If you are active in the SQL Server newsgroups and forums as I am,
you will notice that the same questions keep popping up all the time. I
picked ten of them which I see daily.
1) Selecting all the values from a table for a particular date
2) Search all columns in all the tables in a database for a specific value
3) Splitting string values
4) Select all rows from one table that don’t exist in another table
5) Getting all rows from one table and only the latest from the child table
6) Getting all characters until a specific character
7) Return all rows with NULL values in a column
8) Row values to column (PIVOT)
9) Pad or remove leading zeroes from numbers
10) Concatenate Values From Multiple Rows Into One Column
1 Selecting all the values from a table for a particular date
This is a very popular question and people sometimes answer that you
need to use between. There is a problem with between if you happen to
have a value at exactly midnight. Let us take a look, first create this
table.
| T-SQL |
1
| CREATE TABLE SomeDates (DateColumn DATETIME)
|
Insert 2 values
| T-SQL |
1
2
| INSERT INTO SomeDates VALUES('2008-10-02 00:00:00.000')
INSERT INTO SomeDates VALUES('2008-10-01 00:00:00.000')
|
Return everything between ‘2008-10-01’ and ‘2008-10-02’
| T-SQL |
1
2
3
4
| SELECT *
FROM SomeDates
WHERE DateColumn BETWEEN '20081001' AND '20081002'
ORDER BY DateColumn
|
This works without a problem, we get this returned
(results)
2008-10-01 00:00:00.000
2008-10-02 00:00:00.000
Let’s add some more dates including the time portion
| T-SQL |
1
2
3
4
5
6
| INSERT INTO SomeDates VALUES('2008-10-02 00:01:00.000')
INSERT INTO SomeDates VALUES('2008-10-02 00:00:59.000')
INSERT INTO SomeDates VALUES('2008-10-02 00:00:01.000')
INSERT INTO SomeDates VALUES('2008-10-01 00:01:00.000')
INSERT INTO SomeDates VALUES('2008-10-01 00:12:00.000')
INSERT INTO SomeDates VALUES('2008-10-01 23:00:00.000')
|
Return everything between ‘2008-10-01’ and ‘2008-10-02’
| T-SQL |
1
2
3
4
| SELECT *
FROM SomeDates
WHERE DateColumn BETWEEN '20081001' AND '20081002'
ORDER BY DateColumn
|
(results)
2008-10-01 00:00:00.000
2008-10-01 00:01:00.000
2008-10-01 00:12:00.000
2008-10-01 23:00:00.000
2008-10-02 00:00:00.000
Here is where it goes wrong; for 2008-10-02 only the midnight value is returned the other ones are ignored
Now if we change 2008-10-02 to 2008-10-03 we get what we want
| T-SQL |
1
2
3
4
| SELECT *
FROM SomeDates
WHERE DateColumn BETWEEN '20081001' AND '20081003'
ORDER BY DateColumn
|
(results)
2008-10-01 00:00:00.000
2008-10-01 00:01:00.000
2008-10-01 00:12:00.000
2008-10-01 23:00:00.000
2008-10-02 00:00:00.000
2008-10-02 00:00:01.000
2008-10-02 00:00:59.000
2008-10-02 00:01:00.000
Now insert a value for 2008-10-03 (midnight)
| T-SQL |
1
| INSERT INTO SomeDates VALUES('2008-10-03 00:00:00.000')
|
Run the query again
| T-SQL |
1
2
3
4
| SELECT *
FROM SomeDates
WHERE DateColumn BETWEEN '20081001' AND '20081003'
ORDER BY DateColumn
|
(results)
2008-10-01 00:00:00.000
2008-10-01 00:01:00.000
2008-10-01 00:12:00.000
2008-10-01 23:00:00.000
2008-10-02 00:00:00.000
2008-10-02 00:00:01.000
2008-10-02 00:00:59.000
2008-10-02 00:01:00.000
2008-10-03 00:00:00.000
We get back 2008-10-03 00:00:00.000, between will return the date if it is exactly midnight
If you use >= and < then you get exactly what you need
| T-SQL |
1
2
3
4
| SELECT *
FROM SomeDates
WHERE DateColumn >= '20081001' AND DateColumn < '20081003'
ORDER BY DateColumn
|
(results)
2008-10-01 00:00:00.000
2008-10-01 00:01:00.000
2008-10-01 00:12:00.000
2008-10-01 23:00:00.000
2008-10-02 00:00:00.000
2008-10-02 00:00:01.000
2008-10-02 00:00:59.000
2008-10-02 00:01:00.000
So be careful when using between because you might get back rows that
you did not expect to get back and it might mess up your reporting if
you do counts or sums
Bonus: how to strip the time off a date?
To strip the time portion of a datetime, you can do this and still return a datetime
| T-SQL |
1
2
| Select * ,DATEADD(dd, DATEDIFF(dd, 0, DateColumn), 0) as stripped
from SomeDates
|
you can also do a convert to varchar with a style value of 112 and then converting back to datetime
| T-SQL |
1
2
| Select * ,convert(datetime,Convert(varchar(8),DateColumn,112)) as stripped
from SomeDates
|
Both methods return this
———————————————–
DateColumn stripped
2008-10-02 00:00:00.000 2008-10-02 00:00:00.000
2008-10-01 00:00:00.000 2008-10-01 00:00:00.000
2008-10-02 00:01:00.000 2008-10-02 00:00:00.000
2008-10-02 00:00:59.000 2008-10-02 00:00:00.000
2008-10-02 00:00:01.000 2008-10-02 00:00:00.000
2008-10-01 00:01:00.000 2008-10-01 00:00:00.000
2008-10-01 00:12:00.000 2008-10-01 00:00:00.000
2008-10-01 23:00:00.000 2008-10-01 00:00:00.000
2008-10-03 00:00:00.000 2008-10-03 00:00:00.000
2 Search all columns in all the tables in a database for a specific value
How can I search all the columns in every table in my database for a specific value?
This question keeps popping up all the time. Here is a variation of this question discussed
MSDN Forum T-SQL. Unfortunately there is no easy way to do this, you have to loop over all the tables in the database and search for the value.
As you can imagine if you have a huge database with a lot of tables this can take a really long time.
There are a couple of checks going on in this proc, if the data is not
numeric then we skip all the numeric columns and do not search those
columns. This code was created by our own gmmastros and what he did was
create 3 different stored procedures, one for dates, numbers and
strings. Then he created a 4th stored proc named FindMyData, this proc
will only call the appropriate child procs if the datatype is right.
We only list the table name and the column name where the value is stored, we do not return that data!
First create the following 3 stored procedures
Just the date searching proc
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
| CREATE PROCEDURE FindMyData_Date
@DataToFind DATETIME
AS
SET NOCOUNT ON
DECLARE @Temp TABLE(RowId INT IDENTITY(1,1), SchemaName sysname, TableName sysname, ColumnName SysName, DataType VARCHAR(100), DataFound BIT)
DECLARE @ISDATE BIT
INSERT INTO @Temp(TableName,SchemaName, ColumnName, DataType)
SELECT C.Table_Name,C.TABLE_SCHEMA, C.Column_Name, C.Data_Type
FROM Information_Schema.Columns AS C
INNER Join Information_Schema.Tables AS T
ON C.Table_Name = T.Table_Name
AND C.TABLE_SCHEMA = T.TABLE_SCHEMA
WHERE Table_Type = 'Base Table'
And (Data_Type = 'DateTime'
Or (Data_Type = 'SmallDateTime' And @DataToFind >= '19000101' And @DataToFind < '20790607'))
DECLARE @i INT
DECLARE @MAX INT
DECLARE @TableName sysname
DECLARE @ColumnName sysname
DECLARE @SchemaName sysname
DECLARE @SQL NVARCHAR(4000)
DECLARE @PARAMETERS NVARCHAR(4000)
DECLARE @DataExists BIT
DECLARE @SQLTemplate NVARCHAR(4000)
SELECT @SQLTemplate = 'If Exists(Select *
From ReplaceTableName
Where [ReplaceColumnName]
= ''' + CONVERT(VARCHAR(30), @DataToFind, 126) + '''
)
Set @DataExists = 1
Else
Set @DataExists = 0',
@PARAMETERS = '@DataExists Bit OUTPUT',
@i = 1
SELECT @i = 1, @MAX = MAX(RowId)
FROM @Temp
WHILE @i <= @MAX
BEGIN
SELECT @SQL = REPLACE(REPLACE(@SQLTemplate, 'ReplaceTableName', QUOTENAME(SchemaName) + '.' + QUOTENAME(TableName)), 'ReplaceColumnName', ColumnName)
FROM @Temp
WHERE RowId = @i
PRINT @SQL
EXEC SP_EXECUTESQL @SQL, @PARAMETERS, @DataExists = @DataExists OUTPUT
IF @DataExists =1
UPDATE @Temp SET DataFound = 1 WHERE RowId = @i
SET @i = @i + 1
END
SELECT SchemaName,TableName, ColumnName
FROM @Temp
WHERE DataFound = 1
go
|
If you want to test just this proc, try this
| T-SQL |
1
| exec FindMyData_Date '20070615'
|
This is the string proc
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
| CREATE PROCEDURE FindMyData_String
@DataToFind NVARCHAR(4000),
@ExactMatch BIT = 0
AS
SET NOCOUNT ON
DECLARE @Temp TABLE(RowId INT IDENTITY(1,1), SchemaName sysname, TableName sysname, ColumnName SysName, DataType VARCHAR(100), DataFound BIT)
INSERT INTO @Temp(TableName,SchemaName, ColumnName, DataType)
SELECT C.Table_Name,C.TABLE_SCHEMA, C.Column_Name, C.Data_Type
FROM Information_Schema.Columns AS C
INNER Join Information_Schema.Tables AS T
ON C.Table_Name = T.Table_Name
AND C.TABLE_SCHEMA = T.TABLE_SCHEMA
WHERE Table_Type = 'Base Table'
And Data_Type In ('ntext','text','nvarchar','nchar','varchar','char')
DECLARE @i INT
DECLARE @MAX INT
DECLARE @TableName sysname
DECLARE @ColumnName sysname
DECLARE @SchemaName sysname
DECLARE @SQL NVARCHAR(4000)
DECLARE @PARAMETERS NVARCHAR(4000)
DECLARE @DataExists BIT
DECLARE @SQLTemplate NVARCHAR(4000)
SELECT @SQLTemplate = CASE WHEN @ExactMatch = 1
THEN 'If Exists(Select *
From ReplaceTableName
Where Convert(nVarChar(4000), [ReplaceColumnName])
= ''' + @DataToFind + '''
)
Set @DataExists = 1
Else
Set @DataExists = 0'
ELSE 'If Exists(Select *
From ReplaceTableName
Where Convert(nVarChar(4000), [ReplaceColumnName])
Like ''%' + @DataToFind + '%''
)
Set @DataExists = 1
Else
Set @DataExists = 0'
END,
@PARAMETERS = '@DataExists Bit OUTPUT',
@i = 1
SELECT @i = 1, @MAX = MAX(RowId)
FROM @Temp
WHILE @i <= @MAX
BEGIN
SELECT @SQL = REPLACE(REPLACE(@SQLTemplate, 'ReplaceTableName', QUOTENAME(SchemaName) + '.' + QUOTENAME(TableName)), 'ReplaceColumnName', ColumnName)
FROM @Temp
WHERE RowId = @i
PRINT @SQL
EXEC SP_EXECUTESQL @SQL, @PARAMETERS, @DataExists = @DataExists OUTPUT
IF @DataExists =1
UPDATE @Temp SET DataFound = 1 WHERE RowId = @i
SET @i = @i + 1
END
SELECT SchemaName,TableName, ColumnName
FROM @Temp
WHERE DataFound = 1
GO
|
If you want to test just this proc, try this
| T-SQL |
1
| exec FindMyData_string 'google', 0
|
Just the number proc
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
| CREATE PROCEDURE FindMyData_Number
@DataToFind NVARCHAR(4000),
@ExactMatch BIT = 0
AS
SET NOCOUNT ON
DECLARE @Temp TABLE(RowId INT IDENTITY(1,1), SchemaName sysname, TableName sysname, ColumnName SysName, DataType VARCHAR(100), DataFound BIT)
DECLARE @IsNumber BIT
DECLARE @ISDATE BIT
IF ISNUMERIC(CONVERT(VARCHAR(20), @DataToFind)) = 1
SET @IsNumber = 1
ELSE
SET @IsNumber = 0
INSERT INTO @Temp(TableName,SchemaName, ColumnName, DataType)
SELECT C.Table_Name,C.TABLE_SCHEMA, C.Column_Name, C.Data_Type
FROM Information_Schema.Columns AS C
INNER Join Information_Schema.Tables AS T
ON C.Table_Name = T.Table_Name
AND C.TABLE_SCHEMA = T.TABLE_SCHEMA
WHERE Table_Type = 'Base Table'
And Data_Type In ('float','real','decimal','money','smallmoney','bigint','int','smallint','tinyint','bit')
DECLARE @i INT
DECLARE @MAX INT
DECLARE @TableName sysname
DECLARE @ColumnName sysname
DECLARE @SQL NVARCHAR(4000)
DECLARE @PARAMETERS NVARCHAR(4000)
DECLARE @DataExists BIT
DECLARE @SQLTemplate NVARCHAR(4000)
SELECT @SQLTemplate = CASE WHEN @ExactMatch = 1
THEN 'If Exists(Select *
From ReplaceTableName
Where Convert(VarChar(40), [ReplaceColumnName])
= ''' + @DataToFind + '''
)
Set @DataExists = 1
Else
Set @DataExists = 0'
ELSE 'If Exists(Select *
From ReplaceTableName
Where Convert(VarChar(40), [ReplaceColumnName])
Like ''%' + @DataToFind + '%''
)
Set @DataExists = 1
Else
Set @DataExists = 0'
END,
@PARAMETERS = '@DataExists Bit OUTPUT',
@i = 1
SELECT @i = 1, @MAX = MAX(RowId)
FROM @Temp
WHILE @i <= @MAX
BEGIN
SELECT @SQL = REPLACE(REPLACE(@SQLTemplate, 'ReplaceTableName', QUOTENAME(SchemaName) + '.' + QUOTENAME(TableName)), 'ReplaceColumnName', ColumnName)
FROM @Temp
WHERE RowId = @i
PRINT @SQL
EXEC SP_EXECUTESQL @SQL, @PARAMETERS, @DataExists = @DataExists OUTPUT
IF @DataExists =1
UPDATE @Temp SET DataFound = 1 WHERE RowId = @i
SET @i = @i + 1
END
SELECT SchemaName,TableName, ColumnName
FROM @Temp
WHERE DataFound = 1
go
|
If you want to test just this proc, try this
| T-SQL |
1
| exec FindMyData_Number '562', 1
|
The mother of all procs!
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
| CREATE PROCEDURE FindMyData
@DataToFind NVARCHAR(4000),
@ExactMatch BIT = 0
AS
SET NOCOUNT ON
CREATE TABLE #Output(SchemaName sysname, TableName sysname, ColumnName sysname)
IF ISDATE(@DataToFind) = 1
INSERT INTO #Output EXEC FindMyData_Date @DataToFind
IF ISNUMERIC(@DataToFind) = 1
INSERT INTO #Output EXEC FindMyData_Number @DataToFind, @Exactmatch
INSERT INTO #Output EXEC FindMyData_String @DataToFind, @ExactMatch
SELECT SchemaName,TableName, ColumnName
FROM #Output
ORDER BY SchemaName,TableName, ColumnName
go
|
Here are some proc calls to test it out
| T-SQL |
1
2
3
| exec FindMyData 'google', 0
exec FindMyData 1, 0
exec FindMyData '20081201', 0
|
| T-SQL |
1
| exec FindMyData 'sysobjects', 0
|
3 Splitting string values
The fastest way to split a comma delimited string is by using a
number table. If you do not have a number table in your database then
use this code to create one
| T-SQL |
1
2
3
4
5
6
7
8
| -- Create our Pivot table ** do this only once
CREATE TABLE NumberPivot (NumberID INT PRIMARY KEY)
INSERT INTO NumberPivot
SELECT number FROM master..spt_values
where type = 'P'
GO
|
Now you can run the following code to return each of the values in the comma delimited string in its own row
| T-SQL |
1
2
3
4
5
6
7
8
9
| DECLARE @SplitString VARCHAR(1000)
SELECT @SplitString ='1,4,77,88,4546,234,2,3,54,87,9,6,4,36,6,9,9,6,4,4,68,9,0,5,3,2,'
SELECT SUBSTRING(',' + @SplitString + ',', NumberID + 1,
CHARINDEX(',', ',' + @SplitString + ',', NumberID + 1) - NumberID -1)AS VALUE
FROM NumberPivot
WHERE NumberID <= LEN(',' + @SplitString + ',') - 1
AND SUBSTRING(',' + @SplitString + ',', NumberID, 1) = ','
GO
|
You can also return distinct values by using DISTINCT
| T-SQL |
1
2
3
4
5
6
7
8
| DECLARE @SplitString VARCHAR(1000)
SELECT @SplitString ='1,4,77,88,4546,234,2,3,54,87,9,6,4,36,6,9,9,6,4,4,68,9,0,5,3,2'
SELECT DISTINCT SUBSTRING(',' + @SplitString + ',', NumberID + 1,
CHARINDEX(',', ',' + @SplitString + ',', NumberID + 1) - NumberID -1)AS VALUE
FROM NumberPivot
WHERE NumberID <= LEN(',' + @SplitString + ',') - 1
AND SUBSTRING(',' + @SplitString + ',', NumberID, 1) = ','
|
You now can dump the result into a table and then you can join one of
your real tables with that table which will execute much faster
4 Select all rows from one table that don’t exist in another table
There are at least 5 ways to return data from one table which is not
in another table. Two of these are SQL Server 2005 and greater only
NOT IN
NOT EXISTS
LEFT and RIGHT JOIN
OUTER APPLY (2005+)
EXCEPT (2005+)
First Create these two tables
| T-SQL |
1
2
3
4
5
6
7
8
| CREATE TABLE testnulls (ID INT)
INSERT INTO testnulls VALUES (1)
INSERT INTO testnulls VALUES (2)
INSERT INTO testnulls VALUES (null)
CREATE TABLE testjoin (ID INT)
INSERT INTO testjoin VALUES (1)
INSERT INTO testjoin VALUES (3)
|
NOT IN
| T-SQL |
1
| SELECT * FROM testjoin WHERE ID NOT IN(SELECT ID FROM testnulls)
|
What happened? Nothing gets returned! The reason is because the subquery returns a NULL and you can’t compare a NULL to anything
Now run this
| T-SQL |
1
| SELECT * FROM testjoin WHERE ID NOT IN(SELECT ID FROM testnulls WHERE ID IS NOT NULL)
|
That worked because we eliminated the NULL values in the subquery
NOT EXISTS
NOT EXISTS doesn’t have the problem that NOT IN has. Run the following code
| T-SQL |
1
2
3
4
| SELECT * FROM testjoin j
WHERE NOT EXISTS (SELECT n.ID
FROM testnulls n
WHERE n.ID = j.ID)
|
Everything worked as expected
LEFT and RIGHT JOIN
Plain vanilla LEFT and RIGHT JOINS
| T-SQL |
1
2
3
4
5
6
7
| SELECT j.* FROM testjoin j
LEFT OUTER JOIN testnulls n ON n.ID = j.ID
WHERE n.ID IS NULL
SELECT j.* FROM testnulls n
RIGHT OUTER JOIN testjoin j ON n.ID = j.ID
WHERE n.ID IS NULL
|
OUTER APPLY (SQL 2005 +)
OUTER APPLY is something that got added to SQL 2005
| T-SQL |
1
2
3
4
5
| SELECT j.* FROM testjoin j
OUTER APPLY
(SELECT id FROM testnulls n
WHERE n.ID = j.ID) a
WHERE a.ID IS NULL
|
EXCEPT(SQL 2005 +)
EXCEPT is something that got added to SQL 2005. It basically returns
everything from the top table which is not in the bottom table
| T-SQL |
1
2
3
| SELECT * FROM testjoin
EXCEPT
SELECT * FROM testnulls
|
INTERSECT
INTERSECT returns what ever is in both tables(like a regular join)
| T-SQL |
1
2
3
| SELECT * FROM testjoin
INTERSECT
SELECT * FROM testnulls
|
5 Getting all rows from one table and only the latest from the child table
First create this table
| T-SQL |
1
2
3
4
5
6
7
8
9
| CREATE TABLE #MaxVal(id INT,VALUE INT,SomeDate datetime)
INSERT #MaxVal VALUES(1,1,'20010101')
INSERT #MaxVal VALUES(1,2,'20020101')
INSERT #MaxVal VALUES(1,3,'20080101')
INSERT #MaxVal VALUES(2,1,'20010101')
INSERT #MaxVal VALUES(2,2,'20060101')
INSERT #MaxVal VALUES(2,3,'20080101')
INSERT #MaxVal VALUES(3,1,'20010101')
INSERT #MaxVal VALUES(3,2,'20080101')
|
If you just need the max value from a column you can just do a group by
| T-SQL |
1
2
3
| SELECT id,MAX(SomeDate) AS VALUE
FROM #MaxVal
GROUP BY id
|
If you need the whole row back then the query below is one way of doing this
| T-SQL |
1
2
3
4
5
6
| SELECT t.* FROM(
SELECT id,MAX(SomeDate) AS MaxValue
FROM #MaxVal
GROUP BY id) x
JOIN #MaxVal t ON x.id =t.id
AND x.MaxValue =t.SomeDate
|
We have another blog post on our site which has a lot more detail
about doing this, as a matter of fact that blog post describes 5 ways to
do it. These 5 ways are:
Key Search
1. Correlated Subquery
2. Derived Table
Number and Filter
3. Windowing Function – Two Stage
4. Windowing Function – One Stage
Simulate MS Access
5. Compound Key, aka Packed Values
That post can be found here: Including an Aggregated Column’s Related Values
6 Getting all characters until a specific character (charindex + left)
There are two built in functions that you can use in SQL Server to
return the first position in a column of the character you are looking
for. These functions are PATINDEX and CHARINDEX. Let’s take a look at
how PATINDEX works. First create this table.
| T-SQL |
1
2
3
4
5
6
7
8
| CREATE TABLE #SomeTable2
(SomeValue VARCHAR(49))
INSERT INTO #SomeTable2 VALUES ('one two three')
INSERT INTO #SomeTable2 VALUES ('1 2 3')
INSERT INTO #SomeTable2 VALUES ('abc def ghi')
INSERT INTO #SomeTable2 VALUES ('one two')
INSERT INTO #SomeTable2 VALUES ('one two three four')
INSERT INTO #SomeTable2 VALUES ('one two three four five')
|
Now we want to return everything up to the first space and also everything after the last space. Here is how we do that
| T-SQL |
1
2
3
| select *,left(SomeValue,patindex('% %',SomeValue)-1),
right(SomeValue,patindex('% %',(reverse(SomeValue)))-1)
from #SomeTable2
|
(results)
| one two three |
one |
three |
| 1 2 3 |
1 |
3 |
| abc def ghi |
abc |
ghi |
| one two |
one |
two |
| one two three four |
one |
four |
| one two three four five |
one |
five |
As you can see to get the last value you can just use the REVERSE function with the RIGHT function.
7 Return all rows with NULL values in a column
People have a hard time with nulls, the first problem with NULLs is
that your WHERE clause is not the same as if you would select a blank or
a value
First create this simple table
| T-SQL |
1
2
3
4
5
| CREATE TABLE #SomeTableNull(SomeValue VARCHAR(49))
INSERT INTO #SomeTableNull VALUES ('1')
INSERT INTO #SomeTableNull VALUES ('')
INSERT INTO #SomeTableNull VALUES ('a')
INSERT INTO #SomeTableNull VALUES (NULL)
|
To search a column your WHERE clause uses the = sign
| T-SQL |
1
2
3
| SELECT * FROM #SomeTableNull WHERE SomeValue = ''
SELECT * FROM #SomeTableNull WHERE SomeValue = '1'
SELECT * FROM #SomeTableNull WHERE SomeValue = 'a'
|
Now let us try that to find a NULL value
| T-SQL |
1
| SELECT * FROM #SomeTableNull WHERE SomeValue = NULL
|
What just happened? We got nothing back? The reason is that you cannot compare a NULL value according to ANSI standards.
| T-SQL |
1
2
3
4
5
| Take a look at this
IF NULL = NULL
print 'equal'
else
print 'not so equal'
|
The bottom print statement got printed, NULL is unknown and you do not know if two unknows are the same
The correct way to return the NULL value is the following
| T-SQL |
1
| SELECT * FROM #SomeTableNull WHERE SomeValue IS NULL
|
Just so that you know this, if you turn off ansi nulls then you can use =
| T-SQL |
1
2
3
4
5
6
| SET ansi_nulls off
SELECT * FROM #SomeTableNull WHERE SomeValue = NULL
SET ansi_nulls on
|
However I do not recommend doing that ever, the default is ON and I would leave it like that
8 Row values to column (PIVOT)
A very frequent request is how to pivot/transpose/crosstab a query.
SQL server 2005 introduced PIVOT, this makes life a lot easier compared
to the SQL 2000 days. so let’s see how this works. First create this
table
| T-SQL |
1
2
| CREATE TABLE #SomeTable
(SomeName VARCHAR(49), Quantity INT)
|
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
| INSERT INTO #SomeTable VALUES ('Scarface', 2)
INSERT INTO #SomeTable VALUES ('Scarface', 4)
INSERT INTO #SomeTable VALUES ('LOTR', 5)
INSERT INTO #SomeTable VALUES ('LOTR', 6)
INSERT INTO #SomeTable VALUES ( 'Jaws', 2)
INSERT INTO #SomeTable VALUES ('Blade', 5)
INSERT INTO #SomeTable VALUES ('Saw', 6)
INSERT INTO #SomeTable VALUES ( 'Saw', 2)
INSERT INTO #SomeTable VALUES ( 'Jaws', 12)
INSERT INTO #SomeTable VALUES ('Blade', 5)
INSERT INTO #SomeTable VALUES ('Saw', 6)
INSERT INTO #SomeTable VALUES ( 'Saw', 2)
|
What we want is too list all the movies in a column and the sum of
all quantities for that movie as a value. So in this case we want this
output
| Scarface |
LOTR |
Jaws |
Saw |
Blade
|
| 6 |
11 |
14 |
16 |
10
|
First let’s look how we can do this in SQL Server 2000, this BTW will also work in SQL Server 2005/2008
| T-SQL |
1
2
3
4
5
6
| SELECT SUM(CASE SomeName WHEN 'Scarface' THEN Quantity ELSE 0 END) AS Scarface,
SUM(CASE SomeName WHEN 'LOTR' THEN Quantity ELSE 0 END) AS LOTR,
SUM(CASE SomeName WHEN 'Jaws' THEN Quantity ELSE 0 END) AS Jaws,
SUM(CASE SomeName WHEN 'Saw' THEN Quantity ELSE 0 END) AS Saw,
SUM(CASE SomeName WHEN 'Blade' THEN Quantity ELSE 0 END) AS Blade
FROM #SomeTable
|
In SQL Server 2005/2008 we can use PIVOT, here is how we can use it
| T-SQL |
1
2
3
4
5
6
7
8
| SELECT Scarface, LOTR, Jaws, Saw,Blade
FROM
(SELECT SomeName,Quantity
FROM #SomeTable) AS pivTemp
PIVOT
( SUM(Quantity)
FOR SomeName IN (Scarface, LOTR, Jaws, Saw,Blade)
) AS pivTable
|
That looks a little bit neater than the SQL 2000 version.
If you are interested in Column To Row take a look at Column To Row (UNPIVOT) on our wiki
9 Pad or remove leading zeroes from numbers
To pad a number with leading zeroes you use as many zeroes as you
want the output to be, then use concatenation plus the right function to
display it
For example if you want to pad a 6 digit column with zeroes you would do something like this
RIGHT(‘000000’ + CONVERT(VARCHAR(6),ColumnName),6)
Here is some code that will show this, first create this table
USE tempdb
go
| T-SQL |
1
2
3
4
5
| CREATE TABLE testpadding(id VARCHAR(50),id2 INT)
INSERT testpadding VALUES('000001',1)
INSERT testpadding VALUES('000134',134)
INSERT testpadding VALUES('002232',2232)
INSERT testpadding VALUES('000002',2)
|
Now run
| T-SQL |
1
2
| SELECT RIGHT('000000' + CONVERT(VARCHAR(6),id2),6)
FROM testpadding
|
(results)
000001
000134
002232
000002
what about the id columns and stripping the zeroes from that?
No problem do this
| T-SQL |
1
2
| SELECT CONVERT(INT,id)
FROM testpadding
|
(results)
1
134
2232
2
Beautiful right? Not so fast, insert this row
| T-SQL |
1
| INSERT testpadding VALUES('02222222222222222222200002',2)
|
| T-SQL |
1
2
| SELECT CONVERT(INT,id)
FROM testpadding
|
Server: Msg 248, Level 16, State 1, Line 1
The conversion of the varchar value ‘02222222222222222222200002’ overflowed an int column.
Okay we can do a bigint instead
| T-SQL |
1
2
| SELECT CONVERT(bigINT,id)
FROM testpadding
|
Server: Msg 8115, Level 16, State 2, Line 1
Arithmetic overflow error converting expression to data type bigint.
Nope, even that doesn’t fit, now what?
| T-SQL |
1
2
| select replace(ltrim(replace(id,'0',' ')),' ','0')
FROM testpadding
|
(results)
1
134
2232
2
2222222222222222222200002
So how does that work? First you replace all the zeroes with spaces,
then you trim it to get rid of the leading spaces, after you replace the
space back to zeroes again, the leading zeroes don’t exists anymore
because they were trimmed
10 Concatenate Values From Multiple Rows Into One Column
If you want to concatenate values from multiple rows into one and you
want to order it then you have to use FOR XML PATH. The ORDER BY is
-not- guaranteed to be processed before the concatenation occurs, if you
use that technique. However, it -is- guaranteed if you use FOR XML
PATH(”).
Let’s take a look. First create these tables
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| USE TEMPDB
GO
CREATE TABLE Authors (Id INT, LastName VARCHAR(100), FirstName VARCHAR(100))
GO
INSERT Authors VALUES(1,'Nielsen','Paul')
INSERT Authors VALUES(2,'King','Stephen')
INSERT Authors VALUES(3,'Stephenson','Neal')
GO
CREATE TABLE Books (ID INT,AuthorID INT, BookName VARCHAR(200))
GO
INSERT Books VALUES(1,1,'SQL Server 2005 Bible')
INSERT Books VALUES(2,1,'SQL Server 2008 Bible')
INSERT Books VALUES(3,2,'It')
INSERT Books VALUES(4,2,'The Stand')
INSERT Books VALUES(5,2,'Thinner')
INSERT Books VALUES(6,2,'Salems Lot')
INSERT Books VALUES(7,3,'Snow Crash')
INSERT Books VALUES(8,3,'The Diamond Age')
INSERT Books VALUES(9,3,'Cryptonomicon')
GO
|
This is the old style function, it will run on SQL Server 2000 and up
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
| CREATE FUNCTION fnGetBooks2 (@AuthorID INT)
RETURNS VARCHAR(8000)
AS
BEGIN
DECLARE @BookList VARCHAR(8000)
SELECT @BookList = ''
SELECT @BookList = @BookList + BookName +','
FROM Books
WHERE AuthorID = @AuthorID
AND BookName IS NOT NULL
ORDER BY BookName --yes we can sort
RETURN LEFT(@BookList,(LEN(@BookList) -1))
END
GO
|
This is the same function using XML Path, so SQL Server 2005 and up
| T-SQL |
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
| CREATE FUNCTION fnGetBooks (@AuthorID INT)
RETURNS VARCHAR(8000)
AS
BEGIN
DECLARE @BookList VARCHAR(8000)
SELECT @BookList =(
SELECT BookName + ', ' AS [text()]
FROM Books
WHERE AuthorID = @AuthorID
AND BookName IS NOT NULL
ORDER BY BookName
FOR XML PATH('') )
RETURN LEFT(@BookList,(LEN(@BookList) -1))
END
GO
|
Here are the calls to the functions
| T-SQL |
1
2
| SELECT *,dbo.fnGetBooks(id) AS Books
FROM Authors
|
| T-SQL |
1
2
| SELECT *,dbo.fnGetBooks2(id) AS Books
FROM Authors
|
(results)
| Id |
LastName |
FirstName |
Books
|
| 1 |
Nielsen |
Paul |
SQL Server 2005 Bible,SQL Server 2008 Bible
|
| 2 |
King |
Stephen |
It,Salems Lot,The Stand,Thinner
|
| 3 |
Stephenson |
Neal |
Cryptonomicon,Snow Crash,The Diamond Age
|
The results are the same but I have been assured by people in the
private SQL Server MVP group that FOR XML PATH will preserve the order
while the other function might not
No comments:
Post a Comment