Question

DECLARE @run_date DATETIME

SELECT @run_date = '2016/05/30'

SELECT PAYMENT_LINE.active,
       PAYMENT_LINE.alloc_date,
       PAYMENT_LINE.allocation_ind,
       PAYMENT_LINE.amount,
       PAYMENT_LINE.fund_charge_id,
       PAYMENT_LINE.paid_td,
       PAYMENT_LINE.payment_id,
       PAYMENT_LINE.payment_line_id,
       PAYMENT_LINE.payment_type_code,
       PAYMENT_LINE.period,
       PAYMENT_LINE.reference_id,
       PAYMENT_LINE.reversal_date,
       PAYMENT_LINE.src_id

       INTO #pym
FROM   PAYMENT_LINE
WHERE  alloc_date IS NOT NULL
       AND alloc_date < @run_date

/*DELETE FROM #pym WHERE reversal_date IS NULL*/



DELETE 
FROM   #pym
WHERE  reversal_date > @run_date



SELECT payment_id,
       Total = SUM(amount)

       INTO #pym1
FROM   #pym
GROUP BY
       payment_id



SELECT PAYMENT.payment_id,
       PAYMENT.payment_amount,
       Total

       INTO #pym2
FROM   #pym1,
       PAYMENT
WHERE  PAYMENT.payment_id = #pym1.payment_id
GROUP BY
       PAYMENT.payment_id
HAVING Total > PAYMENT.payment_amount



SELECT MANUFACTURER.manufacturer_id,
       MANUFACTURER.manufacturer_name,
       alloc_date,
       PAYMENT.payment_id,
       PAYMENT.payment_amount,
       Total

       INTO #pym3
FROM   #pym2,
       PAYMENT,
       MANUFACTURER,
       PAYMENT_LINE
WHERE  Total > PAYMENT.payment_amount



SELECT payment_id,
       Total_payment = SUM(payment_amount - Total)

       INTO #pym4
FROM   #pym3
GROUP BY
       payment_id



SELECT MANUFACTURER.manufacturer_id,
       MANUFACTURER.manufacturer_name,
       alloc_date,
       PAYMENT.payment_id,
       PAYMENT.payment_amount,
       Total,
       Total_payment
FROM   #pym4,
       #pym2,
       PAYMENT,
       MANUFACTURER,
       PAYMENT_LINE

Answer 1

我已经看了一下这个查询，如果没有更多的信息，它真的没有多大意义。前三个查询;

SELECT PAYMENT_LINE.active,
   PAYMENT_LINE.alloc_date,
   PAYMENT_LINE.allocation_ind,
   PAYMENT_LINE.amount,
   PAYMENT_LINE.fund_charge_id,
   PAYMENT_LINE.paid_td,
   PAYMENT_LINE.payment_id,
   PAYMENT_LINE.payment_line_id,
   PAYMENT_LINE.payment_type_code,
   PAYMENT_LINE.period,
   PAYMENT_LINE.reference_id,
   PAYMENT_LINE.reversal_date,
   PAYMENT_LINE.src_id

   INTO #pym
FROM   PAYMENT_LINE
WHERE  alloc_date IS NOT NULL
   AND alloc_date < @run_date

DELETE 
FROM   #pym
WHERE  reversal_date > @run_date



SELECT payment_id,
   Total = SUM(amount)

   INTO #pym1
FROM   #pym
GROUP BY
   payment_id

可以全部合并到此;

SELECT payment_id,
       Total = SUM(amount)
       INTO #pym
FROM   PAYMENT_LINE
WHERE  alloc_date IS NOT NULL
       AND alloc_date < @run_date
       AND reversal_date <= @run_date
GROUP BY payment_id

你真的需要阅读正确的连接语法，你在这里使用了非常古老的语法;

SELECT PAYMENT.payment_id,
   PAYMENT.payment_amount,
   Total

   INTO #pym2
FROM   #pym1,
   PAYMENT
WHERE  PAYMENT.payment_id = #pym1.payment_id
GROUP BY
   PAYMENT.payment_id
HAVING Total > PAYMENT.payment_amount

请阅读Aaron Bertrand的文章

Bad habits to kick : using old-style JOINs

你还需要注意笛卡尔连接及其所需要的内容，你的最终查询会以多种方式搞乱你的数据;

SELECT MANUFACTURER.manufacturer_id,
   MANUFACTURER.manufacturer_name,
   alloc_date,
   PAYMENT.payment_id,
   PAYMENT.payment_amount,
   Total,
   Total_payment
FROM   #pym4,
   #pym2,
   PAYMENT,
   MANUFACTURER,
   PAYMENT_LINE

您还应该阅读使用表别名，这些将使您的角色在长期内变得更加简单;

When to use SQL Table Alias

我认为你需要重温你正在尝试做的事情，并问自己你所使用的临时表逻辑是否必要。对我来说，看起来你的数据将会变得非常混乱。考虑是否可以将所有这些都放入一个查询中，每个部分都是您加入的子选择，以便检索所需的数据;

http://www.techrepublic.com/article/use-sql-subselects-to-consolidate-queries/

这个存储过程需要很长时间才能运行？

1 个答案: