NULL
 107 www.winrunner.com NULL
 Note that a left outer join may return extra rows from the first (left) table that do not satisfy the join
condition. In those extra rows, columns from the second (right) table will be given null values.
 The extra rows returned from the left outer join in this example represents links that have no rates
in the above example.
 How To Write a Query with a Right Outer Join?
 If you want to query from two tables with a right outer join, you can use the RIGHT OUTER
JOIN ... ON clause in the FROM clause. The following query returns output with a right outer join
from two tables: fyi_links and fyi_rates. The join condition is that the id in the fyi_links table equals
to the id in the fyi_rates table:
 RIGHT OUTER JOIN fyi_rates r ON l.id = r.id
 GO
 id url comment
 NULL NULL Number 1
 NULL NULL Not bad
 NULL NULL Good job
 NULL NULL Nice tool
 Note that a right outer join may return extra rows from the second (right) table that do not satisfy
the join condition. In those extra rows, columns from the first (left) table will be given null values.
 The extra rows returned from the right outer join in this example represents rates that have no
links in the above example.
How To Write a Query with a Full Outer Join?
If you want to query from two tables with a full outer join, you can use the FULL OUTER JOIN ... ON
clause in the FROM clause. The following query returns output with a full outer join from two tables:
fyi_links and fyi_rates. The join condition is that the id in the fyi_links table equals to the id in the
fyi_rates table:
SELECT l.id, l.url, r.comment FROM fyi_links l
FULL OUTER JOIN fyi_rates r ON l.id = r.id
GO
id url comment
101 dev.fyicenter.com The best
102 dba.fyicenter.com Well done
103 sqa.fyicenter.com Thumbs up
104 www.mysql.com NULL
105 www.oracle.com NULL
106 www.php.net NULL
107 www.winrunner.com NULL
NULL NULL Number 1
NULL NULL Not bad
NULL NULL Good job
NULL NULL Nice tool
As you can see, an full outer join returns 3 groups of rows:

 The rows from both tables that satisfy the join condition.
 The rows from the first (left) table that do not satisfy the join condition.
 The rows from the second (right) table that do not satisfy the join condition
 How To Write an Inner Join with the WHERE Clause?

 If you don't want to use the INNER JOIN ... ON clause to write an inner join, you can put the join
condition in the WHERE clause as shown in the following query example:
 SELECT l.id, l.url, r.comment
 FROM fyi_links l, fyi_rates r WHERE l.id = r.id
 GO
 id url comment
 How To Name Query Output Columns?
 Each column in the query output has a default name. If you don't like the default name, you can
specify a new name for any column in the query output by using the AS clause. The following
statement shows you a good example:
 SELECT tag AS Category, YEAR(created) AS Year,
 COUNT(*) AS Counts FROM fyi_links
 GROUP BY tag, YEAR(created) ORDER BY COUNT(*) DESC
 GO
 Category Year Counts
 SQA 2003 1
 DEV 2004 1
 DBA 2005 1
 DBA 2006 1
 DEV 2006 1
 DBA 2007 1
 SQA 2007 1
What Is a Subquery in a SELECT Query Statement?
A subquery is a SELECT statement used as part of the selection criteria of the main SELECT
statement. The subquery specified in the WHERE clause will be evaluated repeated on each row of
the selection base table. The output of the subquery will be used in the final evaluation of the criteria.
Usually, subqueries are used in the following Boolean operations:
 "expression IN (subquery)" - True if the expression matches one of the returned values from the
subquery.
 "expression NOT IN (subquery)" - True if the expression does not match any of the returned
values from the subquery.
 "EXISTS (subquery)" - True if the subquery returns one or more rows.
 "NOT EXISTS (subquery)" - True if the subquery returns no rows.
 How To Use Subqueries with the IN Operators?

 A subquery can be used with the IN operator as "expression IN (subquery)". The subquery should
return a single column with one or more rows to form a list of values to be used by the IN
operation. The following tutorial exercise shows you how to use a subquery with the IN operator.
It returns all links with ids in the fyi_rates table.
 SELECT id, url, tag, YEAR(created) As year
 FROM fyi_links WHERE id IN (SELECT id FROM fyi_rates)
 GO
 id url tag Year
 101 dev.fyicenter.com DEV 2006
 102 dba.fyicenter.com DBA 2007
 103 sqa.fyicenter.com SQA 2007

 FROM fyi_links
 WHERE id IN (101, 102, 103, 204, 205, 206, 207)
 GO
 id url tag Year
 As you can see, the subquery is equivalent to a list of values
 How To Use Subqueries with the EXISTS Operators?
 A subquery can be used with the EXISTS operator as "EXISTS (subquery)", which returns true if
the subquery returns one or more rows. The following statement is a good example of "EXISTS
(subquery)". It returns rows from fyi_links table that there are rows existing in the fyi_rates table
with the same id.
 FROM fyi_links WHERE EXISTS (
 SELECT * FROM fyi_rates
 WHERE fyi_rates.id = fyi_links.id)
 GO
 id url tag Year
 Note that the subquery uses columns from the source table of the outer query.
 How To Use Subqueries in the FROM Clause?
 If you have a query returning many rows of data, and you want to perform another query on those
rows, you can put the first query as a subquery in the FROM clause of the second query. A
subquery used in this way become a temporary table, and you must provide a table alias name
for the subquery as in "SELECT ... FROM (SELECT ...) aliasName". The following statement
shows you how to use a subquery as base table for the main query:
 SELECT * FROM (SELECT l.id, l.url, r.comment
 FROM fyi_links l LEFT OUTER JOIN fyi_rates r
 ON l.id = r.id) WHERE url LIKE '%er%'
 GO
 Incorrect syntax near the keyword 'WHERE'.

 SELECT * FROM (SELECT l.id, l.url, r.comment
 FROM fyi_links l LEFT OUTER JOIN fyi_rates r
 ON l.id = r.id) s WHERE s.url LIKE '%er%'
 GO
 107 www.winrunner.com NULL
 The error on the first query is caused by the missing alias name to name output of the subquery
as a temporary table.
 How To Count Groups Returned with the GROUP BY Clause?
 If you use the COUNT(*) function on groups returned with the GROUP BY clause, it will count the
number of rows within each group, not the number of groups. If you want to count the number of
groups, you can put the GROUP BY query into a subquery and apply the COUNT(*) function on
the main query as shown in the following tutorial exercise:
 COUNT(*) AS Counts FROM fyi_links GROUP BY tag,
 YEAR(created)
 GO
 Category Year Counts
 SQA 2003 1
 DEV 2004 1
 DBA 2005 1
 DBA 2006 1
 DEV 2006 1
 DBA 2007 1
 SQA 2007 1

 SELECT COUNT(*) FROM (
 COUNT(*) AS Counts FROM fyi_links GROUP BY tag,
 YEAR(created) ) groups
 GO
 7
 How To Return the Top 5 Rows from a SELECT Query?
 If you want the query to return only the first 5 rows, you can use the "TOP 5" clause. The TOP
clause takes one parameter to indicate how many top rows to return. The following statements
returns the first 5 rows and 3 rows from the fyi_links:
 SELECT TOP 5 id, url, counts, tag FROM fyi_links
 ORDER BY counts DESC
 GO
 id url counts tag
 102 dba.fyicenter.com 972 DBA
 105 www.oracle.com 960 DBA
 107 www.winrunner.com 828 SQA
 103 sqa.fyicenter.com 728 SQA
 106 www.php.net 439 DEV

 SELECT TOP 3 id, url, counts, tag FROM fyi_links
 ORDER BY counts DESC
 GO
 id url counts tag
 102 dba.fyicenter.com 972 DBA
 105 www.oracle.com 960 DBA
 107 www.winrunner.com 828 SQA
 How To Return the Second 5 Rows?
 If you want to display query output in multiple pages with 5 rows per page, and the visitor wants
to see the output for the second page, you need to display query output from row 6 to row 10. If
you are using MySQL server, you can use the "LIMIT startRow maxRows".
 But the LIMIT clause is not supported by the SQL server. And there seems to be no easy
workaround. You may consider to return the top 10 rows, skip the first 5 rows, then keep the
second 5 rows.
 How To Use UNION to Merge Outputs from Two Queries Together?
 If you have two queries that returns the same row fields, you can merge their outputs together
with the UNION operator. The following tutorial exercise shows you how to use the UNION
operator:
 SELECT * FROM fyi_links WHERE tag = 'DBA'
 GO
 id url notes counts created tag
 102 dba.fyicenter.com NULL 972 2007-05-19 DBA
 104 www.mysql.com 390 2006-01-01 DBA
 105 www.oracle.com 960 2005-01-01 DBA

 SELECT * FROM fyi_links WHERE tag = 'DEV'
 GO
 101 dev.fyicenter.com NULL 120 2006-04-30 DEV
 106 www.php.net 439 2004-01-01 DEV

 SELECT * FROM fyi_links WHERE tag = 'DBA'
 UNION
 SELECT * FROM fyi_links WHERE tag = 'DEV'
 GO
 102 dba.fyicenter.com NULL 972 2007-05-19 DBA
 104 www.mysql.com 390 2006-01-01 DBA
 105 www.oracle.com 960 2005-01-01 DBA
 101 dev.fyicenter.com NULL 120 2006-04-30 DEV
 106 www.php.net 439 2004-01-01 DEV
How To Use ORDER BY with UNION Operators
If you need to sort the output from two queries grouped together with a UNION operator, you need to
apply the ORDER BY clause at the group level, not at the subquery level.
Note that SQL Server and MySQL react differently to the ORDER BY clause used in a subquery of a
UNION operator:
 SQL Server will give error if ORDER BY is used inside a subquery of a UNION operator.
 MySQL will ignore the ORDER BY clause inside a subquery of a UNION operator.
The following tutorial exercise shows you how to use ORDER BY clause with UNION operator:
(SELECT * FROM fyi_links WHERE tag = 'DBA'
ORDER BY created)
UNION
(SELECT * FROM fyi_links WHERE tag = 'DEV'
ORDER BY created)
GO
Incorrect syntax near the keyword 'ORDER'.
Incorrect syntax near the keyword 'ORDER'.
(SELECT * FROM fyi_links WHERE tag = 'DBA')

UNION
(SELECT * FROM fyi_links WHERE tag = 'DEV')
ORDER BY created
GO
id url notes counts created tag
106 www.php.net 439 2004-01-01 DEV
105 www.oracle.com 960 2005-01-01 DBA
104 www.mysql.com 390 2006-01-01 DBA
101 dev.fyicenter.com NULL 120 2006-04-30 DEV
102 dba.fyicenter.com NULL 972 2007-05-19 DBA
Note that the ORDER BY works nicely. Rows returned from both subqueries are sorted together.
What Is a SELECT Query Statement?
The SELECT statement is also called the query statement. It is the most frequently used SQL
statement in any database application. SELECT statements allows you to retrieve data from one or
more tables or views, with different selection criteria, grouping criteria and sorting orders.
A SELETE statement has the following basic syntax:

SELECT select_list
FROM table_source
WHERE search_condition
GROUP BY group_by_expression
HAVING search_condition
ORDER BY order_by_expression
Here is an example of a SELECT statement with all clauses mentioned above:

SELECT SalesOrderID, SUM(LineTotal) AS TotalPrice
FROM SalesLT.SalesOrderDetail
WHERE ModifiedDate > '2004-05-01'
GROUP BY SalesOrderID
HAVING COUNT(*) > 30
ORDER BY TotalPrice DESC
How To Create a Testing Table with Test Data?
If you want to follow the tutorial exercises presented in this collection, you need to create a testing
table and insert some test data, as shown in the following sample script:
counts INT,
created DATETIME NOT NULL DEFAULT(getdate()))
GO

'dev.fyicenter.com',
NULL,
0,
'2006-04-30');
GO

'dba.fyicenter.com',
NULL,
0,
DEFAULT);
GO

VALUES ('sqa.fyicenter.com', 103);
GO
You should keep this table and its data to practice other tutorial exercises presented in this collection.
How To Select All Columns of All Rows from a Table with a SELECT statement?
The simplest query statement is the one that selects all columns of all rows from a single table:
"SELECT * FROM tableName". The (*) in the SELECT clause tells the query to return all columns.
The missing WHERE clause tells the query to return all rows in specified table. The tutorial exercise
below returns all columns and all rows from table "fyi_links":
How To Select Some Specific Columns from a Table in a Query?
If you want explicitly tell the query to return some specific columns, you can specify the column
names in the SELECT clause. The following select statement returns only three columns, "id",
"created" and "url" from the table "fyi_links":
SELECT id, created, url FROM fyi_links
id created url
101 2006-04-30 dev.fyicenter.com
102 2007-05-19 dba.fyicenter.com
103 2007-05-19 sqa.fyicenter.com
How To Select Some Specific Rows from a Table?
If you don't want select all rows from a table, you can specify a WHERE clause to tell the query to
return only the rows that meets the condition defined in the WHERE clause. The WHERE clause
condition is a normal Boolean expression. If any data from the table needs to be used in the Boolean
expression, column names should be used to represent the table data.
The first select statement below only returns rows that have url names containing the letter "a". The
second select statement returns no rows, because the WHERE clause results FALSE for all rows in
the table.
SELECT * FROM fyi_links WHERE url LIKE '%a%'
GO
SELECT * FROM fyi_links WHERE id < 100

GO
0 row
How To Add More Data to the Testing Table?
If you want to continue with other tutorial exercises in this FAQ collection, you need to add more data
to the testing table. Follow the script below to add a new column and more rows:
ALTER TABLE fyi_links ADD tag VARCHAR(8)
GO
UPDATE fyi_links SET tag = 'DEV' WHERE id = 101

GO
UPDATE fyi_links SET tag = 'DBA' WHERE id = 102
GO
UPDATE fyi_links SET tag = 'SQA' WHERE id = 103
GO

'www.mysql.com', '', '0', '2006-01-01', 'DBA')
GO
'www.oracle.com', '', '0', '2005-01-01', 'DBA')
GO
'www.php.net', '', '0', '2004-01-01', 'DEV')
GO
'www.winrunner.com', '', '0', '2003-01-01', 'SQA')
GO
UPDATE fyi_links
SET counts = CAST(LOG(id)*1000000 AS INT) % 1000
GO

GO
103 sqa.fyicenter.com NULL 728 2007-05-19 SQA
104 www.mysql.com 390 2006-01-01 DBA
106 www.php.net 439 2004-01-01 DEV
107 www.winrunner.com 828 2003-01-01 SQA
How To Sort the Query Output with ORDER BY Clauses?
If you want the returning rows to be sorted, you can specify a sorting expression in the ORDER BY
clause. The simplest sort expression is column name who's values will be sorted by. The following
select statement returns rows sorted by the values in the "counts" column:
SELECT * FROM fyi_links ORDER BY counts
GO
104 www.mysql.com 390 2006-01-01 DBA
106 www.php.net 439 2004-01-01 DEV
103 sqa.fyicenter.com NULL 728 2007-05-19 SQA
107 www.winrunner.com 828 2003-01-01 SQA
The output is sorted by the values of the "counts" column.
Can the Query Output Be Sorted by Multiple Columns?
You can specifying multiple columns in the ORDER BY clause as shown in the following example
statement, which returns rows sorted by "tag" and "counts" values:
SELECT tag, counts, url, created
FROM fyi_links ORDER BY tag, counts
GO
tab counts url created
DBA 390 www.mysql.com 2006-01-01
DBA 960 www.oracle.com 2005-01-01
DBA 972 dba.fyicenter.com 2007-05-19
DEV 120 dev.fyicenter.com 2006-04-30
DEV 439 www.php.net 2004-01-01
SQA 728 sqa.fyicenter.com 2007-05-19
SQA 828 www.winrunner.com 2003-01-01
How To Sort Query Output in Descending Order?
If you want to sort a column in descending order, you can specify the DESC keyword in the ORDER
BY clause. The following SELECT statement first sorts the "tag" in descending order, then sorts the
"counts" in ascending order:
FROM fyi_links ORDER BY tag DESC, counts
GO
tab counts url created
SQA 828 www.winrunner.com 2003-01-01
DEV 439 www.php.net 2004-01-01
DBA 390 www.mysql.com 2006-01-01
DBA 960 www.oracle.com 2005-01-01
How To Count Rows with the COUNT(*) Function?
If you want to count the number of rows, you can use the COUNT(*) function in the SELECT clause.
The following tutorial exercise shows you some good examples:
GO
7
WHERE url LIKE '%fyi%'
GO
3
So there are 7 rows in total in table "fyi_links", and 3 rows that have 'fyi' as part of their url names.
Can SELECT Statements Be Used on Views?
Select (query) statements can be used on views in the same way as tables. The following tutorial
exercise helps you creating a view and running a query statement on the view:
CREATE VIEW myLinks AS SELECT * FROM fyi_links
WHERE url LIKE '%fyi%'
GO

FROM myLinks ORDER BY counts DESC
GO
tag counts url created
How To Filter Out Duplications in the Returning Rows?
If there are duplications in the returning rows, and you want to remove the duplications, you can use
the keyword DISTINCT in the SELECT clause. The DISTINCT applies to the combination of all data
fields specified in the SELECT clause. The tutorial exercise below shows you how DISTINCT works:
CREATE TABLE fyi_team (first_name VARCHAR(8),
last_name VARCHAR(8))
GO
INSERT INTO fyi_team VALUES ('John', 'Gate')

GO
INSERT INTO fyi_team VALUES ('John', 'Russell')
GO
INSERT INTO fyi_team VALUES ('John', 'Seo')
GO
GO
INSERT INTO fyi_team VALUES ('James', 'Gate')
GO
INSERT INTO fyi_team VALUES ('Peter', 'Gate')
GO
GO
SELECT * FROM fyi_team
GO
first_name last_name
John Gate
John Russell
John Seo
John Gate
James Gate
Peter Gate
John Gate
SELECT DISTINCT * FROM fyi_team

GO
first_name last_name
James Gate
John Gate
John Russell
John Seo
Peter Gate
SELECT DISTINCT last_name FROM fyi_team

Gate
Russell
Seo
Remember that * in select list represents all columns.
What Are Group Functions in Query Statements?
Group functions are functions applied to a group of rows. Examples of group functions are:
 COUNT(*) - Returns the number of rows in the group.
 MIN(exp) - Returns the minimum value of the expression evaluated on each row of the group.
 MAX(exp) - Returns the maximum value of the expression evaluated on each row of the group.
 AVG(exp) - Returns the average value of the expression evaluated on each row of the group.
 SUM(exp) - Returns the sum value of the expression evaluated on each row of the group.
 COUNTDISTINCT(exp) - Returns the number of distinct values of the expression evaluated on
each row of the group.
 How To Use Group Functions in the SELECT Clause?

 If group functions are used in the SELECT clause, all rows that meet the criteria defined in the
WHERE clause will be treated as a single group. The group functions will be apply all rows in that
group as a whole. The final output of the SELECT statement is the resulting values of the group
functions, not the rows in the group.
 Here are two good examples of using group functions :
 SELECT COUNT(*), MAX(counts), MIN(created)
 FROM fyi_links
 GO
 COUNT(*) MAX(counts) MIN(created)
 7 972 2003-01-01

 SELECT COUNT(*), MAX(counts), MIN(created)
 FROM fyi_links WHERE tag = 'DBA'
 GO
 COUNT(*) MAX(counts) MIN(created)
 3 972 2005-01-01
 In first case, the group contains all the rows in table fyi_links, because is no WHERE clause. In
the second case, the group contains only 3 rows because of the WHERE clause tag = 'DBA'.
 Can Group Functions Be Mixed with Non-group Selection Fields?
 If a group function is used in the SELECT clause, all other selection fields must be group level
fields. Non-group fields can not be mixed with group fields in the SELECT clause. The script
below gives you an example of invalid SELECT statements with group and non-group selection
fields mixed in a SELECT statement:
 SELECT COUNT(*), url FROM fyi_links
 GO
 Column 'fyi_links.url' is invalid in the select list because
 it is not contained in either an aggregate function or the
 GROUP BY clause.

 SELECT 2*COUNT(*), 2*counts FROM fyi_links
 GO
 Column 'fyi_links.counts' is invalid in the select list
 because it is not contained in either an aggregate function
 or the GROUP BY clause.
 In these examples, COUNT(*) is a group field and "url"/"2*counts" is a non-group field. The error
message also tells that "url"/"counts" is not an aggregate function (group function).
How To Divide Query Output into Multiple Groups with the GROUP BY Clause?
Sometimes, you want to divide the query output into multiple groups, and apply group functions on
each individual groups. Dividing query output into multiple groups can be done with the GROUP BY
clause. Here is the syntax of a SELECT statement with a GROUP BY clause.
SELECT group_level_fields FROM source_tables
 group_by_express - An list of columns to be used as the group criteria - Rows that have the same
combination of values of there columns form a single group.
 group_level_fields - An list of selection expressions that can be evaluated at the group level.
The final output of the SELECT statement is the resulting values of group_level_fields for each group.
The following script gives you a good GROUP BY example with a single column as the
group_by_expression. In this case, rows with the same value of this column will be considered as a
single group.
SELECT tag, COUNT(*), MAX(counts), MIN(created)
FROM fyi_links GROUP BY tag
GO
tag COUNT(*) MAX(counts) MIN(created)
DBA 3 972 2005-01-01
DEV 2 439 2004-01-01
SQA 2 828 2003-01-01
Notice that, column "tag" can also be used in group_level_fields, because it is used as the
group_by_expression, and becomes a constant for any given group.
How To Apply Filtering Criteria at Group Level with The HAVING Clause?
Let's say you have divided the query output into multiple groups with the GROUP BY clause. Now
you are only interested in some of the groups, not all the groups. If you want to filter out some groups
from the query, you can apply filtering criteria at the group level by using the HAVING clause inside
the GROUP BY clause with this syntax:
SELECT group_level_fields FROM source_tables
HAVING group_filtering_condition
Since group_filtering_condition applies to groups, only group level expressions can be used in
group_filtering_condition. The following tutorial exercise gives you some good examples of HAVING
clause:
SELECT tag, COUNT(*), MIN(created), AVG(counts)
FROM fyi_links
GROUP BY tag HAVING AVG(counts) > 300
GO
tag COUNT(*) MIN(created) AVG(counts)
DBA 3 2005-01-01 774
SQA 2 2003-01-01 778
SELECT tag, COUNT(*), MIN(created), AVG(counts)

FROM fyi_links
GROUP BY tag
HAVING AVG(counts) > 300 AND tag = 'DBA'
GO
tag COUNT(*) MIN(created) AVG(counts)
DBA 3 2005-01-01 774
Note that the more criteria you have in the HAVING clause, the less groups you will get.
How To Count Duplicated Values in a Column?
If you have a column with duplicated values, and you want to know what are those duplicated values
are and how many duplicates are there for each of those values, you can use the "GROUP BY ...
HAVING" clause as shown in the following example. It returns how many duplicated first names in the
fyi_team table:
SELECT first_name, COUNT(*) FROM fyi_team
GROUP BY first_name HAVING COUNT(*) > 1
GO
first_name count(*)
John 5
So we have "John" as duplicated first name 5 times in the fyi_team table
Can Multiple Columns Be Used in GROUP BY?
If you want to break your output into smaller groups, you can specify multiple column names or
expressions in the GROUP BY clause. Output in each group must satisfy a specific combination of
the expressions listed in the GROUP BY clause. The more columns or expressions entered in the
GROUP BY clause, the smaller the groups will be.
The tutorial exercise below shows you how to break data into groups per "tag" and per year when
they were created. Then the group function COUNT(*) is applied on each group:
SELECT tag, YEAR(created), COUNT(*)
FROM fyi_links GROUP BY tag, YEAR(created)
GO
tag year(created) count(*)
SQA 2003 1
DEV 2004 1
DBA 2005 1
DBA 2006 1
DEV 2006 1
DBA 2007 1
SQA 2007 1
So there is only one row in each group.
Can Group Functions Be Used in the ORDER BY Clause?
If the query output is aggregated as groups, you can sort the groups by using group functions in the
ORDER BY clause. The following statement returns the maximum "counts" in each group, determined
by a unique combination of tag and year. The group output is sorted by the maximum "counts" in
each group in ascending order:
SELECT tag, YEAR(created), MAX(counts)
FROM fyi_links GROUP BY tag, YEAR(created)
ORDER BY MAX(counts)
GO
tag year(created) max(counts)
DEV 2006 120
DBA 2006 390
DEV 2004 439
SQA 2007 728
SQA 2003 828
DBA 2005 960
DBA 2007 972
What Is SQL Language?
SQL, SEQUEL (Structured English Query Language), is a language for RDBMS (Relational Database
Management Systems).
During the 1970s, a group at IBM's San Jose research center developed a database system "System
R" based upon Codd's model. Structured English Query Language ("SEQUEL") was designed to
manipulate and retrieve data stored in System R. The acronym SEQUEL was later condensed to
SQL.
SQL was adopted as a standard by ANSI (American National Standards Institute) in 1986 and ISO
(International Organization for Standardization) in 1987. Since then, SQL standard has gone through
a number of revisions:
 1986, SQL-86 or SQL-87 - First published by ANSI. Ratified by ISO in 1987.
 1989, SQL-89 - Minor revision.
 1992, SQL-92 or SQL2 - Major revision (ISO 9075).
 1999, SQL:1999 or SQL3 - Added regular expression matching, recursive queries, triggers, non-
scalar types and some object-oriented features.
 2003, SQL:2003 - Introduced XML-related features, window functions, standardized sequences
and columns with auto-generated values (including identity-columns).
 2006, SQL:2006 - ISO/IEC 9075-14:2006 defines ways in which SQL can be used in conjunction
with XML.
What Is Transact-SQL Language?
According to Wikipedia, Transact-SQL, sometimes abbreviated T-SQL, is Microsoft's and Sybase's

proprietary extension to the SQL language. Microsoft's implementation ships in the Microsoft SQL
Server product. Sybase uses the language in its Adaptive Server Enterprise, the successor to Sybase
SQL Server.
In order to make it more powerful, SQL has been enhanced with additional features such as:
 Control-of-flow language - Like BEGIN and END, BREAK, CONTINUE, GOTO, IF and ELSE,
RETURN, WAITFOR, and WHILE statements.
 Local variables - Like DECLARE and SET statements
 Various support functions for string processing, date processing, mathematics, etc. - Like SIN(),
ASIN(), LOG(), etc.
 Improvements to DELETE and UPDATE statements - Both the DELETE and UPDATE
statements allow a FROM clause to be added which allows joins to be included.
What Is a Transact-SQL Statement?
A Transact-SQL statement is a basic execution unit of the Transact-SQL language. Here are some
examples of Transact-SQL statements:
 "CREATE TABLE ...;" - A data definition language statement that defines a table.
 "INSERT INTO ...;" - A data manipulation language statement that inserts data into a table.
 "SELECT * FROM ...;" - A query statement that returns data from a table.
 "DROP TABLE ...;" - A data definition language statement that deletes a table.
 "DECLARE @name data_type;" - A declaration statement that defines a local variable.
 "SET @name = expression;" - An assignment statement that assigns a value to a local variable.
 "BEGIN TRANSACTION;" - A transaction management statement that marks the beginning of a
transaction.
 "ROLLBACK TRANSACTION;" - A transaction management statement that rolls back all changes
of a transaction.
How To Start and End Transact-SQL Statements?
There are simple rule about writing Transact-SQL statements:

 A Transact-SQL statement should be started with a pre-defined statement keyword.
 A Transact-SQL statement should be ended with a new line (/n) or a semicolon (;).
 A Transact-SQL statement can be entered in a single line or multiple lines.
 Transact-SQL statement keywords are case-insensitive.
 Extra white space characters are ignored within Transact-SQL statements.
Here are some good examples of Transact-SQL statements:

counts INT,
created DATETIME NOT NULL DEFAULT(getdate()));
SeLeCT 'Welcome to FYIcenter.com SQL Server!';
Print 'Come visit this page again!'
How To Enter Comments in Transact-SQL Statements?
There are 2 ways to enter comments within Transact-SQL statements:

 Starting comments with two dashes "--": Everything between "--" and the end of the line is treated
as a comment.
 Entering comments between "/*" and "*/": Everything between "/*" and "*/" is treated as a
comment.
Here are some good examples of how to enter comments:

/* The following statement
creates a table
called fyi_links */
CREATE TABLE fyi_links (
id INTEGER PRIMARY KEY, -- the primary key
url VARCHAR(80) NOT NULL, -- address of the link
counts INT, -- number of clicks
created DATETIME NOT NULL DEFAULT(getdate())
);
GO
-- Get rid of this table
DROP TABLE fyi_links;
GO
What Is a Transact-SQL Statement Batch?
A batch is a group of one or more Transact-SQL statements sent at the same time from an
application to SQL Server 2005 for execution. SQL Server compiles the statements of a batch into a
single executable unit, called an execution plan. The statements in the execution plan are then
executed one at a time.
How to define and send a batch is depending on the client application you are using. For example:
 You need to use the GO command in "sqlcmd" tool to end a batch and send it to SQL server.
 SQL Server Management Studio also uses GO command to separate batches when multiple
statements are selected for execution.
 In ADO, a batch is defined as all the statements entered in the CommandText property of a Command
object.
What Happens to a Statement Batch If There Is a Compilation Error?
If a statement batch has multiple statements, and one of them has compilation error, all statements in
the batch will not be executed. The tutorial exercise below gives you some good examples:
SELECT getdate();
SELECT getdates();
SELECT getdate();
GO
'getdates' is not a recognized built-in function name.
As you can see, the compilation error on the second statement stops the execution of all statements.
How To Use GO Command in "sqlcmd"?
"sqlcmd" is a command line client application to run Transact-SQL statements on a target SQL
Server.
When "sqlcmd" is started and connected to a SQL Server, it will start a new batch and prompt you to
enter the first statement of the batch. You can enter one or more statements in one or more lines to
form a Transact-SQL statement batch.
To end a batch of statements and send it to the SQL Server for execution, you need to enter the GO
command. The following "sqlcmd" tutorial session sends two batches to the SQL Server:
C:\>sqlcmd -S localhost\sqlexpress -U sa -P FYIcenter
1> SELECT getdate();
2> SELECT getdates();
4> GO
'getdates' is not a recognized built-in function name.
4> GO
-----------------------
2007-05-19 22:55:07.233
(1 rows affected)
-----------------------
2007-05-19 22:55:07.233
(1 rows affected)
-----------------------
2007-05-19 22:55:07.233
(1 rows affected)
1>QUIT
C:\
How To Create User Messages with PRINT Statements?
Normally, messages are generated by SQL Server and returned to the client applications associated
with the execution result of statement batches. But you can define your messages and ask the SQL
Server to your client session with the PRINT statement, as shown in this tutorial exercise:
C:\>sqlcmd -S localhost\sqlexpress -U sa -P FYIcenter
1> PRINT 'What time is it?';
2> SELECT GETDATE();
3> GO
What time is it?
-----------------------
2007-05-19 21:51:23.797
(1 rows affected)
1> PRINT 'What time is it?';

2> PRINT 'It''s '+CAST(GETDATE() AS NVARCHAR(30));
3> GO
What time is it?
It's May 19 2007 9:55PM
Note that the SELECT statement returns query result, which is very different than messages returned
by the PRINT statement.
How Many Categories of Data Types Used by SQL Server?

A data type is an attribute that specifies the type of data that the object can hold: integer data,
character data, monetary data, date and time data, binary strings, and so on. In SQL Server 2005,
each column, local variable, expression, and parameter has a related data type.
Data types in SQL Server 2005 are organized into the following categories:
 Exact Numerics - BIGINT, INT, SMALLINT, TINYINT, BIT, DECIMAL, NUMERIC, MONEY, and
SMALLMONEY.
 Approximate Numerics - FLOAT, and REAL.
 Date and Time - DATETIME and SMALLDATETIME.
 Character Strings - CHAR, VARCHAR, and TEXT.
 Unicode Character Strings - NCHAR, NVARCHAR, and NTEXT
 Binary Strings - BINARY, VARBINARY, and IMAGE.
 Other Data Types - CURSOR, SQL_VARIANT, TABLE, UNIQUEIDENTIFIER, TIMESTAMP, and
XML.
What Are Exact Numeric Data Types?
Exact numeric data types are used to store numeric values with exact precisions and scales. SQL
Server 2005 supports the following exact numeric data types:
 BIGINT - 8-byte integers in the range of -2^63 (-9,223,372,036,854,775,808) to 2^63-1
(9,223,372,036,854,775,807).
 INT - 4-byte integers in the range of -2^31 (-2,147,483,648) to 2^31-1 (2,147,483,647).
 SMALLINT - 2-byte integers in the range of -2^15 (-32,768) to 2^15-1 (32,767).
 TINYINT - 1-byte integers in the range of 0 to 255.
 BIT - 1-bit integers with 2 values: 0 and 1.
 DECIMAL - Decimal numbers with a fixed precision p and a fixed scale s defined as
DECIMAL(p,s). DECIMAL values are in the range of -10^38+1 to 10^38-1. DECIMAL has a
synonym of DEC.
 NUMERIC - Same as DECIMAL.
 MONEY - Currency values stored in 8 bytes in the range of -922,337,203,685,477.5808 to
922,337,203,685,477.5807
 SMALLMONEY - Currency values stored in 4 bytes in the range of - 214,748.3648 to
214,748.3647
Here are some good examples of exact numeric values:

PRINT 372036854775808; -- BIGINT
PRINT 147483648; -- INT
PRINT 2768; -- SMALLINT
PRINT 250; -- TINYINT
PRINT 1; -- BIT
PRINT 12748.3648; -- DECIMAL(9,2)
PRINT 337203685477.58; -- MONEY
PRINT 748.36; -- SMALLMONEY
GO
What Are Approximate Numeric Data Types?

Approximate numeric data types are used to store numeric values with floating decimal points. SQL
Server 2005 supports the following approximate numeric data types:
 FLOAT - Floating point values with a fixed number of bits n for the mantissa part defined as
FLOAT(n). FLOAT values are in the range of -1.79E+308 to -2.23E-308, 0 and 2.23E-308 to
1.79E+308.
 REAL - Same as FLOAT(24), also called single precision floating point numbers.
The DOUBLE PRECISION data type is supported as a synonym of FLOAT(53).
Here are some good examples of approximate numeric values:

PRINT 9.22337203685e+010; -- FLOAT(53)
PRINT 9.22337e+010; -- FLOAT(24)
GO
What Are Date and Time Data Types?
Date and time data types are used to store an instances of calendar dates and times. SQL Server
2005 supports the following date and time data types:
 DATETIME - Date and time values stored in 8 bytes with 4 bytes for the date and 4 bytes for the
time. DATETIME values are in the range of January 1, 1753 to December 31, 9999 with a
precision of 3.33 milliseconds.
 SMALLDATETIME - Date and time values stored in 4 bytes with 2 bytes for the date and 2 bytes
for the time. SMALLDATETIME values are in the range of January 1, 1900 to June 6, 2079 with a
precision of 1 minute.
Note that SQL Server does not support DATE and TIME data types separately.
Here are some good examples of date and time values:

PRINT '2107-05-19 22:52:51.607'; -- DATETIME
PRINT '2007-05-19 22:52:00'; -- SMALLDATETIME
GO
What Are Character String Data Types?
Character string data types are used to store code page based character strings. SQL Server 2005
supports the following character string data types:
 CHAR - Code page based character strings with a fixed length of n bytes defined as CHAR(n).
The maximum length is 8,000 bytes. CHAR has a synonym of CHARACTER.
 VARCHAR - Code page based character strings with a variable length of n bytes defined as
VARCHAR(n). The maximum length is 8,000 bytes. VARCHAR has two synonyms of CHAR
VARYING and CHARACTER VARYING. VARCHAR also has special form as VARCHAR(MAX),
which can store up to 2^31-1 bytes.
 TEXT - Code page based character strings with a variable length up to 2^31-1 (2,147,483,647)
bytes. TEXT is equivalent to VARCHAR(MAX).
Here are some good examples of character string values:

PRINT 'Hello! '; -- CHAR(10)
PRINT 'Hello!'; -- VARCHAR(10)
GO
What Are Unicode Character String Data Types?
Unicode character string data types are used to store Unicode character strings. SQL Server 2005
supports the following Unicode character string data types:
 NCHAR - Unicode character strings with a fixed length of n characters defined as NCHAR(n). The
maximum length is 4,000 characters. NCHAR has two synonyms of NATIONAL CHAR and
NATIONAL CHARACTER.
 VARCHAR - Unicode character strings with a variable length of n characters defined as
VARCHAR(n). The maximum length is 4,000 characters. NVARCHAR has two synonyms of
NATIONAL CHAR VARYING and NATIONAL CHARACTER VARYING. NVARCHAR also has
special form as NVARCHAR(MAX), which can store up to 2^31-1 bytes.
 NTEXT - Unicode character strings with a variable length up to 2^31-1 (2,147,483,647) bytes.
NTEXT is equivalent to NVARCHAR(MAX). NTEXT has a synonym of NATIONAL TEXT.
What Are Binary String Data Types?
Binary string data types are used to store binary strings. SQL Server 2005 supports the following
binary string data types:
 BINARY - Binary strings with a fixed length of n bytes defined as BINARY(n). The maximum
length is 8,000 bytes.
 VARBINARY - Binary strings with a variable length of n bytes defined as VARBINARY(n). The
maximum length is 8,000 bytes. VARBINARY has a synonym of BINARY VARYING.
VARBINARY also has special form as VARBINARY(MAX), which can store up to 2^31-1 bytes.
 IMAGE - Binary strings with a variable length up to 2^31-1 (2,147,483,647) bytes.
What Are the Differences between CHAR and NCHAR?
Both CHAR and NCHAR are fixed length data types. But they have the following main differences:
 CHAR stores characters based on the code page with 1 byte per character most of the time.
 NCHAR stores characters using Unicode character set with 2 bytes per character most of the
time.
 CHAR(n) specifies a length of n bytes by default.
 NCHAR(n) specifies a length of n characters by default.
What Are the Differences between CHAR and VARCHAR?
CHAR and VARCHAR are both used to store code page based character strings. But they have the
following main differences:
 CHAR(n) stores character strings with a fixed length, n bytes, storage format.
 VARCHAR(n) stores character strings with a variable length, less than n bytes, storage format.
 If the input string of CHAR(n) has less than n bytes, it will be padded with space characters to
become n bytes.
 If the input string of VARCHAR(n) has less than n bytes, it will not be padded with any characters.
 If the input string of CHAR(n) has more than n bytes, it will be truncated to n bytes.
 If the input string of VARCHAR(n) has more than n bytes, it will also be truncated to n bytes.
What Are the Differences between DECIMAL and FLOAT?

DECIMAL and FLOAT are both used to store numerical values. But they have the following main
differences:
 DECIMAL(p,s) stores values with the decimal point fixed at the position of s (scale) digits from the
right. The total number of decimal digits is also fixed as p (precesion).
 FLOAT(n) stores values with the decimal point floating based on the value. The number of bits
used to store the mantissa part is fixed as n.
 If the input value of DECIMAL(p,s) has more digits after the decimal point than the scale s, the
value will be rounded to the scale s.
 If the input value of FLOAT(n) has more total digits (mantissa) than what n bits can store, the
value will be rounded to fit the storage size.
 If the input value of DECIMAL(p,s) has more digits before the decimal point than p-s, SQL Server
will give you an over-flow error.
 If the input value of FLOAT(n) is too big that the exponential part goes over the positive limit, SQL
Server will give you an over-flow error.
 If the input value of FLOAT(n) is too small that the exponential part goes over the negative limit,
SQL Server will give you an under-flow error.
What Are Date and Time Data Types?
Date and time data types are used to store an instances of calendar dates and times. SQL Server
2005 supports the following date and time data types:
 DATETIME - Date and time values stored in 8 bytes with 4 bytes for the date and 4 bytes for the
time. DATETIME values are in the range of January 1, 1753 to December 31, 9999 with a
precision of 3.33 milliseconds.
 SMALLDATETIME - Date and time values stored in 4 bytes with 2 bytes for the date and 2 bytes
for the time. SMALLDATETIME values are in the range of January 1, 1900 to June 6, 2079 with a
precision of 1 minute.
Note that SQL Server does not support DATE and TIME data types separately.
Here are some good examples of date and time values:

PRINT '2107-05-19 22:52:51.607'; -- DATETIME
PRINT '2007-05-19 22:52:00'; -- SMALLDATETIME
GO
What Is a Constant or Literal?
A constant, or literal, is a symbol that represents a specific value of a specific data type. Constants or
literals are used commonly as the default values for table columns, variables, and parameters. The
format of constant or literal depends on the data type of the value it represents. There are x formats
used to specify literals in SQL Server:
 Integer Number Literals - Strings of numbers prefixed with (+/-) as the sign if needed. They are
used for integer data types. For example: -2255.
 Decimal Number Literals - Strings of numbers with a single decimal point and an optional prefix
sign (+/-). They are used for DECIMAL data type. For example: 1894.1204.
 Floating Point Number Literals - Numeric values represented in the scientific notation. They are
used for FLOAT data type. For example: 9.22337e+010.
 Character String Literals - Strings of characters enclosed in single quotes. They are used for
character string data types. . For example: 'Welcome to FYIcenter.com!'.
 Unicode Character String Literals - Strings of Unicode characters enclosed in single quotes with a
prefix of (N). They are used for Unicode character string data types. For example: N'Welcome to
FYIcenter.com'.
 Binary String Literals - Strings of bytes expressed in HEX numbers and prefixed with (0x). They
are used for binary string data types. For example: '0x69048AEFDD010E'.
 Date and Time Literals - Data and time values expressed in several popular formats and
enclosed in single quotes ('). They are used for data and time data types. For example: '2007-05-
19 22:55:07.233'.
 Money Literals - Strings of numbers with a single decimal point with a prefix of ($) and optional
sign (+/-). They are used for money data types. For example: $542023.14.
How To Write Character String Constants or Literals?
Character string literals are used everywhere in Transact-SQL statements. You need to learn some
important rules on writing character string literals:
 Character string literals must be enclosed in single quotes as 'Welcome to FYIcenter.com!'.
 An empty string must be expressed as ''.
 Single quote characters inside a string must be expressed as two single quotes. For example:
'It''s Sunday!'.
 Character strings use non-default character code pages must use the COLLATE clause to
provide the code page through a collate name. For example, 'Français' COLLATE French_CI_AS.
What Is a Collation?
A collation is a set of rules defining a character set and its sorting rules. SQL Server support a large
number of built-in collations. For example:
 Albanian_CI_AI_KS_WS - Albanian, case-insensitive (CI), accent-insensitive (AI), kanatype-
sensitive (KS), width-sensitive (WS).
 Arabic_CI_AS_KS_WS - Arabic, case-insensitive, accent-sensitive, kanatype-sensitive, width-
sensitive.
 French_CI_AI - French, case-insensitive, accent-insensitive, kanatype-insensitive, width-
insensitive.
 Korean_Wansung_BIN - Korean-Wansung, binary sort.
 SQL_Latin1_General_CP1250_CI_AS - Latin1-General, case-insensitive, accent-sensitive,
kanatype-insensitive, width-insensitive.
If you want to get a list of all collations support on the SQL Server, run this statement:
SELECT * FROM fn_helpcollations()
GO
How To Specify the Collation for a Character Data Type?
If you do not want to use the default collation provided by the SQL Server, you can use the
"COLLATE collation_name" clause to specify a different collation to be used at different levels:
 Database Level - Used in CREATE DATABASE or ALTER DATABASE statements to set a new
default collation for the entire database.
 Table Column Level - Used in CREATE TABLE or ALTER TABLE statements to override the
database default collation for this column.
 Character String Expression Level - Used at the end of the expression to cast the expression to
the specified collation.
What Happens If Strings Are Casted into Wrong Code Pages?
In SQL Server, different collations may use different code pages. For example:
 Albanian_CI_AI_KS_WS - Albanian, Code page 1250.
 Arabic_CI_AS_KS_WS - Arabic, Code page 1256.
 French_CI_AI - French, Code page 1252.
 Korean_Wansung_BIN - Korean-Wansung, Code page 949.
 SQL_Latin1_General_CP1250_CI_AS - Latin1-General, Code page 1250.
If you are casting a string of characters from one code page to a different code page, some character
will be converted to similar. For example
PRINT 'Français: eéèê-aà-oô';
-- The default code page
PRINT 'Français: eéèê-aà-oô'

COLLATE French_CI_AI; -- Code page 1252

COLLATE Polish_CI_AS; -- Code page 1250

COLLATE Cyrillic_General_CI_AS; -- Code page 1256
GO
Français: eéèê-aà-oô
Français: eéèê-aà-oô
Français: eéee-aa-oô
Francais: eeee-aa-oo
How find out the default Collation?
How To Find Out What Is the Default Collation in a Database?
The default collation of a database comes from the server if you are not using the COLLATE clause in
the CREATE DATABASE statement. If you are not using the COLLATE clause for character string
column, it will use the default collation from the database.
Using this logic, you can find out the default collation in a database or server very easily. See the
following tutorial exercise:
CREATE DATABASE FyiCenterData
GO
CREATE TABLE faq (Name VARCHAR(80))
GO
SP_HELP FAQ
GO
Column_name ... Collation
Name ... SQL_Latin1_General_CP1_CI_AS
Now we know that the default collation for SQL Server 2005 is SQL_Latin1_General_CP1_CI_AS.
How Fixed Length Strings Are Truncated and Padded?
When the length of the input string does not match the storage size of the fixed length string data type
CHAR(n). SQL Server will:
 If the input string of CHAR(n) has less than n bytes, it will be padded with space characters to
become n bytes.
 If the input string of CHAR(n) has more than n bytes, it will be truncated to n bytes.
The tutorial exercise shows you good examples of truncating and padding fixed length character
strings:
-- Length matches the data type size

DECLARE @msg CHAR(36);
SET @msg = 'Welcome to FYIcenter.com SQL Server!';
PRINT '('+@msg+')';
GO
(Welcome to FYIcenter.com SQL Server!)
-- Length is bigger than the data type size - truncated

PRINT '('+@msg+')';
GO
(Welcome to FYIcenter.com)
-- Length is smaller than the data type size - padded

PRINT '('+@msg+')';
GO
(Welcome to FYIcenter.com SQL Server! )
How To Enter Unicode Character String Literals?
Unicode characters are multi-byte characters. They are very hard to be entered as string literals,
because it requires:
 The SQL client tool to support Unicode characters. The command line tool 'sqlcmd' does not
support Unicode. But SQL Server Management Studio does support Unicode.
 A special Unicode character input tool to allow you to enter multi-byte Unicode characters.
But if you know the code value of a Unicode character, you can enter it with the help of the
NCHAR(int) function. For example:
-- A Unicode string of some interesting characters
PRINT NCHAR(9733)+NCHAR(9734)+NCHAR(9792)+NCHAR(9794);
GO
Run the sample statement with SQL Server Management Studio, you should see some Unicode
characters.
How To Enter Binary String Literals?
Binary string long values are normally generated by client applications by reading input channels, like
image files. But sometimes, you may need to enter some short binary strings as literals. Binary string
literals are entered as a string of bytes expressed in HEX numbers and prefixed with (0x).
Input strings will be truncated or padded to fit the storage size of fixed length binary string data type.
The padding bytes will be the zero byte: 0x00. The tutorial exercise shows you good examples of
truncating and padding fixed length binary strings:
-- Size matches
DECLARE @x BINARY(8);
SET @x = 0x2605260626402642;
PRINT @x;
GO
0x2605260626402642
-- Truncated
SET @x = 0x2605260626402642;
PRINT @x;
GO
0x26052606
-- Padded
SET @x = 0x2605260626402642;
PRINT @x;
GO
0x260526062640264200000000
-- No padding on variable length data type

DECLARE @x VARBINARY(8);
SET @x = 0x; -- Empty binary string
PRINT @x;
GO
0x
-- Padding on fixed length data type

SET @x = 0x; -- Empty binary strings
PRINT @x;
GO
0x0000000000000000
How To Enter Date and Time Literals?
Date and time literals are entered as character strings enclosed in single quotes ('). The string must in
one of the recognizable data and time formats. Some of the most commonly used formats are given
in the following tutorial exercise:
-- Default format for query output
DECLARE @x DATETIME;
SET @x = '2007-05-19 22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
-- Default for string print output

SET @x = 'May 19 2007 10:55PM';
SELECT @x;
GO
2007-05-19 22:55:00.000
-- Europe default format

SET @x = '19-May-2007 22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
-- ISO8601 standard format

SET @x = '2007-05-19T22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
Why I Can Not Enter 0.001 Second in Date and Time Literals?
If you enter milliseconds in data and time literals, they will be rounded up to 10/3 milliseconds
increments, because DATETIME data type uses 4 bytes to store the time of the day. A 4-byte integer
can only give an accuracy of one three-hundredth second, or 3.33 milliseconds.
So if you enter a time with 0.001 second, it will be rounded to 0.000 second. The tutorial exercise
below gives you some good examples of how milliseconds are rounded by the SQL Server.
-- No rounding
SET @x = '2007-05-19 22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
-- Rounded down to 0.000

SET @x = '2007-05-19 22:55:07.001';
SELECT @x;
GO
2007-05-19 22:55:07.000
-- Rounded up to 0.003
SET @x = '2007-05-19 22:55:07.002';
SELECT @x;
GO
2007-05-19 22:55:07.003
-- Rounded up to 0.007
SET @x = '2007-05-19 22:55:07.006';
SELECT @x;
GO
2007-05-19 22:55:07.007
What Happens If Date-Only Values Are Provided as Date and Time Literals?
If only date value is provided in a data and time literal, the SQL Server will pad the time value with a
zero, or '00:00:00.000', representing the midnight time of the day. The tutorial exercise below gives
you some good examples:
-- 'mm/dd/yyyy' format
SET @x = '05/19/2007';
SELECT @x;
GO
2007-05-19 00:00:00.000
-- 'mm.dd.yy' format
SET @x = '05.19.07';
SELECT @x;
GO
2007-05-19 00:00:00.000
-- 'yyyy-mm-dd' format
SET @x = '2007-05-19';
SELECT @x;
GO
2007-05-19 00:00:00.000
-- 'mon dd, yyyy' format

SET @x = 'May 19, 2007';
SELECT @x;
GO
2007-05-19 00:00:00.000
-- 'dd-mon-yyyy' format
SET @x = '19-May-2007';
SELECT @x;
GO
2007-05-19 00:00:00.000
What Happens If Time-Only Values Are Provided as Date and Time Literals?
If only time value is provided in a data and time literal, the SQL Server will pad the date value with a
zero, representing the base date, January 1, 1900. The tutorial exercise below gives you some good
examples:
-- 'hh:mi:ss.mmm' format
SET @x = '22:55:07.233';
SELECT @x;
GO
1900-01-01 22:55:07.233
-- 'hh:mi:ss.mmmAM/PM' format
SET @x = '10:55:07.233PM';
SELECT @x;
GO
1900-01-01 22:55:07.233
-- 'hh:miAM/PM' format
SET @x = '10:55PM';
SELECT @x;
GO
1900-01-01 22:55:00.000
What Are Out-of-Range Errors with Date and Time Literals?
When you enter data and time literals, you may get out-of-range errors due to two common mistakes:
 The date value is a valid calendar date, but it is not in the range covered by DATETIME data
type: from January 1, 1753, to December 31, 9999.
 The date value is not a valid calendar date, for example: 30-Feb-2007.
 The time value does not represent a valid time, for example: 24:55:07.233.
The tutorial exercise below shows you some data and time out-of-range errors:
-- Invalid date
SET @x = '30-Feb-2007 22:55:07.233';
GO
The conversion of a char data type to a datetime data type
resulted in an out-of-range datetime value.
-- Date below the DATETIME limits

SET @x = '19-May-1752 22:55:07.233';
GO
-- Invalid time
SET @x = '19-May-2007 24:55:07.233';
GO
What Happens If an Integer Is Too Big for INT Date Type?
If you are entering an INT data type literal with representing an integer value too big for INT data type
to store, the SQL Server will give you an arithmetic overflow error. The same error will happen on
BIGINT, INT, SMALLINT, and TINYINT data types.
Remember that INT data types uses 4 bytes to an integer value between -2^31 (-2,147,483,648) to
2^31-1 (2,147,483,647). The tutorial exercise below gives an example of arithmetic overflow errors.
-- INT value in the range
DECLARE @x INT;
SET @x = 2147483647;
SELECT @x;
GO
2147483647
-- INT value overflow

DECLARE @x INT;
SET @x = 2147483648;
Arithmetic overflow error converting expression to data type int
How Extra Digits Are Handled with NUMERIC Data Type Literals?
Exact numeric data types defined with NUMERIC(p,s) has two limits defined by two parameters: p
(precision) and s (scale):
 Maximum number of digits of the integer part (digits before the decimal point) is defined as p-s. If
this limit is passed, SQL Server will give you an arithmetic overflow error.
 Maximum number of digits of the decimal part (digits after the decimal point) is defined as s. If
this limit is passed, SQL Server will perform a round operation.
The tutorial exercise below gives an example of arithmetic overflow errors and rounding operations.
-- Exact numeric value
DECLARE @x NUMERIC(9,2); -- NUMERIC(p,s)
SET @x = 1234567.12;
SELECT @x;
GO
1234567.12
-- Overflow error: p-s limit passed

SET @x = 123456789.12;
GO
Arithmetic overflow error converting numeric to data type
numeric.
-- Rounding on extra decimal digits: s limit passed

SET @x = 1234567.12345;
SELECT @x;
GO
1234567.12
How REAL and FLOAT Literal Values Are Rounded?
By definition, FLOAT(n) should store the mantissa of the floating number in n bits. For example,
FLOAT(16) should have a precision one-byte less than FLOAT(24).
However, SQL Server 2005 only supports two precisions for floating numbers:
 Single Precision: FLOAT(24) or REAL, stored in 4 bytes, giving about 7 digits of precision,
covering all types from FLOAT(1) to FLOAT(24),
 Double Precision: FLOAT(53), stored in 8 bytes, giving about 15 digits of precision, covering all
types from FLOAT(25) to FLOAT(53).
The tutorial exercise below shows you some different precision and rounding examples:
-- FLOAT(1) works like FLOAT(24)
DECLARE @x FLOAT(1)
SET @x = 9.234567890E+10;
SELECT @x;
GO
9.234568E+10 -- 7 digits precision
-- Single precision with rounding

DECLARE @x REAL; -- FLOAT(24)
SET @x = 9.234567890E+10;
SELECT @x;
GO
-- FLOAT(25) works like FLOAT(53)

DECLARE @x FLOAT(25);
SET @x = 9.2345678901234567890E+100;
SELECT @x;
GO
-- Double precision with rounding

SET @x = 9.2345678901234567890E+100;
SELECT @x;
GO
What Are the Underflow and Overflow Behaviors on FLOAT Literals?
If you enter a floating number that is too big or too small for the FLOAT data type, SQL Server 2005
will behave as:
 FLOAT(24) Underflow: If a floating number is too small for FLOAT(24), it will be stored as 0
without any warning.
 FLOAT(24) Overflow: If a floating number is too big for FLOAT(24), you will get an arithmetic
overflow error.
 FLOAT(53) Underflow: If a floating number is too small for FLOAT(53), it will be stored as 0 with a
warning.
 FLOAT(53) Overflow: If a floating number is too big for FLOAT(53), you will get a value-out-of-
range error.
The tutorial exercise below some good underflow and overflow examples:
-- Single precision underflow without warning
SET @x = 9.234568E-39;
SELECT @x;
GO
0
-- Single precision overflow error

SET @x = 9.234568E+39;
GO
Arithmetic overflow error for type real,
value = 9234568000000000400000000000000000000000.000000.
-- Double precision underflow with warning

SET @x = 9.23456789012346E-309
SELECT @x;
GO
Warning: the floating point value '9.23456789012346E-309'
is too small. It will be interpreted as 0.
0
-- Double precision overflow error

SET @x = 9.23456789012346E+309
GO
The floating point value '9.23456789012346E+309' is out of
the range of computer representation (8 bytes).
How To Enter Date and Time Literals?
Date and time literals are entered as character strings enclosed in single quotes ('). The string must in
one of the recognizable data and time formats. Some of the most commonly used formats are given
in the following tutorial exercise:
-- Default format for query output
SET @x = '2007-05-19 22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
-- Default for string print output
SET @x = 'May 19 2007 10:55PM';
SELECT @x;
GO
2007-05-19 22:55:00.000
-- Europe default format

SET @x = '19-May-2007 22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
-- ISO8601 standard format

SET @x = '2007-05-19T22:55:07.233';
SELECT @x;
GO
2007-05-19 22:55:07.233
What Is an Expression?
A numeric expression is a combination of identifiers, values, and operators that SQL Server 2005 can
evaluate to obtain a numeric value.
A simple expression could be a constant, a function, a column name, a variable, or a subquery

without any operators.
Complex expressions can be constructed by joining other expressions with operators.
The following tutorial exercise shows you some expression examples:

DECLARE @site VARCHAR(40);
SET @site = 'FYIcenter.com';
SELECT 'Welcome'; -- Expression: constant
SELECT @site; -- Expression: variable
SELECT GETDATE(); -- Expression: function
SELECT 'Welcome to '+@site; -- Expression with an operator
GO
Welcome
FYIcenter.com
2007-05-19 18:42:09.077
Welcome to FYIcenter.com
DECLARE @rate NUMERIC(5,2);

DECLARE @deposit MONEY;
DECLARE @value MONEY;
SET @rate = 5.25;
SET @deposit = 4000.00;
-- Expression with multiple operators.

SET @value = @deposit*(1.0+@rate/100)*(1.0+@rate/100);
PRINT @value;
GO
4431.03
What Are Arithmetic Operators?
An arithmetic operator performs an arithmetic operation on two expressions of numeric data types.
SQL Server supports 5 arithmetic operators:
 + (Add): Addition
 - (Subtract): Subtraction
 * (Multiply): Multiplication
 / (Divide): Division
 % (Modulo): Modulation
There are several rules on arithmetic operations:

 Arithmetic operations are performed on expressions of numeric data types like INT, NUMERIC,
and FLOAT.
 When arithmetic operations are performed on expressions of the same data types, no data type
conversion is needed.
 When arithmetic operations are performed on expressions of different data types, implicit data
type conversion will be performed before the arithmetic operation on the expression with a lower
data type rank.
 Numeric data type ranks are from low to high: TINYINT, SMALLINT, INT, BIGINT,
SMALLMONEY, MONEY, DECIMAL, and FLOAT.
 The data type of the resulting value of an arithmetic operation is the same as the two input
expressions.
 What Happens to an Arithmetic Operation with Two Different Data Types?

 When two expressions with different data types are put together for an arithmetic operation, the
expression with a lower rank data type will be converted automatically to higher rank data type.
Usually, a data type with a larger storage size has a higher rank. A conversion from a lower rank
data type to a higher rank data type is considered safe, because there is no data lose during the
conversion.
 The tutorial exercise shows you some good examples on automatic conversion during arithmetic
operations:
 -- INT / INT, No conversion
 -- Returning INT
 DECLARE @i INT;
 SET @i = 1;
 SELECT @i/3;
 GO
 0

 -- NUMERIC / INT, INT converted to NUMERIC
 -- Returning NUMERIC
 DECLARE @d NUMERIC(9,6);
 SET @d = 1.0;
 SELECT @d/3;
 GO
 0.33333333

 -- INT * NUMERIC, INT converted to NUMERIC
 -- Returning NUMERIC
 DECLARE @pi NUMERIC(15,7);
 DECLARE @d INT;
 SET @pi = 3.1415927;
 SET @d = 10000000;
 SELECT @d*@pi;
 GO
 31415927.0000000

 -- INT * FLOAT, INT converted to FLOAT
 -- Returning FLOAT
 DECLARE @pi FLOAT(24);
 DECLARE @d INT;
 SET @pi = 3.1415927E+00;
 SET @d = 10000000;
 SELECT @d*@pi; -- @d converted from INT to FLOAT
 GO
 3.141593E+07
How To Convert a Numeric Expression from One Data Type to Another?
There are 4 ways to convert a numeric expression from one data type to another data type:
 Implicit conversion by arithmetic operations - When arithmetic operations are performed on
expressions of different data types, implicit data type conversion will be performed before the
arithmetic operation on the expression with a lower data type rank.
 Implicit conversion by assignment operations - When a numeric expression is assigned to
variable, column, or parameter of different data type, the expression will be converted to match
the data type of the variable, column, or parameter.
 Explicit conversion using the CAST() function - A numeric expression can be explicitly converted
to different data type using the CAST(expression AS data_type) function.
 Explicit conversion using the CONVERT() function - A numeric expression can be explicitly
converted to different data type using the CONVERT(data_type, expression) function.
Some numeric data type conversion examples are provided in the tutorial exercise below:
-- Implicit conversion by an arithmetic operation
-- INT converted to NUMERIC
DECLARE @d NUMERIC(9,6);
SET @d = 1.0;
SELECT @d/3;
GO
0.33333333
-- Implicit conversion by an assignment operation

-- NUMERIC converted to INT
DECLARE @i INT;
DECLARE @d NUMERIC(9,3);
SET @d = 123.456;
SET @i = @d;
SELECT @i;
GO
123
-- Explicit conversion by CAST()

-- FLOAT converted to NUMERIC
DECLARE @pi FLOAT(24);
SET @pi = 3.1415927E+00;
SELECT CAST(@pi AS NUMERIC(5,2));
GO
3.14
-- Explicit conversion by CONVERT()

-- FLOAT converted to NUMERIC
SET @pi = 3.1415927E+00;
SELECT CONVERT(NUMERIC(5,2), @pi);
GO
3.14
How To Convert Numeric Expression Data Types by Assignment Operations?
An assignment operation is used to assign an expression to a variable, a column, or a parameter. If

the data type of the expression does not match the data type of the receiving variable, column, or
parameter, SQL Server will perform an implicit data type conversion on the expression.
Note that implicit data type conversion during assignment operation can convert a higher rank data
type to a lower rank data type, which may resulting in losing data during the conversion. The tutorial
exercise shows you some good examples:
-- INT converted to NUMERIC DECLARE @i INT; DECLARE @d NUMERIC(9,3); SET @i = 123; SET
@d = @i; SELECT @d; GO 123.000 -- INT converted to NUMERIC DECLARE @i INT; DECLARE @d
NUMERIC(9,3); SET @i = 123; SET @d = @i; SELECT @d; GO 123.000 DECLARE @pi FLOAT(24);
DECLARE @dp NUMERIC(5,2); SET @pi = 3.1415927E+00; SET @dp = @pi; SELECT @dp; GO 3.14
How To Convert Numeric Expression Data Types using the CAST() Function?
If you want to convert the data type of a numeric expression to a new data type, you can use the
CAST(expression AS data_type) function. The tutorial exercise below shows you how to use the
CAST() function:
-- FLOAT converted to NUMERIC by CAST()
SET @pi = 3.141592E+00;
SELECT CAST(@pi AS NUMERIC(5,2));
GO
3.14
-- FLOAT converted to NUMERIC by CAST()

SET @x = 12345.12E+00;
SELECT CAST(@x AS NUMERIC(10,5));
GO
12345.12305
-- FLOAT converted to INT by CAST()

SET @x = 12345.12E+00;
SELECT CAST(@x AS INT);
GO
12345
How To Convert Numeric Expression Data Types using the CONVERT() Function?
If you want to convert the data type of a numeric expression to a new data type, you can use the
CONVERT(data_type, expression) function. The tutorial exercise below shows you how to use the
CONVERT() function:
-- FLOAT converted to NUMERIC by CONVERT()
SET @pi = 3.141592E+00;
SELECT CONVERT(NUMERIC(5,2), @pi);
GO
3.14
-- FLOAT converted to NUMERIC by CONVERT()

SET @x = 12345.12E+00;
SELECT CONVERT(NUMERIC(10,5), @x);
GO
12345.12012
-- FLOAT converted to INT by CONVERT()

SET @x = 12345.12E+00;
SELECT CONVERT(INT, @x);
GO
12345
How To Convert Character Strings into Numeric Values?
Sometimes you need to convert numeric values enclosed in character strings back to numeric values
by using the CONVERT() function.
When converting character strings to values with CONVERT(), you need to remember two rules:
 Leading and trailing space characters are ignored.
 The input string must represent a numeric literal that matches the target numeric data type.
 If the input string does not represent a numeric literal that matches the target numeric data type,
you will get a conversion error.
The tutorial exercise below shows you how to use the CONVERT() function to convert strings to
values:
SELECT CONVERT(NUMERIC(10,5), ' 12345.12 ');
GO
12345.12000
-- Input does not match the target data type

SELECT CONVERT(INT, '12345.12');
GO
Conversion failed when converting the varchar
value '12345.12' to data type int.
-- Input does not match the target data type

SELECT CONVERT(NUMERIC(10,5), '12345.12E+00');
GO
Error converting data type varchar to numeric.
-- Double conversion works

SELECT CONVERT(NUMERIC(10,5),
CONVERT(FLOAT(24), ' 12345.12E+00 '));
GO
12345.12012
SELECT CONVERT(FLOAT(24), ' 12345.12E+00 ');

GO
12345.12
What Happens When Converting Big Values to Integers?
If you are converting a numeric expression to an integer data type and the value is too big for integer
storage size, you will get an arithmetic overflow error as shown in the following examples:
-- Overflow error on implicit conversion
DECLARE @y TINYINT;
SET @x = 12345.12E+00;
SET @y = @x;
GO
Arithmetic overflow error for type tinyint,
value = 12345.120117.
-- Overflow error on explicit conversions

SET @x = 12345.12E+00;
SELECT CAST(@x AS TINYINT);
SELECT CONVERT(TINYINT, @x);
GO
value = 12345.120117.
value = 12345.120117.
What Happens When Converting Big Values to NUMERIC Data Types?
If you are converting a numeric expression to a NUMERIC data type and the value is too big for the
storage size, you will get an arithmetic overflow error as shown in the following examples:
-- Overflow error on implicit conversion
DECLARE @y NUMERIC(5,2);
SET @x = 12345.12E+00;
SET @y = @x;
GO
Arithmetic overflow error converting real to data type
numeric.
-- Overflow error on explicit conversions

SET @x = 12345.12E+00;
SELECT CAST(@x AS NUMERIC(5,2));
SELECT CONVERT(NUMERIC(5,2),@x);
GO
numeric.
numeric.
What Are the Mathematical Functions Supported by SQL Server 2005?
SQL Server 2005 supports 23 mathematical functions: ABS, ACOS, ASIN, ATAN, ATN2, CEILING,
COS, COT, DEGREES, EXP, FLOOR, LOG, LOG10, PI, POWER, RADIANS, RAND, ROUND, SIGN,
SIN, SQRT, SQUARE, and TAN.
The return data types of mathematical functions are determined by two rules:
 Arithmetical functions: ABS, CEILING, DEGREES, FLOOR, POWER, RADIANS, and SIGN are
returning the same data type as the input values. No implicit data type conversions are
performed.
 Other mathematical functions: COS, COT, EXP, LOG, LOG10, SIN, SQRT, SQUARE and TAN
are taking only FLOAT data type as input and returning FLOAT data type. Implicit data type
conversion will be performed if you are calling these functions with a non-FLOAT data type.
The tutorial exercise gives you some good examples on how to use mathematical functions:
-- ABS retuns the same data type as the input
DECLARE @z INT;
SET @x = -12345.123456789E+20;
SET @y = -12345.12;
SET @z = -12345
SELECT ABS(@x);
SELECT ABS(@y);
SELECT ABS(@z);
GO
1.2345123456789E+24
12345.12
12345
-- SQRT converts input to FLOAT(53) first

DECLARE @z INT;
SET @x = 12345.123456789E+20;
SET @y = 12345.12;
SET @z = 12345
SELECT SQRT(@x);
SELECT SQRT(@y);
SELECT SQRT(@z);
GO
1111086110829.8
111.108595527079
111.108055513541
How To Convert Numeric Values to Integers?
Sometimes you need to round a numeric value into an integer. SQL Server 2005 offers you a number
of ways to do this:
 FLOOR(value) - Returning the largest integer less than or equal to the input value. The returning
data type is the same as the input value.
 CEILLING(value) - Returning the smallest integer greater than or equal to the input value. The
returning data type is the same as the input value.
 ROUND(value, 0, 0) - Returning the integer most close to the input value. The returning data type
is the same as the input value.
 CAST(value AS INT) - Returning the largest integer less than or equal to the input value. The
returning data type is INT.
 CONVERT(INT, value) - Returning the largest integer less than or equal to the input value. The
returning data type is INT.
The tutorial exercise below gives some good examples of converting numeric values to
integers:
SELECT FLOOR(1234.5678);
SELECT CEILING(1234.5678);
SELECT ROUND(1234.5678, 0, 0);
SELECT CAST(1234.5678 AS INT);
SELECT CONVERT(INT, 1234.5678);
GO
1234
1235
1235.0000
1234
1234
How To Round a Numeric Value To a Specific Precision?
Sometimes you need to round a numeric value to a specific precision. For example, you may want to
round values in your financial statement to the precision of 1000.00. This can be done by the
ROUND() function with the following syntax:
ROUND(value, precision, type)
value: The input value to be rounded.
precision: The location of the precision digit relative

to the decimal point.
type: 0 - Round to nearest value;

1 - Truncate to a lower value.
The tutorial exercise below gives some good examples of how to use the ROUND() function:
SELECT ROUND(1234.5678, 0, 0);
SELECT ROUND(1234.5678, -3, 0);
SELECT ROUND(1234.5678, -4, 0);
SELECT ROUND(1234.5678, 3, 0);
SELECT ROUND(1234.5678, 3, 1);
GO
1235.0000
1000.0000
0.0000
1234.5680
1234.5670
How To Generate Random Numbers with the RAND() Function?
Random numbers are very useful for generating test data, passwords, or other security related data.
SQL Server 2005 offers you the random number generator function RAND in two format:
 RAND(seed) - Starting a new sequence of random numbers based on the given integer "seed"
and returning the first random number in FLOAT(53) from the sequence.
 RAND() - Returning the next random number in FLOAT(53) from the current sequence. If there
has been no current sequence, SQL Server will start a new sequence with a random "seed".
Note that calling RAND(seed) with the same seed will start the same sequence and return the same
number. To avoid this repeating pattern, you should always call RAND() without any seed and let the
server to randomly pickup a sequence.
The tutorial exercise below shows some good examples on how to generate random numbers:
SELECT RAND(100), RAND(), RAND(); -- new sequence
SELECT RAND(100), RAND(), RAND(); -- same sequence again
SELECT RAND(), RAND(), RAND();
SELECT RAND(), RAND(), RAND();
GO
0.715436657367485 0.28463380767982 0.0131039082850364
0.715436657367485 0.28463380767982 0.0131039082850364
0.28769876521071 0.100505471175005 0.292787286982702
0.868829058415689 0.370366365964781 0.58334760467751
-- Random integer between 0 and 100

SELECT FLOOR(100*RAND());
GO
68
29
20
82
How To Concatenate Two Character Strings Together?

Concatenating two character strings together is most commonly used string operation. SQL Server
2005 allows to concatenate two character strings into a single string with the (+) operator. The
following tutorial exercise shows you some string concatenation examples:
DECLARE @site VARCHAR(40);
SET @site = 'FYIcenter.com';
SELECT 'Welcome to '+@site;
SELECT 'Current date and time is '
+CONVERT(VARCHAR(20), GETDATE());
GO
Welcome to FYIcenter.com
Current date and time is May 19 2007 5:18PM
DECLARE @start INT, @end INT, @total INT;

SET @start = 21;
SET @end = 30;
SET @total = 728;
SELECT 'Search result '
+ CONVERT(VARCHAR(20),@start)
+ ' - '
+ CONVERT(VARCHAR(20),@end)
+ ' of '
+ CONVERT(VARCHAR(20),@total);
GO
Search result 21 - 30 of 728
What Happens When Unicode Strings Concatenate with Non-Unicode Strings?
If a Unicode string NVARCHAR is concatenated with a non-Unicode string VARCHAR, SQL Server
will implicitly convert the non-Unicode string to Unicode string for concatenation.
DECLARE @regcode VARCHAR(40);
DECLARE @unicode NVARCHAR(40);
SET @regcode = 'Some Unicode characters: '
SET @unicode = NCHAR(9733)+NCHAR(9734)+NCHAR(9792)
+NCHAR(9794);
SELECT @regcode + @unicode;
SELECT DATALENGTH(@regcode);
SELECT DATALENGTH(@unicode);
SELECT DATALENGTH(@regcode + @unicode);
Some Unicode characters: ????
25
8
58
Note that the non-Unicode string @regcode has been converted to a Unicode string. The number of
bytes of @regcode changed from 25 to 50. With 8 bytes from @unicode, the number of bytes of the
concatenated string becomes 58.
How To Convert a Unicode Strings to Non-Unicode Strings?
Since Unicode character set is different than code page based (non-Unicode) character set,
converting Unicode strings to non-Unicode strings may result in wrong characters or missing
characters. So you should avoid converting Unicode strings to non-Unicode strings.
If you really want to, there are 3 ways to convert a Unicode string to a non-Unicode string:
 Implicit conversion by assignment operations - When a Unicode string is assigned to a variable, a
column, or a parameter of a non-Unicode string data type, SQL Server will implicitly convert the
Unicode string to a non-Unicode string.
 Explicit conversion using the CAST() function - A Unicode string can be explicitly converted to
non-Unicode string using the CAST(Unicode_string AS VARCHAR(size)) function.
 Explicit conversion using the CONVERT() function - A Unicode string can be explicitly converted
to non-Unicode string using the CONVERT(VARCHAR(size), Unicdoe_string) function.
Some numeric data type conversion examples are provided in the tutorial exercise below:
-- Implicit conversion by an assignment operation
DECLARE @regcode VARCHAR(40);
SET @regcode = N'Some Unicode characters: '
+ NCHAR(9733)+NCHAR(9734)+NCHAR(9792)+NCHAR(9794);
SELECT @regcode;
GO
-- Explicit conversion by CAST()

SELECT CAST(N'Some Unicode characters: '
+ NCHAR(9733)+NCHAR(9734)+NCHAR(9792)+NCHAR(9794)
AS VARCHAR(40));
GO
-- Explicit conversion by CONVERT()

SELECT CONVERT(VARCHAR(40), N'Some Unicode characters: '
+ NCHAR(9733)+NCHAR(9734)+NCHAR(9792)+NCHAR(9794));
GO
What Are the Character String Functions Supported by SQL Server 2005?
SQL Server 2005 supports 23 character string functions:

 ASCII(char) - Returning the code value of a non-Unicode character.
 CHAR(int) - Returning the non-Unicode character of a code value.
 CHARINDEX(word, string, start_location) - Returning the location of the searched "word" in a
string.
 DIFFERENCE(string1, string2) - Returning an integer value that indicates the difference between
the SOUNDEX values of two strings.
 LEFT(string, length) - Returning a substring of "length" character from the left hand side.
 LEN(string) - Returning the number of characters in the string.
 LOWER(string) - Returning the same string with all upper case characters converted to lower
case.
 LTRIM(string) - Returning the same string with leading spaces removed.
 NCHAR(int) - Returning the Unicode character of a code value.
 PATINDEX(pattern, string) - Returning the location of the "pattern" in a string.
 QUOTENAME(string, quote) - Returning the same string enclosed in "quote".
 REPLACE(string, old, new) - Returning the same string with "old" replaced by "new".
 REPLICATE(string, number) - Returning the same string repeated "number" of times.
 REVERSE(string) - Returning the same string with all character positions reversed.
 RIGHT(string, length) - Returning a substring of "length" characters from the right hand side.
 RTRIM(string) - Returning the same string with trailing spaces removed.
 SOUNDEX(string) - Returning a four-character (SOUNDEX) code to evaluate the similarity of two
strings.
 SPACE(length) - Returning a string of "length" space characters.
 STR(number, precision, scale) - Returning a string representation of a numeric value.
 STUFF(string, start, length, new) - Returning the same string with one part being replaced.
 SUBSTRING(string, start, length) - Returning a substring.
 UNICODE(nchar) - Returning the code value of a Unicode character.
 UPPER(string) - Returning the same string with all lower case characters converted to upper
case.
How To Insert New Line Characters into Strings?
If you want to break a string into multiple lines, you need to insert new line characters into the string.
With some client tools like SQL Server Management Studio, it is not so easy to insert a new line
character.
One work around is to use the CHAR(int) function to generated new line character and other special
characters with their code values:
 CHAR(9) - Generates the tab character.
 CHAR(10) - Generates the line feed (new line) character.
 CHAR(13) - Generates the carriage return character.
The tutorial examples below gives you a good example

PRINT 'Welcome to '+CHAR(10)+'FYIcenter.com';
PRINT CHAR(10);
PRINT 'Current date and time is '
+CONVERT(VARCHAR(20), GETDATE());
GO
Welcome to
FYIcenter.com
Current date and time is May 19 2007 7:30PM
How To Locate and Take Substrings with CHARINDEX() and SUBSTRING() Functions?
Transact-SQL is not a language designed for manipulating strings, but it does have two simple
functions to locate and take substrings: CHARINDEX() and SUBSTRING().
The tutorial exercise below assumes two given strings: 'Pages: 18' and 'Words: 3240'. The objective
is to calculate the number of words per page. Read the script below to see how this is done by using
CHARINDEX() and SUBSTRING() functions:
DECLARE @sPages VARCHAR(40), @sWords VARCHAR(40);
SET @sPages = 'Pages: 18';
SET @sWords = 'Words: 3240';
SET @sPages = SUBSTRING(@sPages, CHARINDEX(':', @sPages)+1, 20);
SET @sWords = SUBSTRING(@sWords, CHARINDEX(':', @sWords)+1, 20);
PRINT 'Number of words per page: '
+ CONVERT(VARCHAR(20), CONVERT(INT, @sWords)/CONVERT(INT, @sPages));
GO
Number of words per page: 180
If you are a PHP developer, you can get this done in a much quick way
How To Concatenate Two Binary Strings Together?
SQL Server 2005 allows to concatenate two binary strings into a single string with the (+) operator.
The following tutorial exercise shows you some binary string concatenation examples:
-- Concatenating two binary string literals
SELECT 0x57656C636F6D6520746F20
+ 0x46594963656E7465722E636F6D;
GO
0x57656C636F6D6520746F2046594963656E7465722E636F6D
-- Watch out: This is not a binary string concatenation

SELECT '0x57656C636F6D6520746F20'
+ '0x46594963656E7465722E636F6D';
GO
0x57656C636F6D6520746F200x46594963656E7465722E636F6D
-- Concatenating two binary strings

SELECT CONVERT(VARBINARY(40),'Welcome to ')
+ CONVERT(VARBINARY(40),'FYIcenter.com');
GO
0x57656C636F6D6520746F2046594963656E7465722E636F6D
-- Binary strings can not be concatenated

with character strings
SELECT 'Welcome to '
+ 0x46594963656E7465722E636F6D;
GO
The data types varchar and varbinary are incompatible
in the add operator.
Can Binary Strings Be Used in Arithmetical Operations?
Can binary strings be used in arithmetical operations? The answer is yes. But there are two simple
rules you need to remember:
 If an arithmetical operation has one binary string operand and one integer data type operand, the
binary string operand will be converted to a integer data type to match the other operand. The
operation will be performed as an integer operation.
 A + operator with two binary strings will be performed as binary string concatenation.
 A -, *, or / operator with two binary strings will be performed as binary string concatenation.
The tutorial exercise below shows you some good examples:

SELECT 0x66 + 44
GO
146
SELECT 0x66 - 44
GO
58
SELECT 0x66 * 44
GO
4488
SELECT 0x66 / 44
GO
2
SELECT 0x66 + 0x44

GO
0x6644
SELECT 0x66 - 0x44

GO
Operand data type varbinary is invalid
for subtract operator.
SELECT 0x66 * 0x44

GO
for multiply operator.
SELECT 0x66 / 0x44

GO
for divide operator.
How To Convert Binary Strings into Integers?
Binary strings and integers are convertible implicitly and explicitly. But there several rules you need to
remember:
 Binary strings will be implicitly converted into an integer data type, if it is involved in an
arithmetical operation with another integer data type operand.
 Binary strings will be implicitly converted into an integer data type, if it is assigned to a variable, a
column or a parameter of an integer data type.
 Binary strings will be explicitly converted into integer data types using CAST() and CONVERT()
functions.
 When converting binary strings that have more bytes than the target data type size, bytes on the
left hand side will be truncated.
 When converting binary strings that have less bytes than the target data type size, 0x00 will be
padded on the left hand side.
Examples showing in the tutorial exercise below will help you remembering those rules.
SELECT 0x66 + 44
GO
146
DECLARE @integer INT;

SET @integer = 0x66;
SELECT @integer + 44
GO
146
SELECT CAST(0x66 AS INT) + 44

GO
146
SELECT CONVERT(INT, 0x66) + 44

GO
146
-- Only last 4 bytes are used for INT conversion

SELECT 0x7700000066 + 44
GO
146
-- 8 bytes will be used for BIGINT conversion

SELECT 0x7700000066 + CONVERT(BIGINT,44)
GO
511101108370
-- 0x00 padded on the left hand side

SELECT 0x66 + 44
SELECT 0x00000066 + 44
GO
146
146
Can Binary Strings Be Converted into NUMERIC or FLOAT Data Types?
Can binary strings be converted into numeric or float data types? The answer is no. Binary strings
can not be converted implicitly or explicitly into NUMERIC, DECIMAL, REAL, or FLOAT data types.
The tutorial exercise gives you some examples of errors when converting binary strings to NUMERIC
or FLOAT data types:
-- Implicit conversion to NUMERIC
SELECT 0x66 + 0.44;
GO
Error converting data type varbinary to numeric.
-- Explicit conversion to NUMERIC

SELECT CONVERT(NUMERIC(9,2), 0x66) + 0.44;
GO
Error converting data type varbinary to numeric.
-- Implicit conversion to REAL

DECLARE @real REAL;
SET @real = 0x66;
Operand type clash: varbinary is incompatible with real
-- Implicit conversion to FLOAT

DECLARE @float FLOAT(53);
SET @float = 0x66;
Operand type clash: varbinary is incompatible with float
Can Binary Strings Be Converted into Character Strings?
Binary strings and character strings are convertible. But there are several rules you need to
remember:
 Binary strings can be converted implicitly to character strings by assignment operations.
 Binary strings can not be converted implicitly to character strings by concatenation operations.
 Binary strings can be converted explicitly to character strings by CAST() and CONVERT()
functions.
 Character strings can not be converted implicitly to binary strings by assignment operations.
 Character strings can not be converted implicitly to binary strings by concatenation operations.
 Character strings can be converted explicitly to binary strings by CAST() and CONVERT()
functions.
For examples, see the tutorial exercise below:

-- Implicit conversion to character strings
DECLARE @char_string VARCHAR(40);
SET @char_string = 0x46594963656E7465722E636F6D;
SELECT @char_string;
GO
FYIcenter.com
-- Implicit conversion to binary strings

DECLARE @binary_string VARBINARY(40);
SET @binary_string = 'FYIcenter.com';
GO
Implicit conversion from data type varchar to varbinary is
not allowed. Use the CONVERT function to run this query.
-- Explicit conversion to binary strings

SET @binary_string = CONVERT(VARBINARY(40),'FYIcenter.com');
SELECT @binary_string;
GO
0x46594963656E7465722E636F6D
-- Implicit conversion in concatenation operation

with character strings
SELECT 'Welcome to '
+ 0x46594963656E7465722E636F6D;
GO
The data types varchar and varbinary are incompatible
in the add operator.
Can Binary Strings Be Converted into Unicode Character Strings?
Can binary strings be converted into Unicode character strings? The answer is yes. But you need to
know how Unicode characters are represented in a binary format. Remember the following simple
rules:
 An ASCII character is stored in 1 byte.
 A Unicode character is stored in 2 bytes.
 When an ASCII character is mapped to a Unicode character, the first type equals to the same
code value as the ASCII character, and the second type equals to 0x00.
The tutorial exercise below shows you some interesting conversion behaviors:
-- 1. Binary representation of ASCII characters
SELECT CONVERT(VARBINARY(40),'FYIcenter.com');
GO
0x46594963656E7465722E636F6D
-- 2. Binary representation of Unicdode characters

SELECT CONVERT(VARBINARY(40),N'FYIcenter.com');
GO
0x460059004900630065006E007400650072002E0063006F006D00
-- 3. Converting a binary string back to Unicode characters

SELECT CONVERT(NVARCHAR(40),
0x460059004900630065006E007400650072002E0063006F006D00);
GO
FYIcenter.com
-- 4. Converting a binary string back to ASCII characters

SELECT CONVERT(VARCHAR(40),
0x460059004900630065006E007400650072002E0063006F006D00);
GO
F
-- 5. Converting a binary string back to Unicode characters

SELECT CONVERT(NVARCHAR(40),0x46594963656E7465722E636F6D);
GO
??????m
-- 6. Converting a binary string back to ASCII characters

SELECT CONVERT(VARCHAR(40),0x46594963656E7465722E636F6D);
GO
FYIcenter.com
Test 4 seems to be giving a wrong result. It should be something like "F Y I c e n t e r . c o m ". But
the client tool does not know how to display the 0x00 ASCII character and stops showing the rest of
the characters.
Result of test 5 shows real characters on the SQL Server Management Studio windows. But they can
not be represented on this Web page, because it is not set to handle Unicode characters. This is why
you see (?) characters.
How To Convert Binary Strings into Hexadecimal Character Strings

When a query returns a binary string value to the client tool, it will be displayed by the client tool as
hexadecimal digits in a character string format. But they are hexadecimal digits on the display
window, not in the database storage.
Converting binary strings into hexadecimal character strings in storage is not an easy task, since
there is not built-in functions to help you. You need to write a conversion algorithm yourself. This
tutorial exercise, called bin2hex, shows you a simple example:
-- bin2hex
DECLARE @hex_string VARCHAR(40);
DECLARE @position INT; -- Loop index
DECLARE @char INT; -- Character code value
DECLARE @half_char INT; -- Character code half value
-- Initial values
SET @binary_string = CONVERT(VARBINARY(40),'FYIcenter.com');
SET @hex_string = ''; -- Output hex strings
SET @position = 1;
-- Loop on each byte

WHILE @position <= DATALENGTH(@binary_string) BEGIN
SET @char
= ASCII(SUBSTRING(@binary_string, @position, 1));
-- Taking the first half byte

SET @half_char = @char / 16; -- left half
IF @half_char < 10
SET @hex_string = @hex_string
+ CHAR(ASCII('0') + @half_char);
ELSE
+ CHAR(ASCII('A') + @half_char-10);
-- Taking the second half byte

SET @half_char = @char % 16; -- left half
IF @half_char < 10
+ CHAR(ASCII('0') + @half_char);
ELSE
+ CHAR(ASCII('A') + @half_char-10);
-- Going the next iteration

SET @position = @position+1;
END
SELECT @hex_string;
GO
46594963656E7465722E636F6D
What Are Bitwise Operations?
Bitwise operations are binary operations performed on one or two binary strings. SQL Server
supports 4 bitwise operations:
 & (Bitwise AND) - Performing the single bit Boolean operation "AND" on each bit position.
 | (Bitwise OR) - Performing the single bit Boolean operation "OR" on each bit position.
 ^ (Bitwise XOR) - Performing the single bit Boolean operation "XOR" on each bit position.
 ~ (Bitwise NOT) - Performing the single bit Boolean operation "NOT" on each bit position.
The table below shows you how single bit Boolean operations work:
Input 1: 0 0 1 1
Input 2: 0 1 0 1
----------------
& (AND) 0 0 0 1
Input 1: 0 0 1 1
Input 2: 0 1 0 1
----------------
| (OR) 0 1 1 1
Input 1: 0 0 1 1
Input 2: 0 1 0 1
----------------
^ (XOR) 0 1 1 0
Input 1: 0 0 1 1
----------------
~ (NOT) 1 1 0 0
How To Add or Remove Days on Date and Time Values?
SQL Server 2005 only support two operators on data and time values:
 + (Add): Adding days to a date and time value.
 - (Subtract): Removing days from a date and time value.
Note that to add or subtract days on a date and time value, the other operand must be a numeric
value: an integer or a decimal number.
Here are some good examples of adding days to and subtracting days from date and time values:
-- Leap year is supported
DECLARE @birth_date DATETIME;
SET @birth_date = '29-Feb-2000 03:00:00.000';
SELECT @birth_date + 1;
SELECT @birth_date - 1;
GO
2000-03-01 03:00:00.000
2000-02-28 03:00:00.000
-- Partial days are supported

SELECT @birth_date + 0.5; -- adding a half day
SELECT @birth_date + 1.0/3;
GO
2000-03-01 15:00:00.000
2000-02-29 10:59:59.970
Can Date and Time Values Be Converted into Integers?
Can date and time values be converted into integers? The answer is yes. The resulting integer will be
the number days relative the base date: Jan 1, 1900. The time of the day will be rounded to days. But
there several rules you need to follow:
 Date and time values can not be converted to integers implicitly using assignment operations.
 Date and time values can be converted to integers explicitly using CAST() or CONVERT()
functions.
Here are some good examples:

-- Implicit conversion is not allowed
DECLARE @days INT;
SET @days = @birth_date;
GO
Implicit conversion from data type datetime to int is
-- Explicit conversion is supported

SELECT CONVERT(INT, @birth_date);
GO
36583
-- 36583 days since the base date?

SET @birth_date = '01-Jan-1900 00:00:00.000';
SELECT @birth_date + 36583;
GO
2000-02-29 00:00:00.000
-- Time of the day will be rounded to days
SELECT CONVERT(INT, @birth_date);
GO
36584
Can Integers Be Converted into Date and Time Values?
Can integers be converted into date and time values? The answer is yes. The input integer will be
used as the number of days relative to the base date: Jan 1, 1900. The time of the day will be set to
0, midnight of the day. But there several rules you need to follow:
 Integers can be converted to date and time values implicitly using assignment operations.
 Integers can be converted to date and time values explicitly using CAST() or CONVERT()
functions.
Here are some good examples:

SET @birth_date = CONVERT(DATETIME, 36583);
SELECT @birth_date;
GO
2000-02-29 00:00:00.000

SET @birth_date = CONVERT(DATETIME, 36583);
SELECT @birth_date;
GO
2000-02-29 00:00:00.000
Are DATETIME and NUMERIC Values Convertible?
Are datetime and numeric value convertible? The answer is yes. Here are the main rules on
DATATIME and NUMERIC value conversions:
 During the conversion a DATETIME value will be treated as a NUMERIC value with the number
of days relative the base date, Jan 1, 1900 being the integer part, and the time of the day being
the decimal part.
 DATETIME values can not be converted to NUMERIC values implicitly using assignment
operations.
 NUMERIC values can be converted to DATETIME values implicitly using assignment operations.
 DATETIME values and DATETIME values can be converted to each other explicitly using CAST()
or CONVERT() functions.

-- Implicit conversion NUMERIC to DATETIME
SET @birth_date = 36583.25;
SELECT @birth_date;
GO
2000-02-29 06:00:00.000
-- Explicit conversion NUMERIC to DATETIME

SET @birth_date = CONVERT(DATETIME, 36583.75);
SELECT @birth_date;
GO
2000-02-29 18:00:00.000
-- Implicit conversion DATETIME to NUMERIC

DECLARE @birth_days NUMERIC(9,2);
SET @birth_date = '2000-02-29 06:00:00.000';
SET @birth_days = @birth_date;
GO
Implicit conversion from data type datetime to numeric is
-- Explicit conversion DATETIME to NUMERIC

DECLARE @birth_days NUMERIC(9,2);
SET @birth_date = '2000-02-29 18:00:00.000';
SET @birth_days = CONVERT(NUMERIC(9,2), @birth_date);
SELECT @birth_days;
GO
36583.75
Can a DATETIME Value Be Subtracted from Another DATETIME Value?
Can a datetime value be subtracted from another datetime value? The answer is yes. The subtract
operation can be performed by the subtract operator (-) as:
 datetime1 - datetime2: Returning a DATETIME value calculated as CONVERT(DATETIME,
CONVERT(NUMERIC(18,9),datetime1) - CONVERT(NUMERIC(18,9),datetime2)).

DECLARE @today_date DATETIME;
SET @birth_date = '2000-02-29 16:00:00.000';
SET @today_date = '2007-05-19 22:00:00.000';
SELECT @today_date - @birth_date;
GO
1907-03-22 06:00:00.000

DECLARE @today_date DATETIME;
DECLARE @age NUMERIC(9,2);
SET @birth_date = '2000-02-29 16:00:00.000';
SET @today_date = '2007-05-19 22:00:00.000';
SET @age = CONVERT(NUMERIC(9,2), @today_date - @birth_date);
SELECT @age;
GO
2636.25
What Are the Date and Time Functions Supported by SQL Server 2005?
SQL Server 2005 supports n character string functions:

 DATEADD(datepart, number, date) - Returning the same date and time value with one part
added with the "number".
 DATEDIFF(datepart, startdate, enddate) - Returning an integer representing the difference of the
specified part between two date and time values.
 DATENAME(datepart, date) - Returning the name string of the specified part of the input date.
 DATEPART(datepart, date) - Returning an integer representing the specified part of the input
date.
 DAY(date) - Returning an integer representing the day part of the input date.
 GETDATE() - Returning a date and time value representing the current date and time.
 GETUTCDATE() - Returning a date and time value representing the current date and time in UTC
(Coordinated Universal Time or Greenwich Mean Time).
 MONTH(date) - Returning an integer representing the month part of the input date.
 YEAR(date) - Returning an integer representing the year part of the input date.
 How To Increment or Decrement Parts of DATETIME Values?

 If you want to increment or decrement one part of a date and time value, you can use the
DATEADD() function in the following format:
 DATEADD(datepart, number, date) returns DATETIME:
 "date" - the input date
 "number" - the incrementing amount
 "datepart" - one of predefined date part names
 Valid date part names are:
 year, yy, yyyy - The year part
 quarter, qq, q - The quarter part
 month, mm, m - The month part
 dayofyear, dy, y - The day-of-year part
 day, dd, d - The day part
 week, wk, ww - The week part
 weekday, dw, w - The weekday part
 hour, hh - The hour part
 minute, mi, n - The minute part
 second, ss, s - The second part
 millisecond, ms - The millisecond part
 For example, DATEADD(year, 1, @birth_date) will increment 1 year to @birth_date.
 How To Use DATEADD() Function?
 DATEADD() is a very useful function for manipulating date and time values. The following tutorial
exercise shows you some good DATEADD() usage examples:
 -- Incrementing 1 year on a leap date
 DECLARE @birth_date DATETIME;
 SET @birth_date = '2000-02-29 16:00:00.000';
 SET @birth_date = DATEADD(year, 1, @birth_date);
 SELECT @birth_date;
 GO
 2001-02-28 16:00:00.000

 -- Decrementing 1 month on a leap date
 SET @birth_date = '2000-02-29 16:00:00.000';
 SET @birth_date = DATEADD(month, -1, @birth_date);
 GO
 2000-01-29 16:00:00.000

 -- Incrementing 2 milliseconds
 SET @birth_date = '2000-02-29 16:00:00.000';
 SET @birth_date = DATEADD(millisecond, 2, @birth_date);
 GO
 2000-02-29 16:00:00.003
 How To Calculate DATETIME Value Differences Using the DATEDIFF() Function?
 If you want to calculate the difference between two date and time values, you can use the
DATEDIFF() function in the following format:
 DATEDIFF(datepart, startdate, enddate) returns INT:
 "startdate" - the start date
 "enddate" - the end date
 See other tutorials for the list of names
 Note that SQL Server manual says DATEDIFF() returns the number of boundaries crossed by the
specified part between two specified dates. This can be interpreted as truncating all parts smaller
than the specified part on both input date and time values, then calculating the difference. The
following tutorial exercise shows you some good DATEDIFF() usage examples:
 -- Difference in days - truncating all parts
 -- smaller than day
 DECLARE @today_date DATETIME;
 SET @birth_date = '2000-02-29 16:10:00.000';
 SET @today_date = '2000-03-02 18:30:00.000';
 SELECT DATEDIFF(day, @birth_date, @today_date);
 GO
 2

 -- Difference in hours - truncating all parts
 -- smaller than hour
 DECLARE @today_date DATETIME;
 SET @birth_date = '2000-02-29 16:10:00.000';
 SET @today_date = '2000-03-02 18:30:00.000';
 SELECT DATEDIFF(hour, @birth_date, @today_date);
 GO
 50
How To Calculate Age in Days, Hours and Minutes?
On many Web sites, news items or blog articles are displayed with a publishing data and time
represented in days, hours and minutes. To do this you can use the DATEDIFF() function with a
simple algorithm as shown in the tutorial exercise below:
-- Age in days
DECLARE @submit_date DATETIME;
SET @submit_date = '2007-05-11 16:10:41.403';
SELECT 'Posted '
+ CONVERT(VARCHAR(40),
DATEDIFF(day, @submit_date, GETDATE()))
+ ' days ago.';
GO
Posted 8 days ago.
-- Age in days, hours and minutes - Wrong

SET @submit_date = '2007-05-11 16:10:41.403';
SELECT 'Posted '
DATEDIFF(day, @submit_date, GETDATE()))
+ ' days, '
DATEDIFF(hour, @submit_date, GETDATE()))
+ ' hours, and '
DATEDIFF(minute, @submit_date, GETDATE()))
+ ' minutes ago.';
GO
Posted 7 days, 186 hours, and 11202 minutes ago.
-- Age in days, hours and minutes - Correct

SET @submit_date = '2007-05-11 16:10:41.403';
SELECT 'Posted '
DATEDIFF(minute, @submit_date, GETDATE())/(24*60))
+ ' days, '
DATEDIFF(minute, @submit_date, GETDATE())%(24*60)/60)
+ ' hours, and '
DATEDIFF(minute, @submit_date, GETDATE())%60)
+ ' minutes ago.';
GO
Posted 7 days, 18 hours, and 42 minutes ago.
Notice how the correct number of days, hours, and minutes are calculated using the concepts of
whole day, remaining partial day, whole hour, and remaining partial hour as shown below:
|...................|...................|...+...+..
|<------- number of whole days -------->|<------->|

remaining partial day ---^
|<----------- number of whole hours ----------->|-|

remaining partial hour ---^
|<----------- total number of minutes ----------->|

 Number of whole days: DATEDIFF(minute, @submit_date, GETDATE())/(24*60), where 24*60 is
the number of minutes in a whole day.
 Number of minutes in the remaining partial day: DATEDIFF(minute, @submit_date, GETDATE())
%(24*60).
 Number of minutes in the remaining partial hour: DATEDIFF(minute, @submit_date, GETDATE())
%60.
 How To Get Month and Weekday Names from DATETIME Values?

 In DATETIME values, the month part and weekday part have names. You can use the
DATENAME() function to get the month name and weekday name from a DATETIME value in the
following format:
 DATENAME(datepart, date) returns NVARCHAR:
 "datepart" - "month" or "weekday"
 Here is a simple example:
 DECLARE @submit_date DATETIME;
 SET @submit_date = '2007-05-19 16:10:41.403';
 SELECT 'Posted in the month of '
 + DATENAME(month, @submit_date)
 + ' on a '+DATENAME(weekday, @submit_date);
 GO
 Posted in the month of May on a Saturday
 How To Get Parts of DATETIME Values as Integers?
 Sometimes, you want to one specific part of a DATETIME value as an integer to be used in your
own date and time calculations. This can be done by using the DATEPART() function in the
following format:
 DATENAME(datepart, date) returns INT:
 See other tutorials for the list of names
 In following tutorial example shows you how display a birth date in a format of yyyy.m.d:
 SET @birth_date = '1987-05-19 16:10:41.403';
 SELECT 'You were born on '
 + CONVERT(VARCHAR(10),DATEPART(year, @birth_date))
 + '.'
 + CONVERT(VARCHAR(10),DATEPART(month, @birth_date))
 + '.'
 + CONVERT(VARCHAR(10),DATEPART(day, @birth_date));
 GO
 You were born on 1987.5.19
 How To Get Year, Month and Day Out of DATETIME Values?
 You can use DATEPART() to get any part of the DATETIME value. But to get year, month and
day, you can use 3 specific functions: YEAR(), MONTH() and DAY(). These functions are
equivalent to DATEPART() as:
 YEAR(date) = DATENAME(year, date)
 MONTH(date) = DATENAME(month, date)
 DAY(date) = DATENAME(day, date)
 In following tutorial example shows you how display a birth date in a format of yyyy.m.d:
 SET @birth_date = '1987-05-19 16:10:41.403';
 + CONVERT(VARCHAR(10),YEAR(@birth_date))
 + '.'
 + CONVERT(VARCHAR(10),MONTH(@birth_date))
 + '.'
 + CONVERT(VARCHAR(10),DAY(@birth_date));
 GO
 You were born on 1987.5.19
 What Is the Difference Between GETDATE() and GETUTCDATE()?
 The difference between GETDATE() and GETUTCDATE() is time zone number of the SQL
Server machine. The tutorial exercise below gives you a good example:
 DECLARE @local_time DATETIME;
 DECLARE @gmt_time DATETIME;
 SET @local_time = GETDATE();
 SET @gmt_time = GETUTCDATE();
 SELECT 'Server local time: '
 + CONVERT(VARCHAR(40),@local_time);
 SELECT 'Server GMT time: '
 + CONVERT(VARCHAR(40),@gmt_time);
 SELECT 'Server time zone: '
 + CONVERT(VARCHAR(40),
 DATEDIFF(hour,@gmt_time,@local_time));
 GO
 Server local time: Jun 2 2007 10:06PM
 Server GMT time: Jun 2 2007 4:21PM
 Server time zone: 6
 Note that the local time is from SQL server machine, not your local machine if you are connected
to the server remotely. The local time is also adjusted by day-light savings.
 How To Format Time Zone in +/-hh:mm Format?
 From the previous tutorial, you know how to calculate the time zone value by using GETDATE()
and GETUTCDATE() functions. But how can you format that time zone value in a nice looking
format like +05:45 or -04:00?
 Unfortunately, SQL Server 2005 offers no built-in functions to convert a date and time value in
your own specific format. You have to do this by yourself as shown in the tutorial exercise:
 -- time_zone_formatter
 DECLARE @gmt_time DATETIME;
 SET @local_time = 'May 19 2007 10:06PM';
 SET @gmt_time = 'May 19 2007 4:21PM';
 SELECT 'Server local time: '
 + CONVERT(VARCHAR(40),@local_time);
 SELECT 'Server GMT time: '
 + CONVERT(VARCHAR(40),@gmt_time);
 SELECT 'Server time zone: '
 + CASE
 WHEN DATEDIFF(minute,@gmt_time,@local_time)>= 0 THEN
 '+'
 ELSE
 '-'
 END
 + RIGHT(CONVERT(VARCHAR(40),
 100+ABS(DATEDIFF(minute,@gmt_time,@local_time)/60)),
 2)
 + ':'
 + RIGHT(CONVERT(VARCHAR(40),
 100+ABS(DATEDIFF(minute,@gmt_time,@local_time)%60)),
 2);
 GO
 Server local time: Jun 2 2007 10:06PM
 Server GMT time: Jun 2 2007 4:21PM
 Server time zone: +05:45
 By the way, if you are wondering which part of the world has a time zone of +05:45. It is in the
area of Kathmandu, capital city of Nepal.
 How To Format DATETIME Values to Strings with the CONVERT() Function?
 SQL Server 2005 offers no functions to format DATETIME values in your own format patterns.
But it does provide you a number of pre-defined format patterns that you can use with the
CONVERT(char_type, date, pattern_code) function. Commonly used pattern codes are:
 Code Name Format
 100 Default mon dd yyyy hh:miAM/PM
 101 U.S. mm/dd/yyyy
 102 ANSI yyyy.mm.dd
 103 British/French dd/mm/yyyy
 104 German dd.mm.yy
 105 Italian dd-mm-yy
 107 Mon dd, yyyy
 108 hh:mi:ss
 110 USA mm-dd-yy
 111 Japan yy/mm/dd
 121 ODBC canonical yyyy-mm-dd hh:mi:ss.mmm
 For examples, see the tutorial exercise below:
 SET @birth_date = '1987-05-19 16:10:41.403';
 + CONVERT(VARCHAR(40),@birth_date,107)
 + ', at '
 + CONVERT(VARCHAR(40),@birth_date,108);
 GO
 You were born on May 19, 1987, at 16:10:41
 How To Truncate DATETIME Values to Dates without Time?
 Assuming that you have some date and time, DATETIME, values, and you want to work with
dates only. You want to truncate them to dates without time, or with time of 00:00:00.000.
 Again, SQL Server offers no simple solution, you need to write the formula yourself. The tutorial
exercise below shows you how to do this with CONVERT() and FLOOR() functions:
 -- truncating to dates
 SET @local_time = 'May 19 2007 10:06PM';
 SELECT 'Server local date and time: ';
 SELECT @local_time;
 SELECT 'Server local date without time: ';
 SELECT CONVERT(DATETIME,
 FLOOR(CONVERT(NUMERIC(18,9),@local_time)));
 GO
 Server local date and time:
 2007-05-19 22:06:00.000
 Server local date without time:
 2007-05-19 00:00:00.000
 How To Set Different Parts of a DATETIME Value?
 In SQL Server, you can get different parts of a DATETIME value with the DATEPART() functions.
But there is no function that allows you to set different parts to a DATETIME value. For example,
you a date_of_birth column as DATETIME in a table, you want to set all rows to the noon time in
this column.
 The tutorial exercise below shows how to set year, month, day, hour, minute and second to any
give DATETIME value:
 -- set parts to a DATETIME value DECLARE @date_of_birth DATETIME; SET @date_of_birth =
GETDATE(); SET @date_of_birth = DATEADD(YEAR, 1987-
DATEPART(YEAR,@date_of_birth), @date_of_birth); SET @date_of_birth =
DATEADD(MONTH, 05-DATEPART(MONTH,@date_of_birth), @date_of_birth); SET
@date_of_birth = DATEADD(DAY, 19-DATEPART(DAY,@date_of_birth), @date_of_birth); SET
@date_of_birth = DATEADD(HOUR, 12-DATEPART(HOUR,@date_of_birth), @date_of_birth);
SET @date_of_birth = DATEADD(MINUTE, 00-DATEPART(MINUTE,@date_of_birth),
@date_of_birth); SET @date_of_birth = DATEADD(SECOND, 00-
DATEPART(SECOND,@date_of_birth), @date_of_birth); SELECT 'You were born on ' +
CONVERT(VARCHAR(40),@date_of_birth,107) + ', at ' +
CONVERT(VARCHAR(40),@date_of_birth,108); GO You were born on May 19, 1987, at
12:00:00
 What Is a Boolean Value?
 A Boolean value indicates a condition in a state of TRUE or FALSE. In Transact-SQL language,
there is not storage data type defined to store a Boolean value. So Boolean values can only exist
temporarily as part of the execution of Transact-SQL expression or statement.
 Boolean values are returned from comparison operations. The tutorial exercise below shows you
some examples of how Boolean values are used:
 -- Boolean values in a CASE expression
 SELECT CASE WHEN 1>0 THEN
 'True'
 ELSE
 'False'
 END;
 GO
 True

 -- Boolean values in an IF statement
 IF 1>0
 SELECT 'True';
 ELSE
 SELECT 'False';
 GO
 True

 -- Stand alone Boolean value not allowed
 SELECT 1>0;
 GO
 Incorrect syntax near '>'.
 What Are Conditional Expressions?
 A conditional expression returns one of the given expressions based a specific condition. SQL
Server 2005 offers the CASE operator to present a conditional expression with two syntaxes:
 1. CASE with simple conditions
 CASE test_value
 WHEN value_1 THEN expression_1
 WHEN value_2 THEN expression_2
 ...
 WHEN value_n THEN expression_n
 ELSE expression_o
 END
 -- Returns "expression_x"
 if "test_value" equals to "value_x".

 2. CASE with complex conditions
 CASE
 WHEN condition_1 THEN expression_1
 WHEN condition_2 THEN expression_2
 ...
 WHEN condition_n THEN expression_n
 ELSE expression_o
 END
 -- Returns "expression_x"
 if "condition_x" is TRUE.
 Here are two examples on how to use the CASE operator:
 DECLARE @command VARCHAR(10);
 SET @command = 'S';
 SELECT CASE @command
 WHEN 'A' THEN 'Add'
 WHEN 'S' THEN 'Save'
 WHEN 'Q' THEN 'Quit'
 ELSE 'Unknown command.'
 END;
 GO
 Save

 DECLARE @command VARCHAR(10);
 SET @command = 'Q';
 SELECT CASE
 WHEN @command = 'A' THEN 'Add'
 WHEN @command = 'S' THEN 'Save'
 WHEN @command = 'Q' THEN 'Quit'
 ELSE 'Unknown command.'
 END;
 GO
 Quit
What Are Comparison Operations?
Comparison operations return Boolean values by comparing the relative positions of two operands.
SQL server supports the following comparison operations:
 = (Equal To): Returns Boolean true, if two operands are equal in value.
 > (Greater Than): Returns Boolean true, if the left operand has a value greater than the right
operand.
 < (Less Than): Returns Boolean true, if the left operand has a value less than the right operand.
 >= (Greater Than or Equal To): Returns Boolean true, if the left operand has a value greater than
or equal to the right operand.
 <= (Less Than or Equal To): Returns Boolean true, if the left operand has a value less than or
equal to the right operand.
 <> (Not Equal To): Returns Boolean true, if two operands are not equal in value.
There are several rules you need to remember about comparison operations:
 Comparison operations can be performed on all data types: exact numbers, approximate
numbers, character strings, and binary strings.
 If two operands have different data types, the operand with a lower data type rank will be
implicitly converted to the data type of the other operand.
 Character string comparison will be performed based the rules defined in the associated collation.
 How To Perform Comparison on Exact Numbers?
 Comparison operations on exact numbers of data types: BIT, INT, NUMERIC, etc., are very easy
to understand. Here are some simple examples:
 -- BIT value comparison
 DECLARE @x BIT, @y BIT;
 SET @x = 0;
 SET @y = 1;
 SELECT CASE WHEN @x > @y THEN 'True' ELSE 'False' END;
 GO
 False

 DECLARE @x INT, @y INT;
 SET @x = -5;
 SET @y = 9;
 GO
 False

 DECLARE @x NUMERIC(9,2), @y NUMERIC(9,2);
 SET @x = -5.25;
 SET @y = -5.15;
 GO
 False
 How To Perform Comparison on Floating Point Numbers?
 Comparison operations on approximate (floating point) numbers are also easy to understand.
Just watch out rounding operations performed during conversions. Here are two examples of
floating point number comparisons:
 -- Rounding during implicit conversion makes them equal
 DECLARE @x FLOAT(24), @y FLOAT(24);
 SET @x = 0.999999999E+00;
 SET @y = 0.99999999E+00;
 SELECT CASE WHEN @x <> @y THEN 'True' ELSE 'False' END;
 GO
 False

 -- Floating point number literals have double precision
 -- So no rounding
 SELECT CASE WHEN 0.999999999E+00 <> 0.99999999E+00 THEN
 'True'
 ELSE
 'False'
 END;
 GO
 True
 How To Perform Comparison on Date and Time Values?
 Comparison operations on date and time values can be performed in the same way as numeric
values. The comparison rule is simple: a date and time value in the future has a greater value
than a date and time value in the past. The tutorial exercise below shows you a good example:
 -- Future date has a greater value
 DECLARE @his_birth_date DATETIME, @her_birth_date DATETIME;
 SET @his_birth_date = '1987-05-19';
 SET @her_birth_date = '1988-07-07';
 SELECT CASE WHEN @his_birth_date < @her_birth_date THEN
 'He is older.'
 ELSE
 'She is older.'
 END;
 GO
 He is older.
 How To Perform Comparison on Character Strings?
 Comparison operations on character strings are performed based on the associated collation.
Each collation defines rules on how characters are ordered, how character cases and accents are
treated, etc. The tutorial exercise below shows you some character string comparison examples
using the default collation: SQL_Latin1_General_CP1_CI_AS.
 -- Case insensitive
 DECLARE @x VARCHAR(40), @y VARCHAR(40);
 SET @x = 'FYIcenter.com';
 SET @y = 'fyicenter.COM';
 SELECT CASE WHEN @x = @y THEN 'True' ELSE 'False' END;
 GO
 True

 -- digits has less values than letters
 SET @x = '1234';
 SET @y = 'abcd';
 SELECT CASE WHEN @x < @y THEN 'True' ELSE 'False' END;
 GO
 True

 -- Trailing spaces are ignored
 SET @x = 'FYIcenter.com';
 SET @y = 'fyicenter.COM ';
 SELECT CASE WHEN @x = @y THEN 'True' ELSE 'False' END;
 GO
 True

 -- Longer strings have greater values
 SET @x = 'abcd';
 SET @y = 'abcde';
 SELECT CASE WHEN @x < @y THEN 'True' ELSE 'False' END;
 GO
 True
 What To Test Value Ranges with the BETWEEN Operator?
 Sometimes you want to compare a value against a value range. You can do this with two regular
comparison operations. But you can also use the special comparison operator BETWEEN to get
it done with the following syntaxes:
 1. Inclusively in the range test
 test_value BETWEEN range_start_value AND range_end_value

 - Returns the same results as the following expression
 test_value >= range_start_value
 AND test_value <= range_end_value

 2. Exclusively out of the range test
 test_value NOT BETWEEN range_start_value AND range_end_value

 test_value < range_start_value
 OR test_value > range_end_value
 Here are two examples of using the BETWEEN operator:
 DECLARE @my_age INT;
 SET @my_age = 17;
 SELECT CASE WHEN
 @my_age BETWEEN 11 AND 19
 THEN 'You are a teenager.'
 ELSE 'You are not a teenager.'
 END;
 GO
 You are a teenager.

 DECLARE @my_age INT;
 SET @my_age = 27;
 @my_age NOT BETWEEN 11 AND 19
 THEN 'You are not a teenager.'
 ELSE 'You are a teenager.'
 END;
 GO
 You are not a teenager.
 What To Test Value Lists with the IN Operator?
 Sometimes you want to test a given value against a list of values. You can do this in a loop with
the regular "equal" operator. But you can also use the special comparison operator IN to get it
done with the following syntaxes:
 1. In-the-list test
 test_value IN (value_1, value_2, ..., value_n)

 test_value = value_1
 OR test_value = value_2
 ...
 OR test_value = value_n

 2. Not-in-the-list test
 test_value NOT IN (value_1, value_2, ..., value_n)

 test_value <> value_1
 AND test_value <> value_2
 ...
 AND test_value <> value_n
 Here are two examples of using the IN operator:
 DECLARE @my_number INT;
 SET @my_number = 767;
 @my_number IN (525, 272, 532, 767, 150, 637)
 THEN 'You are a winner.'
 ELSE 'You are not a winner.'
 END;
 GO
 You are a winner.

 DECLARE @my_number INT;
 SET @my_number = 676;
 @my_number NOT IN (525, 272, 532, 767, 150, 637)
 THEN 'You are not a winner.'
 ELSE 'You are a winner.'
 END;
 GO
 You are not a winner.
What To Perform Pattern Match with the LIKE Operator?
Pattern match is a very important operation for search records base on character string columns.
SQL Server 2005 offers the LIKE operator to perform pattern match operations in two formats:
target_string LIKE pattern
-- Returns TRUE
if the target string matches the pattern
target_string NOT LIKE pattern

-- Returns TRUE
if the target string does not match the pattern
Pattern match is a powerful operation. But you need to remember several rules:
 Pattern may contain predefined wildcard characters, similar to Unix Regular Expressions. But
they are not the same.
 '%' is a wildcard character that matches any string of zero or more characters.
 '_' is a wildcard character that matches any single character.
 '_' is a wildcard character that matches any single character.
 '[abc]' is a wildcard character that matches any character listed inside the brackets.
 '[a-c]' is a wildcard character that matches any character in the range defined in the brackets.
 '[^abc]' is a wildcard character that matches any character not listed inside the brackets.
 '[^a-c]' is a wildcard character that matches any character not in the range defined in the brackets.
Here is a simple example of LIKE operator:
SELECT CASE WHEN
'FYIcenter.com' LIKE 'FYI%'
THEN 'Pattern matched.'
ELSE 'Pattern not matched.'
END;
GO
Pattern matched.
How To Use Wildcard Characters in LIKE Operations?
Wildcard character '%' can be used in the pattern string for the LIKE operator to match any string of
zero or more characters. The following example uses '%Sport% Store' to search all company names
that has a partial word 'Sport' and ended with the word 'Store'. The sample database
AdventureWorksLT provided by Microsoft is used.
USE adventureWorksLT
GO
SELECT c.CompanyName FROM SalesLT.Customer c

WHERE c.CompanyName LIKE '%Sport% Store'
GO
CompanyName
-------------------------
Specialty Sports Store
Camping and Sports Store
Vigorous Sports Store
Our Sporting Goods Store
Sports Store
Sports Products Store
How To Use Wildcard Characters in LIKE Operations?
Wildcard character '%' can be used in the pattern string for the LIKE operator to match any string of
zero or more characters. The following example uses '%Sport% Store' to search all company names
that has a partial word 'Sport' and ended with the word 'Store'. The sample database
GO
SELECT c.CompanyName FROM SalesLT.Customer c

WHERE c.CompanyName LIKE '%Sport% Store'
GO
CompanyName
-------------------------
Specialty Sports Store
Camping and Sports Store
Vigorous Sports Store
Our Sporting Goods Store
Sports Store
Sports Products Store
How To Test Values Returned by a Subquery with the IN Operator?
Normally, the comparison operator IN is used against a list of specifie values as in the format of:
"test_value IN (value_1, value_2, ..., value_n)". But you can also replace the list of values by a
subquery as the following formats:
test_value IN (SELECT column FROM ...)
-- Returns TRUE if the test_value equals to one of
the values returned from the subquery
test_value NOT IN (SELECT column FROM ...)

-- Returns TRUE if the test_value does not equal to any of
the values returned from the subquery
The following tutorial exercise shows you two examples of IN operators. The sample database
GO
SELECT COUNT(*) FROM SalesLT.Customer c

WHERE c.CustomerID IN (
SELECT s.CustomerID FROM SalesLT.SalesOrderHeader s
)
GO
32
SELECT COUNT(*) FROM SalesLT.Customer c

WHERE c.CustomerID NOT IN (
SELECT s.CustomerID FROM SalesLT.SalesOrderHeader s
)
GO
408
What Are Logical/Boolean Operations?
Logical (Boolean) operations are performed on Boolean values with logical operators like 'AND', 'OR',
or 'NOT'. Logical operations return Boolean values. SQL Server 2005 supports the following logical
operations:
 AND - Returns TRUE if both operands are TRUE.
 OR - Returns TRUE if one of the operands is TRUE.
 NOT - Returns TRUE if the only operand is FALSE.
Logical operations are commonly used to combine Boolean values resulted from comparison
operations. The following tutorial exercise shows you a good example:
DECLARE @income MONEY;
DECLARE @marriage_status VARCHAR(10);
SET @income = 55000.00;
SET @marriage_status = 'Single';
SELECT CASE WHEN
(@marriage_status = 'Married' AND @income < 65000)
OR (@marriage_status = 'Single' AND @income < 35000)
THEN 'Qualified for the benefit.'
ELSE 'Not qualified for the benefit.'
END;
GO
Not qualified for the benefit.
How To Use "IF ... ELSE IF ... ELSE ..." Statement Structures?
"IF ... ELSE IF ... ELSE ..." statement structure is used to select one of the specified statements to be
executed based on pecified Boolean conditions. Here is the syntax of "IF ... ELSE IF ... ELSE ..."
statement structure:
IF condition_1 statement_1;
ELSE IF condition_2 statement_2;
...
ELSE IF condition_n statement_n;
ELSE statement_o;
-- Executes statement_x is
if condition_x results in Boolean TRUE
The tutorial exercise below shows you how to use an IF ... ELSE statement structure to selectively
execute one of the CREATE TABLE statements:
USE FyiCenterData
GO
DECLARE @site_name VARCHAR(40);

SET @site_name = 'SQA';
IF @site_name = 'DBA'
CREATE TABLE dba_links (url VARCHAR(256));
ELSE IF @site_name = 'SQA'
CREATE TABLE sqa_links (url VARCHAR(256));
ELSE
PRINT 'Unknown site name: '+@site_name;
GO
Command(s) completed successfully.
SELECT name FROM sys.tables WHERE name LIKE '%links';

GO
sqa_links
How To Use "BEGIN ... END" Statement Structures?
"BEGIN ... END" statement structure is used to group multiple statements into a single statement
block, which can be used in other statement structures as a single statement. For example, a
statement block can be used in an "IF ... ELSE ..." statement structure as a single statement.
The tutorial exercise below shows you how to use "BEGIN ... END" statement structures to place
multiple statements into an "IF ... ELSE" statement structure:
DECLARE @site_name VARCHAR(40);
SET @site_name = 'SQA';
IF @site_name = 'DBA'
BEGIN
PRINT 'Dropping table: dba_links';
DROP TABLE dba_links;
END
ELSE IF @site_name = 'SQA'
BEGIN
PRINT 'Dropping table: sqa_links';
DROP TABLE sqa_links;
END
ELSE
PRINT 'Unknown site name: '+@site_name;
GO
Dropping table: sqa_links
How To Use WHILE Loops?
WHILE statement structure is used to create a loop to execute a statement or a statement block
repeatedly under a specific condition. WHILE statement structure has the following syntax formats:
1. Loop with a single statement
WHILE condition repeating_statement
2. Loop with a statement block

WHILE condition BEGINE
statement_1
statement_2
...
statement_n
END
The tutorial exercise below shows you how to use a WHILE statement structure to execute a
statement block repeatedly:
-- Counting number of days in 2000
DECLARE @date DATETIME;
DECLARE @count INT;
SET @date = '2000-01-01';
SET @count = 0;
WHILE DATEPART(YEAR, @date) = 2000 BEGIN
SET @count = @count + 1;
SET @date = DATEADD(DAY, 1, @date);
END
SELECT @count;
366
-- 2000 is a leap year!
How To Stop a Loop Early with BREAK Statements?
If you want to stop a WHILE loop early, you can use the BREAK statement in the loop statement
block.
The tutorial exercise below shows you how to use a BREAK statement to stop a WHILE loop early:
-- Counting number of days in 2000
DECLARE @count INT;
SET @date = '2000-01-01';
SET @count = 0;
WHILE 1=1 BEGIN
IF DATEPART(YEAR, @date) > 2000 BREAK;
END
SELECT @count;
366
-- 2000 is a leap year!
How To Skip Remaining Statements in a Loop Block Using CONTINUE Statements?
If you want to skip the remaining statements in a loop block, you can use the CONTINUE statement.
The tutorial exercise below shows you how to use a CONTINUE statement to skip the remaining
statements and continue the next iteration:
-- Printing first 7 Sundays in 2000
DECLARE @count INT;
SET @date = '1999-12-31';
SET @count = 0;
WHILE DATEPART(YEAR, @date) < 2001 BEGIN
IF @count = 7 BREAK;
IF DATENAME(WEEKDAY, @date) <> 'Sunday' CONTINUE;
PRINT CONVERT(VARCHAR(40),@date,107);
END
GO
Jan 02, 2000
Jan 09, 2000
Jan 16, 2000
Jan 23, 2000
Jan 30, 2000
Feb 06, 2000
Feb 13, 2000
What Samples and Sample Databases Are Provided by Microsoft?
In order to help you to learn SQL Server, Microsoft provides several free sample scripts and sample
databases.
 SqlServerSamples.msi - 25,469 KB: Sample scripts.
 AdventureWorksDB.msi - 28,053 KB: Sample OLTP database: AdventureWorks, case-sensitive
collation version.
 AdventureWorksDBCI.msi - 29,177 KB: Sample OLTP database: AdventureWorks, case-
insensitive collation version.
 AdventureWorksBI.msi - 7,393 KB: Sample data warehouse database: AdventureWorks, case-
sensitive collation version.
 AdventureWorksBICI.msi - 16,764 KB: Sample data warehouse database: AdventureWorks,
case-insensitive collation version.
 AdventureWorksLT.msi - 2,251 KB: The scaled-down sample database AdventureWorks for
beginners.
 How to download and install SQL Server 2005 Sample Scripts?
 If you want to learn from sample scripts provided by Microsoft, you should follow this tutorial to
download and install them:
 1. Go to the SQL Server 2005 Samples and Sample Databases download page.
 2. Go to the x86 section in the Instructions section, Click "SqlServerSamples.msi -- 25,469 KB" to
download the sample file. Save the download file to c:\temp.
 3. Double click on the downloaded file: c:\temp\SqlServerSamples.msi. The installation setup
window shows up. Follow the instructions to finish the installation.
 When the installation is done, all sample scripts will be installed in the C:\Program Files\Microsoft
SQL Server\90\Samples directory:
 .\Analysis Services
 .\Engine
 .\Integrated Samples
 .\Integration Services
 .\Notification Services
 .\Replication
 .\Reporting Services
 How to download and install the scaled-down database AdventureWorksLT?
 If you want to practice you DBA skills with the scaled-down version of the sample database
AdventureWorksLT provided by Microsoft, you should follow this tutorial to download and install it
first:
 1. Go to the SQL Server 2005 Samples and Sample Databases download page.
 2. Go to the x86 section in the Instructions section, Click "AdventureWorksLT.msi -- 2,251 KB" to
download the sample file. Save the download file to c:\temp.
 3. Double click on the downloaded file: c:\temp\AdventureWorksLT.msi. The installation setup
window shows up. Follow the instructions to finish the installation.
 When the installation is done, two physical database files are installed in the data directory:
C:\Program Files\Microsoft SQL Server\MSSQL.1\MSSQL\Data:
 AdventureWorksLT_Data.mdf 5,120 KB
 AdventureWorksLT_Log.ldf 2,048 KB
 How to attach AdventureWorksLT physical files to the server?
 After installed the sample database AdventureWorksLT, you need to attach it to your SQL server
to make it available by follow this tutorial:
 EXEC sp_attach_db @dbname=N'AdventureWorksLT',
 @filename1=N'C:\Program Files\Microsoft SQL Server'
 +'\MSSQL.1\MSSQL\Data\AdventureWorks_Data.mdf',
 @filename2=N'C:\Program Files\Microsoft SQL Server'
 +'\MSSQL.1\MSSQL\Data\AdventureWorks_Log.ldf'
 GO

 USE AdventureWorksLT
 GO

 SELECT TOP 10 CustomerID, FirstName, LastName, Phone
 FROM SalesLT.Customer
 GO
 CustomerID FirstName LastName Phone
 1 Orlando Gee 245-555-0173
 2 Keith Harris 170-555-0127
 3 Donna Carreras 279-555-0130
 4 Janet Gates 710-555-0173
 5 Lucy Harrington 828-555-0186
 6 Rosmarie Carroll 244-555-0112
 7 Dominic Gash 192-555-0173
 10 Kathleen Garza 150-555-0127
 11 Katherine Harding 926-555-0159
 12 Johnny Caprio 112-555-0191
 Looks like the sample database AdventureWorksLT is ready for you play
 How AdventureWorksLT tables are related?
 There are 12 user tables defined in AdventureWorksLT:

 GO

 SELECT s.name, t.name, t.type_desc
 FROM sys.tables t, sys.schemas s
 WHERE t.schema_id = s.schema_id
 ORDER BY s.name
 GO
 s.name t.name t.type_desc
 dbo BuildVersion USER_TABLE
 dbo ErrorLog USER_TABLE
 SalesLT Address USER_TABLE
 SalesLT Customer USER_TABLE
 SalesLT CustomerAddress USER_TABLE
 SalesLT Product USER_TABLE
 SalesLT ProductCategory USER_TABLE
 SalesLT ProductDescription USER_TABLE
 SalesLT ProductModel USER_TABLE
 SalesLT ProductModelProductDescription USER_TABLE
 SalesLT SalesOrderDetail USER_TABLE
 SalesLT SalesOrderHeader USER_TABLE
 Here is an ER (Entity Relation) diagram presented on Jasmin's photo galary:
.
 How to add an address record into AdventureWorksLT?
 To find out if we can add data into AdventureWorksLT or not, you can try to add an address
record into the "SalesLT.Address" in AdventureWorksLT:
 GO

 INSERT SalesLT.Address (AddressLine1, City, StateProvince,
 CountryRegion, PostalCode)
 VALUES('1 Main Street', 'Java', 'Oracle',
 'FYI', 'Center')
 GO

 SELECT * FROM SalesLT.Address
 WHERE CountryRegion = 'FYI'
 GO
 AddressID/AddressLin1/AddressLine2/City/StateProvince
 /CountryRegion/PostalCode/rowguid
 /ModifiedDate

 11384/1 Main Street/NULL/Java/Oracle
 /FYI/Center/6073DFAE-9803-4B4F-B60E-D9742C0EED1D
 /2007-05-19 19:24:44.140
 What Are Indexes?
 An index is a secondary database object associated with a table to improve the retrieval
performance of rows from that table.
 An index can be defined for a single column or multiple columns of a given table. If an index is
defined on a single column of a table, the index can be viewed as ordered map between all
values in that column and their row locations in the table. For examples, if you have table with a
column called "company_num" and you created an index for that column. The contents of the
table and the index may look like what is presented in the following picture:
 As shown in the above picture, locating the rows of company_num = 17 can be done much faster
through the index.
 How To Create an Index on an Existing Table?
 If you want to an index on an existing table, you can use the CREATE INDEX statement in a
simple syntax:
 CREATE INDEX index_name ON table_name (column_name)
 The tutorial exercise below shows you how to add an index for the "in" column of the "fyi_links"
table:
 USE FyiCenterData;
 GO

 -- Drop the old table, if needed
 DROP TABLE fyi_links;
 GO

 -- Create a fresh new table
 CREATE TABLE fyi_links (
 id INT NOT NULL,
 url VARCHAR(80) NOT NULL,
 notes VARCHAR(1024),
 counts INT,
 created DATETIME NOT NULL DEFAULT(getdate())
 );
 GO

 -- Create an index for "id"
 CREATE INDEX fyi_links_id ON fyi_links (id);
 GO

 -- Create an index for "url"
 CREATE INDEX fyi_links_url ON fyi_links (url);
 GO
 How To View Existing Indexes on an Given Table using SP_HELP?
 If you want to know how many indexes have been defined for a given table, you can use the
SP_HELP built-in stored procedure in the following syntax:
 EXEC SP_HELP table_name
 -- Returns all database objects related the given table
 The tutorial exercise shows you how many indexes were defined from the previous tutorial on
table "fyi_links":
 EXEC SP_HELP fyi_links;
 GO
 ...

 index_name index_description index_keys
 ------------ ------------------------------- ----------
 fyi_links_id nonclustered located on PRIMARY id
 fyi_links_url nonclustered located on PRIMARY url
 ...
 How To View Existing Indexes on an Given Table using sys.indexes?
 Another way to view existing indexes defined for a given table is to use the system view called
"sys.indexes". The tutorial exercise shows you how many indexes were defined from the previous
tutorial on table "fyi_links":
 GO

 SELECT * FROM sys.indexes WHERE object_id = (
 SELECT object_id FROM sys.tables WHERE name = 'fyi_links'
 );
 GO
 object_id name index_id type_desc is_unique
 --------- ------------- -------- ---------- ---------
 421576540 NULL 0 HEAP 0
 421576540 fyi_links_id 2 NONCLUSTERED 0
 421576540 fyi_links_url 3 NONCLUSTERED 0
 The extra line in the query result is not a real index at this moment. It will be explained in another
tutorial.
 How To Drop Existing Indexes?
 For some reason, if you want remove an existing index, you can use the DROP INDEX statement
with following syntax:
 CREATE INDEX table_name.index_name
 The tutorial exercise below shows you how to remove the index "fyi_links_id":
 GO

 );
 GO
 --------- ------------- -------- ---------- ---------
 421576540 NULL 0 HEAP 0
 421576540 fyi_links_id 2 NONCLUSTERED 0

 DROP INDEX fyi_links.fyi_links_id;
 GO

 );
 GO
 --------- ------------- -------- ---------- ---------
 421576540 NULL 0 HEAP 0
 Is the PRIMARY KEY Column of a Table an Index?
 If you define a primary key on a table, an index for the primary key column will be created by
default. The tutorial exercise below shows you the index created as part of the primary key
column of "fyi_links":
 GO

 GO

 -- Create a table with primary key
 id INT PRIMARY KEY,
 counts INT,
 );
 GO

 -- Create an index for column "url"
 GO

 -- View indexes
 GO
 index_name index_description keys
 ----------------------- -------------------------- ----
 fyi_links_url nonclustered located url
 on PRIMARY

 PK__fyi_links__239E4DCF clustered, unique, primary
 key located on PRIMARY id
 Notice that the index created as part of the primary key is named by SQL Server as
"PK__fyi_links__239E4DCF".
 Does the UNIQUE Constraint Create an Index?
 If you add the UNIQUE constraint on a column, SQL Server will automatically add a non-
clustered index for that column. The tutorial exercise below shows you the index created as part
of the UNIQUE column, "id", of "fyi_links":
 GO

 GO

 -- Create a table with a UNIQUE constraint
 id INT UNIQUE,
 counts INT,
 );
 GO

 -- Create an index for column "url"
 GO

 -- View indexes
 GO
 ----------------------- -------------------------- ----
 fyi_links_url nonclustered located url
 on PRIMARY

 UQ__fyi_links__4222D4EF nonclustered, unique
 key located on PRIMARY id
 Notice that the index created as part of the UNIQUE constraint is named by SQL Server as
"UQ__fyi_links__4222D4EF".
What Is the Difference Between Clustered and Non-Clustered Indexes?
SQL Server 2005 supports two types of indexes: clustered index and non-clustered index. Here are
the main differences between them:
 One table can only have only one clustered index.
 One table can only have many non-clustered index.
 A clustered index requires no separate storage than the table storage. It forces the rows to be
stored sorted on the index key.
 A non-clustered index requires separate storage than the table storage to store the index
information.
 A table with a clustered index is called clustered table. Its rows are stored in a B-Tree structure
sorted.
 A table without any clustered indexes is called non-clustered table. Its rows are stored in heap
structure unsorted.
 The default index created as part of the primary key column is a clustered index
 How To Create a Clustered Index?

 If there is no primary key in a table, you can add one clustered index to that table with CREATE
CLUSTERED INDEX statement. The tutorial exercise below shows you how to create a clustered
index:
 GO

 GO

 id INT,
 counts INT,
 );
 GO

 -- Create a clustered index
 CREATE CLUSTERED INDEX fyi_links_url ON fyi_links (url);
 GO

 -- Create a non-clustered index
 CREATE INDEX fyi_links_created ON fyi_links (created);
 GO

 -- View indexes
 GO
 ----------------- -------------------------- ----
 fyi_links_created nonclustered located on PRIMARY created
 fyi_links_url clustered located on PRIMARY url
 How To Create an Index for Multiple Columns?
 An index for multiple columns works similarly to a sorting process on multiple columns. If an index
is defined for two columns, the index key is composed by values from those two columns.
 A multi-column index will be used to speed up the search process based on the same columns in
the same order as the index definition. For example, if you define an index called "combo_index"
for "url" and "counts". "combo_index" will be used only when searching or sorting rows by "url"
and "counts".
 The tutorial exercise below shows you how to create an index for two columns:
 -- Create an index for two columns
 CREATE INDEX combo_index ON fyi_links (url, counts);
 GO

 -- View indexes
 GO
 ----------------- -------------------------- ----
 fyi_links_created nonclustered located on PRIMARY created
 fyi_links_url clustered located on PRIMARY url
 combo_index nonclustered located on PRIMARY url, counts

 -- Drop the index
 DROP INDEX fyi_links.combo_index;
 GO
 How To Create a Large Table with Random Data for Index Testing?
 If you want to see how index can be used to improve data search performance, you have to build
some large tables, which requires large amount of random data. This tutorial exercise helps you
to build a large table with pure random data:
 GO

 DROP TABLE fyi_random;
 GO

 CREATE TABLE fyi_random (
 id INT,
 rand_integer INT,
 rand_number numeric(18,9),
 rand_datetime DATETIME,
 rand_string VARCHAR(80)
 );
 GO

 -- Insert rows with random values
 DECLARE @row INT;
 DECLARE @string VARCHAR(80), @length INT, @code INT;
 SET @row = 0;
 WHILE @row < 100000 BEGIN
 SET @row = @row + 1;

 -- Build the random string
 SET @length = ROUND(80*RAND(),0);
 SET @string = '';
 WHILE @length > 0 BEGIN
 SET @length = @length - 1;
 SET @code = ROUND(32*RAND(),0) - 6;
 IF @code BETWEEN 1 AND 26
 SET @string = @string + CHAR(ASCII('a')+@code-1);
 ELSE
 SET @string = @string + ' ';
 END

 -- Ready for the record
 SET NOCOUNT ON;
 INSERT INTO fyi_random VALUES (
 @row,
 ROUND(2000000*RAND()-1000000,0),
 ROUND(2000000*RAND()-1000000,9),
 CONVERT(DATETIME, ROUND(60000*RAND()-30000,9)),
 @string
 )
 END
 PRINT 'Rows inserted: '+CONVERT(VARCHAR(20),@row);
 GO
 Rows inserted: 100000

 SELECT TOP 10 * FROM fyi_random;
 GO
 id integer number date string
 1 609596 896530.556031080 1891-04-22 rbxddlaayxq x...
 2 401483 375107.107911612 1941-10-15 ou pdep rjixd...
 3 -350570 103724.712787200 1921-11-01 uwylic paxtlp...
 4 -125377 204467.582749559 1931-11-11 xoikcdv gxixb...
 5 -222733 -967023.477922098 1934-10-13 vj pqkyj ujbn...
 6 377398 263410.255290027 1960-04-02 btstgqqx dy...
 7 -422529 43452.274555387 1911-04-19 qyoxqqo zuvxh...
 8 -900104 357566.072355041 1833-10-19 d xyqfotmvbnd...
 9 -724738 -396821.456353329 1939-09-27 rx ...
 10 398355 -817318.223094167 1874-03-25 g on oekx bc ...
 How To Measure Performance of INSERT Statements?
 When indexes are defined for a table, each time a new row is inserted to the table, all the indexes
must be updated. This extra update work could slow down the insert statement.
 To find out how much slower the INSERT statements will be on tables with indexes, you need to
measure the performance of INSERT statements on tables without indexes. This tutorial shows
you how to measure the time spend on inserting 100000 rows to fyi_links without indexes:
 GO

 GO

 -- Create a table with no indexes
 id INT,
 counts INT,
 );
 GO

 -- Empty the table if needed
 DELETE FROM fyi_links;
 GO

 -- Performance test of INSERT
 DECLARE @start_time DATETIME, @end_time DATETIME;
 SET @start_time = GETDATE();
 INSERT INTO fyi_links
 SELECT id, rand_string, REVERSE(rand_string),
 rand_integer, rand_datetime
 FROM fyi_random
 SET @end_time = GETDATE();
 PRINT 'Milliseconds used: '+CONVERT(VARCHAR(20),
 DATEDIFF(MILLISECOND,@start_time,@end_time));
 GO

 -- First time
 (100000 row(s) affected)
 Milliseconds used: 18013

 -- Second time

 -- Third time
 Does Index Slows Down INSERT Statements?
 If you want to see the impact of indexes on INSERT statements, you can repeat the same insert
script on the table "fyi_links" of the same structure with two indexes: one non-clustered index on
column "url" and one non-clustered index on column "counts". See the tutorial exercise below:
 GO

 DROP TABLE fyi_links_indexed;
 GO

 -- Create a table
 CREATE TABLE fyi_links_indexed (
 id INT,
 counts INT,
 );
 GO

 -- Create two indexes
 CREATE INDEX fyi_links_url ON fyi_links_indexed (url);
 CREATE INDEX fyi_links_counts ON fyi_links_indexed (counts);
 GO

 -- Empty the table if needed
 DELETE FROM fyi_links_indexed;
 GO

 -- Performance test of INSERT
 INSERT INTO fyi_links_indexed
 SELECT id, rand_string, REVERSE(rand_string),
 rand_integer, rand_datetime
 FROM fyi_random
 GO

 -- First time

 -- Second time

 -- Third time
 Comparing the result between this tutorial and the previous tutorial, two indexes make 100000
insert statements about 4 times slower. Note that the first measurement of both tests seems to be
affected by the database engine caching.
 Does Index Speed Up SELECT Statements?
 If you want to see the impact of indexes on SELECT statements, you can run the same SELECT
statement on "fyi_links" and "fyi_links_indexed" tables. See the tutorial exercise below:
 GO

 -- Run SELECT on the table without indexes
 SELECT TOP 3 counts, url FROM fyi_links
 WHERE url LIKE 'a%'
 ORDER BY counts DESC;
 GO
 counts url
 ----------- ------------------------------------------------
 999417 a ihgu migox mlqvi gvs n nmabwdk iehu ezjjv n l
 999008 agqbr gkjwfpjvp z c cq k inen j ppjsu x iuk uhp
 998471 a yiu squqco eih


 -- Run SELECT on the table with indexes
 SELECT TOP 3 counts, url FROM fyi_links_indexed
 WHERE url LIKE 'a%'
 ORDER BY counts DESC;
 GO
 counts url
 ----------- ------------------------------------------------
 999417 a ihgu migox mlqvi gvs n nmabwdk iehu ezjjv n l
 999008 agqbr gkjwfpjvp z c cq k inen j ppjsu x iuk uhp
 998471 a yiu squqco eih

 The test result shows indexes on "url" and "counts" did improve the SELECT statement from 46
milliseconds to almost 0 millisecond.
 What Happens If You Add a New Index to Large Table?
 An index can be added when you create a new table. New rows will be indexed as they are
inserted into the table. But you can also add a new index to an existing table with the same
CREATE INDEX statement. The existing rows will be indexed as part of the CREATE INDEX
statement.
 If you add a new index to an existing table with a large number of rows. The CREATE INDEX
statement could take some time to finish. See the tutorial exercise below:
 GO

 -- Drop indexes if needed
 DROP INDEX fyi_links_indexed.fyi_links_url;
 DROP INDEX fyi_links_indexed.fyi_links_counts;
 GO

 SELECT COUNT(*) FROM fyi_links_indexed;
 GO
 100000

 -- Create two indexes
 CREATE INDEX fyi_links_url ON fyi_links_indexed (url);
 CREATE INDEX fyi_links_counts ON fyi_links_indexed (counts);
 GO

 -- First time

 -- Second time

 -- Third time
 You can see creating two indexes on a table of 100000 rows costs about 12 seconds.
 What Is the Impact on Other User Sessions When Creating Indexes?
 If you are creating a new index on a table with existing data, all existing rows will be indexed as
part of the CREATE INDEX statement. If the table is large, the indexing process could take some
time. The impact of this indexing process on other user sessions is based whether SQL server is
using the Offline mode or Online mode.
 Be default, SQL Server performs indexing operations in Offline mode, where table locks are
applied for the duration of the index operation. An offline index operation that creates, rebuilds, or
drops a clustered index, or rebuilds or drops a nonclustered index, acquires a Schema
modification (Sch-M) lock on the table. This prevents all user access to the underlying table for
the duration of the operation. An offline index operation that creates a nonclustered index
acquires a Shared (S) lock on the table. This prevents updates to the underlying table but allows
read operations, such as SELECT statements.
 SQL Server Enterprise Edition supports indexing operations in Online mode, where other user
sessions will not be impacted.
 However, SQL Server Express Edition does no support the Online mode. If you try it, you will get
an error as shown below:
 CREATE INDEX fyi_links_url ON fyi_links_indexed (url)
 WITH (ONLINE = ON);
 GO
 Online index operations can only be performed in Enterprise
 edition of SQL Server.
 What Is Index Fragmentation?
 Index fragmentation is a phenomena where index contents are no longer stored continuously in
the storage. When index contents become scattered in the storage, fragmented, performance on
index will degrade.
 If you want to see the fragmentation level of an index, you can use the system function called
sys.dm_db_index_physical_stats() in the following format:
 SELECT * FROM sys.dm_db_index_physical_stats(
 database_id, table_id, index_id, DEFAULT, DEFAULT
 )
 The tutorial exercise below shows you how to view the fragmentation level of all indexes on table
"fyi_links_indexed":
 GO

 GO
 100000

 SELECT i.index_id, i.name, s.avg_fragmentation_in_percent
 FROM sys.dm_db_index_physical_stats (
 DB_ID(N'FyiCenterData'),
 OBJECT_ID(N'fyi_links_indexed'),
 DEFAULT, DEFAULT, DEFAULT) s, sys.indexes i
 WHERE s.object_id = i.object_id
 AND s.index_id = i.index_id;
 GO
 index_id name avg_fragmentation_in_percent
 0 NULL 0.574712643678161
 2 fyi_links_url 0
 3 fyi_links_counts 0
 Since there is no clustered index (index_id=0) on this table, the first row in the return is just a
place holder.
 The other two rows show that "fyi_links_url" and "fyi_links_counts" have zero fragmentation,
since they were created not long ago.
 What Causes Index Fragmentation?
 Index fragmentation is usually caused by deleting of existing rows or updating existing values of
the indexed column. Inserting new rows should not cause any index fragmentation.
 This tutorial exercise shows you how update statements of 50000 rows on the table
"fyi_link_indexed" with 100000 rows caused the index fragmented 84%:
 GO

 GO
 100000

 GO
 0 NULL 0.574712643678161
 3 fyi_links_counts 0

 UPDATE fyi_links_indexed SET url = REVERSE(url)
 WHERE id <=50000;
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 84.053862508859
 3 fyi_links_counts 0.448430493273543
 Note that only the index on "url" fragmented. Index on "counts" did not affected, because the
update statement only touched the "url" column
 How To Defragment Table Indexes?
 When a table index is fragmented to a certain percentage, you need to defragment the index to
maintain its performance level. There are 3 ways to defragment:
 1. "ALTER INDEX index_name ON table_name REORGANIZE" - Defragmenting the specified
index performed in online mode. No locks applied on affected table. Used for indexes with a small
fragmentation percentage.
 2. "ALTER INDEX index_name ON table_name REBUILD" - Defragmenting the specified index
performed in offline mode by default. It can be performed in online mode. Used for indexes with a
large fragmentation percentage.
 3. "CREATE INDEX ... WITH (DROP_EXISTING = ON)" - Re-creating the specified index. The
definition of the index can be changed.
 How To Defragment Indexes with ALTER INDEX ... REORGANIZE?
 When an index is defragmented to a small percentage, like < 30%, you can use the "ALTER
INDEX ... REORGANIZE" statement to defragment the index. Here is a tutorial exercise on
defragmenting indexes:
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 84.053862508859
 3 fyi_links_counts 0.448430493273543

 ALTER INDEX fyi_links_url ON fyi_links_indexed REORGANIZE;
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 1.87590187590188
 3 fyi_links_counts 0.448430493273543
 The fragmentation level has been reduced from 84.05% to 1.88%.
 How To Rebuild Indexes with ALTER INDEX ... REBUILD?
 When an index is defragmented to a large percentage, like > 30%, you can use the "ALTER
INDEX ... REBUILD" statement to rebuild the index. Here is a tutorial exercise on rebuilding
indexes:
 GO

 UPDATE fyi_links_indexed SET url = REVERSE(url)
 WHERE id <=50000;
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 85.0142045454545
 3 fyi_links_counts 0.448430493273543

 ALTER INDEX fyi_links_url ON fyi_links_indexed REBUILD;
 GO

 GO
 0 NULL 0.574712643678161
 3 fyi_links_counts 0.448430493273543
 Rebuilding an index brings the fragmentation level to 0% in this case.
 How To Rebuild All Indexes on a Single Table?
 If you have several indexes on a single table and want to rebuild all of them, you may use the
"ALTER INDEX ALL ON table_name REBUILD" statement as shown in the tutorial exercise
below:
 GO

 UPDATE fyi_links_indexed
 SET url = REVERSE(url), counts = -counts
 WHERE id <=50000;
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 85.750315258512
 3 fyi_links_counts 84.040404040404

 ALTER INDEX ALL ON fyi_links_indexed REBUILD;
 GO

 GO
 0 NULL 0.574712643678161
 2 fyi_links_url 0.12987012987013
 3 fyi_links_counts 0.448430493273543
 Both indexes were defragmented to a very low level now
 How To Recreate an Existing Index?
 If you want to change the definition of an existing index, you can use the "DROP INDEX"
statement to drop the index first. Then use the "CREATE INDEX" statement to create it again with
the new definition.
 But you can also combine those two statements into one:
 CREATE INDEX ... WITH (DROP_EXISTING = ON)
 The tutorial exercise below recreates fyi_links_url with a change to index columns:
 GO

 CREATE INDEX fyi_links_url ON fyi_links_indexed (url, counts)
 WITH (DROP_EXISTING = ON);
 GO

 SP_HELP fyi_links_indexed;
 GO
 index_name index_description index_keys
 ---------------- -------------------------------- ------------
 fyi_links_counts nonclustered located on PRIMARY counts
 fyi_links_url nonclustered located on PRIMARY url, counts
What Are Views?
A view is a database object that represents the data in one or more tables in the same structure as a
separate table. Here are some basic rules about views:
 Tables store real data.
 Views do not store real data.
 Views must have underlying tables to provide data.
 Each view is based on a single SELECT statement to control what data to collect from tables, and
how data should be represented.
 View's columns can be mapped directly to columns in underlying tables.
 View's columns can be created expressions based multiple columns in underlying tables.
 Views can be used in same way as tables in queries.
 How To Create a View on an Existing Table?
 If you want to a view on an existing table, you can use the CREATE VIEW statement in a simple
syntax:
 CREATE VIEW view_name AS SELECT ...
 The tutorial exercise below shows you how to create a view to represent sub set of data stored in
fyi_links table:
 GO

 CREATE VIEW fyi_links_top AS
 SELECT id, counts, url FROM fyi_links
 WHERE counts > 100;
 GO

 SELECT TOP 10 * FROM fyi_link_top;
 GO
 id counts url
 ------ ------- -----------------------------------------
 7600 237946 eyfndw jdt lee ztejeyx l q jdh k
 19437 222337 eypx u x
 55924 1877 eyq ntohxe i rtnlu riwaskzp cucoa dva c
 63742 121330 ezdaeh mmgmo vaai meytbjjv f jixfsdjw pw
 92455 945262 ezlmyenrw dyeb
 36391 41386 f
 87433 977726 f
 7180 559314 f kqbqlej s xixuurcgg lh r dqqvqsstxw
 2252 702033 f bkh jy sqrkttuoarxmfp idqyhyy tme d
 1228 146283 f m asukh
How To See Existing Views?
If you want to know how many views you have created in a database, you use the system view called
sys.views to get a list of views defined in the current database. The tutorial exercise shows you how
many views in database FyiCenterData:
USE FyiCenterData;
GO
CREATE VIEW fyi_links_view AS

SELECT * FROM fyi_links;
GO
SELECT * FROM sys.views;

GO
name object_id schema_id type type_desc
--------------- ----------- ---------- ---- ----------
fyi_links_view 1189579276 1 V VIEW
fyi_links_top 1205579333 1 V VIEW
(2 row(s) affected)
How To Drop Existing Views from a Database?
If you don't need a specific view any more, you can use the DROP VIEW statement to delete it from
the database. The following tutorial exercise shows you how to delete the view, fyi_links_view:
USE FyiCenterData;
GO

GO
--------------- ----------- ---------- ---- ----------
fyi_links_view 1189579276 1 V VIEW
(2 row(s) affected)
DROP VIEW fyi_links_view;

GO

GO
--------------- ----------- ---------- ---- ----------
(1 row(s) affected)
How To Get a List of Columns in a View using "sys.columns"?
If you have an existing view, but you don't remember what are the columns defined in the view, you
can use the "sys.columns" system view to get a list of all columns of all views in the current database.
In order to a list of columns of a single view, you need to join sys.columns and sys.views as shown in
the tutorial example below:
SELECT * FROM sys.columns c, sys.views v
WHERE c.object_id = v.object_id
AND t.name = 'fyi_links_top'
GO
object_id name column_id user_type_id max_length
----------- ------- ---------- ------------ ----------
1205579333 id 1 56 4
1205579333 counts 2 56 4
1205579333 url 3 167 80
You can see the column names easily from the sys.columns view. But you can only see the column
type IDs. This requires another join to get the column type names. You may try the "sp_columns"
stored procedure to get a better list of columns shown in the next tutorial.
How To Get a List of Columns in a View using the "sp_columns" Stored Procedure?
If you have an existing table, but you don't remember what are the columns defined in the view, you
can use the "sp_columns" stored procedure to get a list of all columns of the specified view. The
following tutorial script shows you a good example:
EXEC SP_COLUMNS fyi_links_top;
GO
TABLE_OWNER TABLE_NAME COLUMN_NAME TYPE_NAME LENGTH
----------- ------------- ----------- --------- ------
dbo fyi_links_top id int 4
dbo fyi_links_top counts int 4
dbo fyi_links_top url varchar 80
(3 row(s) affected)
The "sp_columns" stored procedure returns a long list of properties for each column of the specified
view.
How To Get a List of Columns in a View using the "sp_help" Stored Procedure?
Another way to get a list of columns from a view is to use the "sp_help" stored procedure. "sp_help"
returns more than just a list of columns. It returns: the view information, the column information, the
identity column, the row GUID column. The tutorial exercise belwo shows you what you will get with
sp_help:
EXEC SP_HELP fyi_links_top;
GO
Name Owner Type Created_datetime
-------------- ------ ----- -----------------------
fyi_links_top dbo view 2007-05-19 13:43:46.983
Column_name Type Computed Length Prec Scale Nullable

------------ -------- --------- ------- ----- ----- --------
id int no 4 10 0 yes
counts int no 4 10 0 yes
url varchar no 80 no
Identity
---------------------------
No identity column defined.
RowGuidCol
-----------------------------
No rowguidcol column defined.
How To Generate CREATE VIEW Script on an Existing View?
If you want to know how an existing view was created, you can use SQL Server Management Studio
to automatically generate a "CREATE VIEW" script The following tutorial shows you how to do this:
1. Run SQL Server Management Studio and connect to SQL server.
2. On the Object Explorer window, follow the object tree: Databases > FyiCenterData > Views >
dbo.fyi_links_top.
3. Click right mouse button on dbo.fyi_links_top. The context menu shows up.
4. Select "Script Table as" > "CREATE to" > "New Query Editor Window". The following script will be
displayed:
USE [FyiCenterData]
GO
/****** Object: View [dbo].[fyi_links_top]
Script Date: 05/19/2007 15:07:27 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
CREATE VIEW [dbo].[fyi_links_top] AS
SELECT id, counts, url FROM fyi_links
WHERE counts > 100;
How To Get the Definition of a View Out of the SQL Server?
If you want get the definition of an existing view back from the SQL Server, you can use the system
view called sys.sql_modules, which stores defitions of views and procedures.
The sys.sql_modules holds view definitions identifiable by the object id of each view. The tutorial
exercise below shows you how to retrieve the definition of view, "fyi_link_view" by joining
sys.sql_modules and sys.views:
USE FyiCenterData;
GO
SELECT m.definition
FROM sys.sql_modules m, sys.views v
WHERE m.object_id = v.object_id
AND v.name = 'fyi_links_top';
GO
definition
-------------------------------------------
CREATE VIEW fyi_links_top (LinkText) AS
SELECT CONVERT(VARCHAR(20),id)
+ ' - ' + CONVERT(VARCHAR(20),counts)
+ ' - ' + url
FROM fyi_links WHERE counts > 1000
(1 row(s) affected)
Can You Create a View with Data from Multiple Tables?

Can You Create a View with Data from Multiple Tables? The answer is yes. A view can be created
with a SELECT statement to join data from multiple tables.
It is a common practice to normalize data into multiple tables. Then using a view to de-normalize
them into a single output.
The tutorial exercise below shows you how to create a view to normalize data from two tables
SalesOrderHeader and Customer in the sample database AdventureWorksLT.
USE AdventureWorksLT;
GO
CREATE VIEW SalesOrderView AS

SELECT o.SalesOrderNumber, o.OrderDate, o.TotalDue,
c.FirstName, c.LastName, c.CompanyName
FROM SalesLT.SalesOrderHeader o, SalesLT.Customer c
WHERE o.CustomerID = c.CustomerID
GO
SELECT TOP 10 SalesOrderNumber, TotalDue, CompanyName

FROM SalesOrderView;
GO
SalesOrderNumber TotalDue CompanyName
---------------- ----------- ------------------------------
SO71915 2361.6403 Aerobic Exercise Company
SO71938 98138.2131 Bulk Discount Store
SO71783 92663.5609 Eastside Department Store
SO71899 2669.3183 Coalition Bike Company
SO71898 70698.9922 Instruments and Parts Company
SO71902 81834.9826 Many Bikes Store
SO71832 39531.6085 Closest Bicycle Store
SO71776 87.0851 West Side Mart
SO71797 86222.8072 Riding Cycles
SO71895 272.6468 Futuristic Bikes
Can You Create a View using Data from Another View?
Can You Create a View with Data from Another View? The answer is yes. A view can be used as a
table to build other views. The tutorial exercise below shows you how to create a view using data
from another view:
USE AdventureWorksLT;
GO
CREATE VIEW SalesOrderTop AS

SELECT SalesOrderNumber, TotalDue, CompanyName
FROM SalesOrderView
WHERE TotalDue > 10000.0
GO
SELECT TOP 10 * FROM SalesOrderTop;
GO
SalesOrderNumber TotalDue CompanyName
---------------- ----------- ------------------------------
SO71780 42452.6519 Nearby Cycle Shop
SO71782 43962.7901 Professional Sales and Service
SO71783 92663.5609 Eastside Department Store
SO71784 119960.824 Action Bicycle Specialists
SO71796 63686.2708 Extreme Riding Supplies
SO71797 86222.8072 Riding Cycles
SO71832 39531.6085 Closest Bicycle Store
SO71845 45992.3665 Trailblazing Sports
SO71858 15275.1977 Thrilling Bike Tours
SO71897 14017.9083 Paints and Solvents Company
What Happens If You Delete a Table That Is Used by a View?
Assuming that you have a table which is used by a view, and you try to delete that table. SQL Server
will let you delete the table without any trouble.
But that view will become invalid. The tutorial exercise below shows you what happens to the view,
when the underlying table is deleted:
USE FyiCenterData;
GO
SELECT * INTO fyi_links_copy

FROM fyi_links WHERE counts > 0;
GO
CREATE VIEW fyi_links_view AS

SELECT * FROM fyi_links_copy;
GO
SELECT COUNT(*) FROM fyi_links_view;

GO
50015
DROP TABLE fyi_links_copy;

GO
SELECT COUNT(*) FROM fyi_links_view;

GO
Invalid object name 'fyi_links_copy'.
Could not use view or function 'fyi_links_view'
because of binding errors.
Can You Use ORDER BY When Defining a View?
Sometimes you want the data in a view to be sorted and try to use the ORDER BY clause in the
SELECT statement to define the view.
But SQL Server will not allow you to use ORDER BY to define a view without the TOP clause. The
tutorial exercise below shows you what error you will get when using ORDER BY in a CREATE VIEW
statement:
USE FyiCenterData;
GO
CREATE VIEW fyi_links_top AS

SELECT id, counts, url FROM fyi_links
WHERE counts > 100
ORDER BY counts DESC;
GO
Msg 1033, Level 15, State 1, Procedure fyi_links_top, Line 4
The ORDER BY clause is invalid in views, inline functions,
derived tables, subqueries, and common table expressions,
unless TOP or FOR XML is also specified.
CREATE VIEW fyi_links_top AS

SELECT TOP 100 id, counts, url FROM fyi_links
WHERE counts > 100
GO
SELECT TOP 3 * FROM fyi_links_top;

GO
id counts url
------ ------- -------------------------------------------
36470 999966 dgqnvmy pyjqd toqcoupuxortasdtzvcae jonfb
12292 999953 qebmw v qqmywe q kza wskxqns jnb
6192 999943 p o qisvrakk hk od
How To Modify the Underlying Query of an Existing View?
If you have an existing view, and want to change the underlying SELECT statement, you can use the
"ALTER VIEW ..." statement to redefine the view. The tutorial exercise below shows you how modify
an existing view:
USE FyiCenterData;
GO
ALTER VIEW fyi_links_top AS

WHERE counts > 100
GO
SELECT * FROM fyi_links_top;

GO
id counts url
------ ------- -------------------------------------------
Can You Insert Data into a View?
Can you insert data into a view? The answer is no.
But if the question is "Can you insert data into the underlying table through view?" The answer is then
yes. SQL Server will allow you to insert data into the underlying table through a view with a condition:
 The insert columns must be limited to columns of a single underlying table.
The tutorial exercise below shows you how to insert data into a underlying table through a view:
USE FyiCenterData;
GO
ALTER VIEW fyi_links_top AS

WHERE counts > 100
GO
INSERT INTO fyi_links_top

VALUES(100001, 1000001, 'dba.fyicenter.com');
GO

GO
id counts url
------ ------- -------------------------------------------
100001 1000001 dba.fyicenter.com
SELECT TOP 1 * FROM fyi_links ORDER BY counts DESC;

GO
------ ----------------- ----- ----------- ----------
Can You Update Data in a View?
Can you update data in a view? The answer is no.
But if the question is "Can you update data in the underlying table through view?" The answer is then
yes. SQL Server will allow you to update data in the underlying table through a view. The tutorial
exercise below is a good example:
UPDATE fyi_links_top SET url = REVERSE(url)
WHERE id = 100001;
GO

GO
id counts url
------ ------- -------------------------------------------
100001 1000001 moc.retneciyf.abd

GO
------ ----------------- ----- ----------- ----------
Can You Delete Data from a View?
Can you delete data in a view? The answer is no.
But if the question is "Can you delete data from the underlying table through view?" The answer is
then yes. SQL Server will allow you to delete data from the underlying table through a view. The
tutorial exercise below is a good example:
DELETE FROM fyi_links_top WHERE id = 100001;
GO

GO

GO
id url ...
------ ------------------------------------------- ...
36470 dgqnvmy pyjqd toqcoupuxortasdtzvcae jonfb ...
How To Assign New Column Names in a View?
By default, column names in a view are provided by the underlying SELECT statement.
But sometimes, the underlying SELECT statement can not provide names for output columns that
specified as expressions with functions and operations. In this case, you need to assign new names
for the view's columns. The tutorial exercise below creates a view to merge several table columns into
a single view column with a format called CSV (Comma Separated Values):
CREATE VIEW fyi_links_dump AS
+ ', ' + CONVERT(VARCHAR(20),counts)
+ ', ''' + url + ''''
GO
Msg 4511, Level 16, State 1, Procedure fyi_links_dump,
Line 2
Create View or Function failed because no column name
was specified for column 1.
CREATE VIEW fyi_links_dump (Line) AS

+ ', ''' + url + ''''
GO
SELECT TOP 3 * FROM fyi_links_dump

GO
Line
------------------------------------------------------------
7600, 237946, ' eyfndw jdt lee ztejeyx l q jdh k '
19437, 222337, ' eypx u x'
55924, 1877, ' eyq ntohxe i rtnlu riwaskzp cucoa dva c rc'
The first CREATE VIEW gives you an error, because the SELECT statement returns no column for
the concatenated value, and no view column name is specified explicitly
How To Assign New Column Names in a View?
By default, column names in a view are provided by the underlying SELECT statement.
But sometimes, the underlying SELECT statement can not provide names for output columns that
specified as expressions with functions and operations. In this case, you need to assign new names
for the view's columns. The tutorial exercise below creates a view to merge several table columns into
a single view column with a format called CSV (Comma Separated Values):
CREATE VIEW fyi_links_dump AS
+ ', ''' + url + ''''
GO
Msg 4511, Level 16, State 1, Procedure fyi_links_dump,
Line 2
Create View or Function failed because no column name
was specified for column 1.
CREATE VIEW fyi_links_dump (Line) AS

+ ', ''' + url + ''''
GO
SELECT TOP 3 * FROM fyi_links_dump

GO
Line
------------------------------------------------------------
7600, 237946, ' eyfndw jdt lee ztejeyx l q jdh k '
19437, 222337, ' eypx u x'
55924, 1877, ' eyq ntohxe i rtnlu riwaskzp cucoa dva c rc'
The first CREATE VIEW gives you an error, because the SELECT statement returns no column for
the concatenated value, and no view column name is specified explicitly
How To Bind a View to the Schema of the Underlying Tables?
By default, views are not bound to the schema of the underlying tables. This means that SQL Server
will allow you to change underlying table's schema any time. For example, you can drop the
underlying table while keep the view. Of course, this will make the view not valid any more.
If you don't want anyone to change underlying table's schema once a view has been defined, you can
create a binding view with the "WITH SCHEMABINDING" clause to bind the view to the schema of
underlying tables. A binding view has the following features:
 Changing of underlying table's schema is not allowed as long as there exists one binding view.
 Indexes can be created only on binding views.
The tutorial exercise below shows you how to create a binding with "WITH SCHEMABINDING":
GO
CREATE VIEW fyi_links_view (ID, DateString, CountUrl)

WITH SCHEMABINDING
AS SELECT id, CONVERT(VARCHAR(16), created, 107),
CONVERT(VARCHAR(20),counts)+' - '+url
FROM fyi_links_copy WHERE counts > 1000;
GO
Msg 4512, Level 16, State 3, Procedure fyi_links_view,
Line 3
Cannot schema bind view 'fyi_links_view' because name
'fyi_links_copy' is invalid for schema binding.
Names must be in two-part format and an object
cannot reference itself.
CREATE VIEW fyi_links_view (ID, DateString, CountUrl)

WITH SCHEMABINDING
AS SELECT id, CONVERT(VARCHAR(16), created, 107),
CONVERT(VARCHAR(20),counts)+' - '+url
FROM dbo.fyi_links_copy WHERE counts > 1000;
GO
DROP TABLE fyi_links_copy;

GO
Cannot DROP TABLE 'fyi_links_copy' because it is being
referenced by object 'fyi_links_view'.
The error received in the first CREATE VIEW statement says that you must prefix the table name with
the schema name.
The error received in the DROP TABLE statement proves that you can not change the underlying
tables if there is binding view.
How To Create an Index on a View?
If you need to search and sort data in a view with a large number of row, you may want to create an
index on the view to speed up your search process.
The tutorial exercise below shows you how to create a unique clustered index on a view.
GO
CREATE VIEW fyi_links_view (ID, UrlReversed)

AS SELECT id, REVERSE(url)
FROM fyi_links_copy WHERE counts > 1000;
GO
CREATE UNIQUE CLUSTERED INDEX date_string

ON fyi_links_view (ID);
GO
Cannot create index on view 'fyi_links_view'
because the view is not schema bound.
ALTER VIEW fyi_links_view (ID, UrlReversed)

WITH SCHEMABINDING
AS SELECT id, REVERSE(url)
FROM dbo.fyi_links_copy WHERE counts > 1000;
GO
CREATE UNIQUE CLUSTERED INDEX date_string

ON fyi_links_view (ID);
GO
EXEC SP_HELP fyi_links_view;

GO
index_name index_description index_keys
----------- ------------------------------------ ----------
date_string clustered, unique located on PRIMARY ID
What Are Stored Procedures?
A stored procedure is a collection of Transact-SQL statements that stored in the SQL Server. A
stored procedure can be executed later with an EXEC statement. SQL Server supports stored
procedures with the following features:
1. Stored procedures can be divided into 3 groups based their accessing scopes:
 Permanent Procedures - Created for permanent use.
 Global Temporary Procedures - Created for temporary use within all sessions.
 Local Temporary Procedures - Created for temporary use within the same session.
2. Stored procedures can take input parameters.
3. Stored procedures can take output parameters.
4. Stored procedures can be mapped to references to Microsoft .NET Framework common language
runtime (CLR) methods
How To Create a Simple Stored Procedure?
If you want to create a simple stored procedure with no input and output parameters, you can use the
"CREATE PROCEDURE" command with a statement batch in a simple format as shown in below:
CREATE PROCEDURE procedure_name AS

statement_1;
statement_2;
...
statement_n;
GO
The following tutorial exercise shows you how to create a simple stored procedure:
USE FyiCenterData;
GO
CREATE PROCEDURE Hello AS

SELECT 'Welcome to:';
SELECT ' FYIcenter.com';
GO
EXEC Hello;
GO
-----------
Welcome to;
(1 row(s) affected)
----------------
FYIcenter.com
(1 row(s) affected)
How To Execute a Stored Procedure?
If you want execute a stored procedure created previously, you can use the EXECUTE statement in
the following formats:
EXEC procedure_name;
EXECUTE procedure_name;
The key word EXEC is actually optional. So you can execute a stored procedure by just entering the
procedure name as the statement. See examples in the following tutorial exercise:
USE FyiCenterData;
GO
-- create a quick procedure

CREATE PROCEDURE date AS
PRINT CONVERT(VARCHAR(20),GETDATE(),107);
GO
-- execute with EXEC

EXEC date;
GO
May 19, 2007
-- execute with EXEC

date;
GO
May 19, 2007
-- using a reserved keyword as procedure name

CREATE PROCEDURE datetime AS PRINT GETDATE();
GO
datetime;
GO
May 19, 2007 11:35PM
Looks like SQL Server allows you to reserved keywords as stored procedure names.
How To List All Stored Procedures in the Current Database?
If you want to see a list of stored procedures in your current database, you can use the system view,
sys.procedures as shown in this tutorial exercise:
USE FyiCenterData;
GO
SELECT * FROM sys.procedures;

GO
Name object_id schema_id type type_desc
---------- ----------- ----------- ---- --------------------
Hello 1621580815 1 P SQL_STORED_PROCEDURE
date 1653580929 1 P SQL_STORED_PROCEDURE
datetime 1669580986 1 P SQL_STORED_PROCEDURE
(3 row(s) affected)
How To Drop an Existing Stored Procedure?
If you have an existing procedure that you don't want to use it anymore, you should delete it from the
SQL Server by using the "DROP PROCEDURE" statement as shown in the tutorial example below:
USE FyiCenterData;
GO
DROP PROCEDURE datetime;

GO
SELECT * FROM sys.procedures;

GO
Name object_id schema_id type type_desc
---------- ----------- ----------- ---- --------------------
Hello 1621580815 1 P SQL_STORED_PROCEDURE
date 1653580929 1 P SQL_STORED_PROCEDURE
(2 row(s) affected)
Stored procedure "datetime" is no longer in the database.
How To Create a Stored Procedure with a Statement Block?
If you are creating a stored procedure with multiple statements, it's better to use "BEGIN ... END" to
group all statements into a single statement block.

USE FyiCenterData;
GO
CREATE PROCEDURE Show AS BEGIN

SELECT name, type_desc FROM sys.tables;
SELECT name, type_desc FROM sys.views;
SELECT name, type_desc FROM sys.procedures;
END;
GO
EXEC Show;
GO
name type_desc
------------------- ---------------------
fyi_random USER_TABLE
fyi_links_indexed USER_TABLE
fyi_links USER_TABLE
fyi_links_copy USER_TABLE
name type_desc
------------------- ---------------------
fyi_links_top VIEW
fyi_links_dump VIEW
fyi_links_view VIEW
name type_desc
------------------- ---------------------
Hello SQL_STORED_PROCEDURE
date SQL_STORED_PROCEDURE
Show SQL_STORED_PROCEDURE
How To End a Stored Procedure Properly?

Where the end of the "CREATE PROCEDURE" statement structure? The answer is simple, the end
of the statement batch.
Even if you are using a "BEGIN ... END" statement block, the stored procedure structure is not going
to end at the end of the statement block. It will continue to the end of the statement batch, usually the
GO command. The tutorial exercise gives you a good example:
USE FyiCenterData;
GO
DROP PROCEDURE ShowFaq;

DROP TABLE Faq;
GO
-- How this statement batch will be executed?

CREATE PROCEDURE ShowFaq AS BEGIN
PRINT 'Number of questions:';
SELECT COUNT(*) FROM Faq;
PRINT 'First 5 questions:'
SELECT TOP 5 * FROM Faq;
END;
CREATE TABLE Faq (Question VARCHAR(80));
GO
EXEC ShowFaq;
GO
Number of questions:
Msg 208, Level 16, State 1, Procedure ShowFaq, Line 3
Invalid object name 'Faq'.
What happened here was that the "CREATE TABLE" statement was not execueted. It was included
as part of the stored procedure "ShowFaq". This is why you were getting the error "Invalid object
name 'Faq'."
How To Generate CREATE PROCEDURE Script on an Existing Stored Procedure?
If you want to know how an existing stored procedure was created, you can use SQL Server
Management Studio to automatically generate a "CREATE PROCEDURE" script The following
tutorial shows you how to do this:
1. Run SQL Server Management Studio and connect to SQL server.
2. On the Object Explorer window, follow the object tree: Databases > FyiCenterData >
Programmability > Stored Procedures > dbo.ShowFaq.
3. Click right mouse button on dbo.ShowFaq. The context menu shows up.
4. Select "Script Stored Procedure as" > "CREATE to" > "New Query Editor Window". The following
script will be displayed:
USE [FyiCenterData]
GO
/****** Object: StoredProcedure [dbo].[ShowFaq]
Script Date: 05/19/2007 21:31:35 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
CREATE PROCEDURE [dbo].[ShowFaq] AS BEGIN
END;
How To Get the Definition of a Stored Procedure Back?
If you want get the definition of an existing stored procedure back from the SQL Server, you can use
the system view called sys.sql_modules, which stores definitions of views and stored procedures.
The sys.sql_modules holds stored procedure definitions identifiable by the object id of each view. The
tutorial exercise below shows you how to retrieve the definition of stored procedure, "ShowFaq" by
joining sys.sql_modules and sys.procedures:
USE FyiCenterData;
GO
SELECT m.definition
FROM sys.sql_modules m, sys.procedures p
WHERE m.object_id = p.object_id
AND p.name = 'ShowFaq';
GO
definition
-----------------------------------------
END;
(1 row(s) affected)
How To Modify an Existing Stored Procedure?
If you find a mistake in an existing stored procedure previously created, you can drop (delete) it and
create it again correctly. But dropping a stored procedure may affect other database objects who are
depending on this stored procedure.
So the best way to correct a mistake in an existing stored procedure is to use the "ALTER
PROCEDURE" statement as shown in the following tutorial example:
USE FyiCenterData;
GO
-- Finding a mistake - the last line is wrong

SELECT m.definition
GO
definition
-----------------------------------------
END;
(1 row(s) affected)
-- Modifying the stored procedure

ALTER PROCEDURE ShowFaq AS BEGIN
END;
GO
-- Confirmation of the correction

SELECT m.definition
GO
definition
-----------------------------------------
END;
(1 row(s) affected)
How To Create Stored Procedures with Parameters?
Very often, you need to create a stored procedure with one or more parameters. You only supply
values to those parameters at the time of executing the stored procedure.
Stored procedures with parameters can be created with the following syntax:
CREATE PROCEDURE procedure_name
@parameter_1 datatype,
@parameter_2 datatype,
...
@parameter_n datatype
AS
statement_1;
statement_2;
...
statement_n;
GO
The following tutorial exercise shows you how to create a stored procedure with one parameter called
@url:
USE FyiCenterData;
GO
DROP PROCEDURE Hello;

GO
CREATE PROCEDURE Hello

@url nvarchar(40)
AS
PRINT 'Welcome to ' + @url;
GO
EXEC Hello 'dba.fyicenter.com';

GO
Welcome to dba.fyicenter.com
How To Provide Values to Stored Procedure Parameters?
If a stored procedure is created with parameters, you need pass values to those parameters when
calling the stored procedure with one of two formats listed below:
-- Passing values only
EXEC procedure_name value_1, value_2, ... value_n;
-- Passing name-value pairs

EXEC procedure_name
@parameter_1 = value_1,
@parameter_2 = value_2,
...
@parameter_n = value_n;
The tutorial exercise below shows 2 ways to pass values to stored procedure parameters:
GO

@url nvarchar(40)
AS
GO
EXEC Hello 'dba.fyicenter.com';

GO
Welcome to dba.fyicenter.com
EXEC Hello @url='dev.fyicenter.com';

GO
Welcome to dev.fyicenter.com
What Are the Advantages of Passing Name-Value Pairs as Parameters?
When calling a stored procedure defined with parameters, you can pass values to those parameters
in two ways:
 Passing only values in the same order as parameters defined in the stored procedure.
 Passing name-value pairs in any order.
The advantages of passing name-value pairs to stored procedure parameters are:

 Makes the calling statement more readable - You know which value is passed to which
parameter.
 Makes it possible to pass values in an order different than how parameters are defined.
The tutorial exercise shows you some good examples of passing name-value pairs as parameters:
CREATE PROCEDURE diff_in_days
@start_date DATETIME,
@end_date DATETIME
AS BEGIN
PRINT CONVERT(VARCHAR(20),@end_date,107)
+ ' - '
+ CONVERT(VARCHAR(20),@start_date,107)
+ ' = '
+ STR(DATEDIFF(DAY, @start_date, @end_date));
END;
GO
EXEC diff_in_days
'01-Jan-2007',
'19-May-2007';
GO
May 19, 2007 - Jan 01, 2007 = 138
EXEC diff_in_days
@start_date='01-Jan-2007',
@end_date='19-May-2007';
GO
May 19, 2007 - Jan 01, 2007 = 138
-- Name-value pairs can be given in any order

EXEC diff_in_days
@end_date='19-May-2007',
@start_date='01-Jan-2007';
GO
May 19, 2007 - Jan 01, 2007 = 138
Can You Pass Expressions to Stored Procedure Parameters?
Can you pass expressions to stored procedure parameters? The answer is no.
When executing stored procedures, all input values must be entered as data literals, which can be
specified within single quotes ('), or without them if they cause no confusion. The tutorial exercise
below shows you how input values should be specified:
CREATE PROCEDURE area_of_circle @radius REAL
AS BEGIN
PRINT 'Radius = ' + STR(@radius,9,3);
PRINT 'Area = ' + STR(3.14*@radius*@radius,9,3);
END;
GO
-- Input value without quotes

EXEC area_of_circle 1.5;
GO
Radius = 1.500
Area = 7.065
-- Input value with quotes

EXEC area_of_circle '1.5';
GO
Radius = 1.500
Area = 7.065
-- Expressions are not allowed
EXEC area_of_circle 1.0+0.5;
GO
Incorrect syntax near '+'.
How To Provide Default Values to Stored Procedure Parameters?
If you add a parameter when creating a stored procedure, you can provide a default value so that the
execution statement is not required to pass input value to this parameter.
To provide a default value to a parameter, you should use this format: "@parameter_name data_type
= default_value". The tutorial exercise below shows you how provide default values to stored
procedure parameters:
USE FyiCenterData;
GO
DROP PROCEDURE diff_in_days;

GO

@end_date DATETIME = '19-May-2007'
AS BEGIN
PRINT CONVERT(VARCHAR(20),@end_date,107)
+ ' - '
+ ' = '
END;
GO
-- Default value is used

EXEC diff_in_days
@start_date='01-Jan-2007';
GO
May 19, 2007 - Jan 01, 2007 = 138
-- Default value is not used

EXEC diff_in_days
GO
May 11, 2007 - Jan 01, 2007 = 130
-- Input value must be supplied for a parameter
-- without a default value
EXEC diff_in_days
GO
Msg 201, Level 16, State 4, Procedure diff_in_days, Line 0
Procedure or Function 'diff_in_days' expects
parameter '@start_date', which was not supplied.
How To Define Output Parameters in Stored Procedures?
Sometime a stored procedure not only want to take input values from the calling statement batch, but
it also want to send output values back to the calling statement batch. This can be done by defining
output parameters in the CREATE PROCEDURE statement.
To define an output parameter, you should use this format: "@parameter_name data_type OUTPUT",
as shown in the following tutorial exercise:
DROP PROCEDURE diff_in_days;
GO
-- Defining an output parameter

@end_date DATETIME = '19-May-2007',
@days VARCHAR(40) OUTPUT
AS BEGIN
SET @days = CONVERT(VARCHAR(20),@end_date,107)
+ ' - '
+ ' = '
END;
GO
EXEC diff_in_days
@start_date='01-Jan-2007'
GO
Msg 201, Level 16, State 4, Procedure diff_in_days, Line 0
Procedure or Function 'diff_in_days' expects
parameter '@days', which was not supplied.
How To Receive Output Values from Stored Procedures?
If an output parameter is defined in a stored procedure, the execution statement must provide a
variable to receive the output value in the format: "@variable_name OUTPUT" or "@parameter_name
= @variable_name OUTPUT". The following tutorial exercise gives you a good example:
-- Using @variable format
DECLARE @message VARCHAR(40);
EXECUTE diff_in_days
'01-Jan-2007',
'11-May-2007',
@message OUTPUT;
PRINT @message;
GO
May 11, 2007 - Jan 01, 2007 = 130
-- Using @parameter = @variable format

EXEC diff_in_days
@days = @message OUTPUT;
PRINT @message;
GO
May 11, 2007 - Jan 01, 2007 = 130
-- Mixed formats are not allowed

EXEC diff_in_days
@message OUTPUT;
PRINT @message;
GO
Must pass parameter number 2 and subsequent parameters
as '@name = value'. After the form '@name = value' has
been used, all subsequent parameters must be passed in
the form '@name = value'.
How To Create a Local Temporary Stored Procedure?
A local temporary stored procedure is a special stored procedure that:

 Is created like a normal (permanent) stored procedure with the name prefixed with a number sign
(#).
 Are only valid in the same client session where it was created.
 Will be deleted when creating session is terminated.
This tutorial exercise here creates two stored procedures, one is permanent and the other is local
temporary:
DROP PROCEDURE #Hello;
GO

@url nvarchar(40)
AS
PRINT 'Welcome to ' + REVERSE(@url);
GO
CREATE PROCEDURE #Hello

@url nvarchar(40)
AS
GO
EXECUTE Hello 'fyicenter.com';

GO
Welcome to moc.retneciyf
EXECUTE #Hello 'fyicenter.com';

GO
Welcome to fyicenter.com
Can Another User Execute Your Local Temporary Stored Procedures?
Can another user execute your local temporary stored procedures? The answer is no.
To test this out, continue with the exercise from the previous tutorial. Keep that user session running
and open a new client session with sqlcmd.exe. Then run the tutorial script below:
-- Executing permanent procedure
-- created by another user session
EXECUTE Hello 'fyicenter.com';
GO
Welcome to moc.retneciyf
-- Executing local temporary procedure

-- created by another user session
EXECUTE #Hello 'fyicenter.com';
GO
Could not find stored procedure '#Hello'.
SQL Server FAQs - Transact-SQL Understanding User_Defined_Functions
A collection of 17 FAQs on SQL Server Transact-SQL user defined functions. Clear answers are
provided with tutorial exercises on creating and dropping user defined functions; defining and passing
parameters into functions; providing and using parameter default values; defining and using table-
valued functions.
What Are User Defined Functions?
A user defined function is a collection of Transact-SQL statements that stored in the SQL Server. A
user defined function will return data when executed.
A user defined function works in the same way as a system function. It can be used as a scalar
expression or a temporary table depending on the turning data format. A system function is provided
as part of the SQL Server. But a user defined function must be created and managed by yourself.
If you are using a group of statements to calculate the same data repeatedly in different places, you
should consider to create a user defined function for that group of statements.
User defined functions are also called functions.
What Are the Differences between User Defined Functions and Stored Procedures?
Differences between user defined functions and stored procedures are:

 Stored procedures does not return any data and they can not be used in expressions.
 User defined functions does return data and they can be used in expressions.
 Stored procedures only takes data constants as input parameters.
 User defined functions can take data constants and expressions as input parameters.
 Stored procedures support output parameters.
 User defined functions do not support output parameters.
 Stored procedures take a parameter's default value, if it is not specified in the calling statement.
 User defined functions take a parameter's default value, if it is specified with the keyword
DEFAULT in the calling statement.
 Stored procedures are called with no parenthesis to include parameters.
 User defined functions are called with parenthesis to include parameters. Empty parenthesis is
needed if no parameters.
 Stored procedures can be created locally and temporarily.
 User defined functions can not be created locally and temporarily.
 To Create a Simple User Defined Function?

 If you want to create a simple user defined function, you can use the "CREATE FUNCTION"
command with a statement block in a simple format as shown in below:
 CREATE FUNCTION function_name()
 RETURNS data_type
 AS BEGIN
 statement_1;
 statement_2;
 ...
 statement_n;
 RETURN expression;
 END;
 GO
 The following tutorial exercise shows you how to create a simple user defined function:
 GO

 CREATE FUNCTION Welcome()
 RETURNS VARCHAR(40)
 AS BEGIN
 RETURN 'Welcome to FYIcenter.com';
 END;
 GO

 PRINT dbo.Welcome();
 GO
 Welcome to FYIcenter.com
 How To Use User Defined Functions in Expressions?
 An user defined function must return a value, which can be used in any expression as long as the
return value data type matches the expression.
 To execute a user defined function and use its return value in an expression, you just need to
enter the schema name and the function name as a value in the expression. The tutorial exercise
below shows you how use a user defined function in an expression:
 GO

 -- Calling a function without schema name
 PRINT 'Hi there, '+Welcome();
 GO
 'Welcome' is not a recognized built-in function name.

 -- Calling a function with schema name
 PRINT 'Hi there, '+dbo.Welcome();
 GO
 Hi there, Welcome to FYIcenter.com
 How To List All User Defined Functions in the Current Database?
 If you want to see a list of all user defined functions in your current database, you can use the
system view, sys.objects as shown in this tutorial exercise:
 GO

 -- Number of Sundays in this year
 CREATE FUNCTION Sundays()
 RETURNS INT
 AS BEGIN
 DECLARE @date DATETIME;
 DECLARE @count INT;
 SET @date = '2006-12-31';
 SET @count = 0;
 WHILE DATEPART(YEAR, @date) <= 2008 BEGIN
 SET @date = DATEADD(DAY, 1, @date);
 IF DATENAME(WEEKDAY, @date) = 'Sunday'
 SET @count = @count + 1;
 END;
 RETURN @count;
 END;
 GO


 SELECT * FROM sys.objects WHERE type_desc LIKE '%FUNCTION';
 GO
 name object_id schema_id type type_desc
 -------- ----------- ---------- ---- -------------------
 Welcome 2085582468 1 FN SQL_SCALAR_FUNCTION
 Sundays 2117582582 1 FN SQL_SCALAR_FUNCTION
 sys.objects contains all types of objects in the current database. You need select only the
FUNCTION object type.
 How To Drop an Existing User Defined Function?
 If you have an existing user defined function that you don't want to use it anymore, you should
delete it from the SQL Server by using the "DROP FUNCTION" statement as shown in the tutorial
example below:
 GO

 DROP FUNCTION Welcome;
 GO

 SELECT * FROM sys.objects WHERE type_desc LIKE '%FUNCTION';
 GO
 name object_id schema_id type type_desc
 -------- ----------- ---------- ---- -------------------
 Sundays 2117582582 1 FN SQL_SCALAR_FUNCTION
 User defined function "Welcome" is no longer in the database.
 How To Generate CREATE FUNCTION Script on an Existing Function?
 If you want to know how an existing user defined function was created, you can use SQL Server
Management Studio to automatically generate a "CREATE FUNCTION" script The following
tutorial shows you how to do this:
 2. On the Object Explorer window, follow the object tree: Databases > FyiCenterData >
Programmability > Functions > Scalar-valued Functions > dbo.Sundays.
 3. Click right mouse button on dbo.Sundays. The context menu shows up.
 4. Select "Script Function as" > "CREATE to" > "New Query Editor Window". The following script
will be displayed:
 USE [FyiCenterData]
 GO
 /****** Object: UserDefinedFunction [dbo].[Sundays]
 Script Date: 05/19/2007 23:24:36 ******/
 SET ANSI_NULLS ON
 GO
 SET QUOTED_IDENTIFIER ON
 GO
 CREATE FUNCTION [dbo].[Sundays]()
 RETURNS INT
 AS BEGIN
 SET @date = '2006-12-31';
 SET @count = 0;
 END;
 RETURN @count;
 END;
 How To Get the Definition of a User Defined Function Back?
 If you want get the definition of an existing user defined function back from the SQL Server, you
can use the system view called sys.sql_modules, which stores definitions of functions, stored
procedures, and views.
 The sys.sql_modules holds user defined function definitions identifiable by the object id of each
function. The tutorial exercise below shows you how to retrieve the definition of stored procedure,
"Sundays" by joining sys.sql_modules and sys.objects:
 GO

 SELECT m.definition
 FROM sys.sql_modules m, sys.objects o
 WHERE m.object_id = o.object_id
 AND o.name = 'Sundays';
 GO
 definition
 -----------------------------------------------
 CREATE FUNCTION Sundays()
 RETURNS INT
 AS BEGIN
 SET @date = '2006-12-31';
 SET @count = 0;
 SET @date = DATEADD(DAY, 1,
 How To Modify an Existing User Defined Function?
 If you find a mistake in an existing function previously created, you can drop (delete) it and create
it again correctly. But dropping a function may affect other database objects who are depending
on this function.
 So the best way to correct a mistake in an existing function is to use the "ALTER FUNCTION"
statement as shown in the following tutorial example:
 GO

 -- Modifying an existing function
 ALTER FUNCTION Sundays()
 RETURNS INT
 AS BEGIN
 SET @date = '2006-12-31';
 SET @count = 0;
 WHILE DATEPART(YEAR, @date) < 2008 BEGIN
 END;
 RETURN @count;
 END;
 GO
 Do you know what correction has been made on this function?
 How To Create User Defined Functions with Parameters?
 Very often, you need to create a function with one or more parameters so that the function can be
more generic. You only supply values to those parameters at the time of executing the function.
 User defined functions with parameters can be created with the following syntax:
 CREATE FUNCTION function_name (
 @parameter_1 data_type,
 @parameter_2 data_type,
 ...
 @parameter_n data_type
 )
 RETURNS data_type
 AS BEGIN
 statement_1;
 statement_2;
 ...
 statement_n;
 END;
 The following tutorial exercise shows you how to create a function with one parameter called
@url:
 GO

 GO

 CREATE FUNCTION Welcome(@url VARCHAR(40))
 AS BEGIN
 RETURN 'Welcome to '+@url;
 END;
 GO

 PRINT 'Hi there, '+dbo.Welcome('dba.FYIcenter.com');
 GO
 Hi there, Welcome to dba.FYIcenter.com
 How To Provide Values to User Defined Function Parameters?
 If a user defined function is created with parameters, you need pass values to those parameters
when calling the function with one of two formats listed below:
 expression... function_name(value_1, value_2, ... value_n)...
 The tutorial exercise below shows you how to pass values to function parameters:
 GO

 GO

 CREATE FUNCTION Welcome(@url VARCHAR(40))
 AS BEGIN
 RETURN 'Welcome to '+@url;
 END;
 GO

 PRINT 'Hi there, '+dbo.Welcome('dba.FYIcenter.com');
 GO
 Hi there, Welcome to dba.FYIcenter.com

 PRINT 'Hi there, '+dbo.Welcome('dev.FYIcenter.com');
 GO
 Hi there, Welcome to dev.FYIcenter.com
 Can You Pass Expressions to Function Parameters?
 Can you pass expressions to stored procedure parameters? The answer is yes.
 When executing functions, input values can be written as expressions. But the resulting value
data type must match the parameter. The tutorial exercise below shows you how input values
should be specified:
 GO

 CREATE FUNCTION Area(@radius REAL)
 RETURNS REAL
 AS BEGIN
 RETURN 3.14*@radius*@radius;
 END;
 GO

 -- Input value data matches the parameter
 PRINT 'Area of a circle: '+STR(dbo.Area(1.5),9,3);
 GO
 Area of a circle: 7.065

 -- Input value data does not match the parameter
 PRINT 'Area of a circle: '+STR(dbo.Area('1.5'),9,3);
 GO
 Incorrect syntax near 'Area'.

 -- Expressions are allowed
 PRINT 'Area of a circle: '+STR(dbo.Area(1.0+0.5),9,3);
 GO
 Area of a circle: 7.065
How To Provide Default Values to Function Parameters?
If you add a parameter when creating a stored procedure, you can provide a default value so that the
execution statement is not required to pass input value to this parameter:
 To define a default value to a parameter when creating the function, you should use this format:
"@parameter_name data_type = default_value".
 To use the default value of a parameter when executing the function, you should use the keyword
DEFAULT as the input value for that parameter.
The tutorial exercise below shows you how provide default values to function parameters:
USE FyiCenterData;
GO
CREATE FUNCTION Age_In_Days (

@birth_date DATETIME,
@today DATETIME = NULL
)
RETURNS INT
AS BEGIN
IF @today IS NULL SET @today = GETDATE();
RETURN DATEDIFF(DAY, @birth_date, @today);
END;
GO
-- Default value is used

PRINT 'Age in days: '+STR(
dbo.Age_In_Days('01-Jan-2007', DEFAULT));
GO
Age in days: 138
-- Default value is not used

dbo.Age_In_Days('01-Jan-2007', '11-May-2007'));
GO
Age in days: 130
-- Can not skip input values

-- even for parameters with default values
dbo.Age_In_Days('01-Jan-2007'));
GO
An insufficient number of arguments were supplied
for the procedure or function dbo.Age_In_Days.
How Many Categories of Functions based Their Return Modes?
SQL Server supports 2 categories of user defined functions based on their return modes:
1. Scalar-valued Functions - A function that returns a single value. Scalar-valued functions can be
used in scalar expressions. Below are some scalar-valued functions:
PRINT GETDATE();
GO
May 19 2007 1:26PM
PRINT 'URL reversed: '+REVERSE('dba.fyicenter.com');

GO
URL reversed: moc.retneciyf.abd
2. Table-valued Functions - A function that returns data in rows and columns like a table. Table-
valued functions can be used in table expressions like the FROM clause of SELECT statements
Below are some scalar-valued functions:
SELECT * FROM fn_helpcollations() WHERE name LIKE 'French_CI%'
GO
name
-------------------
French_CI_AI
French_CI_AI_WS
French_CI_AI_KS
French_CI_AI_KS_WS
French_CI_AS
French_CI_AS_WS
French_CI_AS_KS
French_CI_AS_KS_WS
(8 row(s) affected)
SELECT i.index_id, i.name, s.avg_fragmentation_in_percent

FROM sys.dm_db_index_physical_stats (
DB_ID(N'FyiCenterData'),
OBJECT_ID(N'fyi_links_indexed'),
DEFAULT, DEFAULT, DEFAULT) s, sys.indexes i
WHERE s.object_id = i.object_id
AND s.index_id = i.index_id;
GO
index_id name avg_fragmentation_in_percent
0 NULL 0.574712643678161
2 fyi_links_url 0
3 fyi_links_counts 0
How Many Ways to Create Table-Valued Functions?
SQL Server supports two syntaxes of creating table-valued functions:
1. Inline Table-valued Functions - A table-valued function created with a single SELECT statement:
CREATE FUNCTION function_name(
@parameter_1 data_type,
...
@parameter_n data_type
)
RETURNS TABLE
AS
RETURN (select_statement);
2. Multi-statement Table-valued Functions - A table-valued function created with a local temporary

table and a statement block:
CREATE FUNCTION function_name(
...
@parameter_n data_type
)
RETURNS @table_variable_name TABLE (
column_definition_list)
AS BEGIN
statement_1;
statement_2;
...
statement_n;
RETURN
END
How To Create an Inline Table-Valued Function?
To create an inline table-valued function, you need to use the "RETURNS TABLE" clause in the
"CREATE FUNCTION" statement. There should be no function body, except for a RETURN
statement with a SELECT subquery:
An inline table-valued function can be viewed as a select statement with parameters, see the
example showing in this tutorial exercise:
USE FyiCenterData;
GO
CREATE FUNCTION Top_Links(@level INT)

RETURNS TABLE
AS
RETURN (SELECT * FROM fyi_links WHERE counts > @level);
GO
SELECT counts, id, url FROM Top_Links(999900) ORDER BY counts DESC;

GO
counts id url
----------- ----------- -----------------------------------
999966 36470 dgqnvmy pyjqd toqcoupuxortasdtzvc
999953 12292 qebmw v qqmywe q kza wskxqns j
999923 79161 kvwwg g
999920 19124 prlg fzoio
999909 90930 xqmeal ikv isx y r
(6 row(s) affected)
How To Create an Multi-Statement Table-Valued Function?
To create a multi-statement table-valued function, you need to define a temporary table as the
returning table in the function. INSERT statements should be used to insert data into the returning
table.
The tutorial exercise below shows you a simple example of how to build a temporary table and make
it as the returning table of a function:
USE FyiCenterData;
GO
CREATE FUNCTION Yearly_Stats(

@start_year INT, @end_year INT)
RETURNS @stats TABLE (year INT,
min INT, max INT, counts INT)
AS BEGIN
DECLARE @year INT;
SET @year = @start_year;
WHILE @year <= @end_year BEGIN
INSERT INTO @stats
SELECT @year AS year, MIN(counts) AS min,
MAX(counts) AS max, COUNT(*) AS counts
FROM fyi_links WHERE DATEPART(YEAR, created) = @year;
SET @year = @year + 1;
END;
RETURN;
END
GO
SELECT * FROM dbo.Yearly_Stats(1900, 1905);

GO
year min max counts
----------- ----------- ----------- -----------
1900 -999932 996991 638
1901 -997138 991874 592
1902 -996779 997315 594
1903 -995476 998520 636
1904 -994838 998956 635
1905 -993178 996249 595
(6 row(s) affected)
SQL Server FAQs - Creating and Managing Schemas
A collection of 11 FAQs on SQL Server on creating and managing schemas. Clear answers are
provided with tutorial exercises on listing all schemas and all objects in a schema; transferring tables
from one schema to another schema; changing ownership of a schema
What Is a Schema in SQL Server 2005?
A schema is a container of database objects with the following interesting related rules:
 A schema may contain different object types, like tables, indexes, views, procedures, functions,
etc.
 A database user can be assigned with a default schema.
 Object names must be prefixed with schema names when referencing schemas outside your
default schema.
 Every schema has a single owner (a database user). A database user may own multiple
schemas.
 If you login name is mapped to the owner of a schema, you have full permissions on this schema.
 To drop a schema, you need to empty the schema first.
 How To Create a New Schema in a Database?

 If you want to create a new schema in an existing database, you can use the "CREATE
SCHEMA" statement as shown in the tutorial example below:
 GO

 CREATE SCHEMA fyi;
 GO
 A new schema called "fyi" has been created in "FyiCenterData" database. "fyi" is an empty
schema at this moment since no objects has been moved into "fyi" yet.
 How To List All Schemas in a Database?
 If you want see all existing schemas in the current database, you can use view sys.schemas as
shown in the example below:
 GO

 SELECT * FROM sys.schemas;
 GO
 name schema_id principal_id
 -------------------- ----------- ------------
 dbo 1 1
 guest 2 2
 INFORMATION_SCHEMA 3 3
 sys 4 4
 fyi 5 1
 db_owner 16384 16384
 db_accessadmin 16385 16385
 db_securityadmin 16386 16386
 db_ddladmin 16387 16387
 db_backupoperator 16389 16389
 db_datareader 16390 16390
 db_datawriter 16391 16391
 db_denydatareader 16392 16392
 db_denydatawriter 16393 16393
 All schemas, except "fyi", in the list were created by SQL Server.
 How To Create a New Table in a Given Schema?
 When you create a new table, you can specify in which schema you want this table to be located
by prefixing the table name with the schema name. In the tutorial example below, a new table
"test" is created in schema "fyi":
 GO

 CREATE TABLE fyi.test (id INT);
 GO

 SELECT t.name AS table_name, t.type_desc,
 s.name AS schema_name
 AND t.name = 'test';
 GO
 table_name type_desc schema_name
 ----------- ----------- ------------
 test USER_TABLE fyi
 The last query confirms that table "test" is inside schema "fyi".
 How To Transfer an Existing Table from One Schema to Another Schema?
 If you want to move an existing table from one schema to another schema, you can use the
"ALTER SCHEMA ... TRANSFER ..." statement as shown in the tutorial exercise below:
 -- Login with "sa"

 GO

 -- Confirming that "fyi_links" is in "dbo"
 AND t.name = 'fyi_links';
 GO
 ----------- ----------- ------------
 fyi_links USER_TABLE dbo

 -- Moving a table to a new schema
 ALTER SCHEMA fyi TRANSFER fyi_links;
 GO

 -- Confirming that "fyi_links" is moved to "fyi"
 AND t.name = 'fyi_links';
 GO
 ----------- ----------- ------------
 fyi_links USER_TABLE fyi
 The last query confirms that table "fyi_links" is now in schema "fyi".
 How To List All Objects in a Given Schema?
 If you are wonder what objects are stored in a given schema as an object contrainer, you can use
view "sys.objects" to get a list of all objects in a schema. The tutorial exercise shows you how to
list all objects in schema "fyi" and "dbo":
 -- Login with 'sa'

 GO

 -- What is "fyi"?
 SELECT o.name, o.schema_id, o.type_desc
 FROM sys.objects o, sys.schemas s
 WHERE o.schema_id = s.schema_id
 AND s.name = 'fyi';
 GO
 name schema_id type_desc
 ------------------------------ ---------- ------------------
 test 5 USER_TABLE
 fyi_links 5 USER_TABLE
 UQ__fyi_links__4222D4EF 5 UNIQUE_CONSTRAINT
 DF__fyi_links__creat__4316F928 5 DEFAULT_CONSTRAINT

 SELECT o.name, o.schema_id, o.type_desc
 FROM sys.objects o, sys.schemas s
 WHERE o.schema_id = s.schema_id
 AND s.name = 'dbo';
 GO
 name s_id type_desc
 ---------------------------- ---- --------------------------
 Welcome 1 SQL_SCALAR_FUNCTION
 Age_In_Days 1 SQL_SCALAR_FUNCTION
 Top_Links 1 SQL_INLINE_TABLE_VALUED...
 Yearly_Stats 1 SQL_TABLE_VALUED_FUNCTION
 fyi_links_indexed 1 USER_TABLE
 DF__fyi_links__creat__403A8C 1 DEFAULT_CONSTRAINT
 fyi_links_copy 1 USER_TABLE
 fyi_links_dump 1 VIEW
 fyi_links_view 1 VIEW
 Show 1 SQL_STORED_PROCEDURE
 Hello 1 SQL_STORED_PROCEDURE
 QueryNotificationErrorsQueue 1 SERVICE_QUEUE
 EventNotificationErrorsQueue 1 SERVICE_QUEUE
 ServiceBrokerQueue 1 SERVICE_QUEUE
 DoubleX 1 SQL_SCALAR_FUNCTION
 ...
 Looks like we have a lots of different types objects stored in schema "dbo".
What Is the Default Schema of Your Login Session?
When you login to a SQL Server and select a database to use, SQL Server will assign your login
session a default schema. The schema name can be omitted when you refer to objects in the default
schema. Here is what you should remember about default schema:
 The default schema of your login session in the current database is the default schema assigned
to the current database level principal - database user.
 If you are referring to an object in the default schema, you do not need to specify the schema
name.
 If you are referring to an object outside the default schema, you must specify the schema name.
The tutorial exercise below shows you how to verify your default schema:
-- Login with "fyi_login"
USE FyiCenterData;
GO
Changed database context to 'FyiCenterData'.
PRINT User_Name();
GO
Fyi_User
SELECT name, default_schema_name

FROM sys.database_principals WHERE type = 'S';
GO
name default_schema_name
-------------------- --------------------
dbo dbo
guest guest
INFORMATION_SCHEMA NULL
sys NULL
Fyi_User dbo
The last query shows that the default schema for "fyi_login" in "FyiCenterData" is "dbo".
Who Is the Owner of a Schema?
When you create a schema in a database, SQL Server will assign a owner (a database user) to this
schema. If your login name is mapped to the owner of a schema at the database level, you have the
full permission on all objects in this schema.
The following tutorial exercise shows you how to see who is the owner of a schema:
-- Login with "sa"
USE FyiCenterData;
GO
SELECT s.name, u.name AS owner

FROM sys.schemas s, sys.database_principals u
WHERE s.principal_id = u.principal_id;
GO
name owner
------------------- --------------------
dbo dbo
fyi dbo
guest guest
...
The last query shows that schame "fyi" is owned by "dbo".
How To Change the Ownership of a Schema?
If you want to change the owner of a schema, you can use the "ALTER AUTHORIZATION" statement
using the following syntax:
ALTER AUTHORIZATION ON SCHEMA::schema_name TO user_name
The following tutorial example shows you how to change ownership of schema "fyi" to "fyi_user":
-- Login with "sa"
USE FyiCenterData;
GO
ALTER AUTHORIZATION ON SCHEMA::fyi TO fyi_user

GO
SELECT s.name, u.name AS owner
FROM sys.schemas s, sys.database_principals u
WHERE s.principal_id = u.principal_id;
GO
name owner
------------------- --------------------
dbo dbo
fyi Fyi_User
guest guest
...
What Happens If You Are Trying to Access a Schema Not Owned by You?
In general, if you are trying to access an object in schema owned by another database user, you will
get a "permission denied" error, unless that you have been granted access permission to that object
explicitly. Here is a tutorial example showing you the permission error:
-- Login with "fyi_login"
USE FyiCenterData;
GO
PRINT User_Name();
GO
Fyi_User
SELECT COUNT(*) FROM dbo.fyi_random;

GO
SELECT permission denied on object 'fyi_random',
database 'FyiCenterData', schema 'dbo'
How To Drop an Existing Schema?
If you want to delete a schema, you need to move all objects out of that schema, then use the "DROP
SCHEMA" statement to delete the schema. The tutorial exercise below shows you how to drop
schema "fyi":
-- Login with "sa"
USE FyiCenterData;
GO
-- Drop failed because schema is not empty

DROP SCHEMA fyi;
GO
Cannot drop schema 'fyi' because it is being referenced
by object 'DF__fyi_links__creat__4316F928'.
-- Move one table out

ALTER SCHEMA dbo TRANSFER fyi.fyi_links;
GO
-- Delete one table

DROP TABLE fyi.test;
GO
-- Dropped ok
DROP SCHEMA fyi;
GO
SQL Server FAQs - Introduction to Triggers
A collection of 20 FAQs on creating and managing triggers. Clear explanations and tutorial exercises
are provided on creating DML and DDL triggers; altering, dropping, enabling and disabling triggers;
accessing inserted and deleted records; overriding DML statements.
What Are Triggers?
A trigger is a special kind of stored procedure that automatically executes when an event occurs in
the database server. A trigger is really an event handler. SQL Server allows users to create triggers
(event handlers) for 3 types of events:
 DML Event - Occurs when a DML (Data Manipulation Language) statement: INSERT, UPDATE
or DELETE, is executed.
 DDL Event - Occurs when a DDL (Data Definition Language) statement: CREATE, ALTER, or
DROP, is executed.
 Logon Event - Occurs when a user logins to the Server.
There are 3 different types of triggers (event handlers) based on the types of events they are
triggered by:
 DML Trigger - Executes in response to a DML event.
 DDL Trigger - Executes in response to a DDL event.
 Logon Trigger - Executes in response to a logon event.
What Are the Basic Features of a Trigger?
Since a SQL Server trigger is a really an event handler, it has the following basic features similar to
event handlers in other programming languages:
 Event Type - It must be declared to handle a specific event, like a DELETE event.
 Object Scope - It must be declared to handle events in a specific database object scope, like a
specific table.
 Statement Body - It must have a statement body, a batch of statements to be executed when the
specified event occurs in specified database object scope. An event handler (trigger) with an
empty statement body is useless.
 Access of Event Attributes - It must have access to some attributes of the event, so it can have
different logics for different instances of the event.
For example, you can implement a trigger to send a security alert message to each user whenever
his or her password is changed. This trigger should have the following features:
 Event Type - It must be declared to handle the UPDATE event.
 Object Scope - It must be declared to handle the UPDATE event on the user password table only.
 Statement Body - It must have a statement body to determine if the password is really changed or
not. If it is changed, send an email to the user's email address.
 Access of Event Attributes - It must have access to some attributes of the event instance, like the
old value and the new value of the password, and the user email address.
 How To Create a Simple Table to Test Triggers?

 If you want to follow other tutorial examples included in this collection, you need to run this SQL
script to create a simple table called fyi_users:
 GO

 DROP TABLE fyi_users;
 GO

 CREATE TABLE fyi_users (
 id INTEGER IDENTITY NOT NULL,
 name VARCHAR(80) NOT NULL,
 email VARCHAR(80) NULL,
 password VARCHAR(32) NULL
 );

 INSERT INTO fyi_users (name) VALUES ('John King');
 INSERT INTO fyi_users (name) VALUES ('Nancy Greenberg');
 GO
 fyi_users is created now with 2 records.
 How To Create a DML Trigger using CREATE TRIGGER Statements?
 A DML trigger is a trigger declared to handle a DML event, which occurs when an INSERT,
UPDATE or DELETE statement is executed. If you want to create a DML trigger, you should use
the "CREATE TRIGGER" statement in the following format:
 CREATE TRIGGER trigger_name ON table_name
 AFTER INSERT, UPDATE, DELETE
 AS
 statements
 GO
 The tutorial exercise below shows you a very simple DML trigger defined on the fyi_users table. It
does nothing but printing a simple static message.
 GO

 CREATE TRIGGER dml_message ON fyi_users
 AS
 PRINT 'Records are inserted, updated,'
 + ' or deleted in fyi_users';
 GO
 A simple DML trigger is defined on fyi_users now.
 How To Test a DML Trigger?
 To test a DML trigger defined on a table, you just need to execute several INSERT, UPDATE and
DELETE statements on that table as shown in this tutorial example:
 GO

 INSERT INTO fyi_users (name) VALUES ('FYI Admin');
 GO
 Records are inserted, updated, or deleted in fyi_users

 UPDATE fyi_users SET email='root@fyicenter'
 WHERE name = 'FYI Admin';
 GO

 DELETE FROM fyi_users WHERE name = 'FYI Admin';
 GO
 The trigger, dml_message, is working as expected
 How To List All Triggers in the Database with sys.triggers?
 If you want to list all triggers defined in the current database, you can use the catalog view,
sys.triggers, as shown in the following tutorial example:
 GO

 CREATE TRIGGER new_user ON fyi_users
 AFTER INSERT
 AS
 PRINT 'New users added.';
 GO

 SELECT * FROM sys.triggers
 GO
 name object_id parent_id type type_desc
 ------------ ----------- ----------- ---- ------------
 dml_message 690101499 674101442 TR SQL_TRIGGER
 new_user 706101556 674101442 TR SQL_TRIGGER
 The result shows that there are 2 triggers defined in FyiCenterData
 How To Modify Existing Triggers using "ALTER TRIGGER"?
 If you want to make changes to an existing trigger, you could use the "ALTER TRIGGER"
statements to refine the trigger again. The tutorial exercise below shows you how to modify the
trigger defined in a previous tutorial:
 GO

 ALTER TRIGGER dml_message ON fyi_users
 AS
 PRINT 'Time: '+CONVERT(VARCHAR(12),GETDATE());
 GO

 UPDATE fyi_users SET email='john@fyicenter' WHERE id = 1;
 GO
 Time: Jul 1 2007
 An extra printing statement is added the trigger.
 How To Delete Existing Triggers using "DROP TRIGGER"?
 If you don't want to use a trigger any more, you should delete it from the database by using the
"DROP TRIGGER" statement as shown in tutorial example:
 GO

 DROP TRIGGER new_user;
 GO

 SELECT * FROM sys.triggers
 GO
 name object_id parent_id type type_desc
 ------------ ----------- ----------- ---- ------------
 dml_message 690101499 674101442 TR SQL_TRIGGER
 Trigger, new_user, is deleted now.
 How To Get the Definition of a Trigger Back?
 If you want get the definition of an existing trigger back from the SQL Server, you can use the
catalog view called sys.sql_modules, which stores definitions of views, stored procedures, and
triggers.
 The sys.sql_modules holds trigger definitions identifiable by the object id of each trigger. The
tutorial exercise below shows you how to retrieve the definition of trigger, "dml_message" by
joining sys.sql_modules and sys.triggers:
 GO

 SELECT m.definition
 FROM sys.sql_modules m, sys.triggers t
 WHERE m.object_id = t.object_id
 AND t.name = 'dml_message';
 GO
 definition
 -------------------------------------------------
 CREATE TRIGGER dml_message ON fyi_users
 AS
 PRINT 'Time: '+CONVERT(VARCHAR(12),GETDATE());
 How To Disable Triggers using "DISABLE TRIGGER"?
 If want to stop the execution of an existing trigger temporarily, you can use the "DISABLE
TRIGGER" statement to disable it. The disabled trigger will be kept in the database.
 If you want to resume the execution of a disabled trigger, you can use the "ENABLE TRIGGER"
statement to enable it.
 The tutorial exercise below shows you how to disable and enable triggers:
 GO

 -- disabling a trigger
 DISABLE TRIGGER dml_message ON fyi_users;
 GO

 INSERT INTO fyi_users (name) VALUES ('Jack Gate');
 GO

 -- enabling a trigger
 ENABLE TRIGGER dml_message ON fyi_users;
 GO

 INSERT INTO fyi_users (name) VALUES ('Roy Bush');
 GO
 Time: Jul 1 2007
 How To Create a Trigger for INSERT Only?
 The trigger, dml_message, provided in previous tutorials was defined to handle all 3 types of DML
statements, INSERT, UPDATE, and DELETE.
 If you do not want the trigger to handle all 3 types of DML statements, you can list only 1 or 2 of
the statement keywords. For example, the following SQL script defines a trigger that only handle
the INSERT statement events:
 GO

 CREATE TRIGGER new_user ON fyi_users
 AFTER INSERT
 AS
 PRINT 'Time: '+CONVERT(VARCHAR(12),GETDATE())
 + ' New users added.';
 GO

 INSERT INTO fyi_users (name) VALUES ('Marc Kumar');
 GO
 Time: Jul 1 2007
 Time: Jul 1 2007 New users added.

 UPDATE fyi_users SET email='marc@fyicenter'
 WHERE name = 'Marc Kumar';
 GO
 Time: Jul 1 2007
 Notice that the INSERT statement triggered two triggers to be executed: dml_message and
new_user. But the UPDATE statement triggered one trigger to be executed: dml_message as
expected.
 It is also interesting to know that when multiple triggers are defined to handle the same event, the
oldest (defined first) will be executed first.
 How To See the Event List of an Existing Trigger using sys.trigger_events?
 If what are the DML events an existing tigger is handling, you can use the catalog view,
sys.trigger_events. You need to join sys.trigger_events and sys.triggers to get a better list as
shown in this tutorial example:
 GO

 SELECT t.name, e.type, e.type_desc
 FROM sys.trigger_events AS e, sys.triggers AS t
 WHERE e.object_id = t.object_id
 GO
 name type type_desc
 -------------- ------ ---------
 dml_message 1 INSERT
 dml_message 2 UPDATE
 dml_message 3 DELETE
 new_user 1 INSERT
 The list clearly shows that dml_message handles 3 events: INSERT, UPDATE and DELETE.
How To Access the Inserted Record of an Event?
When a DML event occurs, SQL Server will prepare a temporary table called "INSERTED", which
contains the new record of the affected row, which is:
 A copy of the inserted row for an INSERT statement.
 A copy of the updated row for an UPDATE statement.
 Empty for a DELETE statement.
The tutorial exercise below shows you how to create a trigger, update_user, to report email changes
on table, fyi_users:
USE FyiCenterData;
GO
DISABLE TRIGGER dml_message ON fyi_users;

GO
CREATE TRIGGER update_user ON fyi_users
AFTER UPDATE
AS
DECLARE @new VARCHAR(80);
SELECT @new = email FROM INSERTED;
PRINT 'Email changed to '+@new;
GO
UPDATE fyi_users SET email='john@fyicenter'

WHERE name = 'John King';
GO
Email changed to john@fyicenter
(1 row(s) affected)
As you can see, the INSERTED table is helpful, if you want the trigger to perform specific logics on
the affected rows.
How To Access the Deleted Record of an Event?
When a DML event occurs, SQL Server will prepare a temporary table called "DELETED", which
contains the old record of the affected row, which is:
 A copy of the deleted row for a DELETE statement.
 A copy of the row to be updated for an UPDATE statement.
 Empty for an INSERT statement.
The tutorial exercise below shows you how to improve the trigger, update_user, to report email
changes on table, fyi_users, with both old and new emails:
USE FyiCenterData;
GO
ALTER TRIGGER update_user ON fyi_users

AFTER UPDATE
AS
DECLARE @old VARCHAR(80);
SELECT @old = email FROM DELETED;
PRINT 'Email changed from '+@old+' to '+@new;
GO
UPDATE fyi_users SET email='king@fyicenter'

GO
Email changed from john@fyicenter to king@fyicenter
(1 row(s) affected)
INSERTED and DELETED are working as expected. The reported message is getting better.
How To Improve the Trigger to Handle NULL Values?
When a NULL value is concatenated with a string, the result will be a null value. So if you want the
trigger to properly report NULL values, you need to enhance the trigger as shown in the following
tutorial example:
USE FyiCenterData;
GO
-- trigger executed but printed a NULL

UPDATE fyi_users SET email=NULL
GO
(1 row(s) affected)

AFTER UPDATE
AS
DECLARE @old VARCHAR(80);
SELECT @old = email FROM DELETED;
SELECT @new = ISNULL(email,'NULL') FROM INSERTED;
SELECT @old = ISNULL(email,'NULL') FROM DELETED;
PRINT 'Email changed from '+@old+' to '+@new;
GO
-- trigger is reporting NULL values now

UPDATE fyi_users SET email=NULL
GO
Email changed from NULL to NULL
(1 row(s) affected)
The trigger is handling NULL values properly now.
What Happens to a Trigger with Multiple Affected Rows?
If there is only one row affected by a DML statement, we know that the DML trigger will be executed
once. But how many times the DML trigger will be executed if the DML statement resulted multiple
affected rows? The answer is still one.
In the case of multiple affected rows, both INSERTED and DELETED tables will contain multiple
rows.
If you want your trigger to report all affected rows, you need to write a loop
USE FyiCenterData;
GO
select * from fyi_users;
-- reporting the first affected row only

UPDATE fyi_users SET email=name;
GO
Email changed from NULL to John King
(5 row(s) affected)
-- reporting all affected rows

AFTER UPDATE
AS
SELECT 'Email changed from '
+ ISNULL(d.email,'NULL')
+ ' to '
+ ISNULL(i.email,'NULL')
FROM INSERTED AS i, DELETED AS d
WHERE i.id = d.id;
GO
UPDATE fyi_users SET email=REVERSE(name);

GO
------------------------------------------------------
Email changed from Marc Kumar to ramuK craM
Email changed from Roy Bush to hsuB yoR
Email changed from Jack Gate to etaG kcaJ
Email changed from Nancy Greenberg to grebneerG ycnaN
Email changed from John King to gniK nhoJ
(5 row(s) affected)
(5 row(s) affected)
How To Override DML Statements with Triggers?
Sometime, you may want to implement some business logics in a DML trigger to cancel the DML
statement. For example, you may want to check the new email address format provided by the
UPDATE statement. If the email address is invalid, you to cancel the UPDATE statement.
There is no easy way to cancel the DML statement in a DML trigger. But there is easy way to override
the DML statement with an "INSTEAD OF" trigger. SQL Server supports 2 options (3 keywords) on
when the defined trigger will be fired:
 AFTER - Trigger fired after the DML statement executed successfully.
 INSTEAD OF - Trigger fired instead of the DML statement execution, allowing the trigger to
decide whether or not, and how, to execute the statement.
 FOR - Same as AFTER.
The tutorial exercise below shows you how define an "INSTEAD OF" trigger on fyi_users to validate
email addresses:
USE FyiCenterData;
GO
CREATE TRIGGER check_email ON fyi_users

INSTEAD OF UPDATE
AS
DECLARE @count INT;
SELECT @count = COUNT(*) FROM INSERTED
WHERE email NOT LIKE '%_@_%';
IF @count = 0
UPDATE fyi_users SET email=i.email
FROM INSERTED AS i
WHERE fyi_users.id = i.id
ELSE
PRINT 'Invalid email(s) found.';
GO
-- invalid email
UPDATE fyi_users SET email='john.king'
GO
Invalid email(s) found.
-- valid email
UPDATE fyi_users SET email='john@fyicenter'
GO
Email changed from gniK nhoJ to john@fyicenter
-- change all
UPDATE fyi_users SET email='dba@fyicenter';
GO
Email changed from ramuK craM to dba@fyicenter
Email changed from hsuB yoR to dba@fyicenter
Email changed from etaG kcaJ to dba@fyicenter
Email changed from grebneerG ycnaN to dba@fyicenter
Email changed from john@fyicenter to dba@fyicenter
The "invalid email" test shows that trigger check_email did stoped the UPDATE statement. John
King's email did not get updated.
The "valid email" test shows that trigger check_email properly updated the email column, if the new
email is valid. The reported message was generated from trigger update_user.
How To Create a DDL Trigger using "CREATE TRIGGER" Statements?
A DDL trigger is defined to handle a DDL statement event, like create, alter and drop tables, views,
indexes, etc. DDL triggers can be used to generate warning messages on database object changes.
The format of creating a DDL trigger should be:
CREATE TRIGGER trigger_name ON DATABASE
AFTER ddl_event_types
AS
statements
GO
-- ddl_event_types are keywords like:
-- CREATE_TABLE, ALTER_TABLE, DROP_TABLE, ...
Below is a simple example of creating a DDL trigger to generate messages on ALTER_TABLE

events:
USE FyiCenterData;
GO
CREATE TRIGGER ddl_message ON DATABASE

AFTER ALTER_TABLE
AS
PRINT 'Someone is changing tables!';
GO
ALTER TABLE fyi_users

ALTER COLUMN id INT NOT NULL;
GO
Someone is changing tables!
Can You Roll Back the DDL Statement in a Trigger?
Can you roll back the DDL statement in a trigger? The answer is yes. Since the DDL statement that
fires the trigger and the statements defined inside the trigger are all executed as a single statement
batch, you can add a ROLLBACK statement in the trigger to rollback the entire batch.
USE FyiCenterData; GO CREATE TRIGGER drop_rollback ON DATABASE AFTER DROP_TABLE AS
PRINT 'Drop table is not allowed!'; ROLLBACK; GO DROP TABLE fyi_users; GO Drop table is not
allowed! Msg 3609, Level 16, State 2, Line 2 The transaction ended in the trigger. The batch has been
aborted.
This trigger is powerful. It will stop you from dropping any tables in FyiCenterData database.
Can You Create a Logon Trigger in SQL Server 2005 Express Edition?
Can you create a logon trigger in SQL Server 2005 Express Edition? The answer is no. LOGON is
not a supported event type in Express Edition. The script below shows you the error message when
you try to create a logon trigger:
CREATE TRIGGER welcome ON ALL SERVER
AFTER LOGON
AS
PRINT 'Welcome to FyiCenterData database!';
GO
Msg 1084, Level 15, State 1, Procedure welcome, Line 2
'LOGON' is an invalid event type.
Or may be there is a configuration setting that disables the logon event type
SQL Server FAQs - Introduction to Cursors
A collection of 9 FAQs on working with cursors. Clear explanations and tutorial exercises are
provided on declaring cursor names and cursor variables; open and close cursors; looping through
cursor result set; backward scrolling; dynamic cursors.
What Are Cursors?
A cursor is a special data type that represents a result set returned by a SELECT query statement.
There are several notes about cursor you need to remember:
 Cursor data type can not be used to define table columns.
 Cursor data type is used on store procedures, functions, and triggers to help you loop through
result sets returned by queries.
 Cursor data type can be used to define cursor variables.
 There are special Transact-SQL statements dedicated to work with cursors variables: OPEN,
FETCH, CLOSE, and DEALLOCATE.
 Cursor variables can be passed as procedure or function parameters.
 There is a special function, CURSOR_STATUS(), for check cursor statuses.
 How To Declare a Cursor with "DECLARE ... CURSOR"?

 If you want to use a cursor to represent the result set of a query, you need to define a cursor
name with a SELECT sub-statement using the "DECLARE ... CURSOR" statement using the
following syntax format:
 DECLARE cursor_name CURSOR FOR
 SELECT ...;
 Note that the DECLARE statement will not actually execute the SELECT sub-statement. It only
attaches the SELECT sub-statement to the cursor.
 A cursor name should be deallocated to free up server resources if the cursor is not needed any
more.
 The tutorial example below shows you how to declare and deallocate a cursor.
 GO

 DECLARE fyi_cursor CURSOR FOR
 SELECT * FROM fyi_links;
 -- other statements
 DEALLOCATE fyi_cursor;
 GO
 How To Execute the Cursor Queries with "OPEN" Statements?
 Once a cursor is declared, you need to execute the query attached to the cursor so that the result
set returned from the query can be accessed through the cursor. To execute the cursor query,
you should use the OPEN statement as in this format:
 OPEN cursor_name;
 When you are done with using the result set attached to a cursor, you should close the result set
to free up server resources.
 The tutorial example below shows you how to open and close a cursor:
 GO

 OPEN fyi_cursor;
 -- result set is ready to use

 -- other statements

 CLOSE fyi_cursor;
 GO
 How To Fetch the Next Row from a Cursor with a "FETCH" Statement?
 When the result set is ready in a cursor, you can use a FETCH statement to retrieve one row
from the result set in the same format as a SELECT statement. The FETCH statement has the
following formats:
 FETCH NEXT FROM cursor_name;
 FETCH PRIOR FROM cursor_name;
 FETCH FIRST FROM cursor_name;
 FETCH LAST FROM cursor_name;
 FETCH ABSOLUTE n FROM cursor_name;
 FETCH RELATIVE n FROM cursor_name;
 The tutorial exercise below shows you how FETCH statements are used to retrieve the first row
and the second row back from a cursor:
 GO


 FETCH NEXT FROM fyi_cursor;
 FETCH NEXT FROM fyi_cursor;

 GO
 id url notes counts time
 ---- ------------------ ----------- ------- -----
 101 dev.fyicenter.com NULL NULL NULL

 id url notes counts time
 ---- ------------------ ----------- ------- -----
 102 dba.fyicenter.com Nice site. 8 NULL
 How To Transfer Data from a Cursor to Variables with a "FETCH" Statement?
 By default, a FETCH statement will display the fetched row on the client program window. If you
want to transfer the output data to variables, you can specify an INTO clause with a list of
variables that matches the list of fields in the result set.
 The tutorial exercise below shows you a good example of using the FETCH statement to transfer
one row of output data from the result set to variables:
 GO

 SELECT id, url, notes, counts, time FROM fyi_links;

 DECLARE @id INT, @url VARCHAR(80), @notes VARCHAR(80),
 @counts INT, @time DATETIME;
 FETCH NEXT FROM fyi_cursor INTO @id, @url, @notes,
 @counts, @time;
 PRINT 'id = '+CONVERT(VARCHAR(20),ISNULL(@id,0));
 PRINT 'url = '+ISNULL(@url,'NULL');
 PRINT 'notes = '+ISNULL(@notes,'NULL');
 PRINT 'counts = '+CONVERT(VARCHAR(20),ISNULL(@counts,0));
 PRINT 'time = '+CONVERT(VARCHAR(20),ISNULL(@time,'2007'));

 GO

 id = 101
 url = dev.fyicenter.com
 notes = NULL
 counts = 0
 time = Jan 1 2007 12:00AM
 How To Loop through the Result Set with @@FETCH_STATUS?
 The FETCH statement only returns one row from the result set. If you want to return all rows, you
need to put the FETCH statement in a loop. A simple way to stop the loop to check the system
variable @@FETCH_STATUS, which returns a status code of the last FETCH statement:
 @status = @@FETCH_STATUS;
 -- Returns 0, the fetch was successful
 -- Returns -1, the fetch failed or end of result set reached
 -- Returns -2, the row fetched is missing
 The tutorial exercise below shows how @@FETCH_STATUS is used to make WHILE loop over
the result set in a cursor:
 GO

 @counts, @time;
 WHILE @@FETCH_STATUS = 0 BEGIN
 PRINT CONVERT(CHAR(5),ISNULL(@id,0))
 +CONVERT(CHAR(18),ISNULL(@url,'NULL'))
 +CONVERT(CHAR(20),ISNULL(@notes,'NULL'))
 +CONVERT(CHAR(4),ISNULL(@counts,0))
 +CONVERT(CHAR(11),ISNULL(@time,'2007'));
 @counts, @time;
 END
 GO
 101 dev.fyicenter.com NULL 0 Jan 1 2007
 102 dba.fyicenter.com Nice site. 8 Jan 1 2007
 1101 moc.retneciyf.ved NULL 0 Jan 1 2007
 202 www.yahoo.com It's another search 0 Jan 1 2007
 2102 dba.fyicenter.com NULL 0 Jan 1 2007
 301 netscape.com Added long time ago!0 Jan 1 1999
 302 myspace.com Added today! 0 Jul 1 2007
 How To Declare and Use Cursor Variables?
 There are two ways to representing a cursor:
 1. A cursor name - A static name representing a cursor object. A cursor name should be linked to
a cursor object in the DECLARE statement.
 2. A cursor variable name - A variable name pointing to a cursor object. A cursor variable name
should be declared with the CURSOR data type. It should be then assigned with a cursor object
using the SET statement.
 The tutorial exercise below shows you how to declare a cursor variable and assign a cursor
object to it:
 GO

 -- declare a cursor variable
 DECLARE @fyi_cursor CURSOR;

 -- assign a cursor object
 SET @fyi_cursor = CURSOR FOR

 -- execute the query and fetch results
 OPEN @fyi_cursor;
 FETCH NEXT FROM @fyi_cursor INTO @id, @url, @notes,
 @counts, @time;
 @counts, @time;
 END
 CLOSE @fyi_cursor;
 DEALLOCATE @fyi_cursor;
 GO

 How To Create a Scrollable Cursor with the SCROLL Option?
 SQL Server offers two scrolling option on cursors:
 1. FORWARD_ONLY - The cursor can only be scrolled forward with "FETCH NEXT" statements.
In another word, you can only loop through the cursor from the first row to the last row.
FORWARD_ONLY is the default option.
 2. SCROLL - The cursor can be scrolled back and forth with "FETCH NEXT", "FETCH PRIOR",
and other fetch options.
 The tutorial example below creates a cursor with the SCROLL option so that the result set can be
looped through backward:
 SET @fyi_cursor = CURSOR SCROLL FOR
 SELECT id, url, notes, counts, time
 FROM fyi_links ORDER BY id;
 FETCH LAST FROM @fyi_cursor INTO @id, @url, @notes,
 @counts, @time;
 FETCH PRIOR FROM @fyi_cursor INTO @id, @url, @notes,
 @counts, @time;
 END
 GO
 How To Create a Dynamic Cursor with the DYNAMIC Option?
 If the underlying table is changed after the cursor is opened, should the changes be reflected in
the cursor result set? The answer is based on the update option used when creating the cursor.
SQL SERVER supports two update options:
 1. STATIC - The result set will be a static copy created when the OPEN statement is executed.
Subsequent updates on the underlying tables will not affect the result set. STATIC is the default
option.
 2. SCROLL - The result set will be dynamically updated each time when a FETCH statement is
executed. In another word, the result set always reflects the latest changes on the underlying
tables.
 The tutorial script below gives you a good example of how dynamic cursors work:
 GO

 SET @fyi_cursor = CURSOR FOR
 @counts, @time;

 -- Changing the underlying table
 UPDATE fyi_links SET url=REVERSE(url);
 @counts, @time;
 END
 GO

 102 moc.retneciyf.abd Nice site. 8 Jan 1 2007
 202 moc.oohay.www It's another search 0 Jan 1 2007
 2102 moc.retneciyf.abd NULL 0 Jan 1 2007
 301 moc.epacsten Added long time ago!0 Jan 1 1999
 302 moc.ecapsym Added today! 0 Jun 30 2007
 The changes are reflected in the result set.
 SQL Server FAQs - Introduction to NULL Values
 A collection of 11 FAQs on working with NULL values. Clear explanations and tutorial exercises
are provided on assigning NULL values to columns, variables and parameters; evaluating
expressions with NULL values; testing NULL values; ISNULL() and NULLIF() functions
What Are NULL Values?
A NULL value is a special value that represents an unknown value. SQL Server supports NULL
values with the following features:
 All data types used for table columns support NULL values. In another word, NULL values can be
stored in database tables.
 Individual table columns may be defined to not allow NULL values. In this case, you can not
assign NULL values to those columns.
 "NULL" is a keyword that represent a NULL value in expressions.
 NULL values can be used directly in SET (assignment) statements.
 If NULL values are involved in an arithmetic operation, the result will be a numeric NULL.
 If NULL values are involved in a string operation, the result will be a string NULL.
 If NULL values are involved in a datetime operation, the result will be a datetime NULL.
 If NULL values are involved in a bitwise operation, the result will be a binary NULL.
 If NULL values are involved in a comparison operation, the result will be a Boolean NULL.
 If NULL values are involved in a Boolean operation, the result could be TRUE, FALSE, or NULL.
 To test NULL values, you need to use two special operators, IS NULL and IS NOT NULL.
 Special functions are available to handle NULL values, like ISNULL(), and NULLIF()
How To Assign NULL Values to Variables or Columns?
The rule for assigning NULL values to variables or table columns is simple: Use keyword "NULL"
directly as normal values. Specificly,
 "NULL" can be used in SET statements to assign NULL values to variables.
 "NULL" can be used in SET clauses in UPDATE statements.
 "NULL" can be used in value lists in INSERT statements.
 "NULL" can be used in parameter lists when calling stored procedures or functions.
The tutorial script below gives you some good examples:

USE FyiCenterData;
GO
-- assign NULL values to variables

SET @birth_date = NULL;
SELECT @birth_date;
GO
-----------------------
NULL
-- assign NULL values to columns

UPDATE fyi_links SET notes = NULL;
GO
(8 row(s) affected)
-- assign NULL values to parameters

EXEC sp_help NULL;
GO
Name
----------------
fyi_links_dump
fyi_links_top
fyi_links_view
...
What Happens If NULL Values Are Involved in Arithmetic Operations?
If NULL values are involved in arithmetic operations, the result will be numeric NULL values. The
following tutorial script shows you some good examples:
SELECT 7+NULL;
GO
-----------
NULL
SELECT 10.02*NULL;
GO
-----------
NULL
SELECT 4.988E+10/NULL;
GO
-----------
NULL
What Happens If NULL Values Are Involved in String Operations?
If NULL values are involved in string operations, the result will be string NULL values. The following
tutorial script shows you some good examples:
SELECT 'FyiCenter'+NULL;
GO
----------
NULL
SELECT LEN(NULL);
GO
----------
NULL
SELECT REVERSE(NULL);
GO
----------
NULL
What Happens If NULL Values Are Involved in Datetime Operations?
If NULL values are involved in datetime operations, the result will be datetime NULL values. The
following tutorial script shows you some good examples:
USE FyiCenterData;
GO
SELECT GETDATE()+NULL;
GO
-----------
NULL
SELECT DATEDIFF(DAY, GETDATE(), NULL);

GO
-----------
NULL
(Continued on next topic
What Happens If NULL Values Are Involved in Bitwise Operations?
If NULL values are involved in bitwise operations, the result will be binary NULL values. The following
tutorial script shows you some good examples:
SELECT 1 | NULL;
GO
-----------
NULL
SELECT 707 & NULL;

GO
-----------
NULL
SELECT ~NULL;
GO
-----------
NULL
What Happens If NULL Values Are Involved in Comparison Operations?
If NULL values are involved in comparison operations, the result will be Boolean NULL values. This
behavior is very interesting because you would expect a comparison operation returns only one of the
two values: TRUE and FALSE. But SQL Server may return you a third value: NULL.
The tutorial script below proves that "1>NULL" is not returning TRUE or FALSE. It is returning a third
value, NULL:
IF 1>NULL PRINT '1>NULL is returning TRUE'
ELSE PRINT '1>NULL is not returning TRUE'
GO
1>NULL is not returning TRUE
IF NOT 1>NULL PRINT '1>NULL is returning FALSE'

ELSE PRINT '1>NULL is not returning FALSE'
GO
1>NULL is not returning FALSE
Another test proves that "'FYI'=NULL" is not returning TRUE or FALSE. It is returning a third value,
NULL:
IF 'FYI'=NULL PRINT '''FYI''=NULL returns TRUE'
ELSE PRINT '''FYI''=NULL does not return TRUE'
GO
'FYI'=NULL does not return TRUE
IF NOT 'FYI'=NULL PRINT '''FYI''=NULL returns FALSE'

ELSE PRINT '''FYI''=NULL does not return FALSE'
GO
'FYI'=NULL does not return FALSE
What Happens If NULL Values Are Involved in Boolean Operations?
If NULL values are involved in Boolean operations, the result will vary depending on the operator
type. For AND operator, FALSE takes precedence over NULL. The result can be summarized in a
table below:
AND TRUE FALSE NULL
TRUE true false null
FALSE false false false
NULL null false null
For OR operator, TRUE takes precedence over NULL. The result can be summarized in a table
below:
OR TRUE FALSE NULL
TRUE true true true
FALSE true false null
NULL true null null
The tutorial script below shows you that NULL AND FALSE returns FALSE:
IF 0=NULL AND 0=1 PRINT 'NULL AND FALSE returns TRUE'
ELSE PRINT 'NULL AND FALSE does not returns TRUE'
GO
NULL AND FALSE does not returns TRUE
IF NOT (0=NULL AND 0=1)

PRINT 'NULL AND FALSE returns FALSE'
ELSE PRINT 'NULL AND FALSE does not returns FALSE'
GO
NULL AND FALSE returns FALSE
How To Test NULL Values Properly?
From previous tutorials, you learned that comparing expressions with NULL will always result NULL.
So you can not use comparison operators to test if an expression represents a NULL value or not.
To test expressions for NULL values, you should use one of the following special operators:
expression IS NULL
-- Returns TRUE if the expression is a NULL value
expression IS NOT NULL

-- Returns TRUE if the expression is not a NULL value
The turorial script below shows you how to use IS NULL and IS NOT NULL:
IF 0 IS NULL PRINT 'TRUE'
ELSE PRINT 'FALSE';
GO
FALSE
IF NULL IS NULL PRINT 'TRUE'

ELSE PRINT 'FALSE';
GO
TRUE
IF 1+NULL IS NULL PRINT 'TRUE'

ELSE PRINT 'FALSE';
GO
TRUE
IF 'NULL' IS NOT NULL PRINT 'TRUE'
ELSE PRINT 'FALSE';
GO
TRUE
How To Replace NULL Values in Expressions using ISNULL()?
As you learned from provious tutorials, NULL values presented in expressions will cause the final
results to be NULL. Sometimes, you want NULL values to be replaced with some default values, like
0, '', or 'NULL', so that expressions can be evaluated properly.
SQL Server offers a built-in function called ISNULL() to help you replacing NULL values in
expressions:
ISNULL(expression, replacement)
-- Returns "expression", if it is not NULL
-- Returns "replacement", if "expression" is NULL
The tutorial example below shows you how to replace possible NULL values in @middle_initial:
USE FyiCenterData;
GO
CREATE PROCEDURE welcome

@first_name VARCHAR(20),
@middle_initial VARCHAR(1),
@last_name VARCHAR(20)
AS
PRINT 'Hello '+@first_name
+ ' '+@middle_initial
+ ' '+@last_name;
PRINT 'Hello '+@first_name
+ ' '+ISNULL(@middle_initial,'')
+ ' '+@last_name;
GO
EXEC welcome 'John', 'W', 'King';

GO
Hello John W King
Hello John W King
EXEC welcome 'John', NULL, 'King';

GO
Hello John King
The first PRINT statement in the second test returns a blank line because of NULL value.
How To Replace Given Values with NULL using NULLIF()?

Sometime you want to hide certain values by replacing them with NULL values. SQL Server offers
you a nice function called NULLIF() to do this:
NULLIF(expression, value)
-- Returns NULL if "expression" equals to value"
-- Returns "expression", otherwise
NULLIF() can be viewed as the reverse function of ISNULL(). The tutorial script below shows you a
good example of using NULLIF():
USE FyiCenterData;
GO
SELECT id, counts FROM fyi_links;

GO
id counts
----------- -----------
101 NULL
102 8
1101 NULL
202 NULL
2101 NULL
2102 NULL
301 NULL
302 NULL
-- converting NULL to 0
UPDATE fyi_links SET counts=ISNULL(counts,0);
GO
GO
id counts
----------- -----------
101 0
102 8
1101 0
202 0
2101 0
2102 0
301 0
302 0
-- converting 0 to NULL
UPDATE fyi_links SET counts=NULLIF(counts,0);
GO
GO
id counts
----------- -----------
101 NULL
102 8
1101 NULL
202 NULL
2101 NULL
2102 NULL
301 NULL
302 NULL
What Is the Security Model Used in SQL Server 2005?
SQL Server 2005 uses a very standard security model involves 3 concepts:
 Securables - Entities representing resources that need to be secured. For example, a database
table is a securable.
 Principals - Entities representing users that request accesses to resources. For example, a login
user is a principal.
 Permissions - Types of accesses associated with securables. Permissions can be granted to or
revoked from principals. For example, "Update" is a permission associated a securable, table "R".
"Update" on "R" can be granted to a principal, user "U". Now user "U" will get "Update" access on
table "R".
In simple terms, a principal answers the security question of "Who are you?"; a securable answers
the security question of "What is your target object?"; a permission answers the security question of
"You are allowed to perform this action on this target object".
SQL Server 2005 supports multiple securables and multiple principals organized into hierarchical
structures as shown in this picture:
(Picture by Microsoft)
What Are Security Principals Used in SQL Server 2005?
SQL Server 2005 supports several basic security principals located at different levels:
 Windows-Level Principals: Windows Local Login and Windows Network Domain Login - Used to
control accesses to SQL Server instances.
 SQL Server-Level Principal: SQL Server Login. - Used to control accesses to SQL Server
instances.
 Database-Level Principal: Database User. - Used to control accesses to database instances.
To access a SQL Server instance, you must use a Windows login or a SQL Server login previously
created in that server instance. Once you logged in, you are represented by this login name as your
security principal at the server level.
If you select a specific database in the server to use, SQL Server will search that database for a
previously user that has been mapped to your login name. If that user is located, you are represented
by this user name as your security principal at the database level.
What Is the Security Principal at the Server Level That Represents Your Session?
Security principal identifies who you are when you interact with the SQL Server. What can do you at
the server level solely depends on the security principal that represents you. So it is very important to
know your security principal at the server level.
What is the server level security principal of your session? The answer is simple - the login name you
used to connect to the server to start the session. For example, if you connects to the server with the
login name "sa", then your security principal is "sa" at the server level.
On a Unix system, there is a nice command called "whoami" which returns your security principal on
the system. SQL Server provides a similar function called Suser_Sname(), which returns your
security principal (a login name) at the server. Try it by following this tutorial example:
-- Find out your security principal at the server level
PRINT Suser_Sname();
GO
sa
This example shows you that you used "sa" as the login name to connect to the server.
What Is the Security Principal at the Database Level That Represents Your Session?
Security principal identifies who you are when you interact with the SQL Server. What can do you at
the database solely depends on the security principal that represents you. So it is very important to
know your security principal at the database level.
What is the database level security principal of your session? The answer is simple - the user name
in the current database that has been mapped to your login name. For example, if you connects to
the server with the login name "sa", and sets "FyiCenterData" as the current database, SQL Server
will map "sa" to a user name defined in "FyiCenterData" based on some logics. That mapped user
name is your security principal for "FyiCenterData" at the database level.
On a Unix system, there is a nice command called "whoami" which returns your security principal on
the system. SQL Server provides a similar function called User_Name(), which returns your security
principal (a user name mapped to the login name) at the database level. Try it by following this tutorial
example:
PRINT Suser_Sname();
GO
sa
USE FyiCenterData;
GO
-- Find out your security principal at the database level

PRINT User_Name();
GO
dbo
This example shows you that "dbo" is user name in "FyiCenterData" that mapped to login name "sa".
How To Create a New Login Name in SQL Server?
In previous tutorials, it is assumed that you use the "sa" (System Administrator) login name to connect
to your SQL Server. But that is not what a normal developer uses to connect to the server, since "sa"
has the ALL permissions granted. You need to create new login names and grant less permissions to
them, and give them to developers.
To create a new login name, you can use the "CREATE LOGIN" statement in a simple syntax like
this:
CREATE LOGIN login_name WITH PASSWORD = 'password'
To run "CREATE LOGIN" statement, you need to connect to the server with a privileged login name
like "sa". See the tutorial example below:
-- Login with 'sa'
-- Create new login names

CREATE LOGIN FYI_DBA WITH PASSWORD = 'ABD_IYF'
GO
CREATE LOGIN Fyi_Login WITH PASSWORD = 'IYF'

GO
How To Verify a Login name with SQLCMD Tool?
The quickest way to verify a login name on a SQL Server is probably to use the SQLCMD tool, which
takes the server name, login name and password in a single command line. Here is how to try it
yourself:
Start a command window and enter the following command:

C:\>SQLCMD -S LOCALHOST\SQLEXPRESS -U FYI_DBA -P ABD_IYF
1> PRINT Suser_Sname();
2> GO
FYI_DBA
1> QUIT
C:\>
This shows FYI_DBA is a valid login and the password is correct. Now try this command:
C:\>SQLCMD -S LOCALHOST\SQLEXPRESS -U FYI_DEV -P ABD_IYF
Msg 18456, Level 14, State 1, Server LOCALHOST\SQLEXPRESS, Line 1
Login failed for user 'FYI_DEV'.
C:\>
This shows FYI_DEV is not a valid login or the password is incorrect.
How To List All Login Names on the Server?

If you want to see a list of all login names defined on the server, you can use the system view,
sys.server_principals as shown in this tutorial exercise:
-- Login with sa
SELECT name, sid, type, type_desc FROM sys.server_principals

WHERE type = 'S';
GO
name sid type type_desc
-------- ----------------------------------- ---- ---------
sa 0x01 S SQL_LOGIN
FYI_DBA 0x348AF32B3B58CB40B67A5F5B7086B96E S SQL_LOGIN
Fyi_Login 0x5EB8701EAEBAA74F86FCF5BD8E37B8C5 S SQL_LOGIN
So you have two login names on the server at this moment.
How To Change the Password of a Login Name?
If a developer lost the password of his or her login name, you can reset the password with the
"ALTER LOGIN" statement as shown in this tutorial example:
-- Login with sa
ALTER LOGIN FYI_DBA WITH PASSWORD = 'fyicenter';

GO
FYI_DBA's password is changed now. Try it with the SQLCMD tool:

C:\>SQLCMD -S LOCALHOST\SQLEXPRESS -U FYI_DBA -P fyicenter
1> QUIT
You can also use the "ALTER LOGIN" to change your own password.
How To Change a Login Name?
If you want to change a login name, you can use the "ALTER LOGIN" statement as shown in this
tutorial example:
-- Login with "sa"
-- Change login name

ALTER LOGIN Fyi_Login WITH NAME = Dba_Login;
GO
-- View login names

WHERE type = 'S';
GO
-------- ----------------------------------- ---- ---------
sa 0x01 S SQL_LOGIN
Dba_Login 0x5EB8701EAEBAA74F86FCF5BD8E37B8C5 S SQL_LOGIN
How To Disable a Login Name?
If you want temporarily disable a login name, you can use the "ALTER LOGIN" statement with a
DISABLE keyword. If you want to enable it later on, you can use the ENABLE keyword. The tutorial
exercise below shows how to disable and enable login name "Dba_Login":
-- Login with "sa"
-- Disable a login
ALTER LOGIN Fyi_Login DISABLE;
-- View login status

SELECT name, type, type_desc, is_disabled
FROM sys.server_principals
WHERE type = 'S';
GO
name type type_desc is_disabled
----------- ---- ---------- -----------
sa S SQL_LOGIN 0
FYI_DBA S SQL_LOGIN 0
Dba_Login S SQL_LOGIN 1
Now try to login with the disabled login name:

C:\>SQLCMD -S LOCALHOST\SQLEXPRESS -U Dba_Login -P IYF
Msg 18470, Level 14, State 1, Server LOCALHOST\SQLEXPRESS
Login failed for user 'Dba_Login'. Reason: The account is
disabled.
C:\>
Run the statements below to enable login name "Dba_Login":

-- Login with "sa"
-- Enable a login
ALTER LOGIN Fyi_Login ENABLE;
How To Delete a Login Name?
If you don't want to keep a login name any more, you should delete it by using the "DROP LOGIN"
statement as shown in this tutorial example:
-- Login with "sa"
DROP LOGIN Dba_Login;

GO
-- View login names
WHERE type = 'S';
GO
-------- ----------------------------------- ---- ---------
sa 0x01 S SQL_LOGIN
How To Create a User Name in a Database?
User names are security principals at the database level. If you want to allow a login name to access
a specific database, you need to create a user name in that database and link it to the login name.
Creating a user name can be done by using the "CREATE USER" statement as shown in this tutorial
exercise:
-- Login with "sa"
-- Create a login
CREATE LOGIN Fyi_Login WITH PASSWORD = 'IYF'
GO
-- Select a database
USE FyiCenterData;
GO
-- Create a user and link it to a login

CREATE USER Fyi_User FOR LOGIN Fyi_Login;
GO
Login name "Fyi_Login" should be able to access database "FyiCenterData" through user name
"Fyi_User".
How To List All User Names in a Database?
If you want to see a list of all user names defined in a database, you can use the system view,
sys.database_principals as shown in this tutorial exercise:
-- Login with sa
-- Select a database
USE FyiCenterData;
GO
-- List all user names

SELECT name, sid, type, type_desc
GO
-------------------- ------------------------ ---- ---------
dbo 0x01 S SQL_USER
guest 0x00 S SQL_USER
INFORMATION_SCHEMA NULL S SQL_USER
sys NULL S SQL_USER
Fyi_User 0x5EB8701EAEBAA74F86F... S SQL_USER
(5 row(s) affected)
As you can see, there are 5 user names defined in "FyiCenterData". 4 are defined by SQL Server
during the database creation process. Only the last one "Fyi_User" was defined by you in the
previous tutorial.
How To Find the Login Name Linked to a Given User Name?
If you know a user name in a database and you want to find out which login name is linked this user
name, you need to check the Security ID (SID) of the user name based on the following rules:
 Each login name is associated a unique SID.
 When a user name is linked to a login name, the login name's SID is copied to the user name.
So the login name linked to a user name must have the SID as the user name. The tutorial exercise
below shows you how to find the login name that is linked to the user name "Fyi_User":
-- Login with sa
USE FyiCenterData;
GO
SELECT u.name AS User_Name, l.name AS Login_Name, u.sid

FROM sys.server_principals l,
sys.database_principals u
WHERE l.sid = u.sid
AND u.name = 'Fyi_User';
GO
User_Name Login_Name sid
---------- ----------- ----------------------------------
Fyi_User Fyi_Login 0x5EB8701EAEBAA74F86FCF5BD8E37B8C5
(1 row(s) affected)
How To Verify a User name with SQLCMD Tool?
The quickest way to verify a user name in a database is probably to use the SQLCMD tool. You can
connect to the server, select the database, and check which user name is linked the current login
name as shown below.
Start a command window and enter the following command:

C:\>SQLCMD -S LOCALHOST\SQLEXPRESS -U Fyi_Login -P IYF
1> USE FyiCenterData;
2> GO
Changed database context to 'FyiCenterData'.
1> PRINT User_Name();

2> GO
Fyi_User
This shows user "Fyi_User" in database "FyiCenterData" is linked to login name "Fyi_Login".
How To Change the Name of a Database User?
If you want to change the name of an existing database user, you can use the "ALTER USER"
statement as shown in the tutorial exercise below:
-- Login with "sa"
USE FyiCenterData;
GO
ALTER USER Fyi_User WITH NAME = Dba_User;

GO

GO
-------------------- ------------------------ ---- ---------
dbo 0x01 S SQL_USER
sys NULL S SQL_USER
Dba_User 0x5EB8701EAEBAA74F86F... S SQL_USER
The of the last user changed from "Fyi_User" to "Dba_User".
How To Delete an Existing Database User?
If you don't want to keep a database user any more, you should delete the user by using the "DROP
USER" statement. This tutorial exercise shows how to delete "Dba_User":
-- Login with "sa"
USE FyiCenterData;
GO
DROP USER Dba_User;

GO
-------------------- ------------------------ ---- ---------
dbo 0x01 S SQL_USER
sys NULL S SQL_USER
User "Dba_User" has been deleted now.
SQL Server FAQs - Configuring ODBC Connection and DSN for SQL Servers
A collection of 13 FAQs on connecting SQL Server via an ODBC DSN setting. Clear explanations
and tutorial exercises are provided on creating new DSN with the SQL Server ODBC driver; starting
SQL Server Browser Service; enabling TCP/IP protocol on SQL Server; verifying TCP/IP port
numbers; testing ODBC DSN settings with MS Access.
What Is Open Database Communication (ODBC)?
ODBC, Open Database Communication, a standard API (application program interface) developed by
Microsoft for Windows applications to communicate with database management servers.
If you want to access a database server through an ODBC driver from an application program, you
need to meet the following requirements:
 An ODBC driver specifically designed for the database server.
 An ODBC DSN (Data Source Name) - an ODBC configuration representing the ODBC driver and
database server.
 An ODBC API (Application Programming Interface) for your application program to interact with
database server through the ODBC driver.
For example, if you want to a SQL Server in a PHP script through an ODBC driver, you need to make
sure that:
 An ODBC driver for MS SQL Server - The ODBC driver is installed as part of the Windows
system.
 An ODBC DSN - You need to create one yourself using the ODBC configuration tool.
 An ODBC API in PHP language - The API is installed as part of the PHP engine.
How To Add a New DSN with the ODBC Driver for SQL Server?
Assuming that the ODBC driver for SQL Server has been installed as part of the Windows system,
the next step of setting up ODBC connection to SQL Server is to create a new DSN (Data Source
Name) with the ODBC Data Source Administrator:
 Go to Control Panel.
 Go to Administrative Tools.
 Run Data Sources (ODBC). The ODBC Data Source Administrator window shows up.
 Go to System DSN tab.
 Click the Add button.
You should a list of all ODBC drivers installed on your system. SQL Server ODBC driver should be on
the list. If you look at the picture below, you will see the SQL Server ODBC driver with version
2000.85.1117.00, file SQLSRV32.DLL, date 8/4/2004:

Select "SQL Server" from the ODBC driver list, and click "Finish" button. The "Create a New Data
Source to SQL Server" wizard window shows up. Continue with the next tutorial to finish up creating a
new DSN.
How To Define the Name and Server for a new DSN?
Continuing from the previous tutorial, on the first screen of the "Create a New Data Source to SQL
Server" wizard, you should enter 3 fields: Name, Description, and Server as suggested below:
Name: FYI_SQL_SERVER
Description: FYIcenter.com SQL Server
Server: LOCALHOST
See the picture below:

Note that the name and description can be decided by yourself. But the server must be the network
machine name where the SQL Server is running. In this example, LOCALHOST is the machine name
for your local system.
Click Next and continue with the next tutorial to finish up creating a new DSN.
How To Provide Login Information for a New ODBC DSN?
Continue from the previous tutorial. After clicking Next on the first screen of the "Create a New Data
Source to SQL Server" wizard, you should see the second screen asking you to select SQL Server
login type, login name and password.
Select the radio button on "With SQL Server authentication using a login ID and password entered by
the user. Also enter in Login ID field: "sa" and in Password field: "FYIcenter". See the picture below:

Remeber that "sa" is the system administrator login name, you probably should use a less privileged
login name here. "FYIcenter" must be the correct password defined in the SQL Server for "sa" login
name.
Click Next and continue with the next tutorial to finish up creating a new DSN
Why Are You Getting Errors When Creating a New ODBC DSN?
Continue from the previous tutorial. After clicking Next on the SQL login information screen, ODBC
manager will try to connect to the SQL Server with the login information you provided.
After a period of waiting time, you may get error message box saying that:
Connection failed:
SQLState: '01000'
SQL Server Error: 53
[Microsoft][ODBC SQL Server Driver][DBNETLIB]
ConnectionOpen(Connect()).
Connection failed:
SQLState: '08001'
SQL Server Error: 17
[Microsoft][ODBC SQL Server Driver][DBNETLIB]
SQL Server does not exist or access denied.
See the picture below:

Three possible reasons for the failing:

 Wrong server name - You provided an incorrect server name.
 SQL Server not configured to take a network connection - You need to check the SQL Server
configuration.
 Wrong login name or password - You provided incorrect login name or password.
The first and third reasons are easy to validate and correct. The second reason requires further
investigation. Continue with the next tutorial to configure your SQL Server to take a network
connection.
What Are the Requirements on SQL Server Network Connections?
By default, SQL Server 2005 Express Edition is installed only one connection protocol enabled:
 Shared Memory - SQL Server connection protocol for applications that are running on the same
machine where the SQL Server is running. For example, if you are running SQLCMD tool on the
SQL Server machine, it will use the "Shared Memory" protocol.
If you want application that are running remotely to connect and access the SQL Server you need
enable the "TCP/IP" protocol.
For applications that are running on the same machine as the SQL Server, but they do not support
"Shared Memory" protocol, you also need to enable the "TCP/IP" protocol for those applications.
ODBC Manager is an application that requires the "TCP/IP" protocol to connect to the SQL Server.
In order for the SQL Server to accept network connections, you also need to run SQL Server Browser
Service on the SQL Server machine.
In summary, there are two requirements for a SQL Server to accept network connections:
 Start SQL Server Browser Service on the SQL Server machine.
 Enable the TCP/IP protocol support on the SQL Server.
 How To Start SQL Server Browser Service?

 SQL Server Browser Service is installed as part of the SQL Server. But it is turned off by default
to reduce the security risk. If you want start SQL Server Browser Service to allow the SQL Server
to accept network connections, you need to follow the steps below:
 1. Go to Control Panel > Administrative Tools.
 2. Double click on "Services". The Services window shows up.
 3. Double click on "SQL Server Browser". The properties dialog box shows up.
 4. Change the "Startup Type" from Disabled to Automatic. Then click the Start button.
 The "SQL Server Browser" service should be running now
 How To Enable TCP/IP Protocol on a SQL Server?
 By default, the TCP/IP protocol is turned off when a SQL Server is installed to reduce security
risk. But if you want applications to connect and access the SQL Server, you need to enable the
TCP/IP protocol on the server by following this tutorial:
 1. Go to Start > All Programs > Microsoft SQL Server 2005 > Configuration Tools > SQL Server
Configuration Manager.
 2. Double click on "SQL Server 2005 Network Configuration" to see the list of SQL Server
instances.
 3. Click on "Protocols for SQLEXPRESS". You will see a list of protocols.
 4. Right-mouse click on TCP/IP and select the "Enable" command. See the picture below:

 5. Click on "SQL Server 2005 Services". You will see two services.
 6. Right-mouse click on "SQL Server (SQLEXPRESS)" and select restart to finish the TCP/IP
protocol setting change.
 Your SQL Server is ready to accept network connection now
 How To Verify the Port Number of the SQL Server?
 When applications use TCP/IP for network communication, you need to know the port number
where the server is listening for connect request from the client.
 If you want to connect to the SQL Server through the TCP/IP, you must know on which port
number the SQL Server is listening for connection requests. Follow this tutorial to find and verify
the port number of your SQL Server.
 1. Go to Start > All Programs > Microsoft SQL Server 2005 > Configuration Tools > SQL Server
Configuration Manager.
 2. Double click on "SQL Server 2005 Network Configuration" to see the list of SQL Server
instances.
 3. Click on "Protocols for SQLEXPRESS". You will see a list of protocols.
 4. Right-mouse click on TCP/IP and select the "Properties" command. The properties dialog box
shows up.
 Click the IP Address tab. The server IP address and port number used by the SQL Server will be
displayed. You should see something like:
 IP Address: 127.0.0.1
 Ports: 1269
 Note that 1433 may also be used as the default port number by your SQL Server.
 How To Configure ODBC DSN with Different Port Numbers?
 If your SQL Server is not using the default port number, like 1269, you need to set the port
number to the correct value during the ODBC DSN creation process, as show in this tutorial:
 1. Start ODBC Data Source Administrator and click System DSN tab.
 2. Click Add button, select SQL Server and click Finish button.
 3. Enter the following and click Next:
 Name: FYI_SQL_SERVER
 Description: FYIcenter.com SQL Server
 Server: LOCALHOST
 4. Select the radio button on "With SQL Server authentication using a login ID and password
entered by the user. Also enter in Login ID field: "sa" and in Password field: "FYIcenter".
 5. Click "Client Configuration..." button. The "Edit Network Library Configuration" dialog box
shows up.
 Check and uncheck "Dynamically determine port" and enter "1269" as the Port Number. See
picture bellow:

 Click OK and continue with the next tutorial to finish up ODBC DSN creation.
 How To Configure and Test ODBC DSN Settings?
 Continue from the previous tutorial. Click Next after you have finished changing the port number.
The ODBC Data Source Administrator will try to connect to the SQL server through the specified
port number and login information.
 If the connection is successful, you will see the next screen asking for selecting default database.
 1. Check "Change the default database to:" check box, and select "FyiCenterData" as the default
database for this DSN setting. See picture bellow:

 2. Click Next to see the last screen of the "Create a New Data Source to SQL Server" wizard.
 3. Change nothing on the last screen and click Finish button. The confirmation screen shows up.
 4. Click "Test Data Source..." button. You should see the test result as:
 Attempting connection
 Connection established
 Verifying option settings
 Disconnecting from server

 TESTS COMPLETED SUCCESSFULLY!
 Your FYI_SQL_SERVER ODBC DSN is ready to use
How To Connect MS Access to SQL Servers through ODBC?
Once you got a DSN defined in the ODBC manager that connects to your SQL Server, you can
connect a normal MS Access document to the Oracle server, and link an Access table to a SQL
Server table. The tutorial below gives you a good example:
 Start MS Access with a new database file.
 Go to File menu.
 Select Get External Data.
 Select Import.... The Import dialog box shows up.
 Select Files of type: ODBC Database(). The Select Data Source dialog box shows up.
 Click the Machine Data Source tab. You should see the DSN name "FYI_SQL_SERVER" you
defined earlier.
 Select "FYI_SQL_SERVER".
 Enter User Name: sa.
 Enter Password: FYIcenter.
Click the OK button to continue. You should see a list of tables available for you to import from the
SQL Server as shown in the picture below:

Select the table you are interested in and click OK. You should the selected table being imported from
the SQL Server to MS Access.
How Can Windows Applications Connect to SQL Servers via ODBC?
One way of connecting a windows application to a SQL Server is to use ODBC drivers. The
requirements to do this is summarized here:
 The SQL Server must have TCP/IP protocol enabled with a specific port number.
 The SQL Server Browser Service must be running on the server machine.
 A ODBC DSN must be created to represent the connection to SQL Server on the specified port
number.
 The local windows application must be able to support ODBC API calls.
The diagram below shows how MS Access can connect to a SQL Server through the ODBC driver:

SQL Server FAQs - PHP ODBC Functions - Connection and Query Execution
A collection of 14 FAQs on using PHP ODBC functions to connect to SQL Server databases. Clear
explanations and tutorial exercises are provided on testing ODBC DSN settings; creating connection
to a SQL Server through a DSN; executing SQL statements; looping through result set; Executing
prepared statements
What Are the Requirements to Use ODBC Connections in PHP Scripts?
If you are planning to use ODBC connections to access SQL Server databases in PHP scripts, you
need to check the following requirements:
 The PHP engine must support ODBC functions. If you install PHP 5.2.2 from The PHP Group, the
ODBC functions are already included in the default installation.
 The SQL Server must have TCP/IP protocol enabled for a specific port.
 The SQL Server Browser Service must be running on the SQL server machine.
 An ODBC DSN must be created on machine where you are running PHP scripts.
 The ODBC DSN connection must be tested to make sure it is working.
You need to following other tutorials provided by FYIcenter.com to install PHP engine, install SQL
Server, configure TCP/IP protocol, start SQL Server Browser Service, create and test ODBC DSN
connections.
What Are Commonly Used ODBC Functions in PHP?
If you look at the PHP 5 manual, you will see a group of functions listed under the ODBC Functions
(Unified) setion. The commonly used ODBC functions are:
 odbc_connect — Establish an OBDC connection.
 odbc_data_source — Returns information about a current connection.
 odbc_close — Close an ODBC connection.
 odbc_exec — Prepare and execute a SQL statement.
 odbc_fetch_row — Fetch a row - moving the pointer to the a new row in a result object.
 odbc_result — Get the value of a specific field from the current row of a result object.
 odbc_fetch_array — Fetch a result row as an associative array.
 odbc_fetch_object — Fetch a result row as an object.
 odbc_num_rows — Number of rows in a result.
 odbc_field_name — Get the name of a specified field index.
 odbc_field_type — Get the data type of a specified field index.
 odbc_next_result — Checks if multiple results are available.
 odbc_free_result — Free resources associated with a result object.
 odbc_prepare — Prepares a statement for execution.
 odbc_execute — Execute a prepared statement.
 odbc_error — Get the last error code.
 odbc_errormsg — Get the last error message.
 odbc_tables — Get the list of table names stored in a specific data source.
 odbc_columns — Lists the column names in specified tables.
 odbc_autocommit — Toggle autocommit behaviour.
 odbc_commit — Commit an ODBC transaction.
 odbc_rollback — Rollback a transaction.
 How To Test ODBC DSN Connection Settings?

 Assuming you have followed other FYIcenter.com tutorials and created an ODBC DSN called
"FYI_SQL_SERVER", and planning to use it your PHP scripts, you should test this ODBC
connection first as shown in this tutorial:
 1. Go to Control Panel > Administrative Tools.
 2. Run Data Sources (ODBC). The ODBC Data Source Administrator window shows up.
 3. Click "System DSN" tab, select "FYI_SQL_SERVER", and click "Configure..." button. The
Microsoft SQL Server DSN Configuration wizard window shows up.
 4. Review the first screen and click Next.
 5. Review the second screen, enter the password as "FYIcenter", and click Next.
 6. Review the third screen and click Next.
 7. Review the fourth screen and click Finish.
 8. Review the confirmation screen and click "Test Data Source...". You should see the test result
as:
 Attempting connection
 Connection established
 Verifying option settings
 Disconnecting from server

 TESTS COMPLETED SUCCESSFULLY!
 Your FYI_SQL_SERVER ODBC DSN is tested and ready to use
 How To Connect to a SQL Server using odbc_connect()?
 If you have an ODBC DSN (Data Source Name) created linking to a SQL Server, you are ready
to connect to the SQL Server through the DSN with ODBC functions. There is no changes
needed in the php.ini configuration file.
 The tutorial script below shows you how to call odbc_connect() and odbc_close() to connect and
disconnect to the SQL Server through the DSN name "FYI_SQL_SERVER":
 <?php
 $con = odbc_connect('FYI_SQL_SERVER','sa','FYIcenter');
 if (!$con) {
 print("There is a problem with SQL Server connection.\n");
 } else {
 print("The SQL Server connection object is ready.\n");
 odbc_close($con);
 }
 ?>
 If you run this script and get this output: "The SQL Server connection object is ready", your PHP
environment and ODBC connection to the SQL Server are working.
 Note that 'sa' and 'FYIcenter' used in this script are system administrator login name and
password. You may use any other login name and password defined on the SQL Server.
 How To List All DSN Entries on Your Local Machine using odbc_data_source()?
 If you are interested to know what DSN entries are available on your local machine, you can use
odbc_data_source($con, SQL_FETCH_FIRST) and odbc_data_source($con,
SQL_FETCH_NEXT) in a loop to list all DSN entries defined on your local machine. The tutorial
script below shows a good example:
 <?php
 if (!$con) {
 print("There is a problem with the connection.\n");
 } else {
 print("The ODBC connection object is ready.\n");
 $list = odbc_data_source($con, SQL_FETCH_FIRST);
 while ($list) {
 foreach ($list as $key => $value) {
 print($key . " = " . $value . "\n");
 }
 $list = odbc_data_source($con, SQL_FETCH_NEXT);
 }
 }
 ?>
 If you run this script, you will get something like this:
 The ODBC connection object is ready.
 server = MS Access Database
 description = Microsoft Access Driver (*.mdb)
 server = dBASE Files
 description = Microsoft dBase Driver (*.dbf)
 server = Excel Files
 description = Microsoft Excel Driver (*.xls)
 server = FYI_SQL_SERVER
 description = SQL Server

 Warning: odbc_data_source(): SQL error: [Microsoft]
 [ODBC SQL Server Driver][SQLServer]Changed database context
 to 'FyiCenterData'., SQL state 01000 in SQLDataSources in
 C:\test\fyi_center.php on line 12
 The error message seems to be very strange. But the result is correct.
 How To Execute a SQL Statement using odbc_exec()?
 Once you have created an ODBC connection object, you can use the odbc_exec() function to
send a SQL statement to the SQL Server linked to the connection object for execution.
 Here is a simple PHP script that creates a new schema and a new table:
 <?php

 # creating a new schema
 $sql = 'CREATE SCHEMA fyi';
 odbc_exec($con, $sql);

 # creating a new table
 $sql = 'CREATE TABLE fyi.center (name VARCHAR(80))';
 odbc_exec($con, $sql);

 ?>
 If you run this script for the first time, it will execute those two statements correctly for you. But if
you run it again, you will some warning messages:
 Warning: odbc_exec(): SQL error: [Microsoft]
 [ODBC SQL Server Driver][SQL Server]
 There is already an object named 'fyi' in the database.,
 SQL state S0001 in SQLExecDirect in C:\test\fyi_center.php
 on line 6

 There is already an object named 'center' in the database.,
 on line 10
 The messages are very clear and easy to understand.
 How To Retrieve Error Messages using odbc_errormsg()?
 When you call odbc_exec() to execute a SQL statement, and the execution failed on the SQL
Server, you can use odbc_error() and odbc_errormsg() to retrieve the error code and error
messages.
 The tutorial script below shows you a good example:
 <?php

 # dropping a table
 $sql = 'DROP TABLE fyi.center';
 $res = odbc_exec($con, $sql);
 if (!$res) {
 print("Execution failed:\n");
 print(" State: ".odbc_error($con)."\n");
 print(" Error: ".odbc_errormsg($con)."\n");
 } else {
 print("Execution was successful.\n");
 }

 ?>
 If you run this script for the first time, you will get this output:
 Execution was successful.
 If you run this script again, the SQL statement will fail on the SQL Server, and you will get:
 Cannot drop the table 'fyi.center', because
 it does not exist or you do not have permission.,
 on line 6
 Execution failed:
 State: S0002
 Error: [Microsoft][ODBC SQL Server Driver][SQL Server]
 Cannot drop the table 'fyi.center', because it does not
 exist or you do not have permission.
 How To Turn Off Warning Messages during PHP Execution?
 If don't want see warning messages generated from the PHP engine when executing PHP scripts,
you can change the error_reporting setting in the php.ini configuration file.
 Open php.ini and change the following lines:
 ;error_reporting = E_ALL & ~E_NOTICE
 error_reporting = E_ALL & ~E_WARNING
 Now run the script in the previous tutorial again, you will not see the warning messages from the
PHP engine:
 Execution failed:
 State: S0002
 Error: [Microsoft][ODBC SQL Server Driver][SQL Server]
 Cannot drop the table 'fyi.center', because it does not
 exist or you do not have permission.
 How To Receive Returning Result from a Query?
 When you execute a SQL SELECT statement with the odbc_exec() function, you can capture the
returning result with a result set object with the following syntax:
 $result_set = odbc_exec($sql_statement);
 #- The returning value could be a Boolean value FALSE,
 #- if the execution failed.
 Data rows and field values in the result set object can be retrieved using other ODBC functions as
shown in the tutorial PHP script below:
 <?php

 $sql = 'SELECT GETDATE()';
 odbc_fetch_row($res);
 $date = odbc_result($res,1);
 print("Database current time: ". $date ."\n");

 ?>
 If you run this script, you will get something like this:
 Database current time: 2007-06-02 22:07:05.110
 How To Loop through Result Set Objects using odbc_fetch_row()?
 If the returning output of a query statement is captured in a result set object, you can use
odbc_fetch_row() to loop through each row in the output.
 The tutorial PHP script below shows you how to list tables in the database:
 <?php

 $sql = "SELECT * FROM sys.objects"
 . " WHERE type_desc='USER_TABLE'";
 print("User Tables:\n");
 while (odbc_fetch_row($res)) {
 print(" ".odbc_result($res,'name')."\n");
 }
 odbc_free_result($res);

 ?>
 If you run this script, you will get something like:
 User Tables:
 fyi_rates
 fyi_team
 fyi_random
 fyi_links_indexed
 fyi_links
 fyi_links_copy
 tipBackup2
 How To Create Prepared Statements using odbc_prepare()?
 If you have a SQL statement that need to executed repeatedly many times with small changes,
you can create a prepared statement object with parameters so it can be executed more
efficiently.
 There are two functions you need to use prepare and execute a prepared statement object:
 $statement_object = odbc_prepare($connection,
 $statement_string);
 #- The $statement_string may have parameters represented
 #- by "?".

 $result_set = odbc_execute($statement_object $array);
 #- The $array is used to supply values to parameters
 #- defined in the statement object.
 The tutorial PHP script below shows you how to insert 3 rows into a table with a prepared
statement object with 2 parameters:
 <?php

 $sql = "INSERT INTO fyi_rates (id, comment) VALUES (?,?)";
 $statement = odbc_prepare($con, $sql);

 $res = odbc_execute($statement, array(301, "Good"));
 $res = odbc_execute($statement, array(302, "Average"));
 $res = odbc_execute($statement, array(303, "Bad"));

 $sql = "SELECT * FROM fyi_rates WHERE id>300";
 print(" ".odbc_result($res,1));
 print(", ".odbc_result($res,2)."\n");
 }

 ?>
 If you this sample script, you will get:
 301, Good
 302, Average
 303, Bad
 How To List All Columns in a Table using odbc_columns()?
 If you want to get a list of all columns in a table, you can use the odbc_columns() function, which
can actually be used to list all columns in all tables and views in the database. The syntax of
odbc_columns() is:
 $result_set = odbc_columns($connection_object,
 $qualifier, # database name for SQL Server
 $schema, # schema name for SQL Server
 $table_name, # table or view name for SQL Server
 $column_name # column name
 )
 #- The returning result set contains 5 fields:
 #- TABLE_QUALIFIER, TABLE_SCHEM, TABLE_NAME, COLUMN_NAME
 #- DATA_TYPE, TYPE_NAME, PRECISION, LENGTH, SCALE, RADIX
 #- NULLABLE, REMARKS
 The schema, table_name and column_name arguments accept search patterns ( '%' to match
zero or more characters and '_' to match a single character).
 The tutorial example below shows you how to get a list of columns in a table called fyi_rates in
the current database, FyiCenterData, which is hard coded in the DSN definition:
 <?php

 # odbc_tables($con, $database, $schema, $table, $column);
 $res = odbc_columns($con, 'FyiCenterData','%','fyi_rates','%');
 print("Column list of fyi_rates:\n");
 print(", ".odbc_result($res,3));
 }

 ?>
 Column list of fyi_rates:
 dbo, fyi_rates, id, 4, int, 10, 4, 0, 10, 1,
 dbo, fyi_rates, comment, 12, varchar, 16, 16, , , 1,
 Now you know that table, fyi_rates, has two columns: "id INT" and "comment VARCHAR(16)".
 How To List All Tables in the Database using odbc_tables()?
 If you want to get a list of all tables in the database, you can use the odbc_tables() function,
which can actually be used to list all tables and views in the database. The syntax of
odbc_tables() is:
 $result_set = odbc_tables($connection_object,
 $owner, # schema name for SQL Server
 $name, # table or view name for SQL Server
 $type # valid type names are TABLE and VIEW
 )
 #- TABLE_QUALIFIER, TABLE_OWNER, TABLE_NAME, TABLE_TYPE,
 #- REMARKS
 The owner and name arguments accept search patterns ('%' to match zero or more characters
and '_' to match a single character).
 The tutorial example below shows you how to get a list of tables in the current database,
FyiCenterData, which is hard coded in the DSN definition:
 <?php

 # odbc_tables($con, $database, $schema, $name, $type);
 $res = odbc_tables($con, 'FyiCenterData','%','%','TABLE');
 }

 ?>
 FyiCenterData, dbo, fyi_links, TABLE,
 FyiCenterData, dbo, fyi_links_copy, TABLE,
 FyiCenterData, dbo, fyi_links_indexed, TABLE,
 FyiCenterData, dbo, fyi_random, TABLE,
 FyiCenterData, dbo, fyi_rates, TABLE,
 FyiCenterData, dbo, fyi_team, TABLE,
 FyiCenterData, dbo, tipBackup2, TABLE,
 How To List All Columns in a Table using odbc_columns()?
 If you want to get a list of all columns in a table, you can use the odbc_columns() function, which
can actually be used to list all columns in all tables and views in the database. The syntax of
odbc_columns() is:
 $result_set = odbc_columns($connection_object,
 $schema, # schema name for SQL Server
 $table_name, # table or view name for SQL Server
 $column_name # column name
 )
 #- TABLE_QUALIFIER, TABLE_SCHEM, TABLE_NAME, COLUMN_NAME
 #- DATA_TYPE, TYPE_NAME, PRECISION, LENGTH, SCALE, RADIX
 #- NULLABLE, REMARKS
 The schema, table_name and column_name arguments accept search patterns ( '%' to match
zero or more characters and '_' to match a single character).
 The tutorial example below shows you how to get a list of columns in a table called fyi_rates in
the current database, FyiCenterData, which is hard coded in the DSN definition:
 <?php

 # odbc_tables($con, $database, $schema, $table, $column);
 $res = odbc_columns($con, 'FyiCenterData','%','fyi_rates','%');
 print("Column list of fyi_rates:\n");
 }

 ?>
 Column list of fyi_rates:
 dbo, fyi_rates, id, 4, int, 10, 4, 0, 10, 1,
 dbo, fyi_rates, comment, 12, varchar, 16, 16, , , 1,
 Now you know that table, fyi_rates, has two columns: "id INT" and "comment VARCHAR(16)".
 How To Create Prepared Statements using odbc_prepare()?
 If you have a SQL statement that need to executed repeatedly many times with small changes,
you can create a prepared statement object with parameters so it can be executed more
efficiently.
 There are two functions you need to use prepare and execute a prepared statement object:
 $statement_object = odbc_prepare($connection,
 $statement_string);
 #- The $statement_string may have parameters represented
 #- by "?".

 $result_set = odbc_execute($statement_object $array);
 #- The $array is used to supply values to parameters
 #- defined in the statement object.
 The tutorial PHP script below shows you how to insert 3 rows into a table with a prepared
statement object with 2 parameters:
 <?php

 $sql = "INSERT INTO fyi_rates (id, comment) VALUES (?,?)";
 $statement = odbc_prepare($con, $sql);

 $res = odbc_execute($statement, array(301, "Good"));
 $res = odbc_execute($statement, array(302, "Average"));
 $res = odbc_execute($statement, array(303, "Bad"));

 $sql = "SELECT * FROM fyi_rates WHERE id>300";
 }

 ?>
 If you this sample script, you will get:
 301, Good
 302, Average
 303, Bad

SQL Server Questions

Uploaded by

SQL Server Questions

Uploaded by

What is Microsoft SQL Server?

Microsoft SQL Server is a relational database management system (RDBMS) developed by

Microsoft SQL Server release history:

 How to download Microsoft SQL Server 2005 Express Edition?

A database is a logical container that contains a set of related database objects:

 What is the simplest way to create a new database?

Where is my database stored on the hard disk?

SELECT type_desc, physical_name, size

ROWS c:\Program Files\Microsoft SQL Server

LOG c:\Program Files\Microsoft SQL Server

Go verify these two files with Windows Explorer.

How to create database with physical files specified?

DROP DATABASE FyiCenterData

CREATE DATABASE FyiCenterData

SELECT type_desc, name, physical_name, size

How to rename databases?

Why I am getting this error when renaming a database?

1. Launch one instance of SQL Server Management Studio and run:

Obviously, the first instance is blocking the "ALTER DATABASE" statement.

 How to set a database state to OFFLINE?

How to set database to be READ_ONLY?

Databases in SQL Server have two update options:

INSERT Links (Name) VALUES ('dba.FYIcenter.com')

ALTER DATABASE FyiCentercomData SET READ_ONLY

INSERT Links (Name) VALUES ('dev.FYIcenter.com')

SELECT * FROM Links

ALTER DATABASE FyiCentercomData SET READ_WRITE

INSERT Links (Name) VALUES ('dev.FYIcenter.com')

Databases in SQL Server have three user access options:

ALTER DATABASE FyiCentercomData SET SINGLE_USER

Now connect to server with another client session and try:

Go back to the first session and re-set the database to MULTI_USER:

What are system databases?

What are DDL (Data Definition Language) statements for tables?

 How to create new tables with "CREATE TABLE" statements?

 How To Generate CREATE TABLE Script on an Existing Table?

What Are DML (Data Manipulation Language) Statements?

(Continued on next topic...)

How To Create a Testing Table with Test Data?

SELECT c.column_id as seq, c.name, x.name as type,

SELECT * FROM fyi_links

The values are stored in the new record nicely.

How To Use Column Default Values in INSERT Statements?

SELECT * FROM fyi_links

How to provide column names in INSERT Statements?

The following tutorial exercise gives you some good examples:

INSERT INTO fyi_links (id) VALUES (110)

SELECT * FROM fyi_links

How To Insert Multiple Rows with One INSERT Statement?

SELECT * FROM fyi_links

How To Update Values in a Table with UPDATE Statements?

UPDATE fyi_links SET counts = 999, notes = 'Good.'

SELECT * FROM fyi_links WHERE id = 101

How To Update Multiple Rows with One UPDATE Statement?

SELECT * FROM fyi_links

SELECT * FROM fyi_links WHERE id >= 500

How to use old values to define new values in UPDATE statements?

UPDATE fyi_links SET id = id+200, counts = id*2

SELECT * FROM fyi_links WHERE id >= 500

Is the Order of Columns in the SET Clause Important?

Here is a good tutorial exercise:

-- Update "id" before "counts"

-- Check the new values

-- Reset to old values

How To Use Values from Other Tables in UPDATE Statements?

-- Insert some rows in the new table

-- Update fyi_links with values from fyi_rates

-- View the updated values

What Happens If the UPDATE Subquery Returns No Rows?

-- view old values

-- make sure there is no matching row in fyi_rates

-- update a subquery returning no rows

-- view new values

What Happens If the UPDATE Subquery Returns Multiple Rows?

-- make sure there are 2 match rows

-- update with subquery that returns 2 rows