sql 存储过程 循环使用

USE [clab]
GO
/****** Object:  StoredProcedure [dbo].[sp_bd_getResultByEcd]    Script Date: 08/06/2014 16:47:55 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
-- =============================================
-- Author:        zhangcy
-- Create date: 2014-07-09
-- Description:    传入病人ID及项目代码获取项目最新的检验结果
-- =============================================
CREATE PROCEDURE [dbo].[sp_bd_getResultByEcd]
    (
    @pat_in_no varchar(50) --病人id
    ,@itm_id_list varchar(500) --项目编码字符串，各项目编码以,进行分割
    )
AS

Declare @NeedParse varchar(500) --参数 没有处理的字符串

if (CharIndex(',', @itm_id_list)=0)
    BEGIN
        SELECT
        a.res_itm_id as 项目id
        ,a.res_id as 报告id
        ,a.res_itm_ecd as 项目代码
        ,a.res_chr as 结果
        ,a.res_date as 检测时间
        ,a.value as 提示
        from
        (SELECT row_number() over (partition by res_itm_id order by res_date DESC) as rownum
        ,pat_in_no
        ,res_id
        ,res_itm_id
        ,res_itm_ecd
        ,res_chr
        ,res_date
        ,value
        from patients with(nolock)
        left join resulto with(nolock) on patients.pat_id=resulto.res_id
        left join dict_res_ref_flag on dict_res_ref_flag.id = resulto.res_ref_flag
        where pat_in_no=@pat_in_no
        and pat_flag in (2,4)
        and res_itm_id=@itm_id_list) as a
        where a.rownum=1
    END
else
    BEGIN
        set @NeedParse =@itm_id_list
        while (charIndex(',', @NeedParse)>0)
            begin
                SELECT
                a.res_itm_id as 项目id
                ,a.res_id as 报告id
                ,a.res_itm_ecd as 项目代码
                ,a.res_chr as 结果
                ,a.res_date as 检测时间
                ,a.value as 提示
                from
                (SELECT row_number() over (partition by res_itm_id order by res_date DESC) as rownum
                ,pat_in_no
                ,res_id
                ,res_itm_id
                ,res_itm_ecd
                ,res_chr
                ,res_date
                ,value
                from patients with(nolock)
                left join resulto with(nolock) on patients.pat_id=resulto.res_id
                left join dict_res_ref_flag on dict_res_ref_flag.id = resulto.res_ref_flag
                where pat_in_no=@pat_in_no
                and pat_flag in (2,4)
                and res_itm_id=SubString(@NeedParse,1,CharIndex(',',@NeedParse)-1)) as a
                where a.rownum=1                    

                set @NeedParse =SubString(@NeedParse,CharIndex(',', @NeedParse)+1,len(@NeedParse)-CharIndex(',', @NeedParse))
            end

            SELECT
                a.res_itm_id as 项目id
                ,a.res_id as 报告id
                ,a.res_itm_ecd as 项目代码
                ,a.res_chr as 结果
                ,a.res_date as 检测时间
                ,a.value as 提示
                from
                (SELECT row_number() over (partition by res_itm_id order by res_date DESC) as rownum
                ,pat_in_no
                ,res_id
                ,res_itm_id
                ,res_itm_ecd
                ,res_chr
                ,res_date
                ,value
                from patients with(nolock)
                left join resulto with(nolock) on patients.pat_id=resulto.res_id
                left join dict_res_ref_flag on dict_res_ref_flag.id = resulto.res_ref_flag
                where pat_in_no=@pat_in_no
                and pat_flag in (2,4)
                and res_itm_id=@NeedParse) as a
                where a.rownum=1

    END
GO

此存储过程为：传入病人ID，项目编码集，其中各编码以逗号分割。

存储过程里面要做的就是循环读取项目编码集里面的单个编码，及结合病人ID获取该病人及该项目的检验结果