1

我正在重建一个搜索查询，因为它在“我所看到的”中变得多余，我想知道什么

(albums_artists, Artists) ()加入吗？是为了提高性能吗？

使用简单内连接的查询，使用旧的 (SQL-89) 隐式连接语法：

    SELECT 
    ma_users.name, 
    ma_users.username, 
    albums.id AS album_id, 
    albums.upc, 
    albums.name AS album_name,
    albums.status, 
    albuminfos.label, 
    DATE_FORMAT(albums.created, '%Y-%m-%d') AS created_date,
    CONCAT(artists.name) AS artist_name,
    COUNT(tracks.id) AS total_tracks,
    albumstatus.description AS album_status
    FROM albums, albuminfos, ma_users , (albums_artists, artists) , tracks ,(albumstatus, albumtypes)
    WHERE 
    albums.id = albuminfos.id 
    AND ma_users.id = albums.account_id
    AND albums.id = albums_artists.artist_id 
    AND albums_artists.artist_id = artists.id
    AND tracks.album_id = albums.id
    AND albums.status = albumstatus.id 
    AND albumtypes.id = albums.albumtype_id
    AND albuminfos.label LIKE '%$keywords%'
    GROUP BY albums.id
    ORDER BY albuminfos.label

也接受我需要预期的改进和错误的意见。

4

2 回答 2

1

据我所知，这些括号是多余的，不利于查询执行——可能有人认为通过将它们包围起来，这些表会得到不同的处理，但我不知道这是真的（甚至在这里需要）。

顺便问一下，是 vma_users 还是 ma_users？（大多数参考是没有 av 的 ma_users）

就我个人而言，我不赞成 MySQL 允许 GROUP BY 是“懒惰”的方式，并且更喜欢列出分组中涉及的所有字段。而且，我也不会坚持那些旧的连接样式。没有看到任何语法错误，对于增强功能，我认为没有足够的信息可以评论。

SELECT
       ma_users.NAME
     , ma_users.username
     , albums.id   AS album_id
     , albums.upc
     , albums.NAME AS album_name
     , albums.STATUS
     , albuminfos.label
     , DATE_FORMAT(albums.created, '%Y-%m-%d') AS created_date
     , CONCAT (artists.NAME) AS artist_name
     , COUNT(tracks.id) AS total_tracks
     , albumstatus.description AS album_status
FROM albums
INNER JOIN albuminfos     ON albums.id = albuminfos.id
/* Vma_users ?? */
INNER JOIN ma_users       ON albums.account_id = ma_users.id
INNER JOIN albums_artists ON albums.id = albums_artists.artist_id
INNER JOIN artists        ON albums_artists.artist_id = artists.id
INNER JOIN tracks         ON albums.id = tracks.album_id
INNER JOIN albumstatus    ON albums.STATUS = albumstatus.id
INNER JOIN albumtypes     ON albums.albumtype_id =albumtypes.id
WHERE albuminfos.label LIKE '%$keywords%'
GROUP BY
       ma_users.NAME
     , ma_users.username
     , albums.id
     , albums.upc
     , albums.NAME
     , albums.STATUS
     , albuminfos.label
     , albums.created /* assuming it's only date with times of 00:00:00 */
     , albumstatus.description
ORDER BY albuminfos.label

好吧：我错过了不一致；我同意那些发现它的人 INNER JOIN albums_artists ON albums.id = albums_artists。艺术家 ID 看起来不正确

于 2013-10-07T08:23:22.280 回答
0

基本上，您的查询应该是这样的，但为什么专辑的 id 会与艺术家的 id 相对应？那一点一定是错的。你一定是说album_id，所以我改变了它：

SELECT ma_users.name
     , ma_users.username
     , albums.id AS album_id
     , albums.upc
     , albums.name AS album_name
     , albums.status
     , albuminfos.label
     , DATE_FORMAT(albums.created, '%Y-%m-%d') AS created_date
     , CONCAT(artists.name) AS artist_name
     , COUNT(tracks.id) AS total_tracks
     , albumstatus.description AS album_status
  FROM albums
  JOIN albuminfos
    ON albuminfos.id = albums.id 
  JOIN ma_users
    ON ma_users.id = albums.account_id
  JOIN tracks
    ON tracks.album_id = albums.id
  JOIN albums_artists
    ON albums_artists.album_id = albums.id 
  JOIN artists
    ON artists.id = albums_artists.artist_id 
  JOIN albumstatus
    ON albumstatus.id = albums.status 
  JOIN albumtypes
    ON albumtypes.id = albums.albumtype_id
 WHERE albuminfos.label LIKE '%$keywords%'
 GROUP 
    BY albums.id
 ORDER 
    BY albuminfos.label
于 2013-10-07T08:18:16.227 回答