Microsoft Docs

General Syntax Rules

T-SQL statements are generally composed of keywords, identifiers, operators, and literals. The following rules apply:

• Statements are typically terminated by a semicolon (;), although it is often optional for single statements.
• Keywords are not case-sensitive, but it is a common practice to write them in uppercase for readability.
• Identifiers (object names, variable names) are generally case-insensitive, depending on the server's collation settings.
• String literals are enclosed in single quotes (').
• Numeric literals are written directly.

Example: Basic SELECT Statement

SELECT
    <column_list>
FROM
    <table_source>
WHERE
    <search_condition>;

Delimiters and Special Characters

Several characters have special meaning in T-SQL:

• ; (Semicolon): Statement terminator.
• '...' (Single Quotes): Encloses string literals.
• -- ... (Double Hyphen): Starts a single-line comment.
• /* ... */ (Slash-Asterisk): Starts a multi-line comment.
• @ (At sign): Precedes variable names (e.g., @myVariable).
• # (Hash): Indicates a temporary table (e.g., #tempTable).
• ## (Double Hash): Indicates a global temporary table.
• [] (Square Brackets): Used to delimit identifiers that contain special characters or reserved words.
• . (Dot): Used to separate schema and object names (e.g., dbo.myTable).

Standard Statement Structure

Most T-SQL statements follow a predictable structure:

• Keywords: Reserved words with specific meanings (e.g., SELECT, INSERT, CREATE, ALTER).
• Identifiers: Names of database objects such as tables, views, stored procedures, and columns.
• Expressions: Combinations of values, operators, and functions that evaluate to a single value.
• Clauses: Optional parts of a statement that modify its behavior (e.g., WHERE, GROUP BY, ORDER BY).

Example: CREATE TABLE Syntax

CREATE TABLE <table_name> (
    <column_name1> <data_type> [<constraints>],
    <column_name2> <data_type> [<constraints>],
    ...
);

Here, <...> denotes placeholders for specific names or values.

Common Programmability Constructs

T-SQL offers rich features for programmability:

Stored Procedures

CREATE PROCEDURE <procedure_name>
    [@parameter1 datatype [= default_value]]
    [, @parameter2 datatype [= default_value]]
AS
BEGIN
    -- T-SQL statements for the procedure body
    -- ...
END;

Functions

CREATE FUNCTION <function_name>
    (@parameter1 datatype, @parameter2 datatype)
RETURNS <return_data_type>
AS
BEGIN
    -- T-SQL statements for the function body
    DECLARE @result <return_data_type>;
    -- ...
    RETURN @result;
END;

Triggers

CREATE TRIGGER <trigger_name>
ON <table_name>
AFTER <INSERT|UPDATE|DELETE>
AS
BEGIN
    -- T-SQL statements for the trigger body
    -- Accessing inserted and deleted tables
    -- ...
END;

Data Type Specification

When defining columns or variables, you must specify a valid T-SQL data type. Common types include:

• INT, BIGINT, SMALLINT, TINYINT (Integer types)
• DECIMAL, NUMERIC, FLOAT, REAL (Numeric types)
• VARCHAR(n), NVARCHAR(n), CHAR(n), NCHAR(n) (String types)
• DATE, DATETIME, DATETIME2, TIME (Date and Time types)
• BIT (Boolean type)
• VARBINARY(n) (Binary data type)
• UNIQUEIDENTIFIER (GUID)

Refer to the T-SQL Data Types documentation for a complete list and detailed descriptions.