Соглашение об именах - это набор правил для выбора последовательности символов, которая будет использоваться для идентификаторов, которые обозначают переменные, типы, функции и другие объекты в исходном коде и документации.
Причины использования соглашения об именах (в отличие от разрешения программистам выбирать любую последовательность символов) включают следующее:
Object | Code | Notation | Length | Plural | Prefix | Suffix | Abbreviation | Char Mask | Example |
---|---|---|---|---|---|---|---|---|---|
Database | UPPERCASE | 30 | No | No | No | Yes | [A-z] | MYDATABASE |
|
Schema | lowercase | 30 | No | No | No | Yes | [a-z][0-9] | myschema |
|
Global Temporary Table | PascalCase | 117 | No | No | No | Yes | ##[A-z][0-9] | ##MyTable |
|
Local Temporary Table | PascalCase | 116 | No | No | No | Yes | #[A-z][0-9] | #MyTable |
|
File Table | PascalCase | 128 | No | FT_ |
No | Yes | [A-z][0-9] | FT_MyTable |
|
Memory-optimized SCHEMA_AND_DATA Table | PascalCase | 128 | No | MT_ |
_SD |
Yes | [A-z][0-9] | MT_MyTable_SD |
|
Memory-optimized SCHEMA_ONLY Table | PascalCase | 128 | No | MT_ |
_SO |
Yes | [A-z][0-9] | MT_MyTable_SO |
|
Temporal Table | PascalCase | 128 | No | No | _TT |
Yes | [A-z][0-9] | MyTable_TT |
|
Disk-Based Table | U | PascalCase | 128 | No | No | No | Yes | [A-z][0-9] | MyTable |
Disk-Based Wide Table - SPARSE Column | U | PascalCase | 128 | No | No | _SPR |
Yes | [A-z][0-9] | MyTable_SPR |
Table Column | PascalCase | 128 | No | No | No | Yes | [A-z][0-9] | MyColumn |
|
Table Column SPARSE | PascalCase | 128 | No | No | _SPR |
Yes | [A-z][0-9] | MyColumn_SPR |
|
Columns Check Constraint | C | PascalCase | 128 | No | CTK_ |
No | Yes | [A-z][0-9] | CTK_MyTable_MyColumn_AnotherColumn |
Column Check Constraint | C | PascalCase | 128 | No | CK_ |
No | Yes | [A-z][0-9] | CK_MyTable_MyColumn |
Column Default Values | D | PascalCase | 128 | No | DF_ |
No | Yes | [A-z][0-9] | DF_MyTable_MyColumn |
Table Primary Key | PK | PascalCase | 128 | No | PK_ |
No | Yes | [A-z][0-9] | PK_MyTableID |
Table Unique (Alternative) Key | UQ | PascalCase | 128 | No | AK_ |
No | Yes | [A-z][0-9] | AK_MyTable_MyColumn_AnotherColumn |
Table Foreign Key | F | PascalCase | 128 | No | FK_ |
No | Yes | [A-z][0-9] | FK_MyTable_ForeignTableID |
Table Clustered Index | PascalCase | 128 | No | IXC_ |
No | Yes | [A-z][0-9] | IXC_MyTable_MyColumn_AnotherColumn |
|
Table Non Clustered Index | PascalCase | 128 | No | IX_ |
No | Yes | [A-z][0-9] | IX_MyTable_MyColumn_AnotherColumn |
|
DDL Trigger | TR | PascalCase | 128 | No | TR_ |
_DDL |
Yes | [A-z][0-9] | TR_LogicalName_DDL |
DML Trigger | TR | PascalCase | 128 | No | TR_ |
_DML |
Yes | [A-z][0-9] | TR_MyTable_LogicalName_DML |
Logon Trigger | TR | PascalCase | 128 | No | TR_ |
_LOG |
Yes | [A-z][0-9] | TR_LogicalName_LOG |
View | V | PascalCase | 128 | No | VI_ |
No | No | [A-z][0-9] | VI_LogicalName |
Indexed View | V | PascalCase | 128 | No | VIX_ |
No | No | [A-z][0-9] | VIX_LogicalName |
Statistic | PascalCase | 128 | No | ST_ |
No | No | [A-z][0-9] | ST_MyTable_MyColumn_AnotherColumn |
|
Stored Procedure | P | PascalCase | 128 | No | usp_ |
No | No | [A-z][0-9] | usp_LogicalName |
Scalar User-Defined Function | FN | PascalCase | 128 | No | udf_ |
No | No | [A-z][0-9] | udf_FunctionLogicalName |
Table-Valued Function | FN | PascalCase | 128 | No | tvf_ |
No | No | [A-z][0-9] | tvf_FunctionLogicalName |
Synonym | SN | camelCase | 128 | No | sy_ |
No | No | [A-z][0-9] | sy_logicalName |
Sequence | SO | PascalCase | 128 | No | sq_ |
No | No | [A-z][0-9] | sq_TableName |
CLR Assembly | PascalCase | 128 | No | CA |
No | Yes | [A-z][0-9] | CALogicalName |
|
CLR Stored Procedures | PC | PascalCase | 128 | No | pc_ |
No | Yes | [A-z][0-9] | pc_CAName_LogicalName |
CLR Scalar User-Defined Function | PascalCase | 128 | No | cudf_ |
No | No | [A-z][0-9] | cudf_CAName_LogicalName |
|
CLR Table-Valued Function | PascalCase | 128 | No | ctvf_ |
No | No | [A-z][0-9] | ctvf_CAName_LogicalName |
|
CLR User-Defined Aggregates | PascalCase | 128 | No | ca_ |
No | No | [A-z][0-9] | ca_CAName_LogicalName |
|
CLR User-Defined Types | PascalCase | 128 | No | ct_ |
No | No | [A-z][0-9] | ct_CAName_LogicalName |
|
CLR Triggers | PascalCase | 128 | No | ctr_ |
No | No | [A-z][0-9] | ctr_CAName_LogicalName |
|
Linked Server | PascalCase | 128 | No | ls_ |
No | No | [A-z][0-9] | ls_LogicalName |
More details about SQL Server data types and mapping it with another databases and program languages you can find here
General Type | Type | ANSI | Recommended | What use instead | Why use or not |
---|---|---|---|---|---|
Exact Numerics | bit | No | Maybe | tinyint | bit convert any number (except 0) to 1, 0 converted to 0 |
Exact Numerics | tinyint | No | Maybe | int | for saving 3 bytes compare to int data type or for replacing bit data type |
Exact Numerics | smallint | Yes | Maybe | int | for saving 2 bytes compare to int data type |
Exact Numerics | int | Yes | Yes | - | |
Exact Numerics | bigint | No | Yes | int | if you work more than 2^31 numbers. |
Exact Numerics | decimal | Yes | Yes | - | |
Exact Numerics | smallmoney | No | Maybe | decimal | possibility to loss precision due to rounding errors |
Exact Numerics | money | No | Maybe | decimal | possibility to loss precision due to rounding errors |
Approximate Numerics | real | Yes | Yes | - | |
Approximate Numerics | float(1-24) | Yes | No | real | SQL Server automatically converts float(1-24) to real data type |
Approximate Numerics | float(24-53) | Yes | Yes | - | |
Date and Time | date | Yes | Yes | - | |
Date and Time | smalldatetime | No | Maybe | date | |
Date and Time | time | Yes | Yes | - | |
Date and Time | datetime2 | No | Yes | - | |
Date and Time | datetime | Yes | Maybe | datetime2 | On the Advantages of DateTime2(n) over DateTime |
Date and time | datetimeoffset | No | Yes | - | |
Character Strings | char | Yes | Maybe | varchar | Save 1 byte from varchar , but be ready for trailing spaces |
Character Strings | varchar | Yes | Yes | - | |
Character Strings | varchar(max) | Yes | Yes | - | |
Character Strings | nchar | Yes | Maybe | nvarchar | |
Character Strings | nvarchar | Yes | Yes | - | |
Character Strings | nvarchar(max) | Yes | Yes | - | |
Character Strings | ntext | No | Deprecated | nvarchar(max) | NVARCHAR(MAX) VS NTEXT in SQL Server |
Character Strings | text | No | Deprecated | varchar(max) | Differences Between Sql Server TEXT and VARCHAR(MAX) Data Type |
Binary Strings | image | No | Deprecated | varbinary(max) | VARBINARY(MAX) Tames the BLOB |
Binary Strings | binary | Yes | Deprecated | varbinary | Conversions between any data type and the binary data types are not guaranteed |
Binary Strings | varbinary | Yes | Yes | - | |
Binary Strings | varbinary(max) | Yes | Yes | - | |
Other Data Types | cursor | No | Yes | - | |
Other Data Types | sql_variant | No | Yes | - | |
Other Data Types | hierarchyid | No | Yes | - | |
Other Data Types | rowversion | No | Maybe | - | |
Other Data Types | timestamp | No | Deprecated | rowversion | it is just synonym to rowversion data type and must be removed |
Other Data Types | uniqueidentifier | No | Yes | - | |
Other Data Types | xml | Yes | Yes | - | |
Other Data Types | table | No | Maybe | - | |
Spatial Data Types | geometry | No | Yes | - | |
Spatial Data Types | geography | No | Yes | - |
This is only recommendations! But it is consistent for choosing only 1 function from possibles alterntives and use only it.
Not Recommended | Recommended | When and Why | More details |
---|---|---|---|
!=
|
<>
|
<> is
ANSI
, != not ANSI , <> and != are identical |
13 |
CONVERT
|
CAST
|
CAST is
ANSI
|
14,15 |
ISNULL
|
COALECSE
|
COALECSE is
ANSI
and supports more than two arguments, ISNULL has dangerous behaviour with possibility to implicit triming string |
16,17 |
DATEDIFF
|
DATEADD
|
The predicate MyDateTime < DATEADD(SECOND, -1, GETUTCDATE()) syntax is
SARGable
|
18,19 |
SELECT
|
SET
|
Using SET (is
ANSI
) instead of SELECT when assigning variables due to properly work with Msg 501 Subquery returned more than 1 value |
20,21,22 |
STR
|
CAST
|
STR is not
ANSI
, extremly slow, don't use more than 15 digits, and has rounding problem - use CAST plus concatenate instead STR |
23 |
ISNUMERIC
|
TRY_CONVERT
|
ISNUMERIC can often lead to data type conversion errors, when importing data. For SQL Server below 2012 use WHERE with LIKE . |
24 |
GETDATE
|
[SYSUTCDATETIME ] |
Daylight Saving Time and other factors can play havoc with our dates and times, rounding to the nearest 3 milliseconds. | 25 |
GETUTCDATE
|
[SYSUTCDATETIME ] |
Daylight Saving Time and other factors can play havoc with our dates and times, rounding to the nearest 3 milliseconds. | 25 |
SYSDATETIME
|
[SYSUTCDATETIME ] |
Daylight Saving Time and other factors can play havoc with our dates and times, rounding to the nearest 3 milliseconds. | 25 |
SQL Server T-SQL Coding Conventions, Best Practices, and Programming Guidelines.
For database objects names in code use only schema plus object name, do not hardcode server and database names in your code:
/* good */
CREATE TABLE dbo.MyTable (MyColumn int);
/* bad*/
CREATE TABLE PRODSERVER.PRODDB.dbo.MyTable (MyColumn int);
CREATE TABLE MyTable (MyColumn int);
Delimiters: spaces (not tabs).
Always use multi-line comment /* */
instead in-line comment --
in production code due to potential formating problems in different tools and programs.
More details here, here, and here.
Never use asterisk (*
) in select statements SELECT *
and INSERT
statements, use explicit column names.
Main problems are: traffic issues, Memory Grants issues, Index usage issues.
Only one exception, see it below.
More details here,
here,
here,
here.
Use asterisk (*
) ONLY in an archiving situation, where rows are being moved to another table that must have the same structure.
INSERT INTO SalesOrderArchive /* Note no column list */
SELECT *
FROM SalesOrder
WHERE OrderDate < @OneYearAgo;
DELETE FROM SalesOrder
WHERE OrderDate < @OneYearAgo;
If a new column is added to SalesOrder
table in the future, but not to SalesOrderArchive
, the INSERT
will fail.
Which sounds bad, but it's actually a really good thing! Because the alternative is much worse.
If all the columns were listed on the INSERT
and the SELECT
, then the INSERT
would succeed, and so would the following DELETE
(which is effectively DELETE *
).
Production code that succeeds doesn't get any attention, and it may be a long time before someone notices that the new column is not being archived, but being silently deleted altogether.
More details here.
No square brackets []
and reserved words in object names and alias, use only Latin symbols [A-z]
and numeric [0-9]
.
Prefer
ANSI
syntax and functions (
CAST
instead
CONVERT
,
COALESE
instead
ISNULL
, etc.).
All finished expressions should have semicolon ;
at the end.
This is
ANSI
standard and Microsoft announced with the SQL Server 2008 release that semicolon statement terminators will become mandatory in a future version so statement terminators other than semicolons (whitespace) are currently deprecated.
This deprecation announcement means that you should always use semicolon terminators in new development.
From Transact-SQL Syntax Conventions (Transact-SQL):
Although the semicolon isn't required for most statements in this version of SQL Server, it will be required in a future version.
More details here, here, and here. Semicolon is mandority for:
WITH CTE
:
When a CTE is used in a statement that is part of a batch, the statement before it must be followed by a semicolon.
Merge
statements:
The MERGE statement requires a semicolon (;) as a statement terminator. Error 10713 is raised when a MERGE statement is run without the terminator.
TROW
exceptions:
The statement before the THROW statement must be followed by the semicolon (;) statement terminator.
If the SEND statement isn't the first statement in a batch or stored procedure, the preceding statement must be terminated with a semicolon (;).
All script files should end with GO
and line break. This is neccesary for batching scripts run throw sqlcmd
or another tools.
Keywords should be in UPPERCASE: SELECT
, FROM
, GROUP BY
etc. This increases the readability of the code.
Data types declaration should be in lowercase: varchar(30)
, int
, real
, nvarchar(max)
etc.
More details here.
All system database and tables must be in lowercase for properly working for Case Sensitive instance: master, sys.tables …
.
Do not use nested transactions. The commit of a nested transaction has absolutely no effect – as the only transaction that really exists as far as SQL Server is concerned is the outer one. More details here.
Whenever you have data modification on non-temporary tables, is to use explicit transactions over autocommit.
SELECT
statement, use autocommit.Avoid using Cross-Database Queries because it increase backup/restore complexity (you restore one database, then realise you don’t have log backups to bring the other database to the same point in time). Also Azure SQL Database does not support cross-database queries and you can not migrate into in future.
Use temp
tables to reduce network trafic, decrease query complexity and also to get better estimates for modification queries. More details here.
INFORMATION_SCHEMA
views only represent a subset of the metadata of an object. The only reliable way to find the schema of a object is to query the sys.objects
catalog view.
When more than one logical operator is used always use parentheses, even when they are not required. This can improve the readability of queries, and reduce the chance of making a subtle mistake because of operator precedence. There is no significant performance penalty in using parentheses. More details here.
SELECT
ProductID
FROM Production.Product
WHERE (ProductModelID = 20 OR ProductModelID = 21)
AND Color = 'Red';
Always use aliases for table names. More details here.
Whenever you’re working with multiple tables in a join (explicit or, like this one, implicit), always specify which table each column comes from. Even if your code works just fine today, just adding an unfortunately named column many years later can break your code. More details here.
SELECT
p.LastName AS "Last Name"
FROM dbo.Person AS p;
More details here. All possible ways using aliases in SQL Server:
/* Recommended due to ANSI */
SELECT SCHEMA_NAME(schema_id) + '.' + "name" AS "Tables" FROM sys.tables;
/* Not recommended but possible */
SELECT SCHEMA_NAME(schema_id) + '.' + [name] AS "Tables" FROM sys.tables;
SELECT Tables = SCHEMA_NAME(schema_id) + '.' + [name] FROM sys.tables;
SELECT "Tables" = SCHEMA_NAME(schema_id) + '.' + [name] FROM sys.tables;
SELECT [Tables] = SCHEMA_NAME(schema_id) + '.' + [name] FROM sys.tables;
SELECT 'Tables' = SCHEMA_NAME(schema_id) + '.' + [name] FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] [Tables] FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] 'Tables' FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] "Tables" FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] Tables FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] AS [Tables] FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] AS 'Tables' FROM sys.tables;
SELECT SCHEMA_NAME(schema_id) + '.' + [name] AS Tables FROM sys.tables;
Always consider using an explicit range condition when comparing dates for properly make use of an index. More details here.
/* bad */
SELECT sale_date
FROM sales
WHERE YEAR(sale_date) >= 1970
AND YEAR(sale_date) < 1971;
/* good */
SELECT sale_date
FROM sales
WHERE sale_date >= CAST('1970-01-01' AS date)
AND sale_date < CAST('1971-01-01', AS date);
The first argument in SELECT
expression should be on the next line:
SELECT
FirstName
Arguments are divided by line breaks, commas should be placed before an argument:
SELECT
FirstName
, LastName
For SQL Server >= 2012 use
FETCH-OFFSET
instead
TOP
.
More details here.
But if you use
TOP
avoid use
TOP
in a SELECT
statement without an ORDER BY
.
More details here.
If you using
TOP
(instead recommended
FETCH-OFFSET
) function with round brackets because
TOP
has supports use of an expression, such as (@Rows*2)
, or a sub query: SELECT TOP(100) LastName …
.
More details here.
Also
TOP
without brackets does not work with UPDATE
and DELETE
statements.
/* Not working without brackets () */
DECLARE @n int = 1;
SELECT TOP@n name FROM sys.objects;
For demo queries use TOP(100)
or lower value because SQL Server uses one sorting method for TOP
1-100 rows, and a different one for 101+ rows.
More details here and here.
ORDER BY
clause as positional representations of the columns in the select list.
The statement with integers is not as easily understood by others compared with specifying the actual column name.
In addition, changes to the select list, such as changing the column order or adding new columns, requires modifying the ORDER BY
clause in order to avoid unexpected results.
More details here.
/* bad */
SELECT ProductID, Name FROM production.Production ORDER BY 2;
/ good / SELECT ProductID, Name FROM production.Production ORDER BY Name;
Avoid wrapping functions around columns specified in the WHERE
and JOIN
clauses.
Doing so makes the columns non-deterministic and prevents the query processor from using indexes.
Use NULL
or NOT NULL
for each column in a temporary table. The
ANSI_NULL_DFLT_ON
option control the way the Database Engine assigns the NULL
or NOT NULL
attributes to columns when these attributes are not specified in a CREATE TABLE
or ALTER TABLE
statement.
If a connection executes a procedure with different settings for these options than the connection that created the procedure, the columns of the table created for the second connection can have different nullability and exhibit different behavior. If NULL
or NOT NULL
is explicitly stated for each column, the temporary tables are created by using the same nullability for all connections that execute the procedure.
Use modification statements that convert nulls and include logic that eliminates rows with null values from queries. Be aware that in Transact-SQL, NULL
is not an empty or "nothing" value. It is a placeholder for an unknown value and can cause unexpected behavior, especially when querying for result sets or using AGGREGATE functions.
Use the UNION ALL
operator instead of the UNION
or OR
operators, unless there is a specific need for distinct values.
The UNION ALL
operator requires less processing overhead because duplicates are not filtered out of the result set.
Avoid using INSERT INTO
a permanent table with ORDER BY
.
More details here.
Avoid using shorthand (wk, yyyy, d
etc.) with date/time operations, use full names: month, day, year
.
More details here.
Avoid ambiguous formats for date-only literals, use CAST('yyyymmdd' AS DATE)
format.
Avoid treating dates like strings and avoid calculations on the left-hand side of the WHERE
clause.
More details here.
Avoid using hints except RECOMPILE
if needed and NOEXPAND
(see next tip).
More details here.
Use
NOEXPAND
hint for indexed views on non enterprise editions and Prior to SQL Server 2016 (13.x) SP1 to let the query optimizer know that we have indexes.
More details here.
Use
LOOP JOIN
and
FAST 1
query hints for deleting huge number of rows with ON DELETE CASCADE
foreign keys specification.
More details here.
If you use hints always use it with WITH
keyword because omitting the WITH
keyword is a deprecated feature and will be removed from future Microsoft SQL Server versions.
One benefit of using the WITH
keyword is that you can specify multiple table hints using the WITH keyword against the same table.
More details here.
Avoid use of SELECT…INTO
for production code, use instead CREATE TABLE
+ INSERT INTO …
approach. More details here.
Use only ISO standard JOINS syntaxes. The old style Microsoft/Sybase JOIN
style for SQL, which uses the =*
and *=
syntax, has been deprecated and is no longer used.
Queries that use this syntax will fail when the database engine level is 10 (SQL Server 2008) or later (compatibility level 100). The ANSI-89 table citation list (FROM tableA, tableB
) is still ISO standard for INNER JOINs
only. Neither of these styles are worth using.
It is always better to specify the type of join you requireINNER
, LEFT OUTER
, RIGHT OUTER
, FULL OUTER
and CROSS
, which has been standard since ANSI SQL-92 was published. While you can choose any supported JOIN
style, without affecting the query plan used by SQL Server, using the ANSI-standard syntax will make your code easier to understand, more consistent, and portable to other relational database systems.
More details here.
Do not use a scalar user-defined function (UDF) in a JOIN
condition, WHERE
search condition, or in a SELECT
list, unless the function is schema-bound.
More details here.
For scalar function use
WITH SCHEMABINDING
option to get a performance boost. More details here
Do not use
INFORMATION_SCHEMA
views to determine the schema of an object.
INFORMATION_SCHEMA
views only represent a subset of the metadata of an object.
The only reliable way to find the schema of a object is to query the
sys.objects
catalog view.
More details here.
Do not use explicit transactions for DML and DDL especially for reorganize index because the locking behavior of this statemnets becomes more restrictive. More details here.
In sqlcmd move the -U
and -P
parameters after the -Q
parameter so that the password is at the end of the command.
More details here
sqlcmd -S MSSQLSERVER$EXPRESS -Q "dbcc checkdb ('master') with DATA_PURITY, NO_INFOMSGS;" -U maintenanceUser -P ""weirdPassword
Use EXISTS
or NOT EXISTS
if referencing a subquery, and IN
or NOT IN
when have a list of literal values.
More details here.
For concatenate unicode strings:
N
;nvarchar(max)
;nvarchar(max)
.
Example:
/* good */
DECLARE @nvcmaxVariable nvarchar(max);
SET @nvcmaxVariable = CAST(N'ಠ russian anomaly ЯЁЪ ಠ ' AS nvarchar(max)) + N'something else' + N'another';
SELECT @nvcmaxVariable;
More details here.
Always specify a length to any text-based data type such as varchar
, nvarchar
, char
, nchar
:
/* bad */
DECLARE @myBadVarcharVariable varchar;
DECLARE @myBadNVarcharVariable nvarchar;
DECLARE @myBadCharVariable char;
DECLARE @myBadNCharVariable nchar;
/* good */
DECLARE @myGoodVarchareVariable varchar(50);
DECLARE @myGoodNVarchareVariable nvarchar(90);
DECLARE @myGoodCharVariable char(7);
DECLARE @myGoodNCharVariable nchar(10);
Use only
ORIGINAL_LOGIN()
function because is the only function that consistently returns the actual login name that we started with regardless of impersonation.
More details here.
Always use IF
statement with BEGIN-END
block to prevent errors with multi line statements:
DECLARE @x int = 0;
DECLARE @y int = 1;
/* bad */
IF @y > @x
SET @x = @x + 1;
SET @y = @y - 1;
ELSE
PRINT(1);
/* Msg 156, Level 15, State 1, Line 8
Incorrect syntax near the keyword 'ELSE'. */
/* good */
IF @y > @x
BEGIN
SET @x = @x + 1;
SET @y = @y - 1;
END;
ELSE
BEGIN
PRINT(1);
END;
FROM, WHERE, INTO, JOIN, GROUP BY, ORDER BY
expressions should be aligned so, that all their arguments are placed under each other (see Example below)
TSQL Example with formating:
WITH CTE_MyCTE AS (
SELECT
t1.Value1 AS Val1
, t1.Value2 AS Val2
, t2.Value3 AS Val3
INNER JOIN dbo.Table3 AS t2
ON t1.Value1 = t2.Value1
WHERE t1.Value1 > 1
AND t2.Value2 >= 101
)
SELECT
t1.Value1 AS Val1
, t1.Value2 AS Val2
, t2.Value3 AS Val3
INTO #Table3
FROM CTE_MyCTE AS t1
ORDER BY t2.Value2;
Recommendations from Microsoft: Stored procedure Best practice
ALTER
statement and start with the object presence check (see example below) for saving GRANTs
on your object.
Also if you use Query Store and plan forcing and DROP
and then CREATE
new object you loosing plan forcing, more details here.
For SQL Server 2016 and higher you can use new CREATE OR ALTER
statement.ALTER
statement should be preceded by 2 line breaksALTER
statement and before AS
keyword should be placed a comment with execution examplesp_
procedures only in master
database - SQL Server will always scan through the system catalog firstBEGIN TRY
and BEGIN CATCH
for error handlingSET NOCOUNT ON;
for stops the message that shows the count of the number of rows affected by a Transact-SQL statement and decreasing network traffic.
More details here and here.SET NOCOUNT OFF;
because it is default behaviorRAISERROR
instead PRINT
if you want to give feedback about the state of the currently executing SQL batch without lags.
More details here and here.<documentation>
<summary>Get all databases meta data using dynamic T-SQL</summary>
<returns>1 data set: temp table #DatabaseInfo.</returns>
<issues>No</issues>
<author>Konstantin Taranov</author>
<created>2018-03-01</created>
<modified>2019-11-14 by Konstantin Taranov</modified>
<version>1.2</version>
<sourceLink>https://github.com/ktaranov/sqlserver-kit/blob/master/Scripts/Databases_Report.sql</sourceLink>
</documentation>
Stored Procedure Example:
IF OBJECT_ID('dbo.usp_StoredProcedure', 'P') IS NULL
EXECUTE('CREATE PROCEDURE dbo.usp_StoredProcedure as SELECT 1');
GO
ALTER PROCEDURE dbo.usp_StoredProcedure(
@parameterValue1 smallint
, @parameterValue2 nvarchar(300)
, @debug bit = 0
)
/*
<documentation>
<summary>Simple example of tsql procedure</summary>
<returns>nothing</returns>
<issues>No</issues>
<author>Konstantin Taranov</author>
<created>2019-01-01</created>
<modified>2019-11-25 by Konstantin Taranov</modified>
<version>1.2</version>
<sourceLink>-</sourceLink>
<example1>EXECUTE dbo.usp_StoredProcedure
@parameterValue1 = 0
, @parameterValue2 = N'BULK'</example1>
</documentation>
*/
AS
SET NOCOUNT ON;
BEGIN TRY
IF (@parameterValue1 < 0 OR @parameterValue2 NOT IN ('SIMPLE', 'BULK', 'FULL'))
RAISERROR('Not valid data parameter!', 16, 1);
IF (@debug) PRINT @parameterValue2;
END TRY
BEGIN CATCH
/* Print error information. */
PRINT 'Error: ' + CAST(ERROR_NUMBER()) AS varchar(50)) +
', Severity: ' + CAST(ERROR_SEVERITY(), varchar(5)) +
', State: ' + CAST(ERROR_STATE(), varchar(5) ) +
', Procedure: ' + COALESCE(ERROR_PROCEDURE(), '-') +
', Line: ' + CAST(ERROR_LINE(), varchar(5)) +
', User name: ' + CAST(ORIGINAL_LOGIN(), sysname);
PRINT ERROR_MESSAGE();
END CATCH;
GO
Highly recommended to read awesome detailed article about dynamic T-SQL by Erland Sommarskog: The Curse and Blessings of Dynamic SQL
Dynamic SQL is a programming technique that allows you to construct SQL statements dynamically at runtime. It allows you to create more general purpose and flexible SQL statement because the full text of the SQL statements may be unknown at compilation. For example, you can use the dynamic SQL to create a stored procedure that queries data against a table whose name is not known until runtime.
More details here.
Do not use nvarchar(max) for your object’s name parameter, use sysname instead (synonym for nvarchar(128) except that, by default, sysname is NOT NULL).
/* Bad */
DECLARE @tableName nvarchar(max) = N'MyTableName';
/* Good */
DECLARE @tableName sysname = N'MyTableName';
Do quote the names of your objects properly.
/* Bad */
DECLARE @tsql nvarchar(max);
DECLARE @tableName sysname = N'My badly named table!';
SET @tsql = N'SELECT object_id FROM ' + @tableName;
/* Good */
DECLARE @tsql nvarchar(max);
DECLARE @tableName sysname = N'My badly named table 111!';
SET @tsql = N'SELECT object_id FROM ' + QUOTENAME(@tableName);
Always use
sp_executesql
instead
EXEC
to prevent sql injection.
Also
sp_executesql
can parameterizing your dynamic statement that means plans can be reused as well (when the value of the dynamic object is the same).
Also
sp_executesql
can even be used to output values as well (see example below).
/* Bad EXEC example with sql injection*/
DECLARE @tsql nvarchar(max);
DECLARE @tableName sysname = N'master.sys.tables; SELECT * FROM master.sys.server_principals;';
SET @tsql = N'SELECT "name" FROM ' + @tableName + N';';
EXEC (@tsql);
/* Good sp_executesql example*/
DECLARE @tsql nvarchar(max);
DECLARE @tableName sysname = N'master.sys.tables';
DECLARE @id int = 2107154552;
SET @tsql = N'SELECT name FROM ' + @tableName +
N' WHERE object_id = ' + CONVERT(nvarchar(max), @id);
EXEC sp_executesql @tsql, N'@ID int', @ID = @id;
/* Good sp_executesql example with OUTPUT */
DECLARE @tsql nvarchar(max);
DECLARE @tableName sysname = N'master.sys.tables';
DECLARE @count bigint;
SET @tsql = N'SELECT @countOUT = COUNT(*) FROM ' + @tableName + N';';
EXEC sp_executesql @tsql, N'@countOUT bigint OUTPUT', @countOUT = @count OUTPUT;
PRINT('@count = ' + CASE WHEN @count IS NULL THEN 'NULL' ELSE CAST(@count AS varchar(30)) END);
Do not use dynamic T-SQL if your statement is not dynamic.
/* Bad */
DECLARE @tsql nvarchar(max);
DECLARE @id int = 2107154552;
SET @tsql = N'SELECT object_id, "name" FROM master.sys.tables WHERE object_id = ' + CAST(@id AS nvarchar(max));
EXEC sp_executesql @tsql;
/* Good */
DECLARE @id int = 2107154552;
SELECT object_id, "name" FROM master.sys.tables WHERE object_id = @id;
@debug
variable to print (or a SELECT
statement if your dynamic T-SQL is over 4000 characters) dynamic statement instead executing it.
See example below.Do take the time to format your dynamic T-SQL.
/* Bad @tsql formating */
DECLARE @tsql nvarchar(max);
DECLARE @sep nvarchar(30) = ' UNION ALL ';
DECLARE @debug bit = 1;
SELECT @tsql = COALESCE(@tsql, N'') +
N'SELECT N' + QUOTENAME(name,'''') +
N' AS DBName, (SELECT COUNT(*) FROM ' +
QUOTENAME(name) + N'.sys.tables) AS TableCount' +
@sep
FROM sys.databases
ORDER BY name;
SET @tsql = LEFT(@tsql, LEN(@tsql) - LEN(@sep));
IF @debug = 1 SELECT @tsql AS "tsql" ELSE EXEC sp_executesql @tsql;
/* Good @tsql formating */
DECLARE @tsql nvarchar(max);
DECLARE @sep nvarchar(30) = ' UNION ALL ';
DECLARE @debug bit = 1;
DECLARE @crlf nvarchar(10) = NCHAR(13) + NCHAR(10);
SELECT @tsql = COALESCE(@tsql, N'') + @crlf +
N'SELECT N' + QUOTENAME(name,'''') + N' AS DBName' + @crlf +
N' , (SELECT COUNT(*) FROM ' + QUOTENAME(name) + N'.sys.tables) AS TableCount' + @crlf +
@sep
FROM sys.databases
ORDER BY name;
SET @tsql = LEFT(@tsql, LEN(@tsql) - LEN(@sep)) + N';';
IF @debug = 1 SELECT @tsql AS "tsql" ELSE EXEC sp_executesql @tsql;