我有2个表TableA和TableB，它们具有相同的列格式，例如，表TableA和TableB都有列
A B C D E F
其中A和B是主键。
如何编写SQL以检查具有相同主键的TableA和TableB在每列中是否包含完全相同的值。
这意味着这两个表具有完全相同的数据。
答案 0 :(得分：56)
您应该能够“MINUS”或“EXCEPT”，具体取决于您的DBMS使用的SQL的风格。
select * from tableA
minus
select * from tableB
如果查询没有返回任何行，则数据完全相同。
答案 1 :(得分：44)
使用关系运算符：
SELECT * FROM TableA
UNION
SELECT * FROM TableB
EXCEPT
SELECT * FROM TableA
INTERSECT
SELECT * FROM TableB;
将Oracle的EXCEPT更改为MINUS。
稍微挑剔点：上面依赖于运算符优先级，根据SQL标准是依赖于实现的，所以YMMV。它适用于SQL Server，优先级为：
INTERSECT EXCEPT和UNION从左到右进行评估。答案 2 :(得分：15)
SELECT col1, col2, col3
FROM
(SELECT * FROM tableA
UNION ALL
SELECT * FROM tableB) data
GROUP BY col1, col2, col3
HAVING count(*)!=2
答案 3 :(得分：5)
SELECT c.ID
FROM clients c
WHERE EXISTS(SELECT c2.ID
FROM clients2 c2
WHERE c2.ID = c.ID);
将返回两个表中所有相同的ID。要获得差异，请将EXISTS更改为NOT EXISTS。
答案 4 :(得分：3)
从oneday获取脚本时，我修改了它以显示每个条目来自哪个表。
DECLARE @table1 NVARCHAR(80)= 'table 1 name'
DECLARE @table2 NVARCHAR(80)= 'table 2 name'
DECLARE @sql NVARCHAR (1000)
SET @sql =
'
SELECT ''' + @table1 + ''' AS table_name,* FROM
(
SELECT * FROM ' + @table1 + '
EXCEPT
SELECT * FROM ' + @table2 + '
) x
UNION
SELECT ''' + @table2 + ''' AS table_name,* FROM
(
SELECT * FROM ' + @table2 + '
EXCEPT
SELECT * FROM ' + @table1 + '
) y
'
EXEC sp_executesql @stmt = @sql
答案 5 :(得分：2)
来源：Use NATURAL FULL JOIN to compare two tables in SQL by Lukas Eder
使用NATURAL FULL JOIN检测两个表之间相同/不同行的巧妙方法。
示例1-状态标记：
SELECT t1.*, t2.*, CASE WHEN t1 IS NULL OR t2 IS NULL THEN 'Not equal' ELSE 'Equal' END
FROM t1
NATURAL FULL JOIN t2;
示例2-过滤行
SELECT *
FROM (SELECT 't1' AS t1, t1.* FROM t1) t1
NATURAL FULL JOIN (SELECT 't2' AS t2, t2.* FROM t2) t2
WHERE t1 IS NULL OR t2 IS NULL -- show differences
--WHERE t1 IS NOT NULL AND t2 IS NOT NULL -- show the same
答案 6 :(得分：1)
只是为了complet，使用except方法存储一个proc来比较2个表，并在3个错误状态，ADD，DEL，GAP的同一个表中给出结果 表必须具有相同的PK，您声明2个表和字段以比较1个或两个表
就这样使用 ps_TableGap'tbl1'，'Tbl2'，'fld1，fld2，fld3'，'fld4'fld5'fld6'（可选）
/****** Object: StoredProcedure [dbo].[ps_TableGap] Script Date: 10/03/2013 16:03:44 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
-- =============================================
-- Author: Arnaud ALLAVENA
-- Create date: 03.10.2013
-- Description: Compare tables
-- =============================================
create PROCEDURE [dbo].[ps_TableGap]
-- Add the parameters for the stored procedure here
@Tbl1 as varchar(100),@Tbl2 as varchar(100),@Fld1 as varchar(1000), @Fld2 as varchar(1000)= ''
AS
BEGIN
-- SET NOCOUNT ON added to prevent extra result sets from
-- interfering with SELECT statements.
SET NOCOUNT ON;
--Variables
--@Tbl1 = table 1
--@Tbl2 = table 2
--@Fld1 = Fields to compare from table 1
--@Fld2 Fields to compare from table 2
Declare @SQL varchar(8000)= '' --SQL statements
Declare @nLoop int = 1 --loop counter
Declare @Pk varchar(1000)= '' --primary key(s)
Declare @Pk1 varchar(1000)= '' --first field of primary key
declare @strTmp varchar(50) = '' --returns value in Pk determination
declare @FldTmp varchar (1000) = '' --temporarily fields for alias calculation
--If @Fld2 empty we take @Fld1
--fields rules: fields to be compare must be in same order and type - always returns Gap
If @Fld2 = '' Set @Fld2 = @Fld1
--Change @Fld2 with Alias prefix xxx become _xxx
while charindex(',',@Fld2)>0
begin
Set @FldTmp = @FldTmp + (select substring(@Fld2,1,charindex(',',@Fld2)-1) + ' as _' + substring(@Fld2,1,charindex(',',@Fld2)-1) + ',')
Set @Fld2 = (select ltrim(right(@Fld2,len(@Fld2)-charindex(',',@Fld2))))
end
Set @FldTmp = @FldTmp + @Fld2 + ' as _' + @Fld2
Set @Fld2 = @FldTmp
--Determinate primary key jointure
--rule: same pk in both tables
Set @nLoop = 1
Set @SQL = 'Declare crsr cursor for select COLUMN_NAME from INFORMATION_SCHEMA.KEY_COLUMN_USAGE where TABLE_NAME = '''
+ @Tbl1 + ''' or TABLE_SCHEMA + ''.'' + TABLE_NAME = ''' + @Tbl1 + ''' or TABLE_CATALOG + ''.'' + TABLE_SCHEMA + ''.'' + TABLE_NAME = ''' + @Tbl1
+ ''' order by ORDINAL_POSITION'
exec(@SQL)
open crsr
fetch next from crsr into @strTmp
while @@fetch_status = 0
begin
if @nLoop = 1
begin
Set @Pk = 's.' + @strTmp + ' = b._' + @strTmp
Set @Pk1 = @strTmp
set @nLoop = @nLoop + 1
end
Else
Set @Pk = @Pk + ' and s.' + @strTmp + ' = b._' + @strTmp
fetch next from crsr into @strTmp
end
close crsr
deallocate crsr
--SQL statement build
set @SQL = 'select case when s.' + @Pk1 + ' is null then ''Del'' when b._' + @Pk1 + ' is null then ''Add'' else ''Gap'' end as TypErr, '''
set @SQL = @SQL + @Tbl1 +''' as Tbl1, s.*, ''' + @Tbl2 +''' as Tbl2 ,b.* from (Select ' + @Fld1 + ' from ' + @Tbl1
set @SQL = @SQL + ' EXCEPT SELECT ' + @Fld2 + ' from ' + @Tbl2 + ')s full join (Select ' + @Fld2 + ' from ' + @Tbl2
set @SQL = @SQL + ' EXCEPT SELECT ' + @Fld1 + ' from ' + @Tbl1 +')b on '+ @Pk
--Run SQL statement
Exec(@SQL)
END
答案 7 :(得分：1)
在Oracle中，使用全部插入和完全外部联接的组合可以找到2个表的差异。在sql中，您可以通过完全外部联接来提取差异，但是在sql中似乎不存在全部插入/首先插入！因此，您必须改为使用以下查询：
select * from A
full outer join B on
A.pk=B.pk
where A.field1!=B.field1
or A.field2!=B.field2 or A.field3!=B.field3 or A.field4!=B.field4
--and A.Date==Date1
尽管不建议在where子句中使用“ OR”，并且通常会降低性能，但是如果表不庞大，仍可以使用上述查询。 如果上述查询有任何结果，则基于字段1,2,3,4的比较，这正是2个表的差异。为了提高查询性能，您还可以按日期对其进行过滤（选中注释部分）
答案 8 :(得分：0)
试试这个
SELECT COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME = 'table1'
intersect
SELECT COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME = 'table2';
答案 9 :(得分：0)
在 SQL Server 中...使用行数，然后将其与相交的行数进行比较：
DECLARE @t1count int = (SELECT COUNT(*) FROM table1)
IF (@t1count = (SELECT COUNT(*) FROM table2))
IF (SELECT COUNT (*) FROM (SELECT * FROM table1 INTERSECT SELECT * FROM table2) AS dT) = @t1count
SELECT 'Equal'
ELSE SELECT 'Not equal'
ELSE
SELECT 'Not equal'
我这样写是为了当表的行数不相等时，完全跳过相交，这将提高这些情况下的性能。
答案 10 :(得分：0)
比较T1（PK，A，B）和T2（PK，A，B）。
首先比较主键集以查找两侧缺少的键值：
SELECT T1.*, T2.* FROM T1 FULL OUTER JOIN T2 ON T1.PK=T2.PK WHERE T1.PK IS NULL OR T2.PK IS NULL;
然后列出所有值不匹配：
SELECT T1.PK, 'A' AS columnName, T1.A AS leftValue, T2.A AS rightValue FROM T1 JOIN T2 ON T1.PK=T2.PK WHERE COALESCE(T1.A,0) != COALESCE(T2.A,0)
UNION ALL
SELECT T1.PK, 'B' AS columnName, T1.B AS leftValue, T2.B AS rightValue FROM T1 JOIN T2 ON T1.PK=T2.PK WHERE COALESCE(T1.B,0) != COALESCE(T2.B,0)
A和B必须为同一类型。您可以使用INFORMATION SCHEMA生成SELECT。不要忘记COALESCE还包括IS NULL结果。您也可以使用FULL OUTER JOIN和COALESCE（T1.PK，0）= COALESCE（T2.PK，0）。
例如，对于varchar类型的列：
SELECT concat('SELECT T1.PK, ''', COLUMN_NAME, ''' AS columnName, T1.', COLUMN_NAME, ' AS leftValue, T2.', COLUMN_NAME, ' AS rightValue FROM T1 JOIN T2 ON T1.PK=T2.PK WHERE COALESCE(T1.',COLUMN_NAME, ',0)!=COALESCE(T2.', COLUMN_NAME, ',0)')
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME='T1' AND DATA_TYPE IN ('nvarchar','varchar');
答案 11 :(得分：0)
我们可以使用下面的简单查询来比较两个DB2表中的数据
步骤1：-从架构（S）的表（T1）中选择我们需要比较的所有列
SELECT T1.col1,T1.col3,T1.col5 from S.T1
步骤2：-使用“减号”关键字比较2个表。
步骤3：-从模式（S）的表（T2）中选择我们需要比较的所有列
SELECT T2.col1,T2.col3,T2.col5 from S.T1
END结果：- 从S.T1中选择T1.col1，T1.col3，T1.col5 减去 从S.T1中选择T2.col1，T2.col3，T2.col5；
如果查询不返回任何行，则数据完全相同。
答案 12 :(得分：0)
我写这篇文章来比较我从Oracle移植到SQL Server的一个非常讨厌的视图的结果。它创建了一对临时表，＃DataVariances和#SchemaVariances，并且（你猜对了）表中的数据和表本身的模式存在差异。
它要求两个表都有一个主键，但如果源表没有主键，则可以将其放入带有标识列的tempdb中。
declare @TableA_ThreePartName nvarchar(max) = ''
declare @TableB_ThreePartName nvarchar(max) = ''
declare @KeyName nvarchar(max) = ''
/***********************************************************************************************
Script to compare two tables and return differneces in schema and data.
Author: Devin Lamothe 2017-08-11
***********************************************************************************************/
set nocount on
-- Split three part name into database/schema/table
declare @Database_A nvarchar(max) = (
select left(@TableA_ThreePartName,charindex('.',@TableA_ThreePartName) - 1))
declare @Table_A nvarchar(max) = (
select right(@TableA_ThreePartName,len(@TableA_ThreePartName) - charindex('.',@TableA_ThreePartName,len(@Database_A) + 2)))
declare @Schema_A nvarchar(max) = (
select replace(replace(@TableA_ThreePartName,@Database_A + '.',''),'.' + @Table_A,''))
declare @Database_B nvarchar(max) = (
select left(@TableB_ThreePartName,charindex('.',@TableB_ThreePartName) - 1))
declare @Table_B nvarchar(max) = (
select right(@TableB_ThreePartName,len(@TableB_ThreePartName) - charindex('.',@TableB_ThreePartName,len(@Database_B) + 2)))
declare @Schema_B nvarchar(max) = (
select replace(replace(@TableB_ThreePartName,@Database_B + '.',''),'.' + @Table_B,''))
-- Get schema for both tables
declare @GetTableADetails nvarchar(max) = '
use [' + @Database_A +']
select COLUMN_NAME
, DATA_TYPE
from INFORMATION_SCHEMA.COLUMNS
where TABLE_NAME = ''' + @Table_A + '''
and TABLE_SCHEMA = ''' + @Schema_A + '''
'
create table #Table_A_Details (
ColumnName nvarchar(max)
, DataType nvarchar(max)
)
insert into #Table_A_Details
exec (@GetTableADetails)
declare @GetTableBDetails nvarchar(max) = '
use [' + @Database_B +']
select COLUMN_NAME
, DATA_TYPE
from INFORMATION_SCHEMA.COLUMNS
where TABLE_NAME = ''' + @Table_B + '''
and TABLE_SCHEMA = ''' + @Schema_B + '''
'
create table #Table_B_Details (
ColumnName nvarchar(max)
, DataType nvarchar(max)
)
insert into #Table_B_Details
exec (@GetTableBDetails)
-- Get differences in table schema
select ROW_NUMBER() over (order by
a.ColumnName
, b.ColumnName) as RowKey
, a.ColumnName as A_ColumnName
, a.DataType as A_DataType
, b.ColumnName as B_ColumnName
, b.DataType as B_DataType
into #FieldList
from #Table_A_Details a
full outer join #Table_B_Details b
on a.ColumnName = b.ColumnName
where a.ColumnName is null
or b.ColumnName is null
or a.DataType <> b.DataType
drop table #Table_A_Details
drop table #Table_B_Details
select coalesce(A_ColumnName,B_ColumnName) as ColumnName
, A_DataType
, B_DataType
into #SchemaVariances
from #FieldList
-- Get differences in table data
declare @LastColumn int = (select max(RowKey) from #FieldList)
declare @RowNumber int = 1
declare @ThisField nvarchar(max)
declare @TestSql nvarchar(max)
create table #DataVariances (
TableKey nvarchar(max)
, FieldName nvarchar(max)
, TableA_Value nvarchar(max)
, TableB_Value nvarchar(max)
)
delete from #FieldList where A_DataType in ('varbinary','image') or B_DataType in ('varbinary','image')
while @RowNumber <= @LastColumn begin
set @TestSql = '
select coalesce(a.[' + @KeyName + '],b.[' + @KeyName + ']) as TableKey
, ''' + @ThisField + ''' as FieldName
, a.[' + @ThisField + '] as [TableA_Value]
, b.[' + @ThisField + '] as [TableB_Value]
from [' + @Database_A + '].[' + @Schema_A + '].[' + @Table_A + '] a
inner join [' + @Database_B + '].[' + @Schema_B + '].[' + @Table_B + '] b
on a.[' + @KeyName + '] = b.[' + @KeyName + ']
where ltrim(rtrim(a.[' + @ThisField + '])) <> ltrim(rtrim(b.[' + @ThisField + ']))
or (a.[' + @ThisField + '] is null and b.[' + @ThisField + '] is not null)
or (a.[' + @ThisField + '] is not null and b.[' + @ThisField + '] is null)
'
insert into #DataVariances
exec (@TestSql)
set @RowNumber = @RowNumber + 1
set @ThisField = (select coalesce(A_ColumnName,B_ColumnName) from #FieldList a where RowKey = @RowNumber)
end
drop table #FieldList
print 'Query complete. Select from #DataVariances to verify data integrity or #SchemaVariances to verify schemas match. Data types varbinary and image are not checked.'
答案 13 :(得分：0)
我在SQL Server中遇到了同样的问题并编写了这个T-SQL脚本来自动化该过程（实际上这是一个淡化版本，我将所有差异写入单个表以便于报告）。
将'MyTable'和'MyOtherTable'更新为您要比较的表的名称。
DECLARE @ColName varchar(100)
DECLARE @Table1 varchar(100) = 'MyTable'
DECLARE @Table2 varchar(100) = 'MyOtherTable'
IF (OBJECT_ID('tempdb..#col') IS NOT NULL) DROP TABLE #col
SELECT IDENTITY(INT, 1, 1) RowNum , c.name
INTO #col
FROM SYS.Objects o
JOIN SYS.columns c on o.object_id = c.object_id
WHERE o.name = @Table1 AND NOT c.Name IN ('List','Columns','YouWantToIgnore')
DECLARE @Counter INT = (SELECT MAX(RowNum) FROM #col)
WHILE @Counter > 0
BEGIN
SET @ColName = (SELECT name FROM #Col WHERE RowNum= @Counter)
EXEC ('SELECT t1.Identifier
,t1.'+@ColName+' AS '+@Table1+@ColName+'
,t2.'+@ColName+' AS '+@Table2+@ColName+'
FROM '+@Table1+' t1
LEFT JOIN '+@Table2+' t2 ON t1.Identifier = t2.Identifier
WHERE t1.'+@ColName+' <> t2.'+@ColName)
SET @Counter = @Counter - 1
END
答案 14 :(得分：0)
SELECT unnest(ARRAY[1,2,2,3,3])
EXCEPT
SELECT unnest(ARRAY[1,1,2,3,3])
UNION
SELECT unnest(ARRAY[1,1,2,3,3])
EXCEPT
SELECT unnest(ARRAY[1,2,2,3,3])
结果为空，但来源不同！
可是：
(
SELECT unnest(ARRAY[1,2,2,3])
EXCEPT ALL
SELECT unnest(ARRAY[2,1,2,3])
)
UNION
(
SELECT unnest(ARRAY[2,1,2,3])
EXCEPT ALL
SELECT unnest(ARRAY[1,2,2,3])
)
作品。
答案 15 :(得分：0)
改善dietbuddha的答案......
select * from
(
select * from tableA
minus
select * from tableB
)
union all
select * from
(
select * from tableB
minus
select * from tableA
)
答案 16 :(得分：-1)
SELECT *
FROM TABLE A
WHERE NOT EXISTS (SELECT 'X'
FROM TABLE B
WHERE B.KEYFIELD1 = A.KEYFIELD1
AND B.KEYFIELD2 = A.KEYFIELD2
AND B.KEYFIELD3 = A.KEYFIELD3)
;
'X'是任何值。
切换表格以查看不同的差异。
确保在表中加入关键字段。
或者仅将MINUS运算符与2条select语句一起使用，但是MINUS只能在Oracle中工作。
答案 17 :(得分：-1)
大多数回复似乎都忽略了卡米尔（Kamil）提出的问题。 （那是表包含相同的行，但是在每个表中重复的行。）不幸的是，由于我在Oracle中，因此我无法使用他的解决方案。我能想到的最好的方法是：
SELECT * FROM
(
SELECT column1, column2, ..., COUNT(*) AS the_count
FROM tableA
GROUP BY column1, column2, ...
MINUS
SELECT column1, column2, ..., COUNT(*) AS the_count
FROM tableB
GROUP BY column1, column2, ...
)
UNION ALL
(
SELECT column1, column2, ..., COUNT(*) AS the_count
FROM tableB
GROUP BY column1, column2, ...
MINUS
SELECT column1, column2, ..., COUNT(*) AS the_count
FROM tableA
GROUP BY column1, column2, ...
)
答案 18 :(得分：-1)
另一种基于dietbuddha＆amp; amp;的回答的增强查询IanMc。 该查询包括有助于显示存在和缺少行的描述。 （注意：对于 SQL Server ）
(
select 'InTableA_NoMatchInTableB' as Msg, * from tableA
except
select 'InTableA_NoMatchInTableB' , * from tableB
)
union all
(
select 'InTableB_NoMatchInTableA' as Msg, * from tableB
except
select 'InTableB_NNoMatchInTableA' ,* from tableA
)
答案 19 :(得分：-1)
在MySQL中，不支持“减号”，并考虑到性能，这是一个快速的
query:
SELECT
t1.id,
t1.id
FROM t1 inner join t2 using (id) where concat(t1.C, t1.D, ...)<>concat(t2.C, t2.D, ...)