Question

Does anyone have a brilliant idea how to obtain the elements with the deepest path from an array with file paths? If this sounds weird, imagine the following array:

/a/b
/a
/1/2/3/4
/1/2
/1/2/3/5
/a/b/c/d/e

What I want to obtain is:

/1/2/3/4
/1/2/3/5
/a/b/c/d/e

Wondering what the fastest method is without having to iterate over the whole array over and over again. Language is PHP (5.2).

Answer 1

Following your clarifications, here’s a function that would do it. It keeps an array of the “deepest paths” found and compares each path against it. The best-case scenario is O(n) (if all paths are subpaths of the largest one) and worst-case scenario is O(n²) (if all paths are completely distinct).

Note that continue 2 means “continue on the outer loop”.

<?php

function getDeepestPaths($array)
{
    $deepestPaths = array();
    foreach ($array as $path)
    {
        $pathLength = strlen($path);
        // look for all the paths we consider the longest
        // (note how we're using references to the array members)
        foreach ($deepestPaths as &$deepPath)
        {
            $deepPathLength = strlen($deepPath);
            // if $path is prefixed by $deepPath, this means that $path is
            // deeper, so we replace $deepPath with $path
            if (substr($path, 0, $deepPathLength) == $deepPath)
            {
                $deepPath = $path;
                continue 2;
            }
            // otherwise, if $deepPath is prefixed by $path, this means that
            // $path is shallower; so we should stop looking
            else if (substr($deepPath, 0, $pathLength) == $path)
            {
                continue 2;
            }
        }
        // $path matches nothing currently in $deepestPaths, so we should
        // add it to the array
        $deepestPaths[] = $path;
    }
    return $deepestPaths;
}

$paths = array('/a/b', '/a', '/1/2/3/4', '/1/2', '/1/2/3/5', '/a/b/c/d/e');
print_r(getDeepestPaths($paths));

?>

If your folder names don’t end with slashes, you’ll want to do an additional check in the two ifs: that the character next to the prefix in the deeper path is a slash, because otherwise a path like /foo/bar will be seen as a “deeper path” than /foo/b (and will replace it).

if (substr($path, 0, $deepPathLength) == $deepPath && $path[$deepPathLength] == '/')
if (substr($deepPath, 0, $path) == $path && $deepPath[$path] == '/')