我在一些人的帮助下编写了这种表结构和查询，它的工作原理非常完美。我想获得两个日期之间的期初和期末余额。我已注释了要获取日期的日期，如果运行该日期，请检查所需的预期输出，如下所示。
以下是结构和示例数据：
DROP TABLE [TransactionMaster];
DROP TABLE [VoucherType];
CREATE TABLE [VoucherType](
[VoucherTypeCode] [tinyint] NOT NULL PRIMARY KEY,
[FullName] [nvarchar](255) NOT NULL
);
INSERT INTO [VoucherType] VALUES (1, 'Cash Payment Voucher');
INSERT INTO [VoucherType] VALUES (2, 'Cash Receipt Voucher');
INSERT INTO [VoucherType] VALUES (3, 'Bank Payment Voucher');
INSERT INTO [VoucherType] VALUES (4, 'Bank Receipt Voucher');
CREATE TABLE [TransactionMaster](
[ID] [bigint] NOT NULL PRIMARY KEY,
[VoucherTypeCode] [tinyint] NOT NULL,
[PayeeName] [varchar](255) NOT NULL,
[TransactionDate] datetime,
[RefNo] [nvarchar](50) NULL
CONSTRAINT [FK_tbl_TransactionMaster_tbl_VoucherType] FOREIGN KEY([VoucherTypeCode])
REFERENCES [VoucherType] ([VoucherTypeCode])
)
INSERT INTO [TransactionMaster] VALUES (1, 2, 'Asim', '2018-03-21', 'CRV-0001-LHR');
INSERT INTO [TransactionMaster] VALUES (2, 4, 'Ali', '2018-03-21', 'BRV-2421-KHI');
INSERT INTO [TransactionMaster] VALUES (3, 1, 'Erick', '2018-03-23', 'CPV-5435-ISL');
INSERT INTO [TransactionMaster] VALUES (4, 3, 'Asim', '2018-03-24', 'BPV-2345-CAN');
INSERT INTO [TransactionMaster] VALUES (5, 2, 'Mehboob', '2018-03-25', 'CRV-2976-PSH');
INSERT INTO [TransactionMaster] VALUES (6, 1, 'Erick', '2018-03-25', 'CPV-2323-KOH');
INSERT INTO [TransactionMaster] VALUES (7, 1, 'Feroze', '2018-03-21', 'CRV-0531-SRG');
INSERT INTO [TransactionMaster] VALUES (8, 3, 'Ali', '2018-03-21', 'BRV-2001-RWP');
CREATE TABLE TransactionDetail
(
ID NUMERIC NOT NULL PRIMARY KEY,
TransactionCode bigint,
DrAmount NUMERIC,
CrAmount NUMERIC
);
INSERT INTO TransactionDetail VALUES (1, 1, '2500', NULL);
INSERT INTO TransactionDetail VALUES (2, 1, NULL, '1500');
INSERT INTO TransactionDetail VALUES (3, 1, NULL, '1000');
INSERT INTO TransactionDetail VALUES (4, 2, '1150', NULL);
INSERT INTO TransactionDetail VALUES (5, 2, NULL, '1150');
INSERT INTO TransactionDetail VALUES (6, 3, '600', NULL);
INSERT INTO TransactionDetail VALUES (7, 3, '400', NULL);
INSERT INTO TransactionDetail VALUES (8, 3, '200', NULL);
INSERT INTO TransactionDetail VALUES (9, 3, NULL, '1200');
INSERT INTO TransactionDetail VALUES (10, 4, '1000', NULL);
INSERT INTO TransactionDetail VALUES (11, 4, NULL, '1000');
INSERT INTO TransactionDetail VALUES (12, 5, '2400', NULL);
INSERT INTO TransactionDetail VALUES (13, 5, NULL, '1200');
INSERT INTO TransactionDetail VALUES (14, 5, NULL, '1000');
INSERT INTO TransactionDetail VALUES (15, 5, NULL, '200');
INSERT INTO TransactionDetail VALUES (16, 6, '2900', NULL);
INSERT INTO TransactionDetail VALUES (17, 6, NULL, '2900');
INSERT INTO TransactionDetail VALUES (18, 7, '700', NULL);
INSERT INTO TransactionDetail VALUES (19, 7, '300', NULL);
INSERT INTO TransactionDetail VALUES (20, 7, '2100', NULL);
INSERT INTO TransactionDetail VALUES (21, 7, NULL, '3100');
INSERT INTO TransactionDetail VALUES (22, 8, '500', NULL);
INSERT INTO TransactionDetail VALUES (23, 8, NULL, '500');
这是查询
with data1 as (
select a.id inid,a.VoucherTypeCode,PayeeName,MAX(c.DrAmount) InAmount,TransactionDate,RefNo,FullName from TransactionMaster a
inner join [VoucherType] b on a.VoucherTypeCode = b.VoucherTypeCode
inner join TransactionDetail c on a.ID = c.TransactionCode
where a.VoucherTypeCode in (1,3)
GROUP BY a.id,a.VoucherTypeCode,PayeeName,TransactionDate,RefNo,FullName
),
data2 as (
select a.id outid,a.VoucherTypeCode,PayeeName,MAX(c.CrAmount) OutAmount,TransactionDate,RefNo,FullName from TransactionMaster a
inner join [VoucherType] b on a.VoucherTypeCode = b.VoucherTypeCode
inner join TransactionDetail c on a.ID = c.TransactionCode
where a.VoucherTypeCode in (2,4)
GROUP BY a.id,a.VoucherTypeCode,PayeeName,TransactionDate,RefNo,FullName
)
select *,COALESCE(a.TransactionDate,b.TransactionDate) as FullDate from data1 a full join data2 b on inid = outid and a.TransactionDate = b.TransactionDate
--WHERE COALESCE(a.TransactionDate,b.TransactionDate) BETWEEN '2018-03-23 00:00:00.000' AND '2018-03-24 00:00:00.000'
order by FullDate
从日期检查中删除评论时，下面提供了预期的输出：
inid VoucherTypeCode PayeeName InAmount TransactionDate RefNo FullName outid VoucherTypeCode PayeeName OutAmount TransactionDate RefNo FullName FullDate Opening
-------------------- --------------- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- --------------------------------------- ----------------------- -------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -------------------- --------------- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- --------------------------------------- ----------------------- -------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -----------------------
3 1 Erick 1200 2018-03-23 00:00:00.000 CPV-5435-ISL Cash Payment Voucher NULL NULL NULL NULL NULL NULL NULL 2018-03-23 00:00:00.000 -50
4 3 Asim 1000 2018-03-24 00:00:00.000 BPV-2345-CAN Bank Payment Voucher NULL NULL NULL NULL NULL NULL NULL 2018-03-24 00:00:00.000 1150
公式为(Opening+InAmount) - Out Amount将是Closing Balance，而Opening将是前一行Closing Balance。
因此，对于第一个记录，打开将为0，关闭将为3100，而对于第二个打开= 3100，将关闭3600，依此类推。
如果VoucherType为1或3，则我必须获得贷方金额CrAmount；如果VoucherType为2或4，则我需要获取借方金额DrAmount。
我之前曾问过这个问题，并且也获得了预期的输出，但是现在数据库结构有所改变，我在这里无法使用该逻辑。
答案 0 :(得分：0)
基本上，它与我在您的其他线程中发布的上一个查询相同。 由于其他表TransactionDetail中的详细信息，您只需要从TransactionMaster对其进行INNER JOIN并执行SUM（）。我没有验证Dr-Cr或Cr-Dr逻辑。请验证自己。
SELECT t.*, v.FullName, o.Opening
FROM [TransactionMaster] t
INNER JOIN [VoucherType] v on t.VoucherTypeCode = v.VoucherTypeCode
OUTER APPLY
(
SELECT Opening = sum(case when m.[VoucherTypeCode] in (1, 3)
then - ISNULL(d.CrAmount, 0)
else + ISNULL(d.CrAmount, 0)
end)
FROM [TransactionMaster] m
INNER JOIN [TransactionDetail] d ON m.ID = d.TransactionCode
WHERE m.TransactionDate < t.TransactionDate
) o
WHERE t.TransactionDate BETWEEN '2018-03-23' AND '2018-03-24'
order by t.TransactionDate
编辑： 您不需要使用FULL JOIN来标识IN和OUT。只需使用以下CASE语句
SELECT inID = case when t.VoucherTypeCode in (1,3) then t.ID end,
inAmount = case when t.VoucherTypeCode in (1,3) then a.Amount end,
outID = case when t.VoucherTypeCode in (2,4) then t.ID end,
outAmount = case when t.VoucherTypeCode in (2,4) then a.Amount end,
t.PayeeName, t.TransactionDate, t.RefNo,
v.FullName, Opening = isnull(o.Opening, 0)
FROM [TransactionMaster] t
INNER JOIN [VoucherType] v on t.VoucherTypeCode = v.VoucherTypeCode
CROSS APPLY
(
SELECT Amount = sum(case when m.[VoucherTypeCode] in (1, 3)
then -CrAmount
else DrAmount
end)
FROM [TransactionMaster] m
INNER JOIN [TransactionDetail] d ON m.ID = d.TransactionCode
WHERE m.ID = t.ID
) a
OUTER APPLY
(
SELECT Opening = sum(case when m.[VoucherTypeCode] in (1, 3)
then -CrAmount
else DrAmount
end)
FROM [TransactionMaster] m
INNER JOIN [TransactionDetail] d ON m.ID = d.TransactionCode
WHERE m.TransactionDate < t.TransactionDate
) o
order by t.TransactionDate