MySQL 5.7：JSON类型支持

在MySQL与PostgreSQL的对比中，PG的JSON格式支持优势总是不断被拿来比较。其实早先MariaDB也有对非结构化的数据进行存储的方案，称为dynamic column，但是方案是通过BLOB类型的方式来存储。这样导致的问题是查询性能不高，不能有效建立索引，与一些文档数据库对比，优势并不大，故在社区的反应证实比较一般。当然，MariaDB的dynamic column功能还不仅限于非结构化数据的存储，但不在本文进行展开。

MySQL 5.7.7版本开始，InnoDB存储引擎已经原生支持JSON格式，该格式不是简单的BLOB类似的替换。原生的JSON格式支持有以下的优势：

JSON数据有效性检查：BLOB类型无法在数据库层做这样的约束性检查。

查询性能的提升：查询不需要遍历所有字符串才能找到数据。

支持索引：通过虚拟列的功能可以对JSON中的部分数据进行索引。

首先我们来看如何在MySQL中使用原生的JSON格式：

mysql> create table user ( uid int auto_increment,data json,primary key(uid)) engine=innodb;
Query OK, 0 rows affected (0.01 sec)
 
mysql> insert into user values (NULL,'{"name":"David","mail":"jiangchengyao@gmail.com","address":"Shangahai"}');
Query OK, 1 row affected (0.00 sec)
 
mysql> insert into user values (NULL,'{"name":"Amy","mail":"amy@gmail.com"}');
Query OK, 1 row affected (0.00 sec)

mysql> create table user ( uid int auto_increment,data json,primary key(uid)) engine=innodb;

Query OK, 0 rows affected (0.01 sec)

mysql> insert into user values (NULL,'{"name":"David","mail":"jiangchengyao@gmail.com","address":"Shangahai"}');

Query OK, 1 row affected (0.00 sec)

mysql> insert into user values (NULL,'{"name":"Amy","mail":"amy@gmail.com"}');

Query OK, 1 row affected (0.00 sec)

可以看到我们新建了表user，并且将列data定义为了JSON类型。这意味着我们可以对插入的数据做JSON格式检查，确保其符合JSON格式的约束，如插入一条不合法的JSON数据会报如下错误：

mysql> insert into user values (NULL,"test");
ERROR 3130 (22032): Invalid JSON text: "Invalid value" at position 2 in value (or column) 'test'.

1 2	mysql> insert into user values (NULL,"test"); ERROR 3130 (22032): Invalid JSON text: "Invalid value" at position 2 in value (or column) 'test'.

此外，正如前面所说的，MySQL 5.7提供了一系列函数来高效地处理JSON字符，而不是需要遍历所有字符来查找，这不得不说是对MariaDB dynamic column的巨大改进：

mysql> select json_extract(data,'$.name'),json_extract(data,'$.address') from user;
+-----------------------------+--------------------------------+
| json_extract(data,'$.name') | json_extract(data,'$.address') |
+-----------------------------+--------------------------------+
| "David"                     | "Shangahai"                    |
| "Amy"                       | NULL                           |
+-----------------------------+--------------------------------+
2 rows in set (0.02 sec)

mysql> select json_extract(data,'$.name'),json_extract(data,'$.address') from user;

+-----------------------------+--------------------------------+

| json_extract(data,'$.name') | json_extract(data,'$.address') |

+-----------------------------+--------------------------------+

| "David" | "Shangahai" |

| "Amy" | NULL |

+-----------------------------+--------------------------------+

2 rows in set (0.02 sec)

当然，最令人的激动的功能应该是MySQL 5.7的虚拟列功能，通过传统的B+树索引即可实现对JSON格式部分属性的快速查询。使用方法是首先创建该虚拟列，然后在该虚拟列上创建索引：

mysql> ALTER TABLE user ADD user_name varchar(128) GENERATED ALWAYS AS (jsn_extract(data,'$.name')) VIRTUAL;
Query OK, 0 rows affected (0.01 sec)
Records: 0 Duplicates: 0 Warnings: 0
 
mysql> select user_name from user;
+-----------+
| user_name |
+-----------+
| "Amy"     |
| "David"   |
+-----------+
2 rows in set (0.00 sec)
 
mysql> alter table user add index idx_username (user_name);
Query OK, 2 rows affected (0.01 sec)
Records: 2  Duplicates: 0  Warnings: 0

mysql> ALTER TABLE user ADD user_name varchar(128) GENERATED ALWAYS AS (jsn_extract(data,'$.name')) VIRTUAL;

Query OK, 0 rows affected (0.01 sec)

Records: 0 Duplicates: 0 Warnings: 0

mysql> select user_name from user;

+-----------+

| user_name |

+-----------+

| "Amy" |

| "David" |

+-----------+

2 rows in set (0.00 sec)

mysql> alter table user add index idx_username (user_name);

Query OK, 2 rows affected (0.01 sec)

Records: 2 Duplicates: 0 Warnings: 0

然后可以通过添加的索引对用户名进行快速的查询，这和普通类型的列查询一样。而通过explain可以验证优化器已经选择了在虚拟列上创建的新索引：

mysql> explain select * from user where user_name='"Amy"'\G
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: user
   partitions: NULL
         type: ref
possible_keys: idx_username
          key: idx_username
      key_len: 131
          ref: const
         rows: 1
     filtered: 100.00
        Extra: NULL
1 row in set, 1 warning (0.00 sec)

mysql> explain select * from user where user_name='"Amy"'\G

*************************** 1. row ***************************

id: 1

select_type: SIMPLE

table: user

partitions: NULL

type: ref

possible_keys: idx_username

key: idx_username

key_len: 131

ref: const

rows: 1

filtered: 100.00

Extra: NULL

1 row in set, 1 warning (0.00 sec)

可以发现MySQL 5.7对于JSON格式堪称完美，相信PostgreSQL阵营需要寻找新的策略来“攻击”MySQL了吧。如无意外，还是会停留在优化器这块，毕竟这块是目前MySQL必须要克服的最大问题，好在MySQL团队已经在重构优化器代码，相信更好的优化器将会在下一个版本中全面爆发。

mysql> set @j='{"age": 18, "sex": "male", "hobby": ["basketball", "football"], "score": [85, 90, 100]}';
Query OK, 0 rows affected (0.00 sec)

1 2	mysql> set @j='{"age": 18, "sex": "male", "hobby": ["basketball", "football"], "score": [85, 90, 100]}'; Query OK, 0 rows affected (0.00 sec)

JSON_CONTAINS

mysql> select JSON_CONTAINS(@j, '"basketball"', '$.hobby');
+----------------------------------------------+
| JSON_CONTAINS(@j, '"basketball"', '$.hobby') |
+----------------------------------------------+
|                                            1 |
+----------------------------------------------+
1 row in set (0.00 sec)

mysql> select JSON_CONTAINS(@j, '"basketball"', '$.hobby');

+----------------------------------------------+

| JSON_CONTAINS(@j, '"basketball"', '$.hobby') |

+----------------------------------------------+

| 1 |

+----------------------------------------------+

1 row in set (0.00 sec)

JSON_SEARCH

mysql> select JSON_SEARCH(@j, 'one', 'basketball', null, '$.hobby') is not null;
+-------------------------------------------------------------------+
| JSON_SEARCH(@j, 'one', 'basketball', null, '$.hobby') is not null |
+-------------------------------------------------------------------+
|                                                                 1 |
+-------------------------------------------------------------------+
1 row in set (0.00 sec)