sql-server - MSSQL Server ',' 附近的语法不正确

Question

我正在使用 sql server 并试图查看如何将变量设置为如下位置列表：

Declare @StartDate datetime,
        @EndDate datetime,
        @EmpName varchar(50),
        @LOCID varchar(25);

Set @StartDate = '2016/01/01';
Set @EndDate = '2019/02/01';
Set @EmpName = 'David Brierton';
Set @LOCID =  ('desoto_tickets','ccc_tickets');

select 
  ID,
  requester_name,
  requester_id,
  COUNT(*) as numberOfTickets
from Tickets
where created_at between @StartDate and @EndDate
  and requester_name = @EmpName
  and ',' + tags + ',' like '%,' + @LOCID + ',%'
group by Id, requester_name, requester_id

我收到一个逗号错误Set @LOCID = ('desoto_tickets','ccc_tickets');

错误状态：

',' 附近的语法不正确。

我在这里做错了什么？

任何帮助将不胜感激！

编辑：在标签列的数据库中，每个标签的列出如下： desoto_counter,desoto_tickets,new_associate_workchecked,new_hire__paperwork_check__transaction_complete___express,newhire_assistance,no_error,ops,walk

每一行在数据库中都有一个这样的标签列表。

新编辑：

    Declare @StartDate datetime,
        @EndDate datetime,
    --  @EmpName varchar(50),
        @LOCID1 varchar(25),
        @LOCID2 varchar(25),
        @LOCID3 varchar(25),
        @LOCID4 varchar(25),
        @LOCID5 varchar(25),
        @LOCID6 varchar(25);

Set @StartDate = '2016/01/01';
Set @EndDate = '2019/02/01';
-- Set @EmpName = 'David Brierton';
Set @LOCID1 =  'desoto_counter';
Set @LOCID2 =  'lwr_counter';
Set @LOCID3 =  'nr_tickets';
Set @LOCID4 =  'ps_tickets';
Set @LOCID5 =  'sdl_tickets';
Set @LOCID5 =  'ccc_tickets';

select 
  ID,
  requester_name,
  requester_id,
  COUNT(*) as numberOfTickets,
  case
when tags like '%,' + @LOCID1 + ',%' then 'CDS'
     tags like '%,' + @LOCID2 + ',%' then 'CLR'
     tags like '%,' + @LOCID3 + ',%' then 'CNR'
     tags like '%,' + @LOCID4 + ',%' then 'CPS'
     tags like '%,' + @LOCID5 + ',%' then 'SDL'
     tags like '%,' + @LOCID6 + ',%' then 'CCC'
else 'Unknown'
end [Location],
from Tickets
where created_at between @StartDate and @EndDate
 -- and requester_name = @EmpName
  and ',' + tags + ',' like '%,' + @LOCID1 + ',%'
  or ',' + tags + ',' like '%,' + @LOCID2 + ',%'
  or ',' + tags + ',' like '%,' + @LOCID3 + ',%'
  or ',' + tags + ',' like '%,' + @LOCID4 + ',%'
  or ',' + tags + ',' like '%,' + @LOCID5 + ',%'
  or ',' + tags + ',' like '%,' + @LOCID6 + ',%'
group by Id, requester_name, requester_id

标签： sql-servertsql

Answer 1

将扩展我的评论。首先，您使用的语法是：

SET @LOCID =  ('desoto_tickets','ccc_tickets');

这不是有效值，因为标量变量只能包含 1 个值。如果要存储分隔字符串，则它需要是 1（文字）值：

SET @LOCID =  'desoto_tickets,ccc_tickets';

但是，WHERE您拥有的条款并不理想，特别是：

AND ',' + tags + ',' LIKE '%,' + @LOCID + ',%'

我建议不要使用这样的语法，而是使用拆分器。如果您使用的是 SQL Server 2016+，则可以使用STRING_SPLIT:

SELECT ID,
       requester_name,
       requester_id,
       COUNT(*) AS numberOfTickets
FROM Tickets T
     CROSS APPLY STRING_SPLIT(@LOCID,',') SS
WHERE created_at BETWEEN @StartDate AND @EndDate
  AND requester_name = @EmpName
  AND SS.[value] = T.Tags --Assumes tags isn't delimited either, if it is, then your data is flawed
GROUP BY Id,
         requester_name,
         requester_id;

如果您使用的是 SQL Server 2008，请搜索delimitedsplit8k，如果您使用的是 2012/2014 ，请搜索delimitedsplit8k_lead。

或者，您可以使用表变量：

DECLARE @Loc table (tag varchar(50));
INSERT INTO @Loc
VALUES('desoto_tickets'),('ccc_tickets');

SELECT ID,
       requester_name,
       requester_id,
       COUNT(*) AS numberOfTickets
FROM Tickets T
     JOIN @Loc L ON T.tags = L.Tag --Assumes tags isn't delimited either, if it is, then your data is flawed
WHERE created_at BETWEEN @StartDate AND @EndDate
  AND requester_name = @EmpName
GROUP BY Id,
         requester_name,
         requester_id;

编辑：好的，OP的数据有缺陷。我将更新此答案以尽快修复他们的数据。

编辑 2：下面“修复”了 OP 的数据。请注意接近尾声的清理和DROP. 对于 OP，请确保在列之前修复所有DROPtags引用：

USE Sandbox;
GO
--Create a sample table
CREATE TABLE dbo.Tickets (ID int IDENTITY,
                          RequesterID int,
                          Tags varchar(MAX));
INSERT INTO dbo.Tickets (RequesterID,
                         Tags)
VALUES (1,'desoto_counter,desoto_tickets,new_associate_workchecked,new_hire__paperwork_check__transaction_complete___express,newhire_assistance,no_error,ops,walk'),
       (2,'newhire_assistance,new_associate_workchecked,no_error');
GO
--Create 2 additional tables
CREATE TABLE dbo.Tags (ID int IDENTITY,
                       TagName varchar(100))

CREATE TABLE dbo.TicketTags (TicketTagID int IDENTITY,
                             TicketID int, --Should be a FOREIGN KEY
                             TagID int); --Should be a FOREIGN KEY

GO
--Get the unique tags. This might be slow if you have a lot of tickets

INSERT INTO dbo.Tags (TagName)
SELECT DISTINCT [value]
FROM tickets t
     CROSS APPLY STRING_SPLIT(Tags,',') SS; --Assumes 2016+ see prior part of answer if not
GO
--Now populate the relational table

INSERT INTO dbo.TicketTags (TicketID,
                            TagID)
SELECT T.ID,
       tags.ID
FROM dbo.tickets T
     CROSS APPLY STRING_SPLIT(Tags,',') SS --Assumes 2016+ see prior part of answer if not
     JOIN dbo.Tags ON SS.[value] = tags.TagName;

GO

--Then drop tbhe old column, which you might want to do LATER
ALTER TABLE dbo.Tickets DROP COLUMN tags;
GO

--And now your "new" query.

DECLARE @Loc table (tag varchar(50));
INSERT INTO @Loc
VALUES('desoto_tickets'),('ccc_tickets');

SELECT *
FROM tickets t
WHERE EXISTS (SELECT 1
              FROM dbo.TicketTags TT
                   JOIN dbo.Tags ON TT.TagID = dbo.Tags.ID
                   JOIN @Loc L ON tags.TagName = L.tag
               WHERE TT.TicketID = T.ID);
GO

--Clean up
DROP TABLE dbo.TicketTags;
DROP TABLE dbo.Tags;
DROP TABLE dbo.Tickets