Mysql 两个子查询之间的减法

我正在尝试减去两个子查询。下面列出了这两个表

COM_PAYMENTS

+++++++++++++++++++
tender_id | amount
+++++++++++++++++++
   10     | 200
   10     | 345
   22     |  18
   25     |  45

COM_RECEIPTS

+++++++++++++++++++
tender_id | amount
+++++++++++++++++++
   10     | 10
   10     | 15
   22     | 13

我试图在查询中实现的是，首先我需要找到COM_付款和COM_收据的投标ID金额之和，并且COM_付款总额需要从COM_收据总额中减去

Example: The out put in this case should be:

tender_id | Total

    10      520 (200 + 345 = 545, 10 + 15 = 25, 545 - 25= 520)
    22      5   (18 - 13)
    25      45 ( Since COM_PAYMENTS doesnt have any recrds)

这是我尝试过的，但如果COM_收据中有数据，则只显示总数，而其他数据则不减去显示空白：

(select
SUM(com_payments.amount)
FROM com_payments
WHERE view_sales_report.tender_id = com_payments.tender_id) -
(select
SUM(com_receipts.rec_amt)
FROM com_receipts
WHERE view_sales_report.tender_id = com_receipts.tender_id)

---

也许你需要这样的东西：

SELECT tender_id, SUM(amount) AS amount
FROM (
    SELECT tender_id, SUM(amount) AS amount
    FROM COM_PAYMENTS
    GROUP BY tender_id
    UNION ALL
    SELECT tender_id, SUM(amount * -1) AS amount
    FROM COM_RECEIPTS
    GROUP BY tender_id
  ) a
GROUP BY tender_id  

---

向前看，我建议您只有一个表，它记录所有交易——付款和收款——有一个交易id和一列记录交易类型。

我认为这里有太多的查询。不起作用：想象一下，在com_收据中有一行投标id=10；金额=-200怎么可能不呢？我们不知道表格是如何填充的是的，我们知道，有了这些，我们只知道表格应该如何填充…@ipiero我认为这在任何情况下都很好。@草莓是的，工会是我的目标；）我同意在缺少主键的情况下，这实际上不是RDBMS意义上的表。不过，我认为我们必须考虑到有人可能会两次提供相同金额的投标。

SELECT SUM(DISTINCT pa.amount)-IFNULL(SUM(DISTINCT re.amount) , 0) FROM com_payments pa
LEFT OUTER JOIN com_receipts re ON pa.`tender_id` = re.`tender_id` GROUP BY pa.tender_id

SELECT SUM(DISTINCT pa.amount)-IFNULL(SUM(DISTINCT re.amount) , 0) FROM com_payments pa
LEFT OUTER JOIN com_receipts re ON pa.`tender_id` = re.`tender_id` GROUP BY pa.tender_id