SQL Server精妙的SQL语句

2022-11-12 09:51:29
内容摘要
这篇文章主要为大家详细介绍了SQL Server精妙的SQL语句,具有一定的参考价值,可以用来参考一下。 对此感兴趣的朋友,看看idc笔记做的技术笔记!说明:复制表(只复制结构,源表名:a新
文章正文

这篇文章主要为大家详细介绍了SQL Server精妙的SQL语句,具有一定的参考价值,可以用来参考一下。

对此感兴趣的朋友,看看idc笔记做的技术笔记!

说明:复制表(只复制结构,源表名:a新表名:b)select*intobfromawhere1<>1说明:拷贝表(拷贝数据,源表名:a目标表名:b)insertintob(a,b,c)selectd,e,ffromb;说明:显示文章、提交人和最后回复时间selecta.title,a.username,b.adddatefromtablea,(selectmax(adddate)adddatefromtablewheretable.title=a.title)b说明:外连接查询(表名1:a表名2:b)selecta.a,a.b,a.c,b.c,b.d,b.ffromaLEFTOUTJOINbONa.a=b.c说明:日程安排提前五分钟提醒select*from日程安排wheredatediff('minute',f开始时间,getdate())>5说明:两张关联表,删除主表中已经在副表中没有的信息deletefrominfowherenotexists(select*frominfobzwhereinfo.infid=infobz.infid)说明:--SQL:SELECTA.NUM,A.NAME,B.UPD_DATE,B.PREV_UPD_DATEFROMTABLE1,(SELECTX.NUM,X.UPD_DATE,Y.UPD_DATEPREV_UPD_DATEFROM(SELECTNUM,UPD_DATE,INBOUND_QTY,STOCK_ONHANDFROMTABLE2WHERETO_CHAR(UPD_DATE,'YYYY/MM')=TO_CHAR(SYSDATE,'YYYY/MM'))X,(SELECTNUM,UPD_DATE,STOCK_ONHANDFROMTABLE2WHERETO_CHAR(UPD_DATE,'YYYY/MM')=TO_CHAR(TO_DATE(TO_CHAR(SYSDATE,'YYYY/MM')||'/01','YYYY/MM/DD')-1,'YYYY/MM'))Y,WHEREX.NUM=Y.NUM(+)ANDX.INBOUND_QTY+NVL(Y.STOCK_ONHAND,0)<>X.STOCK_ONHAND)BWHEREA.NUM=B.NUM说明:--select*fromstudentinfowherenotexists(select*fromstudentwherestudentinfo.id=student.id)and系名称='"&strdepartmentname&"'and专业名称='"&strprofessionname&"'orderby性别,生源地,高考总成绩从数据库中去一年的各单位电话费统计(电话费定额贺电化肥清单两个表来源)SELECTa.userper,a.tel,a.standfee,TO_CHAR(a.telfeedate,'yyyy')AStelyear,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'01',a.factration))ASJAN,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'02',a.factration))ASFRI,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'03',a.factration))ASMAR,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'04',a.factration))ASAPR,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'05',a.factration))ASMAY,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'06',a.factration))ASJUE,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'07',a.factration))ASJUL,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'08',a.factration))ASAGU,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'09',a.factration))ASSEP,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'10',a.factration))ASOCT,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'11',a.factration))ASNOV,SUM(decode(TO_CHAR(a.telfeedate,'mm'),'12',a.factration))ASDECFROM(SELECTa.userper,a.tel,a.standfee,b.telfeedate,b.factrationFROMTELFEESTANDa,TELFEEbWHEREa.tel=b.telfax)aGROUPBYa.userper,a.tel,a.standfee,TO_CHAR(a.telfeedate,'yyyy')说明:四表联查问题select*fromaleftinnerjoinbona.a=b.brightinnerjoincona.a=c.cinnerjoindona.a=d.dwhere.....说明:得到表中最小的未使用的ID号SELECT(CASEWHENEXISTS(SELECT*FROMHandlebWHEREb.HandleID=1)THENMIN(HandleID)+1ELSE1END)asHandleIDFROMHandleWHERENOTHandleIDIN(SELECTa.HandleID-1FROMHandlea)一个SQL语句的问题:行列转换select*fromv_temp上面的视图结果如下:user_namerole_name-------------------------系统管理员管理员feng管理员feng一般用户test一般用户想把结果变成这样:user_namerole_name---------------------------系统管理员管理员feng管理员,一般用户test一般用户===================createtablea_test(namevarchar(20),role2varchar(20))insertintoa_testvalues('李','管理员')insertintoa_testvalues('张','管理员')insertintoa_testvalues('张','一般用户')insertintoa_testvalues('常','一般用户')createfunctionjoin_str(@contentvarchar(100))returnsvarchar(2000)asbegindeclare@strvarchar(2000)set@str=''select@str=@str+','+rtrim(role2)froma_testwhere[name]=@contentselect@str=right(@str,len(@str)-1)return@strendgo--调用:select[name],dbo.join_str([name])role2froma_testgroupby[name]--selectdistinctname,dbo.uf_test(name)froma_test快速比较结构相同的两表结构相同的两表,一表有记录3万条左右,一表有记录2万条左右,我怎样快速查找两表的不同记录?============================给你一个测试方法,从northwind中的orders表取数据。select*inton1fromordersselect*inton2fromordersselect*fromn1select*fromn2--添加主键,然后修改n1中若干字段的若干条altertablen1addconstraintpk_n1_idprimarykey(OrderID)altertablen2addconstraintpk_n2_idprimarykey(OrderID)selectOrderIDfrom(select*fromn1unionselect*fromn2)agroupbyOrderIDhavingcount(*)>1应该可以,而且将不同的记录的ID显示出来。下面的适用于双方记录一样的情况,select*fromn1whereorderidin(selectOrderIDfrom(select*fromn1unionselect*fromn2)agroupbyOrderIDhavingcount(*)>1)至于双方互不存在的记录是比较好处理的--删除n1,n2中若干条记录deletefromn1whereorderIDin('10728','10730')deletefromn2whereorderIDin('11000','11001')--*************************************************************--双方都有该记录却不完全相同select*fromn1whereorderidin(selectOrderIDfrom(select*fromn1unionselect*fromn2)agroupbyOrderIDhavingcount(*)>1)union--n2中存在但在n1中不存的在10728,10730select*fromn1whereOrderIDnotin(selectOrderIDfromn2)union--n1中存在但在n2中不存的在11000,11001select*fromn2whereOrderIDnotin(selectOrderIDfromn1)四种方法取表里n到m条纪录:1.selecttopm*into临时表(或表变量)fromtablenameorderbycolumnname--将topm笔插入setrowcountnselect*from表变量orderbycolumnnamedesc2.selecttopn*from(selecttopm*fromtablenameorderbycolumnname)aorderbycolumnnamedesc3.如果tablename里没有其他identity列,那么:selectidentity(int)id0,*into#tempfromtablename取n到m条的语句为:select*from#tempwhereid0>=nandid0<=m如果你在执行selectidentity(int)id0,*into#tempfromtablename这条语句的时候报错,那是因为你的DB中间的selectinto/bulkcopy属性没有打开要先执行:execsp_dboption你的DB名字,'selectinto/bulkcopy',true4.如果表里有identity属性,那么简单:select*fromtablenamewhereidentitycolbetweennandm如何删除一个表中重复的记录?createtablea_dist(idint,namevarchar(20))insertintoa_distvalues(1,'abc')insertintoa_distvalues(1,'abc')insertintoa_distvalues(1,'abc')insertintoa_distvalues(1,'abc')execup_distinct'a_dist','id'select*froma_distcreateprocedureup_distinct(@t_namevarchar(30),@f_keyvarchar(30))--f_key表示是分组字段﹐即主键字段asbegindeclare@maxinteger,@idvarchar(30),@sqlvarchar(7999),@typeintegerselect@sql='declarecur_rowscursorforselect'+@f_key+',count(*)from'+@t_name+'groupby'+@f_key+'havingcount(*)>1'exec(@sql)opencur_rowsfetchcur_rowsinto@id,@maxwhile@@fetch_status=0beginselect@max=@max-1setrowcount@maxselect@type=xtypefromsyscolumnswhereid=object_id(@t_name)andname=@f_keyif@type=56select@sql='deletefrom'+@t_name+'where'+@f_key+'='+@idif@type=167select@sql='deletefrom'+@t_name+'where'+@f_key+'='+''''+@id+''''exec(@sql)fetchcur_rowsinto@id,@maxendclosecur_rowsdeallocatecur_rowssetrowcount0endselect*fromsystypesselect*fromsyscolumnswhereid=object_id('a_dist')查询数据的最大排序问题(只能用一条语句写)CREATETABLEhard(quchar(11),cochar(11),jenumeric(3,0))insertintohardvalues('A','1',3)insertintohardvalues('A','2',4)insertintohardvalues('A','4',2)insertintohardvalues('A','6',9)insertintohardvalues('B','1',4)insertintohardvalues('B','2',5)insertintohardvalues('B','3',6)insertintohardvalues('C','3',4)insertintohardvalues('C','6',7)insertintohardvalues('C','2',3)要求查询出来的结果如下:qucoje---------------------------A69A24B36B25C67C34就是要按qu分组,每组中取je最大的前2位!!而且只能用一句sql语句!!!select*fromhardawherejein(selecttop2jefromhardbwherea.qu=b.quorderbyje)求删除重复记录的sql语句?怎样把具有相同字段的纪录删除,只留下一条。例如,表test里有id,name字段如果有name相同的记录只留下一条,其余的删除。name的内容不定,相同的记录数不定。有没有这样的sql语句?==============================A:一个完整的解决方案:将重复的记录记入temp1表:select[标志字段id],count(*)intotemp1from[表名]groupby[标志字段id]havingcount(*)>12、将不重复的记录记入temp1表:inserttemp1select[标志字段id],count(*)from[表名]groupby[标志字段id]havingcount(*)=13、作一个包含所有不重复记录的表:select*intotemp2from[表名]where标志字段idin(select标志字段idfromtemp1)4、删除重复表:delete[表名]5、恢复表:insert[表名]select*fromtemp2
12下一页阅读全文

注:关于SQL Server精妙的SQL语句的内容就先介绍到这里,更多相关文章的可以留意

代码注释

作者:喵哥笔记

IDC笔记

学的不仅是技术,更是梦想!