GROUP BY,每个组使用 Key:Value 个过滤器

GROUP BY with Key:Value filters for each group

给定具有以下架构的 table:

记录编号 ProcessId 数据
80830F00-379C-11EA-96B7-372060281AFC 1 4.2
80830F00-379C-11EA-96B7-372060281AFC 2 445588662211448
80830F00-379C-11EA-96B7-372060281AFC 7 1
791F6D30-379C-11EA-96B7-372060281AFC 1 3.3
791F6D30-379C-11EA-96B7-372060281AFC 2 999999999999999
791F6D30-379C-11EA-96B7-372060281AFC 7 2

所以 RecordId 代表一些分组键,每个 ProcessId 都有它的 Data

现在我有一个看起来像这样的过滤器:

filter: [
    { "1": "3.3" },
    { "2": "999999999999999" }
]

所以 keyProcessIdvalueData

问题

如何获得所有满足查询的 RecordId,这意味着每个组(按 RecordId 分组)应包含具有特定数据的特定进程。

我认为应该这样使用:

SELECT RecordId
FROM [SomeTable]
GROUP BY RecordId
HAVING --- the row with ProcessId = 1 AND Data = '3.3'
    AND --- the row with ProcessId = 2 AND Data = '999999999999999'

您可以将 havingcase 一起使用:

SELECT RecordId
FROM [SomeTable]
GROUP BY RecordId
HAVING SUM(CASE WHEN ProcessId = 1 AND Data = '3.3' THEN 1 ELSE 0 END) > 0 AND
       SUM(CASE WHEN ProcessId = 2 AND Data = '999999999999999' THEN 1 ELSE 0 END) > 0;