考虑一个包含名称的数据库表,其中有三行:
Peter
Paul
Mary
有没有一种简单的方法可以把它变成彼得、保罗、玛丽的一串?
考虑一个包含名称的数据库表,其中有三行:
Peter
Paul
Mary
有没有一种简单的方法可以把它变成彼得、保罗、玛丽的一串?
当前回答
对于Oracle DB,请参阅以下问题:如何在Oracle中将多行连接成一行而不创建存储过程?
最好的答案似乎是@Emmanuel,使用了内置的LISTAGG()函数,该函数在Oracle 11g Release 2和更高版本中提供。
SELECT question_id,
LISTAGG(element_id, ',') WITHIN GROUP (ORDER BY element_id)
FROM YOUR_TABLE;
GROUP BY question_id
正如@user762952所指出的,根据Oracle的文档http://www.oracle-base.com/articles/misc/string-aggregation-techniques.php,WM_CONCAT()函数也是一个选项。它看起来很稳定,但Oracle明确建议不要将它用于任何应用程序SQL,因此使用时风险自负。
除此之外,您还必须编写自己的函数;上面的Oracle文档提供了如何做到这一点的指南。
其他回答
在SQL Server中实现这一点的一种方法是将表内容返回为XML(对于XML原始),将结果转换为字符串,然后用“,”替换标记。
PostgreSQL数组非常棒。例子:
创建一些测试数据:
postgres=# \c test
You are now connected to database "test" as user "hgimenez".
test=# create table names (name text);
CREATE TABLE
test=# insert into names (name) values ('Peter'), ('Paul'), ('Mary');
INSERT 0 3
test=# select * from names;
name
-------
Peter
Paul
Mary
(3 rows)
将它们聚合到一个数组中:
test=# select array_agg(name) from names;
array_agg
-------------------
{Peter,Paul,Mary}
(1 row)
将数组转换为逗号分隔的字符串:
test=# select array_to_string(array_agg(name), ', ') from names;
array_to_string
-------------------
Peter, Paul, Mary
(1 row)
DONE
由于PostgreSQL 9.0,引用删除的答案“没有名字的马”更容易:
select string_agg(name, ',')
from names;
以下是实现这一目标的完整解决方案:
-- Table Creation
CREATE TABLE Tbl
( CustomerCode VARCHAR(50)
, CustomerName VARCHAR(50)
, Type VARCHAR(50)
,Items VARCHAR(50)
)
insert into Tbl
SELECT 'C0001','Thomas','BREAKFAST','Milk'
union SELECT 'C0001','Thomas','BREAKFAST','Bread'
union SELECT 'C0001','Thomas','BREAKFAST','Egg'
union SELECT 'C0001','Thomas','LUNCH','Rice'
union SELECT 'C0001','Thomas','LUNCH','Fish Curry'
union SELECT 'C0001','Thomas','LUNCH','Lessy'
union SELECT 'C0002','JOSEPH','BREAKFAST','Bread'
union SELECT 'C0002','JOSEPH','BREAKFAST','Jam'
union SELECT 'C0002','JOSEPH','BREAKFAST','Tea'
union SELECT 'C0002','JOSEPH','Supper','Tea'
union SELECT 'C0002','JOSEPH','Brunch','Roti'
-- function creation
GO
CREATE FUNCTION [dbo].[fn_GetItemsByType]
(
@CustomerCode VARCHAR(50)
,@Type VARCHAR(50)
)
RETURNS @ItemType TABLE ( Items VARCHAR(5000) )
AS
BEGIN
INSERT INTO @ItemType(Items)
SELECT STUFF((SELECT distinct ',' + [Items]
FROM Tbl
WHERE CustomerCode = @CustomerCode
AND Type=@Type
FOR XML PATH(''))
,1,1,'') as Items
RETURN
END
GO
-- fianl Query
DECLARE @cols AS NVARCHAR(MAX),
@query AS NVARCHAR(MAX)
select @cols = STUFF((SELECT distinct ',' + QUOTENAME(Type)
from Tbl
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set @query = 'SELECT CustomerCode,CustomerName,' + @cols + '
from
(
select
distinct CustomerCode
,CustomerName
,Type
,F.Items
FROM Tbl T
CROSS APPLY [fn_GetItemsByType] (T.CustomerCode,T.Type) F
) x
pivot
(
max(Items)
for Type in (' + @cols + ')
) p '
execute(@query)
这也很有用
create table #test (id int,name varchar(10))
--use separate inserts on older versions of SQL Server
insert into #test values (1,'Peter'), (1,'Paul'), (1,'Mary'), (2,'Alex'), (3,'Jack')
DECLARE @t VARCHAR(255)
SELECT @t = ISNULL(@t + ',' + name, name) FROM #test WHERE id = 1
select @t
drop table #test
回报
Peter,Paul,Mary
首先,您应该声明一个表变量并用表数据填充它,然后,使用WHILE循环,逐个选择行并将其值添加到nvarchar(max)变量中。
Go
declare @temp table(
title nvarchar(50)
)
insert into @temp(title)
select p.Title from dbo.person p
--
declare @mainString nvarchar(max)
set @mainString = '';
--
while ((select count(*) from @temp) != 0)
begin
declare @itemTitle nvarchar(50)
set @itemTitle = (select top(1) t.Title from @temp t)
if @mainString = ''
begin
set @mainString = @itemTitle
end
else
begin
set @mainString = concat(@mainString,',',@itemTitle)
end
delete top(1) from @temp
end
print @mainString