How to generate a unique id in a recursive function in PHP? - php

Question

In my recursive function I can't generate a unique id (via an incremented variable) for my multidimensional array, here is what I tried.
I need to assign a unique ID for each directory name I find: $data[] = ['id'=>$i,'text'=>$dir];
function scandir_rec($root, $i)
{
$data = [];
if (!is_dir($root)) {
return;
}
$dirs = scandir($root);
foreach ($dirs as $dir) {
$i++;
if ($dir == '.' || $dir == '..') {
continue;
}
$path = $root . '/' . $dir;
if (is_file($path)) {
continue;
}
if (is_dir($path)) {
$nodes = scandir_rec($path, $i); // <--- unique ID
if (!empty($nodes)) $chunk['children'] = $nodes;
}
$data[] = [ 'id' => $i, 'text' => $dir ]; // <--- unique ID
}
return $data;
}
$rootDir = '/var/www/html';
$i = 0;
$json_data = json_encode(scandir_rec($rootDir, $i), JSON_UNESCAPED_UNICODE|JSON_UNESCAPED_SLASHES); // <--- unique ID
file_put_contents('tree.json', $json_data);
Desired result
[
{
"id":1,
"text":"Root folder",
"children":[
{
"id":2,
"text":"Child folder 1"
},
{
"id":3,
"text":"Child folder 2"
}
]
},
{
"id":4,
"text":"Root folder"
},
{
"id":5,
"text":"Root folder",
"children":[
{
"id":6,
"text":"Child folder 1"
},
{
"id":7,
"text":"Child folder 2"
}
]
}
]

Answer 1

Personally I hate returning from recursive functions, it tends to get weird unless you are returning a status for the caller to do something with. Instead, it is often easier to use byref parameters.
I've commented the code below, hopefully it makes sense. We're passing two parameters byref, one is "the current array to add stuff to" and the other is "the global index". We're also keeping track of depth because you are treating the root differently than children.
function scandir_rec(string $root, array &$current_parent, int &$global_index, int $depth = 0): void
{
// Bail early if we don't have a directory
if (!is_dir($root)) {
return;
}
// Get the children
$dirs = scandir($root);
// This is the local child index and resets for successive calls
$local_index = 1;
foreach ($dirs as $dir) {
// Once again, bail early if we can
if ($dir === '.' || $dir === '..') {
continue;
}
// The next path to scan
$next_path = $root . '/' . $dir;
// Make sure it isn't a file
if (is_file($next_path)) {
continue;
}
// Everyone gets this
$current_node = [
// Set the ID and increment so that the next gets a bigger one
'id' => $global_index++,
// Root node is treated differently thant children
'text' => 0 === $depth ? 'Root node' : "Child node ${local_index}",
];
// If the next node is a directory
if (is_dir($next_path)) {
// Create a temporary holder for the contents
$children = [];
// Scan again, incrementing the depth by one
scandir_rec($next_path, $children, $global_index, ++$depth);
// If the above did something to our array, append to our local (the master)
if ($children) {
$current_node['children'] = $children;
}
}
// Append the current node
$current_parent[] = $current_node;
}
}
To call this, remember that it doesn't return so we need to run it by itself, We also want to pass the global index in by ref so we need to create a variable to hold it.
$current_parent = [];
$global_index = 1;
scandir_rec('/var/www', $current_parent, $global_index);
$json_data = json_encode($current_parent, JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES | JSON_PRETTY_PRINT);
This code is written against PHP 7.4 do you might need to tweak a couple of things if you are on an older version.
edit
In hindsight, I was taking your text too literally, I think. You can probably get rid of the $depth parameter completely and just set your text to:
'text' => $root,
I'm running this from Windows and I just ran a quick test folder which produced:
[
{
"id": 1,
"text": "/Dell",
"children": [
{
"id": 2,
"text": "/Dell/25tgw",
"children": [
{
"id": 3,
"text": "/Dell/25tgw/RealtekHDAudio"
},
{
"id": 4,
"text": "/Dell/25tgw/RealtekHDAudio"
},
{
"id": 5,
"text": "/Dell/25tgw/RealtekHDAudio"
},
{
"id": 6,
"text": "/Dell/25tgw/RealtekHDAudio"
}
]
}
]
},
{
"id": 7,
"text": "/Dell",
"children": [
{
"id": 8,
"text": "/Dell/UpdatePackage"
}
]
},
{
"id": 9,
"text": "/Dell"
}
]

Answer 2

The issue is that each recursive call starts it's own count. Let's say $i is 10 inside the loop at a given time. You start a recursive call that increases it up to 15 but when the call ends and returns to parent function, $i there is still 10.
Perhaps the simplest fix is to pass the counter by reference:
function scandir_rec($root, &$i)
{
}
Another solution is to avoid recursion altogether. PHP has builtin iterators that can walk a three without recursion, e.g.:
$files = new RecursiveIteratorIterator(
new RecursiveDirectoryIterator($rootDir),
RecursiveIteratorIterator::SELF_FIRST
);
foreach ($files as $name => $object) {
echo sprintf("[%s] %s\n", $object->isDir() ? 'DIR' : ' ', $name);
}