<?php
namespace ActiveTwist\SEO\Utils;

class IndexabilityEvaluator
{
    const STATUS_INDEXABLE = 'indexable';
    const STATUS_NOINDEX = 'noindex';
    const STATUS_BLOCKED = 'blocked';
    const STATUS_INDETERMINATE = 'indeterminate';

    /**
     * Determine the indexability status of a URL based on inspection data.
     *
     * @param array $inspection Result from Inspector::inspect_url()
     * @param string $url The outcome URL (permalinK)
     * @return array [ 'status' => string, 'reason' => string ]
     */
    public static function evaluate($inspection, $url)
    {
        // 1. Check HTTP Status
        // If not 200, it's strictly not indexable content (though 3xx redirects index flow)
        if ($inspection['code'] !== 200) {
            return [
                'status' => self::STATUS_NOINDEX,
                'reason' => 'HTTP ' . $inspection['code']
            ];
        }

        // 2. Check Robots.txt (Plugin Controlled)
        if (self::is_robots_blocked($url)) {
            return [
                'status' => self::STATUS_BLOCKED,
                'reason' => 'Blocked by robots.txt'
            ];
        }

        // 3. Check Directives (Meta & Headlines)
        $noindex = self::has_directive($inspection, 'noindex');
        $index = self::has_directive($inspection, 'index');

        if ($noindex && $index) {
            return [
                'status' => self::STATUS_INDETERMINATE,
                'reason' => 'Conflicting "index" and "noindex" directives'
            ];
        }

        if ($noindex) {
            return [
                'status' => self::STATUS_NOINDEX,
                'reason' => 'Explicit "noindex" directive'
            ];
        }

        // Default: Indexable
        return [
            'status' => self::STATUS_INDEXABLE,
            'reason' => 'Indexable (Default)'
        ];
    }

    private static function has_directive($inspection, $directive)
    {
        $sources = array_merge($inspection['robots_meta'], $inspection['x_robots_tag']);
        foreach ($sources as $source) {
            // Parse "noindex, follow" -> ["noindex", "follow"]
            $parts = explode(',', $source);
            foreach ($parts as $part) {
                if (trim(strtolower($part)) === $directive) {
                    return true;
                }
            }
        }
        return false;
    }

    private static function is_robots_blocked($url)
    {
        // Get effective robots.txt content
        $content = \ActiveTwist\SEO\Settings\Options::get('robots_txt_content');
        if (empty($content)) {
            $content = \ActiveTwist\SEO\Features\RobotsController::get_default_content();
        }

        // Parse path from URL
        $path = wp_parse_url($url, PHP_URL_PATH);
        if (empty($path)) {
            $path = '/';
        }

        // Simple User-Agent: * parser
        $lines = explode("\n", $content);
        $ua_applies = true; // Assuming * or matching UA logic is simple here

        foreach ($lines as $line) {
            $line = trim($line);
            if (empty($line) || strpos($line, '#') === 0)
                continue;

            // Very basic parser: User-agent reset
            if (stripos($line, 'User-agent:') === 0) {
                $ua = trim(substr($line, 11));
                $ua_applies = ($ua === '*'); // Only care about global rules for now
                continue;
            }

            if (!$ua_applies)
                continue;

            if (stripos($line, 'Disallow:') === 0) {
                $rule = trim(substr($line, 9));
                if (empty($rule))
                    continue; // Disallow: empty means allow all

                // Check prefix match
                // Rule: /wp-admin/ matches /wp-admin/foo
                if (strpos($path, $rule) === 0) {
                    return true;
                }
            }
        }

        return false;
    }
}