从一个查询的结果设置变量以供另一个查询使用?

时间:2012-09-05 17:58:48

标签: sql sql-server-2008

我有2个查询,其中一个查询包含另一个查询要使用的参数。我做了类似的事情,但没有多个参数。以下是查询:

SELECT [loadJob]
      ,[loadStep]
      ,[parameter]
      ,[value]
  FROM [Admin].[admin].[LoadParameters] where loadJob like 'someJobName%' and (parameter = 'targetDatabaseName' or parameter = 'targetDatasetName' or parameter = 'targetDatasetSchema')

SELECT 
      TABLE_CATALOG as DB,
      TABLE_SCHEMA,
      TABLE_NAME,
      COLUMN_NAME,
      ORDINAL_POSITION,
      COLUMN_DEFAULT,
      IS_NULLABLE,
      DATA_TYPE,
      CHARACTER_MAXIMUM_LENGTH,      
      NUMERIC_PRECISION,
      NUMERIC_PRECISION_RADIX,
      NUMERIC_SCALE,
      DATETIME_PRECISION,      
      CHARACTER_SET_NAME
FROM 
INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_CATALOG = 'TheDataBase' // From the [parameter] with name 'targetDatabaseName' get form the [value] column
AND TABLE_SCHEMA = 'data' // From the [parameter] with name 'targetDatasetSchema' get form the [value] column
AND TABLE_NAME = 'TableName'  // From the [parameter] with name 'targetDatasetName' get form the [value] column

所以如果第一个查询的结果是:

[parameter]           [value]

targetDatabaseName    SomeDataBaseName

targetDatasetSchema   data

targetDatasetName     TheTableName

第二个查询中的where子句将使用这些参数:

FROM 
INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_CATALOG = 'SomeDataBaseName'
AND TABLE_SCHEMA = 'data'
AND TABLE_NAME = 'TheTableName'

我想按表名对所有内容进行分组,以便将其导出为CSV。我尝试了一些事情,但我无法自豪地说会有所作为。

1 个答案:

答案 0 :(得分:0)

如果你真的不需要数据库,那么你可以使用类似的东西:

with FirstQuery as (
    select max(case when parameter = 'TargetDatabaseName' then value end) as DB,
           max(case when parameter = 'TargetDatasetSchema' then value end) as Schema,
           max(case when parameter = 'TargetDataSetName' then value end) as table
    from (<first query goes here>) t
)
selelect <whatever>
from Information_Schema q2 join
     FirstQuery q1
     on q1.Table_Schema = q2.Schema and
        q1.Table_Name = q2.table

我看到的问题是数据库。在SQL Server中,Information_Schema表仅包含有关一个数据库的信息。要一次获取多个数据库,您可能必须使用动态SQL。