Posts tagged as:

T-SQL

SQL Like Statement (or) Condition is basically used for checking if a particular charter string appears in a given string or not. We can use wildcard symbols in the where clause and also in the CASE Statements for pattern matching. Because of this, the SQL Like operator is more powerful than using the equal to (=) or not equal to (! =) symbols for pattern matching and more generically for string comparison. The basic point that we need to keep in mind while using the SQL Like operator is; the result set returned by it is a Boolean value, meaning a TRUE or FALSE.

SQL Like Syntax

The basic syntax for SQL Like is shown below.

match_expression [ NOT ] LIKE pattern [ ESCAPE escape_character ]

SQL Like – Wildcards

The following are the wildcards that can be used with the SQL like condition – %( Percentage symbol), _ (Underscore), [] (Square Parenthesis), ^ (Caret symbol); will show each of them with examples…

Firstly, let us create a temporary table and insert some data into it and see how we can use the SQL Like operator.

CREATE TABLE #SQL_Like (NameID int, ContactName nvarchar(50))

 INSERT INTO #SQL_Like  Values (1,'Amy')
 INSERT INTO #SQL_Like  Values (2,'Anna')
 INSERT INTO #SQL_Like  Values (3,'Brenda')
 INSERT INTO #SQL_Like  Values (4,'Catherine')
 INSERT INTO #SQL_Like  Values (5,'Daniel')
 INSERT INTO #SQL_Like  Values (6,'Emilio')
 INSERT INTO #SQL_Like  Values (7,'Edna')
 INSERT INTO #SQL_Like  Values (8,'Frances')
 INSERT INTO #SQL_Like  Values (9,'Gustavo')
 INSERT INTO #SQL_Like  Values (10,'Hannah')
 INSERT INTO #SQL_Like  Values (11,'Ian')
 INSERT INTO #SQL_Like  Values (12,'John')

Examples for using SQL Like Condition

The most basic form to start with, suppose from the above table if we want to get all the names starting with “A” we can use the below query.

SELECT * FROM #SQL_Like  WHERE ContactName Like 'A%'

The “%” percentage wild card is used for doing a pattern match of any string of zero or more characters. And this is the result that we would get once we are done with executing the above line of T-SQL.

NameID ContactName

1          Amy

2          Anna

If we want to get a list of names that end with “A”, we can use the same wild card, but we will have to put it in the front and the letter that we want to search for at the end, shown like the below.

SELECT * FROM #SQL_Like  WHERE ContactName Like '%A'

And the result for the same would be

NameID ContactName

2          Anna

3          Brenda

7          Edna

Suppose, if there is a requirement where we know that we need to look into the names, in whose name the 2nd letter is “A”, then we have to write a query like the below, the “_” underscore wild card symbol denotes a single character.

SELECT * FROM #SQL_Like  WHERE ContactName Like '_A%'

And the result for the same would be

NameID ContactName

4          Catherine

5          Daniel

10         Hannah

11         Ian

Say, if we want to search for a name, which starts with “A” and in which the second alphabet is something between “A” and “M”, then our result has to be the first name that we had inserted, i.e “Amy”. For this we actually use the “[ ]” square brackets, they are used for searching any single character within the specified range ([a-g]) or set ([abcdefg]).

SELECT * FROM #SQL_Like  WHERE ContactName Like A[A-M]%'

Using SQL NOT Like

Till now we have seen how to directly use the “Like operator”. Now we shall take up an example where we will add “Not” in front of our “Like”. For this, the example that we can take is, say we need a list of all the name’s that do not have the alphabet “A” in them, then our query would resemble the below.

SELECT * FROM #SQL_Like  WHERE ContactName NOT Like '%A%'

Using SQL Like Escape Clause

Everything seems to be fine till now, so, let’s think like this, the wild card symbols are really cool, but what if I need to look into a string for those wild card character’s, puzzled!!!

For this we have these escape clause J. Check the syntax at the beginning of the article
In the below example, my escape character is “!”, so I am actually trying to search for John%, of course practically no names will exist in this way unless there is a type!!! This is just for the sake of taking an example and looking at on how to use it

SELECT * FROM #SQL_Like  WHERE ContactName LIKE '%John!%%' ESCAPE '!'

Let’s take one more example, here we want to look for the underscore symbol, and that’s a wildcard symbol, so decide to make use of the Escape clause over here as well.

SELECT * FROM #SQL_Like  WHERE ContactName LIKE '%John!_%' ESCAPE '!' 

The basic thing over here is that we need to look for a symbol that can be used in the escape clause, means something, that does not exist in our string.

Using SQL Like with Variables and Expressions
We can use the SQL LIKE operator with Variables and Expressions, they can combination of alphabets, alphabets and alpha-numeric, a query with scalar output, a varchar or nvarchar type local variable. But, remember one thing, at the end, everything would be looked at as a string(Not char) by the sql server.

The below is a self-explanatory code snippet. Firstly, we are declaring a variable, then we are setting it to a string “la” and then we are searching for it in the names. In the first, example we are trying to search for characters in the start of the name and in the second example we are trying to search for the characters in the end of the name

DECLARE @Name nvarchar(2)
 SET @Name ='Ia'
 SELECT * FROM #SQL_Like  WHERE ContactName Like @Name +'%'

DECLARE @Name nvarchar(2)
 SET @Name ='N'
 SELECT * FROM #SQL_Like  WHERE ContactName Like '%'+ @Name

Using SQL Like with Case Statement
As stated earlier, the sql like condition returns a true or false, so we can use this in multiple scenarios in conjunction with a CASE statement. Here in the below code snippet we are actually trying to find all the names that start with A, B. We can also construct similar thing where the condition can equal to 0.

SELECT * FROM #SQL_Like
 WHERE CASE
 WHEN ContactName LIKE 'A%' THEN 1
 WHEN ContactName LIKE 'B%' THEN 1
 END = 1

{ Comments on this entry are closed }

People coming from other programming languages to any SQL based language will know how GO TO works in general, even with SQL Server, it is nearly the same.

Go To statement makes the execution of the T-SQL batch to jump and go to a “pre-defined” label. The term pre-defined is important over here, go to statement abruptly makes the code to jump to a label without executing any lines after it, unless the label is next to the Go To line….

We can embed Go to in any of the conditional control-of-flow statements, statement blocks, or procedures but, in practice it is not practical to make much use of this statement as too much of it would not make to code readable and understandable by humans!!! But one thing is, if we have no other choice other than to use this statement then we have to just go ahead using it.

The below is an example from BOL on how to use the GOTO statement, it shows us on how it can be used and how the calls can be made using it, not that the second branch would never be execute because of the condition that is there.

DECLARE @Counter int;

SET @Counter = 1;

WHILE @Counter < 10

BEGIN

SELECT @Counter

SET @Counter = @Counter + 1

IF @Counter = 4 GOTO Branch_One --Jumps to the first branch.

IF @Counter = 5 GOTO Branch_Two  --This will never execute.

END

Branch_One:

SELECT 'Jumping To Branch One.'

GOTO Branch_Three; --This will prevent Branch_Two from executing.

Branch_Two:

SELECT 'Jumping To Branch Two.'

Branch_Three:

SELECT 'Jumping To Branch Three.'

And the below is with a small alter to the above code just showing that we can have a GO TO call/ the Label’s anywhere, there is no precedence.


DECLARE @Counter int;

SET @Counter = 1;

Branch_One:

SELECT 'I am above the GOTO Call'

GOTO Branch_Three; --This will prevent Branch_Two from executing.

WHILE @Counter < 10

BEGIN

SELECT @Counter

SET @Counter = @Counter + 1

IF @Counter = 4 GOTO Branch_One --Jumps to the first branch.

IF @Counter = 5 GOTO Branch_Two  --This will never execute.

END

Branch_Two:

SELECT 'Jumping To Branch Two.'

Branch_Three:

SELECT 'I am below the GOTO Call'

{ Comments on this entry are closed }

I have a table where I do logging for all the SQL Server Jobs, I had a requirement where I had to calculate the idle time between two jobs to find the load on the server and better utilise the idle time of the server for other maintenance related operations, for a Job I track many parameters, but the once that are interested for me now for this query are the Job Start Time and the Job End Time, . That is I had to calculate time difference between the last job end time to the current job start time.

 For example, in the below table, I need to know the time difference between the Job End Time for Job ID 1 and Job Start Time of Job ID 2, that’s 120 minutes.

 JobID   JobStartTime                            JobEndTime

1          2011-01-07 02:55:00.000        2011-01-07 03:55:00.000

2          2011-01-07 05:55:00.000        2011-01-07 06:55:00.000

3          2011-01-07 08:55:00.000        2011-01-07 09:55:00.000

 Let me create a temporary table and show how I did it.

CREATE TABLE #JobLog(

JobID int identity(1,1), JobStartTime DATETIME,JobEndTime DATETIME)

INSERT INTO #JobLog(JobStartTime,JobEndTime)

VALUES('2011-01-07 02:55:00.000','2011-01-07 03:55:00.000')

INSERT INTO #JobLog(JobStartTime,JobEndTime)

VALUES('2011-01-07 05:55:00.000','2011-01-07 06:55:00.000')

INSERT INTO #JobLog(JobStartTime,JobEndTime)

VALUES('2011-01-07 08:55:00.000','2011-01-07 09:55:00.000')

 

Checking for table and it’s records.

SELECT * FROM #JobLog

Time difference between Last Job End Time and Current Job Start Time

WITH CTE as(

SELECT JobID, JobStartTime, JobEndTime, (

    SELECT MAX(T1.JobEndTime)

    FROM #JobLog T1

    WHERE T1.JobEndTime < T2.JobStartTime

  ) AS Previous_JobEndTime

  FROM #JobLog T2

 )

SELECT JobID, JobStartTime, JobEndTime

      , DATEDIFF(MI,Previous_JobEndTime,JobStartTime) AS IdleTime

FROM CTE

 

 And the below is the resuly.

JobID   JobStartTime                            JobEndTime                             IdleTime

1          2011-01-07 02:55:00.000        2011-01-07 03:55:00.000        NULL

2          2011-01-07 05:55:00.000        2011-01-07 06:55:00.000        120

3          2011-01-07 08:55:00.000        2011-01-07 09:55:00.000        120

dropping the temporary table

DROP TABLE #JobLog

{ Comments on this entry are closed }

These are a set of queries that I keep using frequently in the stored procedures that I wrote, they can basically be used for finding the date time of

Start and End of Today

SELECT DATEADD(DD,DATEDIFF(DD,0, GETDATE()),0) as [Start Of Day]
	  ,DATEADD(s,-1,DATEADD(dd, DATEDIFF(dd,0,GETDATE())+1,0)) as [End Of Day]

Start and End of Week

SELECT DATEADD(WK, DATEDIFF(WK,0,GETDATE()),0) as [Start Of Week]
	  ,DATEADD(s,-1,DATEADD(WK, DATEDIFF(WK,0,GETDATE())+1,0)) as [End Of Week]

Start and End of Month

SELECT DATEADD(M, DATEDIFF(M,0,GETDATE()),0) as [Start Of Month]
	  ,DATEADD(s,-1,DATEADD(mm, DATEDIFF(m,0,GETDATE())+1,0)) as [End Of Month]

Start and End of Next Month

SELECT DATEADD(DD,DATEDIFF(DD,0,DATEADD(dd,-(DAY(DATEADD(mm,1,GETDATE()))-1),DATEADD(mm,1,GETDATE()))),0) as [Start of Next Month]
	  ,DATEADD(s,-1,DATEADD(mm, DATEDIFF(m,0,GETDATE())+2,0)) as [End Of Next Month]

Start and End of Quarter

SELECT DATEADD(QQ, DATEDIFF(QQ,0,GETDATE()),0) as [Start Of Qaurter]
	  ,DATEADD(s,-1,DATEADD(QQ, DATEDIFF(QQ,0,GETDATE())+1,0)) as [End Of Quarter]

Start and End of Year

SELECT DATEADD(YY, DATEDIFF(YY,0,GETDATE()),0) as [Start Of Year]
	  ,DATEADD(s,-1,DATEADD(YY, DATEDIFF(YY,0,GETDATE())+1,0)) as [End Of Year]

Please let me know if there are any other suitable better performing methods for accomplishing the above in comments.

{ Comments on this entry are closed }

There are many instances where I need to convert a set of rows in a single column to a comma separated string, and this is how I accomplish it, going with an example, creating a single column table with some string values in it…

DECLARE @MakeStr VARCHAR(200)
CREATE TABLE #TEMPTABLE1(COL1 NVARCHAR(10))

INSERT INTO #TEMPTABLE1 VALUES('ABC')
INSERT INTO #TEMPTABLE1 VALUES('BCD')
INSERT INTO #TEMPTABLE1 VALUES('CDE')
INSERT INTO #TEMPTABLE1 VALUES('EFG')
INSERT INTO #TEMPTABLE1 VALUES('GHI')

Create Comma Seperated String Using COALESCE Method

SELECT @MakeStr = COALESCE(@MakeStr+',' ,'') + COL1
FROM #TEMPTABLE1

SELECT @MakeStr as SingleString

Create Comma Seperated String Using XML Method

SELECT SUBSTRING(
(SELECT ',' + COL1
FROM #TEMPTABLE1
ORDER BY COL1
FOR XML PATH('')),2,200) AS SingleString

And the result for the above would be

SingleString
——————–
ABC,BCD,CDE,EFG,GHI

Dropping the temporary table that we have create.

DROP TABLE #TEMPTABLE1

Please let me know if there are any other suitable better performing methods for accomplishing the above in comments.

{ Comments on this entry are closed }