postgresql - 分析将文件上传到 postgreSQL 中的 API 时产生的错误

Question

我正在尝试分析当我们的客户尝试将他们的文件上传到我们的平板表单时产生的错误，不确定是否可以使用 PostgreSQL。

我的表有三列company_id、import_id 和operation_errors。

我尝试了这个查询，它在有一个错误的行上工作得很好，但在大多数情况下，我每次导入有 3 个以上的错误，我的查询只选择第一个错误，当有更多错误时。

SELECT importable_id ,
       importable_type ,
       substring(lower(replace_operation)
                 FROM 'title:(.+?)detail:') test
FROM
  ( SELECT import_id ,
           importable_id ,
           i.importable_type ,
           operation_errors  as replace_operation
   FROM import_results ir
   JOIN imports i ON ir.import_id = i.id
   WHERE operation_errors IS NOT NULL
     AND i.created_at >= date_trunc('month', CURRENT_DATE :: date)) a
WHERE importable_type = 'Company';

正如我所说，该行可能有一个或多个错误，这里是错误消息的一个示例

      ---
      - !ruby/hash:ActiveSupport::xxxxxxxxxxxxxxxx
        code: 314
      title: xxxxxxxxxxxxxxxxxxxxxxxx.
        detail: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
        xxxx. i.e 21 February, 2018.
     source: !ruby/hash:ActiveSupport::xxxxxxxxxxxxxxxxxxxxxxxxx
      pointer: "/data/attributes/xxxxxxx"
      - !ruby/hash:ActiveSupport::xxxxxxxxxxxxxxxxxxxxxxxxxxx
       code: 343
        title: xxxxxx xxxxx xxxxxxx xxxxxxxx xxxx
     detail: xxxxxx xxxxxxx xxxxxx xxxxxxxx xxxxxxxxxxxxx
        xxxxxx xxxxxxxx xxxxx xxxxxxx xxxx xxxxxxxx xxxxxxxxxxx.
     source: !ruby/hash:ActiveSupport::xxxxxxxxxxxxxxxxx
     pointer: "/data/attributes/xxxxxxxxxxx"
    - !ruby/hash:ActiveSupport::             
     code: 342
      title: xxxxx xxxxx xxxxx xxxxxxx xxxxx xxxxxxxxxx xxxxxxxxx xxx
    detail: xxxxxx xx xxx xxx xxx xxx xxxx xxxxxxx xxxxx
    xxxxxxxxx     x xxxxxxx xxxxxxx xx xxxxxx xx xx.
      source: !ruby/hash:ActiveSupport::xxxxxxxxxxx
     pointer: "/data/attributes/xxxxxxxxxx"

我希望有这样的东西

      importable_id errore1         errore 2        errore 3
       1           title: xxxxx.    title: xxxxx.   title: xxxxx.
       2           title: xxxxx.        
       3           itle: xxxxx.      title: xxxxx.  

我也试过

       split_part(replace_operation, ' ', 1) AS col1

但我最终得到了 40 列，但仍然没有得到某些行中的所有错误

Answer 1

感谢@Andyk，我设法找到了问题的答案

WITH errors AS
  (SELECT importable_id ,
          importable_type,
          created_at ,
          substring(lower(replace_operation)
                    FROM 'code:(.+?)title') code_1 ,
          split_part(replace_operation, 'code', 2) AS code_2 ,
          split_part(replace_operation, 'title', 2) AS sec_part ,
          split_part(split_part(replace_operation, 'title', 2), ':', 2) error_1 ,
          split_part(replace_operation, 'title', 3) AS third_part ,
          split_part(split_part(replace_operation, 'title', 3), ':', 2) error_2 ,
          split_part(replace_operation, 'title', 4) AS f_part ,
          split_part(split_part(replace_operation, 'title', 4), ':', 2) error_3 ,
          split_part(replace_operation, 'title', 5) AS f_part ,
          split_part(split_part(replace_operation, 'title', 5), ':', 2) error_4
   FROM
     ( SELECT import_id ,
              importable_id ,
              i.importable_type ,
              operation_errors AS replace_operation,
              i.created_at
      FROM import_results ir
      JOIN imports i ON ir.import_id = i.id
      WHERE operation_errors IS NOT NULL
        AND i.created_at >= date_trunc('month', CURRENT_DATE :: date)) a
   WHERE importable_type = 'Company')
SELECT importable_id,
       created_at::date ,
       code_1 ,
       error_1 ,
       error_2 ,
       error_3 ,
       error_4
FROM errors