Question

之前我已经问过这个问题了，但是现在需要添加更多字段，因为自连接太多，同样的方法现在需要超时。

这是我当前的查询。当我添加城市和州列时，它很慢，但性能足够，但是当我添加zip列时，它会超时。还有另一种方法吗？

SELECT um.user_id, 
       users.display_name, 
       GROUP_CONCAT(um.meta_value SEPARATOR ' '),
       city.meta_value,
       state.meta_value,
       zip.meta_value
FROM `wp_7bv88z_usermeta` um, 
     `wp_7bv88z_pmpro_memberships_users` u, 
     `wp_7bv88z_users` users,
     `wp_7bv88z_usermeta` lname,
     `wp_7bv88z_usermeta` fname,
     `wp_7bv88z_usermeta` city,
     `wp_7bv88z_usermeta` state,
     `wp_7bv88z_usermeta` zip
WHERE u.user_id = um.user_id 
     and users.id = u.user_id 
     and (lname.user_id = u.user_id and lname.meta_key = 'pmpro_slastname')
     and (fname.user_id = u.user_id and fname.meta_key = 'pmpro_sfirstname')
     and (city.user_id = u.user_id and city.meta_key = 'pmpro_scity')
     and (state.user_id = u.user_id and state.meta_key = 'pmpro_sstate')
     and (zip.user_id = u.user_id and zip.meta_key = 'pmpro_szipcode')
     and (um.meta_key='pmpro_sfirstname' or um.meta_key='pmpro_slastname') 
     and u.status='active'
GROUP BY lname.meta_value, fname.meta_value, city.meta_value, 
         state.meta_value, zip.meta_value, um.user_id

示例数据：

user_meta表

╔════=====╦══════════════==╦══════=======╗
║ user_id ║  meta_key        ║ meta_value  ║
╠════=====╬══════════════==╬══════=======╣
║  1      ║ pmpro_sfirstname ║ Jeff        ║
║  1      ║ pmpro_slasttname ║ Atwood      ║
║  1      ║ pmpro_scity      ║ Chicago     ║
║  1      ║ pmpro_sstate     ║ IL          ║
║  1      ║ pmpro_szip       ║ 60007       ║
║  2      ║ pmpro_sfirstname ║ Geoff       ║
║  2      ║ pmpro_slasttname ║ Davis       ║
║  2      ║ pmpro_scity      ║ Fairfax     ║
║  2      ║ pmpro_sstate     ║ VA          ║
║  2      ║ pmpro_szip       ║ 22030       ║
║  3      ║ pmpro_sfirstname ║ Jarrod      ║
║  3      ║ pmpro_slasttname ║ Minor       ║
║  3      ║ pmpro_scity      ║ Dallas      ║
║  3      ║ pmpro_sstate     ║ TX          ║
║  3      ║ pmpro_szip       ║ 75001       ║
║  4      ║ pmpro_sfirstname ║ Joel        ║
║  4      ║ pmpro_slasttname ║ Brown       ║
║  4      ║ pmpro_scity      ║ Boston      ║
║  4      ║ pmpro_sstate     ║ MA          ║
║  4      ║ pmpro_szip       ║ 02101       ║
╚════=====╩════════════=====╩══════======╝

预期结果：

1, Jeff Atwood, Chicago, IL, 60007
4, Joel Brown, Fairfax, VA, 22030
2, Geoff Davis, Dallas, TX, 75001
3, Jarrod Minor, Boston, MA, 02101

Answer 1

您可以使用聚合从EAV数据结构中选择值

select  user_id,
        max( case when meta_key = 'pmpro_sfirstname' then meta_value else null end) fname,
        max( case when meta_key = 'pmpro_slasttname' then meta_value else null end) lname,
        max( case when meta_key = 'pmpro_scity' then meta_value else null end) city,
        max( case when meta_key = 'pmpro_sstate' then meta_value else null end) state,
        max( case when meta_key = 'pmpro_szip' then meta_value else null end) zip
from user_meta
group by user_id

Demo

基于具有自联接的另一列值按值排序

1 个答案: