Question

我有以下内容，但由于GROUP BY未添加群组用户ID的总收入，因此无法获得正确的总计。

SELECT
  users.id,
  Count(orders.id) AS total_orders,
  SUM(item_orders.item_price * item_orders.quantity) AS total_rev
FROM
  orders 
LEFT JOIN
  users ON orders.user_id = users.id 
LEFT JOIN
  item_orders ON item_orders.order_id = orders.id
WHERE 
  orders.date >= '2015-10-06 00:00:00' AND orders.date <= '2016-03-23 23:59:59'
GROUP BY
  users.id  -- ignores duplicate ids, doesn't sum total_rev for those
ORDER BY
  total_rev DESC 
LIMIT 0,25

我想使用WITH ROLLUP来解决这个问题，但是当我使用WITH ROLLUP时，我无法使用ORDER BY而我无法使用别名GROUP BY订购包含我的每个用户ID总数。

SELECT
  users.id,
  Count(orders.id) AS total_orders,
  SUM(item_orders.item_price * item_orders.quantity) AS total_rev
FROM
  orders 
LEFT JOIN
  users ON orders.user_id = users.id 
LEFT JOIN
  item_orders ON item_orders.order_id = orders.id
WHERE 
  orders.date >= '2015-10-06 00:00:00' AND orders.date <= '2016-03-23 23:59:59'
GROUP BY
  users.id, total_rev DESC WITH ROLLUP -- does not work because total_rev is an alias!
LIMIT 0,25

有关如何使其发挥作用的任何建议？我在DB中这样做有意义吗？我试图通过让数据库预先分配来节省资源，这样我就可以在我的应用程序中进行分页。

编辑：正如下面评论中提到的xjstratedgebx，第一个查询正常，没有理由使用WITH ROLLUP。第一个查询也可以缩短：

SELECT
  orders.user_id,
  Count(orders.id) AS total_orders,
  SUM(item_orders.item_price * item_orders.quantity) AS total_rev
FROM
  orders 
LEFT JOIN
  item_orders ON item_orders.order_id = orders.id
WHERE 
  orders.date >= '2015-10-06 00:00:00' AND orders.date <= '2016-03-23 23:59:59'
GROUP BY
  orders.user_id
ORDER BY
  total_rev DESC 
LIMIT 0,25

获取订单总计并使用1个查询对其进行排序

0 个答案: