SQL相关功能实现.

网上查看行转列.差不多都是对于一个表.今天有时间自己整理下.实现一个相对通用点的.

-- =============================================
-- Author:		xin_zhou
-- Create date: 2010-03-21
-- Description:	相对通用行转列
-- =============================================
Create PROCEDURE RowToColumn 
	@tablename nvarchar(50),
	@group_value nvarchar(50),
	@coulmn_case nvarchar(50),
	@coulmn_value nvarchar(50),
	@defaultvalue nvarchar(50)=null,
	@group_fun nvarchar(10)='max'
AS
BEGIN
	declare @coulmntable table (coulmn nvarchar(50))
	declare @str_coulmntable nvarchar(120)
	declare @str_head nvarchar(60)
	declare @str_tail nvarchar(120)
	declare @str_case nvarchar(4000)
	declare @str_default nvarchar(60)
	set @str_coulmntable = 'select distinct ' + @coulmn_case + ' as coulmn from ' + @tablename
	insert into @coulmntable exec(@str_coulmntable)	
	set @str_head = 'select '+ @group_value + ' , '	
	set @str_case = ''
	if(@defaultvalue is null)
		set @str_default = ' else null end) as '
	else
		set @str_default = ' else ''' + @defaultvalue + ''' end) as '
	select @str_case = @str_case + ' ' + @group_fun + '(case '+ @coulmn_case + ' when ''' + coulmn + ''' then ' 
					+ @coulmn_value + @str_default + '''' + coulmn + ''' , ' from  @coulmntable
	set @str_case = left(@str_case,len(@str_case)-1)	
	set @str_tail = ' from ' + @tablename + ' group by ' + @group_value	
	print @str_head+@str_case+@str_tail
	exec(@str_head+@str_case+@str_tail)
END

整个就是差不多如此.参数意思大家对比下面的图就知道了.(关于参数group_value里可以输入多个栏位.用','分开,如'coulmnA,coulmnB,coulmnC')

至于为什么默认的聚合函数为max,是因为行转列应做到一对一.所以聚合函数在这本没有意义.

而别的聚合函数有的只能对int等来取.所以我首看聚合函数能用的范围大就用那个.

2.查询服务器上的所有数据库相关文件(数据文件与日志文件大小.)

主要注意的是这个本意应用游标来完成.但是心中不怎么愿意用游标.结合@@rowcount来模拟游标的实现.暂时达到目标.

declare @ch nvarchar(50)
set @ch= ''
declare @db_names table([name] nvarchar(50))
insert into @db_names(name) select name from master.dbo.sysdatabases
declare @db_file_size table(dbname nvarchar(50),datafilename nvarchar(50),datafilesize nvarchar(50),logfilename nvarchar(50),logfilesize nvarchar(50))
while @@rowcount >0
begin
	declare @dd table(name nvarchar(50),size nvarchar(50))
	select top 1 @ch=[name]  from @db_names
	insert into @dd exec('select  name,convert(varchar(50),(size/128)) as size from '+@ch+'.dbo.sysfiles')
	declare @x_name nvarchar(50)
	declare @x_size nvarchar(50)
	select @x_name=[name],@x_size=[size] from @dd where right([name],3) <> 'log' 		
	Insert into @db_file_size(dbname,datafilename,datafilesize) values(@ch,@x_name,@x_size)
	select @x_name=[name],@x_size=[size] from @dd  where right([name],3) = 'log' 
	update @db_file_size set logfilename=@x_name,logfilesize=@x_size where dbname = @ch 	
	delete @db_names where name=@ch	
end
select distinct * from @db_file_size

先写到这.

有理解错误的地方请大家指正.