聊聊mysql的树形结构存储及查询

序

本文主要研究一下mysql的树形结构存储及查询

存储parent

这种方式就是每个节点存储自己的parent_id信息

• 建表及数据准备

  CREATE TABLE `menu` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(50) NOT NULL,
  `parent_id` int(11) NOT NULL DEFAULT '0',
  PRIMARY KEY (`id`)
  ) ENGINE=InnoDB;
  
  INSERT INTO `menu` (`id`, `name`, `parent_id`) VALUES
  (1, 'level1a',  0),
  (2, 'level1b', 0),
  (3, 'level2a-1a',1),
  (4, 'level2b-1a',1),
  (5, 'level2a-1b', 2),
  (6, 'level2b-1b', 2),
  (7, 'level3-2a1a', 3),
  (8, 'level3-2b1a', 4),
  (9, 'level3-2a1b', 5),
  (10, 'level3-2b1b', 6);

• 查询

  -- 查询跟节点下的所有节点
  SELECT t1.name AS lev1, t2.name as lev2, t3.name as lev3
  FROM menu AS t1
  LEFT JOIN menu AS t2 ON t2.parent_id = t1.id
  LEFT JOIN menu AS t3 ON t3.parent_id = t2.id
  WHERE t1.name = 'level1a';
  
  +---------+------------+-------------+
  | lev1    | lev2       | lev3        |
  +---------+------------+-------------+
  | level1a | level2a-1a | level3-2a1a |
  | level1a | level2b-1a | level3-2b1a |
  +---------+------------+-------------+
  
  -- 查询叶子节点
  SELECT t1.name FROM
  menu AS t1 LEFT JOIN menu as t2
  ON t1.id = t2.parent_id
  WHERE t2.id IS NULL;
  
  +-------------+
  | name        |
  +-------------+
  | level3-2a1a |
  | level3-2b1a |
  | level3-2a1b |
  | level3-2b1b |
  +-------------+

  存储及修改上比较方便，就是要在sql里头查询树比较费劲，一般是加载到内存由应用自己构造

存储path

这种方式在存储parent的基础上，额外存储path，即从根节点到该节点的路径

• 建表及数据准备

  CREATE TABLE `menu_path` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `name` varchar(50) NOT NULL,
  `parent_id` int(11) NOT NULL DEFAULT '0',
  `path` varchar(255) NOT NULL DEFAULT '',
  PRIMARY KEY (`id`)
  ) ENGINE=InnoDB;
  
  INSERT INTO `menu_path` (`id`, `name`, `parent_id`, `path`) VALUES
  (1, 'level1a', 0, '1/'),
  (2, 'level1b', 0, '2/'),
  (3, 'level2a-1a',1, '1/3'),
  (4, 'level2b-1a',1, '1/4'),
  (5, 'level2a-1b', 2, '2/5'),
  (6, 'level2b-1b', 2, '2/6'),
  (7, 'level3-2a1a', 3, '1/3/7'),
  (8, 'level3-2b1a', 4, '1/4/8'),
  (9, 'level3-2a1b', 5, '2/5/9'),
  (10, 'level3-2b1b', 6, '2/6/10');

• 查询

  -- 查询某个节点的所有子节点
  select * from menu_path where path like '1/%'
  +----+-------------+-----------+-------+
  | id | name        | parent_id | path  |
  +----+-------------+-----------+-------+
  | 1  | level1a     | 0         | 1/    |
  | 3  | level2a-1a  | 1         | 1/3   |
  | 4  | level2b-1a  | 1         | 1/4   |
  | 7  | level3-2a1a | 3         | 1/3/7 |
  | 8  | level3-2b1a | 4         | 1/4/8 |
  +----+-------------+-----------+-------+

  查找某个节点及其子节点比较方面，就是修改比较费劲，特别是节点移动，所有子节点的path都得跟着修改

MPTT(Modified Preorder Tree Traversal)

不存储parent_id，改为存储lft,rgt，它们的值由树的先序遍历顺序决定

• 建表及数据准备

  CREATE TABLE `menu_preorder` (
  `id` int(11) NOT NULL,
  `name` varchar(50) NOT NULL,
  `lft` int(11) NOT NULL DEFAULT '0',
  `rgt` int(11) NOT NULL DEFAULT '0',
  PRIMARY KEY (`id`)
  ) ENGINE=InnoDB;
  
                   1(level1a)14
         2(level2a)7                8(level2b)13
  3(level3a-2a)4 5(level3b-2a)6 9(level3c-2b)10 11(level3d-2b)12
  
  INSERT INTO `menu_preorder` (`id`, `name`, `lft`, `rgt`) VALUES
  (1, 'level1a', 1, 14),
  (2, 'level2a',2, 7),
  (3, 'level2b',8, 13),
  (4, 'level3a-2a', 3, 4),
  (5, 'level3b-2a', 5, 6),
  (6, 'level3c-2b', 9, 10),
  (7, 'level3d-2b', 11, 12);
  
  select * from menu_preorder
  +----+------------+-----+-----+
  | id | name       | lft | rgt |
  +----+------------+-----+-----+
  | 1  | level1a    | 1   | 14  |
  | 2  | level2a    | 2   | 7   |
  | 3  | level2b    | 8   | 13  |
  | 4  | level3a-2a | 3   | 4   |
  | 5  | level3b-2a | 5   | 6   |
  | 6  | level3c-2b | 9   | 10  |
  | 7  | level3d-2b | 11  | 12  |
  +----+------------+-----+-----+

• 查询

  -- 查询某个节点及其子节点，比如level2b
  select * from menu_preorder where lft between 8 and 13
  +----+------------+-----+-----+
  | id | name       | lft | rgt |
  +----+------------+-----+-----+
  | 3  | level2b    | 8   | 13  |
  | 6  | level3c-2b | 9   | 10  |
  | 7  | level3d-2b | 11  | 12  |
  +----+------------+-----+-----+
  
  -- 查询所有叶子节点
  SELECT name
  FROM menu_preorder
  WHERE rgt = lft + 1;
  
  +------------+
  | name       |
  +------------+
  | level3a-2a |
  | level3b-2a |
  | level3c-2b |
  | level3d-2b |
  +------------+
  
  -- 查询某个节点及其父节点
  SELECT parent.*
  FROM menu_preorder AS node,
  menu_preorder AS parent
  WHERE node.lft BETWEEN parent.lft AND parent.rgt
  AND node.name = 'level2b'
  ORDER BY parent.lft;
  
  +----+---------+-----+-----+
  | id | name    | lft | rgt |
  +----+---------+-----+-----+
  | 1  | level1a | 1   | 14  |
  | 3  | level2b | 8   | 13  |
  +----+---------+-----+-----+
  
  -- 树形结构展示
  SELECT CONCAT( REPEAT(' ', COUNT(parent.name) - 1), node.name) AS name
  FROM menu_preorder AS node,
  menu_preorder AS parent
  WHERE node.lft BETWEEN parent.lft AND parent.rgt
  GROUP BY node.name
  ORDER BY node.lft;
  
  +--------------+
  | name         |
  +--------------+
  | level1a      |
  |  level2a     |
  |   level3a-2a |
  |   level3b-2a |
  |  level2b     |
  |   level3c-2b |
  |   level3d-2b |
  +--------------+

好处是通过lft进行范围(该节点的lft,rgt作为范围)查找就可以，缺点就是增删节点导致很多节点的lft及rgt都要修改

小结

• 存储parent的方式最为场景，一般树形结构数据量不大的话，直接在应用层内存构造树形结构和搜索
• 存储path的好处是可以借助path来查找节点及其子节点，缺点就是移动node需要级联所有子节点的path，比较费劲
• MPTT的方式好处是通过lft进行范围(该节点的lft,rgt作为范围)查找就可以，缺点就是增删节点导致很多节点的lft及rgt都要修改

doc

• Managing Hierarchical Data in MySQL
• hierarchical-data-database
• hierarchical-data-database-2
• hierarchical-data-database-3