如何使用for循环预览邻接列表树（非递归解决方案）

Question

I store a tree in MySQL database by using the adjacency list method

when I want to preview the tree, PHP retrieves the whole tree from the database and preview it using recursion.

But iterations is better than recursion in performance, so I want to populate the tree using the for loops to achieve better performance.

I do not want to use any MySQL functions or methods or triggers, I just want to populate the tree by using iterations (for loops)

Answer 1

I stumbled on this old unanswered question, while searching for a link to aide in answering a similar question on IRC.

This is a rough example showing how to build a list of paths from an adjacency list, it requires the data to be sorted by the 'parent' then 'id', this could easily be translated to an SQL SELECT with an ORDER BY parent, id clause:

<?php
$arr = [
    ['id' => 1, 'parent' => 0],
    ['id' => 2, 'parent' => 1],
    ['id' => 3, 'parent' => 2], 
    ['id' => 4, 'parent' => 3], 
    ['id' => 5, 'parent' => 0],
    ['id' => 6, 'parent' => 5],
    ['id' => 7, 'parent' => 5], 
    ['id' => 8, 'parent' => 7],
    ['id' => 9, 'parent' => 8],
    ['id' => 10, 'parent' => 5],
    ['id' => 11, 'parent' => 0],
    ['id' => 12, 'parent' => 5],
    ['id' => 13, 'parent' => 5],
    ['id' => 14, 'parent' => 11],
    ['id' => 15, 'parent' => 13],
    ['id' => 16, 'parent' => 1],
    ['id' => 17, 'parent' => 15],
];

usort($arr, function($a, $b) { return $a['parent'] - $b['parent']?: $a['id'] - $b['id']; });

$paths = [];
foreach($arr as $current) {
    if($current['parent'] > 0) {
        $paths[ $current['id'] ] = array_merge($paths[ $current['parent'] ], [ $current['parent'] ]);
    } else {
        $paths[ $current['id'] ] = [];
    }
}

ksort($paths);
foreach($paths as $k => $v) {
    printf("%s => %s\n", $k, implode('/', array_merge($v, [ $k ])));
}

Output:

1 => 1
2 => 1/2
3 => 1/2/3
4 => 1/2/3/4
5 => 5
6 => 5/6
7 => 5/7
8 => 5/7/8
9 => 5/7/8/9
10 => 5/10
11 => 11
12 => 5/12
13 => 5/13
14 => 11/14
15 => 5/13/15
16 => 1/16
17 => 5/13/15/17