使用PHP排查目录中已使用文件的技术求助

Hey there! Let's tackle this problem step by step. Your current script is a good start, but it's missing some critical pieces that are causing it to miss used files. Here are the key optimizations and fixes to get you a complete list of actively used files in your root directory:

Key Issues in the Original Code #

• Incorrect file paths: You're scanning /public_html/ but trying to read files with just their base name, which might not work if your script isn't running from that directory.
• Missing PHP internal references: The script only looks at HTML tags, but PHP files often use include, require, or their variants to reference other files—these are completely missed.
• Inefficient array checks: Using in_array() for every check slows things down as the list grows; using array keys is faster.
• Unhandled DOM parsing warnings: Invalid HTML (common in real-world projects) can throw warnings that break parsing.
• Missing entry files: The script doesn't include the PHP files you're scanning themselves (e.g., an index.php that's never referenced by another file but is the main entry point).
• Limited tag coverage: It misses script sources, CSS links, and other resource references.

Optimized Code Implementation #

<?php
$rootDir = '/public_html/';
$usedFiles = [];

// First, add all PHP files in root to the list (they're potential entry points)
$files = scandir($rootDir);
foreach ($files as $file) {
    $filePath = $rootDir . $file;
    // Skip directories and non-PHP files for initial entry list
    if (is_file($filePath) && pathinfo($filePath, PATHINFO_EXTENSION) === 'php') {
        $usedFiles[$file] = true; // Use keys for O(1) lookups
    }
}

// Now scan each used file for references
foreach (array_keys($usedFiles) as $fileName) {
    $filePath = $rootDir . $fileName;
    $content = file_get_contents($filePath);
    if (!$content) continue;

    // 1. Parse HTML tags for references
    libxml_use_internal_errors(true); // Suppress HTML parsing warnings
    $dom = new DOMDocument();
    $dom->loadHTML($content);
    libxml_clear_errors();

    // Check <a> tags
    foreach ($dom->getElementsByTagName('a') as $tag) {
        $href = $tag->getAttribute('href');
        $referencedFile = basename($href);
        $referencedPath = $rootDir . $referencedFile;
        if (is_file($referencedPath)) {
            $usedFiles[$referencedFile] = true;
        }
    }

    // Check <form> action attributes
    foreach ($dom->getElementsByTagName('form') as $tag) {
        $action = $tag->getAttribute('action');
        $referencedFile = basename($action);
        $referencedPath = $rootDir . $referencedFile;
        if (is_file($referencedPath)) {
            $usedFiles[$referencedFile] = true;
        }
    }

    // Check <img> sources
    foreach ($dom->getElementsByTagName('img') as $tag) {
        $src = $tag->getAttribute('src');
        $referencedFile = basename($src);
        $referencedPath = $rootDir . $referencedFile;
        if (is_file($referencedPath)) {
            $usedFiles[$referencedFile] = true;
        }
    }

    // Check <script> sources
    foreach ($dom->getElementsByTagName('script') as $tag) {
        $src = $tag->getAttribute('src');
        if (!empty($src)) {
            $referencedFile = basename($src);
            $referencedPath = $rootDir . $referencedFile;
            if (is_file($referencedPath)) {
                $usedFiles[$referencedFile] = true;
            }
        }
    }

    // Check <link> stylesheets
    foreach ($dom->getElementsByTagName('link') as $tag) {
        $href = $tag->getAttribute('href');
        if ($tag->getAttribute('rel') === 'stylesheet' && !empty($href)) {
            $referencedFile = basename($href);
            $referencedPath = $rootDir . $referencedFile;
            if (is_file($referencedPath)) {
                $usedFiles[$referencedFile] = true;
            }
        }
    }

    // 2. Parse PHP include/require references
    $phpPattern = '/(include|require)(_once)?\s*[\'\"](.*?)[\'\"]/i';
    preg_match_all($phpPattern, $content, $matches);
    foreach ($matches[3] as $includePath) {
        $referencedFile = basename($includePath);
        $referencedPath = $rootDir . $referencedFile;
        if (is_file($referencedPath)) {
            $usedFiles[$referencedFile] = true;
        }
    }
}

// Output the list of used files
print_r(array_keys($usedFiles));
?>

Additional Notes #

• Relative paths: The code uses basename() since you only care about the root directory, but if you had subdirectories to consider, you'd need to resolve relative paths properly (e.g., using realpath()).
• Dynamic references: This script won't catch dynamically generated paths (e.g., include $var . '.php'). For those, you'd need more advanced static analysis, but this covers most common cases.
• Test first: Before deleting anything, cross-verify the list with your knowledge of the project—some files might be used in cron jobs, API endpoints, or other places not captured by this scan.

内容的提问来源于stack exchange，提问作者Gabor