似乎每个人都在引用一个REPLACE函数。甚至多次调用REPLACE函数。但是当你有一个未知数量的空格的动态输出时，它就不起作用了。任何经常处理这个问题的人都知道REPLACE只会删除一个空格，而不是全部。LTRIM和RTRIM似乎有同样的问题。留给微软吧。 下面是一个示例输出，使用WHILE循环删除所有CHAR(32)值(空格)。

DECLARE @INPUT_VAL  VARCHAR(8000)
DECLARE @OUTPUT_VAL VARCHAR(8000)

SET @INPUT_VAL = '      C               A                         '
SET @OUTPUT_VAL = @INPUT_VAL
WHILE CHARINDEX(CHAR(32), @OUTPUT_VAL) > 0 BEGIN
    SET @OUTPUT_VAL = REPLACE(@INPUT_VAL, CHAR(32), '')
END

PRINT 'START:' + @INPUT_VAL + ':END'
PRINT 'START:' + @OUTPUT_VAL + ':END'

下面是上面代码的输出:

START:      C               A                         :END
START:CA:END

现在更进一步，在UPDATE或SELECT语句中使用它，将其更改为udf。

CREATE FUNCTION udf_RemoveSpaces (@INPUT_VAL    VARCHAR(8000))
RETURNS VARCHAR(8000)
AS 
BEGIN

DECLARE @OUTPUT_VAL VARCHAR(8000)
SET @OUTPUT_VAL = @INPUT_VAL
-- ITTERATE THROUGH STRING TO LOOK FOR THE ASCII VALUE OF SPACE (CHAR(32)) REPLACE IT WITH BLANK, NOT NULL
WHILE CHARINDEX(CHAR(32), @OUTPUT_VAL) > 0 BEGIN
    SET @OUTPUT_VAL = REPLACE(@INPUT_VAL, CHAR(32), '')
END

RETURN @OUTPUT_VAL
END

然后在SELECT或INSERT语句中使用该函数:

UPDATE A
SET STATUS_REASON_CODE = WHATEVER.dbo.udf_RemoveSpaces(STATUS_REASON_CODE)
FROM WHATEVER..ACCT_INFO A
WHERE A.SOMEVALUE = @SOMEVALUE

INSERT INTO SOMETABLE
(STATUS_REASON_CODE)
SELECT WHATEVER.dbo.udf_RemoveSpaces(STATUS_REASON_CODE)
FROM WHATEVER..ACCT_INFO A
WHERE A.SOMEVALUE = @SOMEVALUE

简单地替换它;

SELECT REPLACE(fld_or_variable, ' ', '')

编辑: 澄清一下;它是一个全局替换，不需要trim()或担心char或varchar的多个空格:

create table #t (
    c char(8),
    v varchar(8))

insert #t (c, v) values 
    ('a a'    , 'a a'    ),
    ('a a  '  , 'a a  '  ),
    ('  a a'  , '  a a'  ),
    ('  a a  ', '  a a  ')

select
    '"' + c + '"' [IN], '"' + replace(c, ' ', '') + '"' [OUT]
from #t  
union all select
    '"' + v + '"', '"' + replace(v, ' ', '') + '"'
from #t 

结果

IN             OUT
===================
"a a     "     "aa"
"a a     "     "aa"
"  a a   "     "aa"
"  a a   "     "aa"
"a a"          "aa"
"a a  "        "aa"
"  a a"        "aa"
"  a a  "      "aa"

如果你需要在所有列中修剪空格，你可以使用这个脚本来动态地做它:

--Just change table name
declare @MyTable varchar(100)
set @MyTable = 'MyTable'

--temp table to get column names and a row id
select column_name, ROW_NUMBER() OVER(ORDER BY column_name) as id into #tempcols from INFORMATION_SCHEMA.COLUMNS 
WHERE   DATA_TYPE IN ('varchar', 'nvarchar') and TABLE_NAME = @MyTable

declare @tri int
select @tri = count(*) from #tempcols
declare @i int
select @i = 0
declare @trimmer nvarchar(max)
declare @comma varchar(1)
set @comma = ', '

--Build Update query
select @trimmer = 'UPDATE [dbo].[' + @MyTable + '] SET '

WHILE @i <= @tri 
BEGIN

    IF (@i = @tri)
        BEGIN
        set @comma = ''
        END
    SELECT  @trimmer = @trimmer + CHAR(10)+ '[' + COLUMN_NAME + '] = LTRIM(RTRIM([' + COLUMN_NAME + ']))'+@comma
    FROM    #tempcols
    where id = @i

    select @i = @i+1
END

--execute the entire query
EXEC sp_executesql @trimmer

drop table #tempcols

从左到右删除字符串中的空格。要消除中间空间，请使用Replace。

您可以使用RTRIM()从右边删除空格，使用LTRIM()从左边删除空格，因此左右空格被删除如下:

SELECT * FROM table WHERE LTRIM(RTRIM(username)) = LTRIM(RTRIM("Bob alias baby"))

参考本博客:

首先，创建示例表和数据:

CREATE TABLE tbl_RemoveExtraSpaces
(
     Rno INT
     ,Name VARCHAR(100)
)
GO

INSERT INTO tbl_RemoveExtraSpaces VALUES (1,'I    am     Anvesh   Patel')
INSERT INTO tbl_RemoveExtraSpaces VALUES (2,'Database   Research and     Development  ')
INSERT INTO tbl_RemoveExtraSpaces VALUES (3,'Database    Administrator     ')
INSERT INTO tbl_RemoveExtraSpaces VALUES (4,'Learning    BIGDATA    and       NOSQL ')
GO

脚本选择字符串没有额外的空格:

SELECT
     [Rno]
    ,[Name] AS StringWithSpace
    ,LTRIM(RTRIM(REPLACE(REPLACE(REPLACE([Name],CHAR(32),'()'),')(',''),'()',CHAR(32)))) AS StringWithoutSpace
FROM tbl_RemoveExtraSpaces

结果:

Rno         StringWithSpace                                 StringWithoutSpace
----------- -----------------------------------------  ---------------------------------------------
1           I    am     Anvesh   Patel                      I am Anvesh Patel
2           Database   Research and     Development         Database Research and Development
3           Database    Administrator                       Database Administrator
4           Learning    BIGDATA    and       NOSQL          Learning BIGDATA and NOSQL

如果是对表的更新，则必须多次运行此更新，直到影响到0行为止。

update tableName
set colName = REPLACE(LTRIM(RTRIM(colName)), '  ', ' ')
where colName like '%  %'