在 MySQL 中计算总计的增值税/税 - 增值税率取决于位置字段

Question

我正在尝试在 MySQL 中创建一个报告（从 Wordpress/Woocommerce 安装中的表格，但是这家商店的纳税方式与 Woocommerce 的做事方式不同，所以我不能使用 Woocommerce 报告。）

计算需要从总额中计算出 20%（英国增值税）的税率 - 但前提是客户位于特定位置。因此，如果客户在“世界”中，税率为 0%。如果客户在英国，税率为 20%。但总收费保持不变（因此出口利润比本地销售多！）

我从：

SELECT TRIM(LEADING 'Order –' FROM post_title) AS Date,
m1.meta_value AS _order_number,
m2.meta_value AS _Location,
m3.meta_value AS _order_total
FROM wp_posts 
LEFT JOIN wp_postmeta AS m1 ON m1.post_id = wp_posts.ID AND m1.meta_key = '_order_number'
LEFT JOIN wp_postmeta AS m2 ON m2.post_id = wp_posts.ID AND m2.meta_key = '_shipping_method_title'
LEFT JOIN wp_postmeta AS m3 ON m3.post_id = wp_posts.ID AND m3.meta_key = '_order_total'
WHERE wp_posts.post_type = 'shop_order'

这给了我这样一张桌子：

Date     |  Order no  | Location    |  Order Total

May 2013 |  123       |World         |1124.00

Jan 2013 |  124       |UK            |163.00

到目前为止一切顺利。因此，我尝试添加一些 SUM 来计算增值税金额。

SELECT TRIM(LEADING 'Order –' FROM post_title) AS Date,
m1.meta_value AS _order_number,
m2.meta_value AS _Location,
m3.meta_value AS _order_total,
SUM(m3.meta_value*20/120) AS _VAT,
SUM(m3.meta_value*100/120) AS _NET
FROM wp_posts 
LEFT JOIN wp_postmeta AS m1 ON m1.post_id = wp_posts.ID AND m1.meta_key = '_order_number'
LEFT JOIN wp_postmeta AS m2 ON m2.post_id = wp_posts.ID AND m2.meta_key = '_shipping_method_title'
LEFT JOIN wp_postmeta AS m3 ON m3.post_id = wp_posts.ID AND m3.meta_key = '_order_total'
WHERE wp_posts.post_type = 'shop_order'

我以为这会给我：

Date       |     Order no | Location  |    Order Total | VAT  |   NET

May 2013  |             123|   World |    1124.00  |    187.33 |   1311.33

Jan 2013  |         124  |    UK     |      163.00   |    27.17 |  190.17

然后我可以继续尝试找出如何按位置更改费率。 不幸的是，它似乎对所有内容求和，而不是逐行给我结果。

1) 我把所有东西都加在一起了，我搞砸了什么？

2) 有关如何仅在 _shipping_method_title=UK 时最好地应用增值税计算的任何想法？

3) 显然，这仅在税率发生变化之前有效。处理该问题的最佳选择是仅在下次修改税率时限制此报告的日期，并为适用新税率的未来订单制作新报告？还是有更聪明的办法？

我很欣赏向客户收取统一费率并支付可变税的做法有点奇怪，但我无法改变这一点，我只是负责提供报告。 :-(

编辑 感谢 Gordon，我已经解决了我的 SUM 问题，现在我的查询看起来像这样并产生了我期望的表格布局，如上所示：

SELECT TRIM(LEADING 'Order –' FROM post_title) AS Date,
m1.meta_value AS _order_number,
m2.meta_value AS _Location,
m3.meta_value AS _order_total,
SUM(m3.meta_value*20/120) AS _VAT,
SUM(m3.meta_value*100/120) AS _NET
FROM wp_posts 
LEFT JOIN wp_postmeta AS m1 ON m1.post_id = wp_posts.ID AND m1.meta_key = '_order_number'
LEFT JOIN wp_postmeta AS m2 ON m2.post_id = wp_posts.ID AND m2.meta_key = '_shipping_method_title'
LEFT JOIN wp_postmeta AS m3 ON m3.post_id = wp_posts.ID AND m3.meta_key = '_order_total'
WHERE wp_posts.post_type = 'shop_order'
GROUP BY wp_posts.ID

但我仍然不确定如何根据位置字段中的值更改乘数。这可能吗，还是我在这里吠错了树？

编辑 2

我已经制定了一个部分解决方案 - 我使用 CASE 代替 SUM 行来为我提供增值税的计算值。

CASE 
        when m2.meta_value = 'World Delivery' then 0
        when m2.meta_value = 'UK Delivery' then  (m3.meta_value*20/120)
        when m2.meta_value = 'European Delivery' then  (m3.meta_value*20/120)
END AS _VAT    

不幸的是，我不能只为 _NET 值重复 CASE 语句（价格 - 增值税，因为它与增值税的 CASE 相同，因此会产生错误。

我想做类似的事情

 SUM(_order_total-_VAT) AS _NET

但 *_order_total* 和 *_VAT* 不被识别为可以计算的列。

最终编辑 再次感谢 Gordon，我的最终查询看起来像这样，我只是添加它以防其他人发现它有帮助。此查询与 Wordpress 导出和报告 wordpress 插件一起使用，因此可以快速集成到 Wordpress 管理中并导出到 Excel，这很漂亮。

SELECT TRIM(LEADING 'Order –' FROM post_title) AS Date,
m1.meta_value AS _order_number,
m2.meta_value AS _Customer_Location,
m3.meta_value AS _order_total,

ROUND(SUM(case when m2.meta_value = 'World Delivery' then 0 else m3.meta_value*20/120 end),2) AS _VAT,
ROUND(SUM(case when m2.meta_value = 'World Delivery' then m3.meta_value else m3.meta_value*100/120 end),2) AS _NET
FROM wp_posts 
LEFT JOIN wp_postmeta AS m1 ON m1.post_id = wp_posts.ID AND m1.meta_key = '_order_number'
LEFT JOIN wp_postmeta AS m2 ON m2.post_id = wp_posts.ID AND m2.meta_key = '_shipping_method_title'
LEFT JOIN wp_postmeta AS m3 ON m3.post_id = wp_posts.ID AND m3.meta_key = '_order_total'
WHERE wp_posts.post_type = 'shop_order' AND wp_comments.comment_post_id = wp_posts.ID  AND wp_comments.comment_content="Order status changed from processing to completed."
GROUP BY wp_posts.ID

Answer 1

你只得到一行的原因是sum() 函数使它成为一个聚合查询。但是，您没有group by。我认为按 postid 分组修复了这个问题，以及对 select 的一些其他调整：

SELECT TRIM(LEADING 'Order –' FROM post_title) AS Date,
max(m1.meta_value) AS _order_number,
max(m2.meta_value) AS _Customer_Location,
max(m3.meta_value) AS _order_total,
SUM(m3.meta_value*20/120) AS _VAT,
SUM(m3.meta_value*100/120) AS _NET
FROM wp_posts 
LEFT JOIN wp_postmeta AS m1 ON m1.post_id = wp_posts.ID AND m1.meta_key = '_order_number'
LEFT JOIN wp_postmeta AS m2 ON m2.post_id = wp_posts.ID AND m2.meta_key = '_shipping_method_title'
LEFT JOIN wp_postmeta AS m3 ON m3.post_id = wp_posts.ID AND m3.meta_key = '_order_total'
WHERE wp_posts.post_type = 'shop_order'

这种类型的数据很难习惯。一条记录的所有数据都拆分为多行。

要根据位置更改乘数，请使用where 子句。这是一个例子：

SUM(case when location = 'UK' then 0 else m3.meta_value*20/120 end) AS _VAT,
SUM(case when location = 'UK' then m3.meta_value else m3.meta_value*100/120 end) AS _NET