我正在开发一个SQL sproc,我想返回每个表的行数。我怎么能重写这个语句,以便列出下面每个表中的行数?
SELECT COUNT(*)
FROM [test_setup_details_form_view] [tsdf]
JOIN [test_setup_header_form_view]
ON [test_setup_header_form_view].[test_setup_header_id]
= [tsdf].[test_setup_header_id]
JOIN [test_header_rv] [th] with(nolock)
ON [th].[test_setup_header_id]
= [test_setup_header_form_view].[test_setup_header_id]
JOIN [test_details_answers_expanded_view] [tdae]
ON [tdae].[test_setup_details_id] = [tsdf].[test_setup_details_id]
AND [th].[test_header_id] = [tdae].[test_header_id]
JOIN [event_log_rv] [e]
ON [e].[event_log_id] = [tdae].[event_log_id]
当我执行这个语句时,它只是给了我所有连接后的总行数。
答案 0 :(得分:3)
如果您试图获取每个表的计数而不管连接如何:
SELECT
OBJECT_SCHEMA_NAME([object_id]),
OBJECT_NAME([object_id]),
c
FROM
(
SELECT [object_id],
c = SUM(row_count)
FROM
sys.dm_db_partition_stats -- no NOLOCK necessary
WHERE
index_id IN (0,1)
AND OBJECT_NAME([object_id]) IN
(
N'test_setup_details_from_view',
N'test_setup_header_from_view',
... etc etc. ...
)
GROUP BY [object_id]
) AS x;
答案 1 :(得分:1)
对每个需要计算的表的唯一列使用count (distinct <columnname>)
。
答案 2 :(得分:-2)
从每张桌子?那么为什么不使用元数据表呢?
您正尝试在元数据表中已存在的代码中执行某些操作:
Select
schema_name(schema_id) + '.' + t.name as TableName
, i.rows
from sys.tables t (nolock)
join sys.sysindexes i (nolock) on t.object_id = i.id
and i.indid < 2