SQL、发票报表、当前和过去的会费

Question

SQL 新手，了解基本查询，尝试编写比我的知识更高级的东西，如果有人可以提供，需要一些帮助。我已经尝试过模板代码并根据我的需要进行定制，但没有成功。

我正在使用 1 个表，并尝试让查询显示客户当前余额 0-30 天，以及 30-60 、 60-90 、 90-120 、 120+ 天之间的任何过期日期。

table: c_inv_h
columns: comp_code, cust_code, inv_num, inv_date, inv_tp, inv_amt, inv_tax1, inv_tax2

示例数据

comp_code, cust_code, inv_num, inv_date, inv_tp, inv_amt, inv_tax1, inv_tax2
M1, 100068, 1, 2016-08-11, RCPT, 103.48, 4.5, 8.98
M1, 100068, 2, 2016-07-11, RCPT, 103.48, 4.5, 8.98
M1, 100068, 3, 2016-06-11, RCPT, 103.48, 4.5, 8.98
M1, 200027, 4, 2016-05-11, RCPT, 103.48, 4.5, 8.98
M1, 200027, 5, 2016-04-11, RCPT, 103.48, 4.5, 8.98

我希望查询显示的内容：

comp_code, cust_code, CurrentBalance, 30-60, 60-90, 90-120, 120+
M1, 100068, 116.96, 233.92, 0, 0, 0
M1, 200027, 0, 233.92, 0, 0, 0
M1, 200027, 0, 0, 0, 0, 233.92

到目前为止我已经尝试过什么

SELECT comp_code, cust_code, inv_num, inv_date, sum(inv_amt) + sum(inv_tax1)
+ sum(inv_tax2) as InvoiceTotal,
DATEDIFF(day, inv_date, sysdate as Numberofdays, 'CurrentBalance' = 
CASE WHEN DATEDIFF(day,inv_date, sysdate < 30 then InvoiceTotal END,
'30_days' = CASE WHEN DATEDIFF(day, inv_date, sysdate between 30 AND 60
Then InvoiceTotal WHEN DATEDIFF(day, inv_date, sysdate > 90 Then
InvoiceTotal END
FROM c_inv_h
WHERE comp_code='M1'
GROUP BY comp_code, cust_code

这可能与我想要实现的目标相去甚远，但我正在努力解决它。

提前非常感谢！！

Answer 1

您对条件聚合的思考是正确的。您只需删除未参与聚合的列并清理语法错误。

SELECT
    comp_code
    ,cust_code
    ,SUM(CASE
          WHEN DATEDIFF(CURDATE(),inv_date) < 30
          THEN COALESCE(inv_amt,0.00) + COALESCE(inv_tax1,0.00) + COALESCE(inv_tax2,0.00)
          ELSE 0
        END
    ) AS CurrentBalance
    ,SUM(CASE
          WHEN DATEDIFF(CURDATE(),inv_date) >= 30 AND DATEDIFF(CURDATE(),inv_date) < 60
          THEN COALESCE(inv_amt,0.00) + COALESCE(inv_tax1,0.00) + COALESCE(inv_tax2,0.00)
          ELSE 0
        END
    ) AS "30-60"
    ,SUM(CASE
          WHEN DATEDIFF(CURDATE(),inv_date) >= 60 AND DATEDIFF(CURDATE(),inv_date) < 90
          THEN COALESCE(inv_amt,0.00) + COALESCE(inv_tax1,0.00) + COALESCE(inv_tax2,0.00)
          ELSE 0
        END
    ) AS "60-90"
    ,SUM(CASE
          WHEN DATEDIFF(CURDATE(),inv_date) >= 90 AND DATEDIFF(CURDATE(),inv_date) < 120
          THEN COALESCE(inv_amt,0.00) + COALESCE(inv_tax1,0.00) + COALESCE(inv_tax2,0.00)
          ELSE 0
        END
    ) AS "90-120"
    ,SUM(CASE
          WHEN DATEDIFF(CURDATE(),inv_date) >= 120
          THEN COALESCE(inv_amt,0.00) + COALESCE(inv_tax1,0.00) + COALESCE(inv_tax2,0.00)
          ELSE 0
        END
    ) AS "120+"
FROM
    c_inv_h
WHERE
    comp_code='M1'
GROUP BY
    comp_code, cust_code
;

我正在使用 COALESCE 来确保存在一个值，因为在添加列时，如果添加 1 + NULL + 1，结果将为 NULL。当使用聚合函数时，例如SUM() 它将忽略 NULL 值。因为您根据您的数据集同时使用这两种方法，所以您可能需要将 0 分配给 COALESCE 提供帮助的 NULL。

接下来我不使用 BETWEEN，因为 BETWEEN 包含比较的双方，所以如果您说一列介于 30 和 60 之间，另一列介于 60 和 90 之间，并且金额是逾期 60 天，它将在两列。

最后一个示例输出与您提供的示例数据不正确，因为日期分别为 5、36、66、97 和 127 天，但我很确定我明白您想要什么。

希望对您有所帮助。