如何从 MySQL JSON 中删除多个值

我需要从 MySQL 中的 JSON 数组中删除多个值

我已经在每个值上尝试过 JSON_REMOVE 和 JSON_SEARCH 但删除元素后 INDEX 发生了变化

JSON 数组：

["1", "5", "18", "68"]

例如：删除“5”和“68”

预期结果 ：

["1", "18"]

EDITED:

我尝试过的代码：

JSON_REMOVE(`can_see`, JSON_UNQUOTE(JSON_SEARCH(`can_see`, 'one', '5')), JSON_UNQUOTE(JSON_SEARCH(`can_see`, 'one', '68')))

查询结果：

["1", "18", "68"]

删除索引 $[0] 处的“5”后，“68”索引更改为 $[2] 但 JSON_SEARCH 从原始 json 返回 $[3]

我也尝试过嵌套：

JSON_REMOVE(JSON_REMOVE(`can_see`, JSON_UNQUOTE(JSON_SEARCH(`can_see`, 'one', '5'))), JSON_UNQUOTE(JSON_SEARCH(JSON_REMOVE(`can_see`, JSON_UNQUOTE(JSON_SEARCH(`can_see`, 'one', '5'))), 'one', '68')))

这可行，但如果我想删除超过 2 个值，它会变得混乱

我需要确保数组已排序

这不是最容易理解的方法，但我认为您必须使用 SQL 数字生成器将 json 数组解析为标记（记录），您可以更轻松地过滤和排序。

Query

SELECT 
  JSON_ARRAYAGG(
     JSON_EXTRACT(records.json, CONCAT('$[', number_generator.number , ']'))
  ) AS json                                    
FROM (

  SELECT 
   @row := @row + 1 AS number
  FROM (
    SELECT 0 UNION SELECT 1 UNION SELECT 2 UNION SELECT 3 UNION SELECT 4 UNION SELECT 5 UNION   SELECT 6 UNION SELECT 7 UNION SELECT 8 UNION SELECT 9
  ) row1
  CROSS JOIN (
    SELECT 0 UNION SELECT 1 UNION SELECT 2 UNION SELECT 3 UNION SELECT 4 UNION SELECT 5 UNION  SELECT 6 UNION SELECT 7 UNION SELECT 8 UNION SELECT 9
  ) row2
  CROSS JOIN (
    SELECT @row := -1 
  ) init_user_params 
) AS number_generator
CROSS JOIN (


SELECT 
    json
  , JSON_LENGTH(records.json) AS json_array_length    
FROM (

  SELECT 
   '["1", "5", "18", "68"]' AS json
  FROM 
   DUAL  
) AS records

) AS records 
WHERE
    number BETWEEN 0 AND  json_array_length - 1   
  AND
    JSON_EXTRACT(records.json, CONCAT('$[', number_generator.number , ']')) NOT IN(5, 68)                   
ORDER BY 
 REPLACE(JSON_EXTRACT(records.json, CONCAT('$[', number_generator.number , ']')), '"', '')                       

Result

| json        |
| ----------- |
| ["1", "18"] |

see demo

另一方面，MySQL 8+ 让它变得更加容易

Query

SELECT 
 JSON_ARRAYAGG (
    result_table.item
 ) AS json
FROM JSON_TABLE(
     '["1", "5", "18", "68"]'
   , "$[*]"

   COLUMNS (
       rowid FOR ORDINALITY
     , item VARCHAR(100) PATH "$"   
   )
) AS result_table
WHERE
 CAST(result_table.item AS UNSIGNED) NOT IN(5, 68) 
ORDER BY 
 CAST(result_table.item AS UNSIGNED) ASC

Result

| json        |
| ----------- |
| ["1", "18"] |

see demo