gpt4 book ai didi

php - 如何将大查询拆分为小查询并将它们作为单个查询附加

转载 作者:行者123 更新时间:2023-11-29 12:54:19 24 4
gpt4 key购买 nike

我有一个大型查询,需要更多时间来执行。我想将查询拆分为小查询并执行它们并将它们附加为单个结果。

select
mm.Member_Id,
(SELECT
SubCategory_Name
From member_subcategory
WHERE mm.Admiral_Type_Id = SubCategory_Id) AS Admiral_Type_Id,
(SELECT
SubCategory_Name
From member_subcategory
WHERE mm.Royalty_Type_Id = SubCategory_Id) AS Royalty_Type_Id,
(SELECT
SubCategory_Name
From member_subcategory
WHERE mm.Parent_Type_Id = SubCategory_Id) AS Parent_Type_Id,
(SELECT
SubCategory_Name
From member_subcategory
WHERE mm.Guest_Type_Id = SubCategory_Id) AS Guest_Type_Id,
mha.First_Name,
mha.Middle_Name,
mha.Last_Name,
mha.Address,
mha.City,
mha.State,
mha.Zip,
mwa.First_Name AS Work_First_Name,
mwa.Middle_Name AS Work_Middle_Name,
mwa.Last_Name AS Work_Last_Name,
mwa.City AS Work_City,
mwa.State AS Work_State,
mwa.Zip AS Work_Zip,
mrm.Year AS Royalty_Year,
mrm.Title AS Royalty_Title,
mrm.First AS Royalty_First,
mrm.Maiden AS Royalty_Maiden,
(CASE WHEN mha.Preferred=0 THEN 'Work Address' WHEN mha.Preferred=1 THEN 'Home Address' END) AS Preferred,
(select
msi.MaidenName AS Spouse_MaidenName
FROM member_spouse_info msi
WHERE msi.Member_Id = mm.Member_Id) as Spouse_MaidenName,
(select
msi.Mobile AS Spouse_Mobile
FROM member_spouse_info msi
WHERE msi.Member_Id = mm.Member_Id) as Spouse_Mobile,
(select
mai.Fleet_Year AS Fleet_Year
FROM member_admiral_info mai
WHERE mai.Member_Id = mm.Member_Id) as Fleet_Year,
mmt.Admiral_Title,
mmt.Spouse_Title,
mmt.Couple_Title,
(select
mdi.Year AS Duchess_Year
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_Year,
(select
mdi.College AS Duchess_College
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_College,
(select
mdi.Major AS Duchess_Major
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_Major,
(select
mdi.Sorority AS Duchess_Sorority
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_Sorority,
(select
mdi.Parent_Name AS Duchess_Parent_Name
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_Parent_Name,
(select
mdi.Escort_Name AS Duchess_Escort_Name
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id) as Duchess_Escort_Name,
(select
dp.Duchess_Note AS Duchess_Note
FROM duchess_payment dp
WHERE dp.Duchess_Id = (select
mdi.Duchess_Id
FROM member_duchess_info mdi
WHERE mdi.Member_Id = mm.Member_Id)) as Duchess_Note,
(select
em.Year AS Escort_Year
FROM escort_master em
WHERE em.Member_Id = mm.Member_Id) as Escort_Year,
(select
em.Name AS Escort_Name
FROM escort_master em
WHERE em.Member_Id = mm.Member_Id) as Escort_Name,
(select
em.Parent_Name AS Escort_Parent_Name
FROM escort_master em
WHERE em.Member_Id = mm.Member_Id) as Escort_Parent_Name,
(select
em.Notes AS Escort_Note
FROM escort_master em
WHERE em.Member_Id = mm.Member_Id) as Escort_Note,
(select
guest.Guest_Name
FROM guest
WHERE guest.Member_Id = mm.Member_Id) as Guest_Name,
(select
(CASE WHEN mpm.Status=0 THEN 'Current' WHEN mpm.Status=1 THEN 'Past due' WHEN mpm.Status=2 THEN 'Paid' END) AS Payment_Status
FROM member_payment_master mpm
WHERE mpm.Member_Id = mm.Member_Id) AS Payment_Status,
(select
mpm.Payment_Date
FROM member_payment_master mpm
WHERE mpm.Member_Id = mm.Member_Id) AS Payment_Date,
(select
mpm.Payment_Method
FROM member_payment_master mpm
WHERE mpm.Member_Id = mm.Member_Id) AS Payment_Method,
(select
mpm.Payment_Amount
FROM member_payment_master mpm
WHERE mpm.Member_Id = mm.Member_Id) AS Payment_Amount
FROM member_master mm,
member_home_address mha,
member_work_address mwa,
member_mailing_title mmt,
member_royalty_master mrm
where mha.Member_Id = mm.Member_Id
AND mwa.Member_Id = mm.Member_Id
AND mmt.Member_Id = mm.Member_Id
AND mrm.Member_Id = mm.Member_Id
ORDER BY mm.Member_Id;

最佳答案

仅出于优化目的而分解查询并将各个部分组合在一起......不是一个好主意。如果您的要求很明确,DBMS 引擎应该能够自行优化。

子查询本身并不邪恶。某些 RDBMS(例如 FireBird)可以处理同一表上的多个子查询,并优化计划以仅访问所需的表一次。不幸的是,并非所有 RDBMS 都这样做。因此,删除子查询可能会有所帮助,特别是在这种情况下,您要从同一个表中获取多个列。

另一个问题是旧的 JOIN 语法(使用 WHERE 条件)。现在建议使用显式 JOIN...ON 语法。这也可能对优化产生影响。

这是您的查询的优化版本,您应该在采取更极端的措施之前尝试:

SELECT
mm.Member_Id,
sca.SubCategory_Name AS Admiral_Type_Id,
scr.SubCategory_Name AS Royalty_Type_Id,
scp.SubCategory_Name AS Parent_Type_Id,
scg.SubCategory_Name AS Guest_Type_Id,
mha.First_Name,
mha.Middle_Name,
mha.Last_Name,
mha.Address,
mha.City,
mha.State,
mha.Zip,
mwa.First_Name AS Work_First_Name,
mwa.Middle_Name AS Work_Middle_Name,
mwa.Last_Name AS Work_Last_Name,
mwa.City AS Work_City,
mwa.State AS Work_State,
mwa.Zip AS Work_Zip,
mrm.Year AS Royalty_Year,
mrm.Title AS Royalty_Title,
mrm.First AS Royalty_First,
mrm.Maiden AS Royalty_Maiden,
(CASE
WHEN mha.Preferred=0 THEN 'Work Address'
WHEN mha.Preferred=1 THEN 'Home Address'
END) AS Preferred,
msi.MaidenName AS Spouse_MaidenName,
msi.Mobile AS Spouse_Mobile,
mai.Fleet_Year AS Fleet_Year,
mmt.Admiral_Title,
mmt.Spouse_Title,
mmt.Couple_Title,
mdi.Year AS Duchess_Year,
mdi.College AS Duchess_College,
mdi.Major AS Duchess_Major,
mdi.Sorority AS Duchess_Sorority,
mdi.Parent_Name AS Duchess_Parent_Name,
mdi.Escort_Name AS Duchess_Escort_Name,
dp.Duchess_Note AS Duchess_Note,
em.Year AS Escort_Year,
em.Name AS Escort_Name,
em.Parent_Name AS Escort_Parent_Name,
em.Notes AS Escort_Note,
g.Guest_Name AS Guest_Name,
(CASE
WHEN mpm.Status=0 THEN 'Current'
WHEN mpm.Status=1 THEN 'Past due'
WHEN mpm.Status=2 THEN 'Paid'
END) AS Payment_Status,
mpm.Payment_Date AS Payment_Date,
mpm.Payment_Method AS Payment_Method,
mpm.Payment_Amount AS Payment_Amount
FROM member_master mm
JOIN member_home_address mha ON mha.Member_Id = mm.Member_Id
JOIN member_work_address mwa ON mwa.Member_Id = mm.Member_Id
JOIN member_mailing_title mmt ON mmt.Member_Id = mm.Member_Id
JOIN member_royalty_master mrm ON mrm.Member_Id = mm.Member_Id
LEFT JOIN member_spouse_info msi ON msi.Member_Id = mm.Member_Id
LEFT JOIN member_admiral_info mai ON mai.Member_Id = mm.Member_Id
LEFT JOIN member_duchess_info mdi ON mdi.Member_Id = mm.Member_Id
LEFT JOIN duchess_payment dp ON dp.Duchess_Id = mdi.Duchess_Id
LEFT JOIN escort_master em ON em.Member_Id = mm.Member_Id
LEFT JOIN guest g ON g.Member_Id = mm.Member_Id
LEFT JOIN member_payment_master mpm ON mpm.Member_Id = mm.Member_Id
LEFT JOIN member_subcategory sca ON sca.SubCategory_Id = mm.Admiral_Type_Id
LEFT JOIN member_subcategory scr ON scr.SubCategory_Id = mm.Royalty_Type_Id
LEFT JOIN member_subcategory scp ON scp.SubCategory_Id = mm.Parent_Type_Id
LEFT JOIN member_subcategory scg ON scg.SubCategory_Id = mm.Guest_Type_Id
ORDER BY mm.Member_Id;

请注意,我使用了很多 LEFT JOINS。这是因为我不知道相关表中是否必须存在这些行。你还没有告诉我们任何有关此事的事情。通过在必须具有匹配行的表上使用 (INNER) JOIN 而不是 LEFT JOIN,可以进一步优化查询性能。

我还希望您在所有这些表上都有外键约束...或者至少有索引来支持这些关系。

关于php - 如何将大查询拆分为小查询并将它们作为单个查询附加,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/24301174/

24 4 0
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com