此学习文是基于MySQL 8.0写的
得益于大神朋友的悉心指导解决不少坑，才写出此文，向大神奉上膝盖

  要在MySQL中存储数据，就必须定义数据库和表结构(schema)，这是一个主要的限制。为了应对这一点，从MySQL 5.7开始，MySQL支恃了 JavaScript对象表示(JavaScriptObject Notation，JSON) 数据类型。之前，这类数据不是单独的数据类型，会被存储为字符串。新的JSON数据类型提供了自动验证的JSON文档以及优化的存储格式。

JSON文档以二进制格式存储，它提供以下功能：

对文档元素的快速读取访问。
当服务器再次读取JSON文档时，不需要重新解析文本获取该值。
通过键或数组索引直接查找子对象或嵌套值，而不需要读取文档中的所有值。
创建一个测试表

1.  
   mysql> create table employees.emp_details (
2.  
       -> emp_no int primary key,
3.  
       -> details json
4.  
       -> );
5.  
   Query OK, 0 rows affected (0.17 sec)
6.  
    
7.  
   mysql> desc employees.emp_details;
8.  
   +---------+---------+------+-----+---------+-------+
9.  
   | Field   | Type    | Null | Key | Default | Extra |
10.  
    +---------+---------+------+-----+---------+-------+
11.  
    | emp_no  | int(11) | NO   | PRI | NULL    |       |
12.  
    | details | json    | YES  |     | NULL    |       |
13.  
    +---------+---------+------+-----+---------+-------+
14.  
    2 rows in set (0.00 sec)

 插入JSON

1.  
   mysql> insert into employees.emp_details (emp_no, details)
2.  
       -> values (‘1‘,
3.  
       -> ‘{"location":"IN","phone":"+11800000000","email":"abc@example.com","address":{"line1":"abc","line2":"xyz street","city":"Bangalore","pin":"560103"}}‘
4.  
       -> );
5.  
   Query OK, 1 row affected (0.13 sec)
6.  
    
7.  
   mysql> select emp_no, details from employees.emp_details;
8.  
   +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
9.  
   | emp_no | details                                                                                                                                                           |
10.  
    +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
11.  
    |      1 | {"email": "abc@example.com", "phone": "+11800000000", "address": {"pin": "560103", "city": "Bangalore", "line1": "abc", "line2": "xyz street"}, "location": "IN"} |
12.  
    +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
13.  
    1 row in set (0.00 sec)

检索JSON
可以使用->和->>运算符检索JSON列的字段：

1.  
   mysql> select emp_no, details -> ‘$.address.pin‘ pin 
2.  
       -> from employees.emp_details;
3.  
   +--------+----------+
4.  
   | emp_no | pin      |
5.  
   +--------+----------+
6.  
   |      1 | "560103" |
7.  
   +--------+----------+
8.  
   1 row in set (0.00 sec)

如果不用引号检索数据，可以使用->> 运算符（推荐此方式）

1.  
   mysql> select emp_no, details ->> ‘$.address.pin‘ pin 
2.  
       -> from employees.emp_details;
3.  
   +--------+--------+
4.  
   | emp_no | pin    |
5.  
   +--------+--------+
6.  
   |      1 | 560103 |
7.  
   +--------+--------+
8.  
   1 row in set (0.00 sec)

JSON函数
MySQL提供了许多处理JSON数据的函数，让我们看看最常用的几种函数。

1. 优雅浏览
想要以优雅的格式显示JSON值，请使用JSON_PRETTY()函数

1.  
   mysql> select emp_no, json_pretty(details) 
2.  
       -> from employees.emp_details\G
3.  
   *************************** 1. row ***************************
4.  
                 emp_no: 1
5.  
   json_pretty(details): {
6.  
     "email": "abc@example.com",
7.  
     "phone": "+11800000000",
8.  
     "address": {
9.  
       "pin": "560103",
10.  
        "city": "Bangalore",
11.  
        "line1": "abc",
12.  
        "line2": "xyz street"
13.  
      },
14.  
      "location": "IN"
15.  
    }
16.  
    1 row in set (0.00 sec)

2. 查找
可以在WHERE子句中使用col ->> path运算符来引用JSON的某一列

1.  
   mysql> select emp_no, details 
2.  
       -> from employees.emp_details 
3.  
       -> where details ->> ‘$.address.pin‘ = "560103";
4.  
   +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
5.  
   | emp_no | details                                                                                                                                                           |
6.  
   +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
7.  
   |      1 | {"email": "abc@example.com", "phone": "+11800000000", "address": {"pin": "560103", "city": "Bangalore", "line1": "abc", "line2": "xyz street"}, "location": "IN"} |
8.  
   +--------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------+
9.  
   1 row in set (0.00 sec)

也可以使用JSON_CONTAINS函数查询数据。如果找到了数据，则返回1，否则返回0

1.  
   mysql> select json_contains(details ->> ‘$.address.pin‘,"560103") 
2.  
       -> from employees.emp_details;
3.  
   +-----------------------------------------------------+
4.  
   | json_contains(details ->> ‘$.address.pin‘,"560103") |
5.  
   +-----------------------------------------------------+
6.  
   |                                                   1 |
7.  
   +-----------------------------------------------------+
8.  
   1 row in set (0.00 sec)

如何查询一个key？使用JSON_CONTAINS_PATH函数检查address. line1是否存在

1.  
   mysql> select json_contains_path(details, ‘one‘, "$.address.line1") 
2.  
       -> from employees.emp_details;
3.  
   +-------------------------------------------------------+
4.  
   | json_contains_path(details, ‘one‘, "$.address.line1") |
5.  
   +-------------------------------------------------------+
6.  
   |                                                     1 |
7.  
   +-------------------------------------------------------+
8.  
   1 row in set (0.00 sec)

one表示至少应该存在一个键，检查address.line1或者address.line2是否存在

1.  
   mysql> select json_contains_path(details, ‘one‘, "$.address.line1", "$.address.line2") 
2.  
       -> from employees.emp_details;
3.  
   +--------------------------------------------------------------------------+
4.  
   | json_contains_path(details, ‘one‘, "$.address.line1", "$.address.line2") |
5.  
   +--------------------------------------------------------------------------+
6.  
   |                                                                        1 |
7.  
   +--------------------------------------------------------------------------+
8.  
   1 row in set (0.00 sec)

如果要检查address.line1或者address.line5是否同时存在，可以使用all，而不是one

1.  
   mysql> select json_contains_path(details, ‘all‘, "$.address.line1", "$.address.line5") 
2.  
       -> from employees.emp_details;
3.  
   +--------------------------------------------------------------------------+
4.  
   | json_contains_path(details, ‘all‘, "$.address.line1", "$.address.line5") |
5.  
   +--------------------------------------------------------------------------+
6.  
   |                                                                        0 |
7.  
   +--------------------------------------------------------------------------+
8.  
   1 row in set (0.00 sec)

3. 修改

可以使用三种不同的函数来修改数据：JSON_SET()、JSON_INSERT()和JSON _REPLACE()。 在MySQL 8之前的版本中，我们还需要对整个列进行完整的更新，这并不是最佳的方法。

3.1. JSON_SET() 替换现有值并添加不存在的值

1.  
   mysql> update employees.emp_details 
2.  
       -> set details = json_set(details, "$.address.pin", "560100", "$.nickname","kai") 
3.  
       -> where emp_no = 1;
4.  
   Query OK, 1 row affected (0.01 sec)
5.  
   Rows matched: 1  Changed: 1  Warnings: 0
6.  
    
7.  
   mysql> select emp_no, json_pretty(details) 
8.  
       -> from employees.emp_details\G
9.  
   *************************** 1. row ***************************
10.  
                  emp_no: 1
11.  
    json_pretty(details): {
12.  
      "email": "abc@example.com",
13.  
      "phone": "+11800000000",
14.  
      "address": {
15.  
        "pin": "560100",
16.  
        "city": "Bangalore",
17.  
        "line1": "abc",
18.  
        "line2": "xyz street"
19.  
      },
20.  
      "location": "IN",
21.  
      "nickname": "kai"
22.  
    }
23.  
    1 row in set (0.00 sec)

3.2. JSON_INSERT() 插入值，但不替换现有值
在这种情况下，$.address.pin不会被更新，只会添加一个新的字段$.address.line4

1.  
   mysql> update employees.emp_details 
2.  
       -> set details = json_insert(details, "$.address.pin", "560132", "$.address.line4","A Wing") 
3.  
       -> where emp_no = 1;
4.  
   Query OK, 1 row affected (0.01 sec)
5.  
   Rows matched: 1  Changed: 1  Warnings: 0
6.  
    
7.  
   mysql> select emp_no, json_pretty(details) 
8.  
       -> from employees.emp_details\G
9.  
   *************************** 1. row ***************************
10.  
                  emp_no: 1
11.  
    json_pretty(details): {
12.  
      "email": "abc@example.com",
13.  
      "phone": "+11800000000",
14.  
      "address": {
15.  
        "pin": "560100",
16.  
        "city": "Bangalore",
17.  
        "line1": "abc",
18.  
        "line2": "xyz street",
19.  
        "line4": "A Wing"
20.  
      },
21.  
      "location": "IN",
22.  
      "nickname": "kai"
23.  
    }
24.  
    1 row in set (0.01 sec)

3.3. JSON_REPLACE()
仅替换现有值
在这种情况下，$.address.line5不会被添加， 只有$.address.pin会被更新

1.  
   mysql> update employees.emp_details 
2.  
       -> set details = json_replace(details, "$.address.pin", "560132", "$.address.line5","Landmark") 
3.  
       -> where emp_no = 1;
4.  
   Query OK, 1 row affected (0.00 sec)
5.  
   Rows matched: 1  Changed: 1  Warnings: 0
6.  
    
7.  
   mysql> select emp_no, json_pretty(details) 
8.  
       -> from employees.emp_details\G
9.  
   *************************** 1. row ***************************
10.  
                  emp_no: 1
11.  
    json_pretty(details): {
12.  
      "email": "abc@example.com",
13.  
      "phone": "+11800000000",
14.  
      "address": {
15.  
        "pin": "560132",
16.  
        "city": "Bangalore",
17.  
        "line1": "abc",
18.  
        "line2": "xyz street",
19.  
        "line4": "A Wing"
20.  
      },
21.  
      "location": "IN",
22.  
      "nickname": "kai"
23.  
    }
24.  
    1 row in set (0.00 sec)

4. 删除 JSON_REMOVE能从JSON文档中删除数据

1.  
   mysql> update employees.emp_details 
2.  
       -> set details = json_remove(details, "$.address.line4") 
3.  
       -> where emp_no = 1;
4.  
   Query OK, 1 row affected (0.01 sec)
5.  
   Rows matched: 1  Changed: 1  Warnings: 0
6.  
    
7.  
   mysql> select emp_no, json_pretty(details) 
8.  
       -> from employees.emp_details\G
9.  
   *************************** 1. row ***************************
10.  
                  emp_no: 1
11.  
    json_pretty(details): {
12.  
      "email": "abc@example.com",
13.  
      "phone": "+11800000000",
14.  
      "address": {
15.  
        "pin": "560132",
16.  
        "city": "Bangalore",
17.  
        "line1": "abc",
18.  
        "line2": "xyz street"
19.  
      },
20.  
      "location": "IN",
21.  
      "nickname": "kai"
22.  
    }
23.  
    1 row in set (0.00 sec)

5. 其他函数

 JSON_KEYS()：获取JSON文档中的所有键

1.  
   mysql> select json_keys(details),json_keys(details ->> "$.address") 
2.  
       -> from employees.emp_details 
3.  
       -> where emp_no= 1;
4.  
   +-------------------------------------------------------+------------------------------------+
5.  
   | json_keys(details)                                    | json_keys(details ->> "$.address") |
6.  
   +-------------------------------------------------------+------------------------------------+
7.  
   | ["email", "phone", "address", "location", "nickname"] | ["pin", "city", "line1", "line2"]  |
8.  
   +-------------------------------------------------------+------------------------------------+
9.  
   1 row in set (0.00 sec)

JSON_LENGTH()：给出JSON文档中的元素数
 

1.  
   mysql> select json_length(details), json_length(details ->> "$.address") 
2.  
       -> from employees.emp_details 
3.  
       -> where emp_no= 1;
4.  
   +----------------------+--------------------------------------+
5.  
   | json_length(details) | json_length(details ->> "$.address") |
6.  
   +----------------------+--------------------------------------+
7.  
   |                    5 |                                    4 |
8.  
   +----------------------+--------------------------------------+
9.  
   1 row in set (0.00 sec)

延伸阅读： https://dev.mysql.com/doc/refman/8.0/en/json-function-reference.html