無限分級的資料查詢是個頭痛的問題,遞迴查詢類別,再組合成字串,用 in 來解決子類所有產品的問題,但是這個效率太低,低的讓人無法接受,在此,有乙個sql的方法,可讓我們提高效率。
-----提取子類的所有類別id
create function getchild (@id int)
returns @t table(id int)
asbegin
insert @t select classid from mproclass where parentid = @id
while @@rowcount > 0
insert @t select a.classid from mproclass as a inner join @t as b
on a.parentid = b.id and a.classid not in(select id from @t)
return
end-----提取子類以及自己id的所有類別id
create function getchildandself (@id int)
returns @t table(id int)
asbegin
insert @t values (@id)
insert @t select classid from mproclass where parentid = @id
while @@rowcount > 0
insert @t select a.classid from mproclass as a inner join @t as b
on a.parentid = b.id and a.classid not in(select id from @t)
return
end以上是最佳方案
查詢該類別的產品資料的sql為:
sql = "select * from mproduct as a inner join (select [id] from getchildandself("+classid+")) as b on a.classid=b.id order by psortid desc,pdate desc";
太爽了!
網上還有其他幾種方法,貼出來大家一起學習學習
一、declare @table table(id int,upperid int)
insert into @table
select 1, 2
union all select 3, 2
union all select 4, 1
union all select 5, 3
declare @upperid int
set @upperid=2;
with result(id,upperid)
as (
select id,upperid from @table where upperid=@upperid
union all
select a.id,a.upperid from @table a inner join result b on a.upperid=b.id
) select*from result
/* id upperid
----------- -----------
1 2
3 2
5 3
4 1
(4 row(s) affected)
*/二、
create table t(id int,upperid int)
insert into t
select 1, 2
union all select 3, 2
union all select 4, 1
union all select 5, 3
select * from t
create function aa(@upperid int)
returns @t table (id int,upperid int,level int)
as begin
declare @i int
set @i=1
insert into @t
select *,@i from t where upperid=@upperid
while @@rowcount> 0
begin
set @i=@i+1
insert into @t
select a.*,@i from t a left join @t b on a.upperid=b.id
where b.level=@i-1
end
return
end
select * from dbo.aa(1)
id upperid level
----------- ----------- -----------
4 1 1
(所影響的行數為 1 行)
select * from dbo.aa(2)
id upperid level
----------- ----------- -----------
1 2 1
3 2 1
4 1 2
5 3 2
三、----建立測試資料
if object_id( 'tbtest ') is not null
drop table tbtest
if object_id( 'spgetchildren ') is not null
drop proc spgetchildren
go create table tbtest(id int, upperid int)
insert tbtest
select 1, 2 union all
select 3, 2 union all
select 4, 1 union all
select 5, 3
go ----建立儲存過程
create proc spgetchildren @id int
as declare @t table(id int)
insert @t select id from tbtest where upperid = @id
while @@rowcount > 0
insert @t select a.id from tbtest as a inner join @t as b
on a.upperid = b.id and a.id not in(select id from @t)
select * from @t
go ----執行儲存過程
declare @upperid int
set @upperid = 2
exec spgetchildren @upperid
----清除測試環境
drop proc spgetchildren
drop table tbtest
/*結果
id
-----------
1 3
4 5
*/
無限分級數量查詢優化
無限分級的資料查詢是個頭痛的問題,遞迴查詢類別,再組合成字串,用 in 來解決子類所有產品的問題,但是這個效率太低,低的讓人無法接受,在此,有乙個sql的方法,可讓我們提高效率。提取子類的所有類別id create function getchild id int returns t table i...
推薦 無限分級數量查詢優化
無限分級的資料查詢是個頭痛的問題,遞迴查詢類別,再組合成字串,用 in 來解決子類所有產品的問題,但是這個效率太低,低的讓人無法接受,在此,有乙個sql的方法,可讓我們提高效率。提取子類的所有類別id create function getchild id int returns t table i...
oracle千萬級資料查詢優化
需求 組合查詢,按條件統計某幾個欄位取前100條記錄 問題 沒建索引導致查詢結果耗時5秒多,不能忍受。解決方法 建索引,在哪個欄位建?在這裡先提下oracle的sql語句的執行。oracle在執行sql語句之前會用優化器optimizer對sql語句進行解析,解析出最優的執行計畫再執行,這樣所花費的...