Question

我无法理解为什么下面的查询应该需要任何组？我不能让这个工作，而其他地方使用string_agg工作

任何建议赞赏

select dv.DocumentVersionID, dv.DocumentID, dv.AuthorUserID1, dv.AuthorUserID2, dv.AuthorUserID3,
(select string_agg(
     ((COALESCE ((', ' || author1.FirstName || ' ' || author1.LastName), ' ')) ||
     (COALESCE ((', ' || author2.FirstName || ' ' || author2.LastName), ' ')) ||
     (COALESCE ((', ' || author3.FirstName || ' ' || author3.LastName), ' ')))  , ' ' ))   as Authors
FROM  DocumentVersion AS dv  INNER JOIN
     "User" AS author1 ON author1.UserID = dv.AuthorUserID1 LEFT OUTER JOIN
     "User" AS author2 ON author2.UserID = dv.AuthorUserID2 LEFT OUTER JOIN
     "User" AS author3 ON author3.UserID = dv.AuthorUserID3  ;

导致： -

ERROR:  column "dv.documentversionid" must appear in the GROUP BY clause or be used in an aggregate function
LINE 1: select dv.DocumentVersionID, dv.DocumentID, dv.AuthorUserID1...

Answer 1

因为select没有from子句所以引用都是外部查询。您有一个聚合函数，因此它们正在聚合。您也有非聚合列，因此需要group by。

您既不需要子查询也不需要string_agg()：

SELECT dv.DocumentVersionID, dv.DocumentID, dv.AuthorUserID1, dv.AuthorUserID2, dv.AuthorUserID3,
       (COALESCE(', ' || author1.FirstName || ' ' || author1.LastName, ' ') ||
        COALESCE(', ' || author2.FirstName || ' ' || author2.LastName, ' ') ||
        COALESCE(', ' || author3.FirstName || ' ' || author3.LastName, ' ')
       )   as Authors
FROM DocumentVersion dv  INNER JOIN
     "User" author1
     ON author1.UserID = dv.AuthorUserID1 LEFT OUTER JOIN
     "User" author2
     ON author2.UserID = dv.AuthorUserID2 LEFT OUTER JOIN
     "User" author3
     ON author3.UserID = dv.AuthorUserID3  ;

如果你想删除前导,：

SELECT dv.DocumentVersionID, dv.DocumentID, dv.AuthorUserID1, dv.AuthorUserID2, dv.AuthorUserID3,
       substr(COALESCE(', ' || author1.FirstName || ' ' || author1.LastName, ' ') ||
              COALESCE(', ' || author2.FirstName || ' ' || author2.LastName, ' ') ||
              COALESCE(', ' || author3.FirstName || ' ' || author3.LastName, ' '), 3
             )   as Authors
FROM DocumentVersion dv  INNER JOIN
     "User" author1
     ON author1.UserID = dv.AuthorUserID1 LEFT OUTER JOIN
     "User" author2
     ON author2.UserID = dv.AuthorUserID2 LEFT OUTER JOIN
     "User" author3
     ON author3.UserID = dv.AuthorUserID3  ;

Answer 2

所有未加入聚合函数的字段必须放在GROUP BY子句中。我只是无法理解您使用string_agg的原因。简单的连接应该适用于您的情况。

SELECT 
 dv.documentversionid, 
 dv.documentid, 
 dv.authoruserid1, 
 dv.authoruserid2, 
 dv.authoruserid3, 
 COALESCE(author1.firstname||' '||author1.lastname,'') 
   ||COALESCE(', '||author2.firstname|| ' '||author2.lastname,'') 
   ||COALESCE(', '||author3.firstname||' '|| author3.lastname,'') 
 AS Authors 
FROM 
 documentversion AS dv 
 INNER JOIN "user" AS author1 ON author1.userid = dv.authoruserid1
 LEFT OUTER JOIN "user" AS author2 ON author2.userid = authoruserid2 
 LEFT OUTER JOIN "user" AS author3 ON author3.userid = authoruserid3

postgres string_agg和GROUP BY子句

2 个答案: