如何在mysql中合并多行数据

Question

我正在尝试将 MySQL 中的数据合并到一个视图或存储过程中，以便写回新表或作为视图访问。我的数据库将客户元数据存储在带有键和值的单独行中。我对如何以有用的形式提取这些数据感到困惑。

我的数据是这样形成的：

 id order       customer    type        key               value
1   42FF86A1    45858007    shipping    address-name      David Customer
2   42FF86A1    45858007    shipping    email-address     david@email.com
3   42FF86A1    45858007    shipping    number            2125551212
4   42FF86A1    45858007    shipping    address-line1     5353 My Street
5   42FF86A1    45858007    shipping    address-line2     #2
6   42FF86A1    45858007    shipping    city              MyCity
7   42FF86A1    45858007    shipping    region            CA
8   42FF86A1    45858007    shipping    postal-code       95555
9   42FF86A1    45858007    shipping    country           US

最终，我希望能够读取数据并将其轻松导出，以便在我的 CRM 或 Excel 中使用。

我已经尝试实现https://stackoverflow.com/a/5967720/3001841，但它对我来说没有意义。

仅在一行中获取数据对我很有用，其中：

客户、订单、地址名称、地址行1、地址行2、城市、地区、邮政编码、国家、电子邮件地址、号码

Answer 1

为了方便起见，这是一个使用视图的示例，但您也可以根据需要将视图查询烘焙到最终查询中。 （我使用了invoice，而你使用了order，对不起。顺便说一句，尽量避免使用mysql保留字作为字段名，记住要反引号会很痛苦。）

create view order_pivot as (
  select customer,  invoice, type, 
  case when `key` = 'address-name' then `value` end as address,
  case when `key` = 'email-address' then `value` end as email,
  case when `key` = 'number' then `value` end as number,
  case when `key` = 'address-line1' then `value` end as addressline1,
  case when `key` = 'address-line2' then `value` end as addressline2,
  case when `key` = 'city' then `value` end as city,
  case when `key` = 'region' then `value` end as region,
  case when `key` = 'postal-code' then `value` end as postalcode,
  case when `key` = 'country' then `value` end as country
  from orders
);

视图将数据转换为列数与我们感兴趣的列数相匹配的行。它不是很灵活，因为它需要您对这些关键字段进行硬编码，但它确实有效。它为我们提供了一个值，以及每一行的一大堆空值 - 所以显然我们需要将它们全部合并到一行中，这就是下一个查询的来源。

select customer, max(address) addr, max(email) email, max(number) number, max(addressline1) a1, max(addressline2) a2, max(city) city, max(region) region, max(postalcode) postcode, max(country) country
  from order_pivot
  group by customer;

我们按客户分组（您可能希望在这里也将where 过滤器放在invoice 上，然后我们可以使用max() 确保我们忽略所有空值，并只抓取具有有效数据的字段。

here's a fiddle

编辑

组合这两个查询，因为您似乎缺乏查看权限：

select customer, max(address) addr, max(email) email, max(number) number, max(addressline1) a1, max(addressline2) a2, max(city) city, max(region) region, max(postalcode) postcode, max(country) country
  from (
  select customer,  invoice, type, 
  case when `key` = 'address-name' then `value` end as address,
  case when `key` = 'email-address' then `value` end as email,
  case when `key` = 'number' then `value` end as number,
  case when `key` = 'address-line1' then `value` end as addressline1,
  case when `key` = 'address-line2' then `value` end as addressline2,
  case when `key` = 'city' then `value` end as city,
  case when `key` = 'region' then `value` end as region,
  case when `key` = 'postal-code' then `value` end as postalcode,
  case when `key` = 'country' then `value` end as country
  from orders
) q
  group by customer;

updated fiddle with combined query