<?php
/**
 * LLM Bot Crawler Logger
 * Logs AI crawler visits to bot-crawl-log.tsv
 * Only receives requests from bots matched in .htaccess RewriteCond
 */

$ua = $_SERVER['HTTP_USER_AGENT'] ?? '';
$path = $_GET['path'] ?? '/';
$uri = '/' . ltrim($path, '/');

// Identify which bot
$bots = [
    // Training crawlers
    'GPTBot', 'ClaudeBot', 'Google-Extended', 'Applebot-Extended',
    'Bytespider', 'Amazonbot', 'meta-externalagent', 'cohere-ai', 'CCBot',
    // Search/citation crawlers
    'OAI-SearchBot', 'ChatGPT-User', 'Claude-SearchBot', 'PerplexityBot',
    'Googlebot', 'bingbot', 'Google-Agent',
    // Other
    'YandexBot', 'DuckDuckBot', 'Baiduspider'
];

$matched = 'unknown';
foreach ($bots as $bot) {
    if (stripos($ua, $bot) !== false) {
        $matched = $bot;
        break;
    }
}

// Log: timestamp, IP, bot name, URI, full UA
$line = date('Y-m-d H:i:s') . "\t"
    . ($_SERVER['REMOTE_ADDR'] ?? '') . "\t"
    . $matched . "\t"
    . $uri . "\t"
    . $ua . "\n";

file_put_contents(__DIR__ . '/bot-crawl-log.tsv', $line, FILE_APPEND | LOCK_EX);

// Serve the original file
$file = __DIR__ . $uri;
if (is_dir($file)) {
    $file = rtrim($file, '/') . '/index.html';
}
// Clean URL support: try appending .html if file not found
if (!file_exists($file) && !pathinfo($file, PATHINFO_EXTENSION)) {
    $htmlFile = rtrim($file, '/') . '.html';
    if (file_exists($htmlFile)) {
        $file = $htmlFile;
    }
}

if (!file_exists($file)) {
    http_response_code(404);
    echo '404 Not Found';
    exit;
}

$ext = strtolower(pathinfo($file, PATHINFO_EXTENSION));
$types = [
    'html' => 'text/html; charset=UTF-8',
    'xml'  => 'application/xml',
    'txt'  => 'text/plain',
    'css'  => 'text/css',
    'js'   => 'application/javascript',
    'svg'  => 'image/svg+xml',
    'json' => 'application/json',
    'png'  => 'image/png',
    'jpg'  => 'image/jpeg',
    'jpeg' => 'image/jpeg',
    'webp' => 'image/webp',
    'ico'  => 'image/x-icon',
    'gif'  => 'image/gif',
    'pdf'  => 'application/pdf',
];

header('Content-Type: ' . ($types[$ext] ?? 'application/octet-stream'));
readfile($file);