sql server 遞歸查詢樹型結(jié)構(gòu)某節(jié)點(diǎn)的所有上級節(jié)點(diǎn)，并且把這些所有上級節(jié)點(diǎn)多行拼接為一行，即合并列數(shù)據(jù)

with eps_root(pk_eps, pk_parent, eps_code, eps_name) as(    SELECT pk_eps, pk_parent, eps_code, eps_name     FROM pm_eps     where enablestate = 2       and pk_eps = "1001A11000000003P62E"    union all    SELECTe.pk_eps, e.pk_parent, e.eps_code, e.eps_name    FROM pm_eps e  inner join eps_root r on e.pk_eps = r.pk_parent    where e.enablestate = 2)select distinct 	eps_code = stuff((select "/" + convert(nvarchar(500), eps_code) from (select distinct eps_code from eps_root) r1 for xml path("")), 1, 1, ""),	--此次需要使用eps_code排序，不然中文的會錯(cuò)亂順序	eps_name = stuff((select "/" + eps_name from (select top 99.99 PERCENT eps_name from eps_root order by eps_code) r2 for xml path("")), 1, 1, "")from eps_root

執(zhí)行sql，如下圖效果：

到此這篇關(guān)于sqlserver 合并列數(shù)據(jù)的實(shí)現(xiàn)的文章就介紹到這了,更多相關(guān)sqlserver 合并列數(shù)據(jù)內(nèi)容請搜索以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持！