sql面试题

Sql常用语法下列语句部分是Mssql语句，不可以在access中使用。SQL分类： DDL数据定义语言(CREATE，ALTER，DROP，DECLARE) DML数据操纵语言(SELECT，DELETE，UPDATE，INSERT) DCL数据控制语言(GRANT，REVOKE，COMMIT，ROLLBACK) 首先,简要介绍基础语句：1、说明：创建数据库CREATE DATABASE database-name 2、说明：删除数据库drop database dbname 3、说明：备份sql server- 创建 备份数据的 deviceUSE masterEXEC sp_addumpdevice disk, testBack, c:mssql7backupMyNwind_1.dat- 开始 备份BACKUP DATABASE pubs TO testBack 4、说明：创建新表create table tabname(col1 type1 not null primary key,col2 type2 not null,.)根据已有的表创建新表： A：create table tab_new like tab_old (使用旧表创建新表)B：create table tab_new as select col1,col2 from tab_old definition only 5、说明：删除新表drop table tabname 6、说明：增加一个列Alter table tabname add column col type注：列增加后将不能删除。DB2中列加上后数据类型也不能改变，唯一能改变的是增加varchar类型的长度。7、说明：添加主键： Alter table tabname add primary key(col) 说明：删除主键： Alter table tabname drop primary key(col) 8、说明：创建索引：create unique index idxname on tabname(col.) 删除索引：drop index idxname注：索引是不可更改的，想更改必须删除重新建。9、说明：创建视图：create view viewname as select statement 删除视图：drop view viewname 10、说明：几个简单的基本的sql语句选择：select * from table1 where 范围插入：insert into table1(field1,field2) s(1,2)删除：delete from table1 where 范围更新：update table1 set field1=1 where 范围查找：select * from table1 where field1 like %1% -like的语法很精妙，查资料!排序：select * from table1 order by field1,field2 desc总数：select count * as totalcount from table1求和：select sum(field1) as sum from table1平均：select avg(field1) as avg from table1最大：select max(field1) as max from table1最小：select min(field1) as min from table1 11、说明：几个高级查询运算词A： UNION 运算符 UNION 运算符通过组合其他两个结果表（例如 TABLE1 和 TABLE2）并消去表中任何重复行而派生出一个结果表。当 ALL 随 UNION 一起使用时（即 UNION ALL），不消除重复行。两种情况下，派生表的每一行不是来自 TABLE1 就是来自 TABLE2。 B： EXCEPT 运算符 EXCEPT 运算符通过包括所有在 TABLE1 中但不在 TABLE2 中的行并消除所有重复行而派生出一个结果表。当 ALL 随 EXCEPT 一起使用时 (EXCEPT ALL)，不消除重复行。 C： INTERSECT 运算符INTERSECT 运算符通过只包括 TABLE1 和 TABLE2 中都有的行并消除所有重复行而派生出一个结果表。当 ALL 随 INTERSECT 一起使用时 (INTERSECT ALL)，不消除重复行。 注：使用运算词的几个查询结果行必须是一致的。 12、说明：使用外连接 A、left outer join： 左外连接（左连接）：结果集几包括连接表的匹配行，也包括左连接表的所有行。 SQL: select a.a, a.b, a.c, b.c, b.d, b.f from a LEFT OUT JOIN b ON a.a = b.cB：right outer join: 右外连接(右连接)：结果集既包括连接表的匹配连接行，也包括右连接表的所有行。 C：full outer join： 全外连接：不仅包括符号连接表的匹配行，还包括两个连接表中的所有记录。不错的sql语句1、说明：复制表(只复制结构,源表名：a 新表名：b) (Access可用)法一：select * into b from a where 11法二：select top 0 * into b from a2、说明：拷贝表(拷贝数据,源表名：a 目标表名：b) (Access可用)insert into b(a, b, c) select d,e,f from b;3、说明：跨数据库之间表的拷贝(具体数据使用绝对路径) (Access可用)insert into b(a, b, c) select d,e,f from b in 具体数据库 where 条件例子：.from b in &Server.MapPath(.)&data.mdb & where.4、说明：子查询(表名1：a 表名2：b)select a,b,c from a where a IN (select d from b ) 或者: select a,b,c from a where a IN (1,2,3)5、说明：显示文章、提交人和最后回复时间select a.title,a.username,b.adddate from table a,(select max(adddate) adddate from table where table.title=a.title) b6、说明：外连接查询(表名1：a 表名2：b)select a.a, a.b, a.c, b.c, b.d, b.f from a LEFT OUT JOIN b ON a.a = b.c7、说明：在线视图查询(表名1：a )select * from (SELECT a,b,c FROM a) T where t.a 1;8、说明：between的用法,between限制查询数据范围时包括了边界值,not between不包括select * from table1 where time between time1 and time2select a,b,c, from table1 where a not between 数值1 and 数值29、说明：in 的使用方法select * from table1 where a not in (值1,值2,值4,值6)10、说明：两张关联表，删除主表中已经在副表中没有的信息 delete from table1 where not exists ( select * from table2 where table1.field1=table2.field1 )11、说明：四表联查问题：select * from a left inner join b on a.a=b.b right inner join c on a.a=c.c inner join d on a.a=d.d where .12、说明：日程安排提前五分钟提醒 SQL: select * from 日程安排 where datediff(minute,f开始时间,getdate()513、说明：一条sql 语句搞定数据库分页select top 10 b.* from (select top 20 主键字段,排序字段 from 表名 order by 排序字段 desc) a,表名 b where b.主键字段 = a.主键字段 order by a.排序字段14、说明：前10条记录select top 10 * form table1 where 范围15、说明：选择在每一组b值相同的数据中对应的a最大的记录的所有信息(类似这样的用法可以用于论坛每月排行榜,每月热销产品分析,按科目成绩排名,等等.)select a,b,c from tablename ta where a=(select max(a) from tablename tb where tb.b=ta.b)16、说明：包括所有在 TableA 中但不在 TableB和TableC 中的行并消除所有重复行而派生出一个结果表(select a from tableA ) except (select a from tableB) except (select a from tableC)17、说明：随机取出10条数据select top 10 * from tablename order by newid()18、说明：随机选择记录select newid()19、说明：删除重复记录Delete from tablename where id not in (select max(id) from tablename group by col1,col2,.)20、说明：列出数据库里所有的表名select name from sysobjects where type=U 21、说明：列出表里的所有的select name from syscolumns where id=object_id(TableName)22、说明：列示type、vender、pcs字段，以type字段排列，case可以方便地实现多重选择，类似select 中的case。select type,sum(case vender when A then pcs else 0 end),sum(case vender when C then pcs else 0 end),sum(case vender when B then pcs else 0 end) FROM tablename group by type显示结果：type vender pcs电脑 A 1电脑 A 1光盘 B 2光盘 A 2手机 B 3手机 C 323、说明：初始化表table1TRUNCATE TABLE table124、说明：选择从10到15的记录select top 5 * from (select top 15 * from table order by id asc) table_别名 order by id descsql技巧如何删除一个表中重复的记录？create table a_dist(id int,name varchar(20)insert into a_dist values(1,abc)insert into a_dist values(1,abc)insert into a_dist values(1,abc)insert into a_dist values(1,abc)exec up_distinct a_dist,idselect * from a_distcreate procedure up_distinct(t_name varchar(30),f_key varchar(30)-f_key表示是分组字段即主键字段asbegindeclare max integer,id varchar(30) ,sql varchar(7999) ,type integerselect sql = declare cur_rows cursor for select +f_key+ ,count(*) from +t_name + group by +f_key + having count(*) 1exec(sql)open cur_rows fetch cur_rows into id,max while fetch_status=0 begin select max = max -1 set rowcount max select type = xtype from syscolumns where id=object_id(t_name) and name=f_keyif type=56select sql = delete from +t_name+ where + f_key+ = + id if type=167select sql = delete from +t_name+ where + f_key+ = + id + exec(sql)fetch cur_rows into id,max end close cur_rows deallocate cur_rowsset rowcount 0endselect * from systypesselect * from syscolumns where id = object_id(a_dist) 查询数据的最大排序问题（只能用一条语句写） CREATE TABLE hard (qu char (11) ,co char (11) ,je numeric(3, 0) insert into hard values (A,1,3)insert into hard values (A,2,4)insert into hard values (A,4,2)insert into hard values (A,6,9)insert into hard values (B,1,4)insert into hard values (B,2,5)insert into hard values (B,3,6)insert into hard values (C,3,4)insert into hard values (C,6,7)insert into hard values (C,2,3)要求查询出来的结果如下：qu co je - - - A 6 9A 2 4B 3 6B 2 5C 6 7C 3 4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 * from hard a where je in (select top 2 je from hard b where a.qu=b.qu order by je) 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(*) into temp1 from 表名group by 标志字段idhaving count(*)12、将不重复的记录记入temp1表:insert temp1 select 标志字段id,count(*) from 表名 group by 标志字段id having count(*)=13、作一个包含所有不重复记录的表：select * into temp2 from 表名 where 标志字段id in(select 标志字段id from temp1)4、删除重复表:delete 表名5、恢复表：insert 表名 select * from temp26、删除临时表:drop table temp1drop table temp2=B:create table a_dist(id int,name varchar(20)insert into a_dist values(1,abc)insert into a_dist values(1,abc)insert into a_dist values(1,abc)insert into a_dist values(1,abc)exec up_distinct a_dist,idselect * from a_distcreate procedure up_distinct(t_name varchar(30),f_key varchar(30)-f_key表示是分组字段即主键字段asbegindeclare max integer,id varchar(30) ,sql varchar(7999) ,type integerselect sql = declare cur_rows cursor for select +f_key+ ,count(*) from +t_name + group by +f_key + having count(*) 1exec(sql)open cur_rows fetch cur_rows into id,max while fetch_status=0 begin select max = max -1 set rowcount max select type = xtype from syscolumns where id=object_id(t_name) and name=f_keyif type=56select sql = delete from +t_name+ where + f_key+ = + id if type=167select sql = delete from +t_name+ where + f_key+ = + id + exec(sql)fetch cur_rows into id,max end close cur_rows deallocate cur_rowsset rowcount 0endselect * from systypesselect * from syscolumns where id = object_id(a_dist) 行列转换-普通 假设有张学生成绩表(CJ)如下 Name Subject Result 张三 语文 80 张三 数学 90 张三 物理 85 李四 语文 85 李四 数学 92 李四 物理 82 想变成 姓名 语文 数学 物理 张三 80 90 85 李四 85 92 82 declare sql varchar(4000) set sql = select Name select sql = sql + ,sum(case Subject when +Subject+ then Result end) +Subject+ from (select distinct Subject from CJ) as a select sql = sql+ from test group by name exec(sql) 行列转换-合并 有表A, id pid 1 1 1 2 1 3 2 1 2 2 3 1 如何化成表B: id pid 1 1,2,3 2 1,2 3 1 创建一个合并的函数 create function fmerg(id int) returns varchar(8000) as begin declare str varchar(8000) set str= select str=str+,+cast(pid as varchar) from 表A where id=id set str=right(str,len(str)-1) return(str) End go -调用自定义函数得到结果 select distinct id,dbo.fmerg(id) from 表A 如何取得一个数据表的所有列名 方法如下：先从SYSTEMOBJECT系统表中取得数据表的SYSTEMID,然后再SYSCOLUMN表中取得该数据表的所有列名。 SQL语句如下： declare objid int,objname char(40) set objname = tablename select objid = id from sysobjects where id = object_id(objname) select Column_name = name from syscolumns where id = objid order by colid 或SELECT * FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME =users 通过SQL语句来更改用户的密码 修改别人的,需要sysadmin role EXEC sp_password NULL, newpassword, User 如果帐号为SA执行EXEC sp_password NULL, newpassword, sa 怎么判断出一个表的哪些字段不允许为空？ select COLUMN_NAME from INFORMATION_SCHEMA.COLUMNS where IS_NULLABLE=NO and TABLE_NAME=tablename 如何在数据库里找到含有相同字段的表？ a. 查已知列名的情况 SELECT b.name as TableName,a.name as columnname From syscolumns a INNER JOIN sysobjects b ON a.id=b.id AND b.type=U AND a.name=你的字段名字 未知列名查所有在不同表出现过的列名 Select o.name As tablename,s1.name As columnname From syscolumns s1, sysobjects o Where s1.id = o.id And o.type = U And Exists ( Select 1 From syscolumns s2 Where s1.name = s2.name And s1.id s2.id ) 查询第xxx行数据 假设id是主键： select * from (select top xxx * from yourtable) aa where not exists(select 1 from (select top xxx-1 * from yourtable) bb where aa.id=bb.id) 如果使用游标也是可以的 fetch absolute number from cursor_name 行数为绝对行数 SQL Server日期计算 a. 一个月的第一天 SELECT DATEADD(mm, DATEDIFF(mm,0,getdate(), 0) b. 本周的星期一 SELECT DATEADD(wk, DATEDIFF(wk,0,getdate(), 0) c. 一年的第一天 SELECT DATEADD(yy, DATEDIFF(yy,0,getdate(), 0) d. 季度的第一天 SELECT DATEADD(qq, DATEDIFF(qq,0,getdate(), 0) e. 上个月的最后一天 SELECT dateadd(ms,-3,DATEADD(mm, DATEDIFF(mm,0,getdate(), 0) f. 去年的最后一天 SELECT dateadd(ms,-3,DATEADD(yy, DATEDIFF(yy,0,getdate(), 0) g. 本月的最后一天 SELECT dateadd(ms,-3,DATEADD(mm, DATEDIFF(m,0,getdate()+1, 0) h. 本月的第一个星期一 select DATEADD(wk, DATEDIFF(wk,0, dateadd(dd,6-datepart(day,getdate(),getdate() ), 0) i. 本年的最后一天 SELECT dateadd(ms,-3,DATEADD(yy, DATEDIFF(yy,0,getdate()+1, 0)。 获取表结构把 sysobjects 替换 成 tablename 即可 SELECT CASE IsNull(I.name, ) When Then Else * End as IsPK, Object_Name(A.id) as t_name, A.name as c_name, IsNull(SubString(M.text, 1, 254), ) as pbc_init, T.name as F_DataType, CASE IsNull(TYPEPROPERTY(T.name, Scale), ) WHEN Then Cast(A.prec as varchar) ELSE Cast(A.prec as varchar) + , + Cast(A.scale as varchar) END as F_Scale, A.isnullable as F_isNullAble FROM Syscolumns as A JOIN Systypes as T ON (A.xType = T.xUserType AND A.Id = Object_id(sysobjects) ) LEFT JOIN ( SysIndexes as I JOIN Syscolumns as A1 ON ( I.id = A1.id and A1.id = object_id(sysobjects) and (I.status & 0x800) = 0x800 AND A1.colid 0 then else end) N主键, b.name N类型, a.length N占用字节数, COLUMNPROPERTY(a.id,a.name,PRECISION) as N长度, isnull(COLUMNPROPERTY(a.id,a.name,Scale),0) as N小数位数, (case when a.isnullable=1 then else end) N允许空, isnull(e.text,) N默认值, isnull(g.value,) AS N字段说明 FROM syscolumns a left join systypes b on a.xtype=b.xusertype inner join sysobjects d on a.id=d.id and d.xtype=U and d.namedtproperties left join syscomments e on a.cdefault=e.id left join sysproperties g on a.id=g.id AND a.colid = g.smallid order by object_name(a.id),a.colorder 快速获取表test的记录总数对大容量表非常有效 快速获取表test的记录总数: select rows from sysindexes where id = object_id(test) and indid in (0,1)update 2 set KHXH=(ID+1)2 2行递增编号update 23 set id1 = No.+right(00000000+id,6) where id not like No% /递增update 23 set id1= No.+right(00000000+replace(id1,No.,),6) /补位递增delete from 1 where (id%2)=1 奇数 替换表名字段update 1 set domurl = replace(domurl,Upload/Imgswf/,Upload/Photo/) where domurl like %Upload/Imgswf/% 截位SELECT LEFT(表名, 5) 截位SELECT LEFT(表名, 5)（MS SQL Server）SQL语句导入导出大全/* 导出到excel EXEC master.xp_cmdshell bcp SettleDB.dbo.shanghu out c:temp1.xls -c -q -SGNETDATA/GNETDATA -Usa -P/* 导入Excel SELECT * FROM OpenDataSource( Microsoft.Jet.OLEDB.4.0, Data Source=c:test.xls;User ID=Admin;Password=;Extended properties=Excel 5.0).xactionsSELECT cast(cast(科目编号 as numeric(10,2) as nvarchar(255)+ 转换后的别名 FROM OpenDataSource( Microsoft.Jet.OLEDB.4.0, Data Source=c:test.xls;User ID=Admin;Password=;Extended properties=Excel 5.0).xactionsselect * from OPENROWSET(MICROSOFT.JET.OLEDB.4.0,Excel 5.0;HDR=YES;DATABASE=c:Book1.xls,Sheet1$)HDR=YES;Excel第一行当成标题行HDR=NO;第一行不当成标题行/* 导入文本文件 EXEC master.xp_cmdshell bcp dbname.tablename in c:DT.txt -c -Sservername -Usa -Ppassword/* 导出文本文件 EXEC master.xp_cmdshell bcp dbname.tablename out c:DT.txt -c -Sservername -Usa -Ppassword 或 EXEC master.xp_cmdshell bcp Select * from dbname.tablename queryout c:DT.txt -c -Sservername -Usa -Ppassword导出到TXT文本，用逗号分开 exec master.xp_cmdshell bcp 库名.表名 out d:tt.txt -c -t ,-U sa -P passwordBULK INSERT 库名.表名 FROM c:test.txt WITH ( FIELDTERMINATOR = ;, ROWTERMINATOR = n )-/* dBase IV文件 select * from OPENROWSET(MICROSOFT.JET.OLEDB.4.0 ,dBase IV;HDR=NO;IMEX=2;DATABASE=C:,select * from 客户资料4.dbf) -*/-/* dBase III文件 select * from OPENROWSET(MICROSOFT.JET.OLEDB.4.0 ,dBase III;HDR=NO;IMEX=2;DATABASE=C:,select * from 客户资料3.dbf) -*/-/* FoxPro 数据库 select * from openrowset(MSDASQL, Driver=Microsoft Visual FoxPro Driver;SourceType=DBF;SourceDB=c:, select * from aa.DBF) -*/*导入DBF文件*/ select * from openrowset(MSDASQL, Driver=Microsoft Visual FoxPro Driver; SourceDB=e:VFP98data; SourceType=DBF, select * from customer where country != USA order by country) go /* 导出到DBF */ 如果要导出数据到已经生成结构(即现存的)FOXPRO表中,可以直接用下面的SQL语句insert into openrowset(MSDASQL, Driver=Microsoft Visual FoxPro Driver;SourceType=DBF;SourceDB=c:, select * from aa.DBF) select * from 表说明: SourceDB=c: 指定foxpro表所在的文件夹 aa.DBF 指定foxpro表的文件名./*导出到Access*/ insert into openrowset(Microsoft.Jet.OLEDB.4.0, x:A.mdb;admin;,A表) select * from 数据库名.B表/*导入Access*/ insert into B表 selet * from openrowset(Microsoft.Jet.OLEDB.4.0, x:A.mdb;admin;,A表)* 导入 xml文件DECLARE idoc int DECLARE doc varchar(1000) -sample XML document SET doc = Customer was very satisfied Important Happy Customer. - Create an internal representation of the XML document. EXEC sp_xml_preparedocument idoc OUTPUT, doc- Execute a SELECT statement using OPENXML rowset provider. SELECT * FROM OPENXML (idoc, /root/Customer/Order, 1) WITH (oid char(5), amount float, comment ntext text() EXEC sp_xml_removedocument idoc/*导整个数据库*/用bcp实现的存储过程/* 实现数据导入/导出的存储过程 根据不同的参数,可以实现导入/导出整个数据库/单个表 调用示例: -导出调用示例 -导出单个表 exec file2table zj,xzkh_sa.地区资料,c:zj.txt,1 -导出整个数据库 exec file2table zj,xzkh_sa,C:docman,1-导入调用示例 -导入单个表 exec file2table zj,xzkh_sa.地区资料,c:zj.txt,0 -导入整个数据库 exec file2table zj,xzkh_sa,C:docman,0*/ if exists(select 1 from sysobjects where name=File2Table and objectproperty(id,IsProcedure)=1) drop procedure File2Table go create procedure File2Table servername varchar(200) -服务器名 ,username varchar(200) -用户名,如果用NT验证方式,则为空 ,password varchar(200) -密码 ,tbname varchar(500) -数据库.dbo.表名,如果不指定:.dbo.表名,则导出数据库的所有用户表 ,filename varchar(1000) -导入/导出路径/文件名,如果tbname参数指明是导出整个数据库,则这个参数是文件存放路径,文件名自动用表名.txt ,isout bit -1为导出,0为导入 as declare sql varchar(8000)if tbname like %.%.% -如果指定了表名,则直接导出单个表 begin set sql=bcp +tbname +case when isout=1 then out else in end + +filename+ /w + /S +servername +case when isnull(username,)= then else /U +username end + /P +isnull(password,) exec master.xp_cmdshell sql end else begin -导出整个数据库,定义游标,取出所有的用户表 declare m_tbname varchar(250) if right(filename,1) set filename=filename+set m_tbname=declare #tb cursor for select name from +tbname+.sysobjects where xtype=U exec(m_tbname) open #tb fetch next from #tb into m_tbname while fetch_status=0 begin set sql=bcp +tbname+.+m_tbname +case when isout=1 then out else in end + +filename+m_tbname+.txt /w + /S +servername +case when isnull(username,)= then else /U +username end + /P +isnull(password,) exec master.xp_cmdshell sql fetch next from #tb into m_tbname end close #tb deallocate #tb end go/*Excel导到Txt*