<?php
if (!defined('ABSPATH')) exit;

/**
 * ADL-024: Duplicate Page Title and H1 Prevention.
 *
 * Keeps the WordPress title as the single H1 for standard pages and prevents
 * AI-generated content from repeating that title as its first heading.
 */
class AIWEST_Page_Title_Heading_Normalizer {
    const META_REPORT = '_aiwest_duplicate_title_heading_last_report';
    protected static $saving = false;

    public static function init() {
        add_action('save_post', [__CLASS__, 'normalize_saved_post'], 20, 3);
    }

    public static function normalize_saved_post($post_id, $post, $update) {
        if (self::$saving || !$post || !in_array($post->post_type, ['page', 'post'], true)) return;
        if (wp_is_post_revision($post_id) || wp_is_post_autosave($post_id)) return;
        if (defined('DOING_AUTOSAVE') && DOING_AUTOSAVE) return;
        if (function_exists('current_user_can') && !current_user_can('edit_post', $post_id)) return;

        $mode = get_post_meta($post_id, '_aiwest_custom_hero_mode', true) === 'yes' ? 'custom_hero' : 'standard';
        $result = self::normalize($post_id, (string)$post->post_content, [
            'post_title' => (string)$post->post_title,
            'post_type' => (string)$post->post_type,
            'mode' => $mode,
        ]);
        self::record_report($post_id, $result['report']);
        if ($result['content'] === (string)$post->post_content) return;

        self::$saving = true;
        wp_update_post([
            'ID' => $post_id,
            'post_content' => $result['content'],
        ]);
        self::$saving = false;
    }

    public static function normalize($post_id, $content, $context = []) {
        $post_id = absint($post_id);
        $context = is_array($context) ? $context : [];
        $post = $post_id ? get_post($post_id) : null;
        $title = trim((string)($context['post_title'] ?? ($post ? $post->post_title : '')));
        $post_type = sanitize_key((string)($context['post_type'] ?? ($post ? $post->post_type : 'page')));
        $mode = sanitize_key((string)($context['mode'] ?? 'standard'));
        if (!in_array($mode, ['standard', 'custom_hero'], true)) $mode = 'standard';

        $theme_title_visible = self::theme_title_is_visible($post_id, $context);
        $content = (string)$content;
        $original = $content;
        $duplicate = false;
        $removed = '';
        $downgraded = 0;
        $first_before = self::first_heading($content);

        if ($title !== '' && $first_before && self::titles_match($title, $first_before['text'])) {
            $duplicate = true;
            if ($theme_title_visible) {
                $removed = $first_before['text'];
                $content = substr_replace($content, '', $first_before['offset'], $first_before['length']);
            } elseif ($mode === 'custom_hero' && intval($first_before['level']) !== 1) {
                $content = self::replace_heading_level_at($content, $first_before, 1);
            }
        }

        if ($theme_title_visible && $mode === 'standard') {
            $content = self::downgrade_body_h1s($content, $downgraded);
        }

        $first_after = self::first_heading($content);
        $manual_action = ($mode === 'custom_hero' && $theme_title_visible && self::contains_h1($content));
        $report = [
            'duplicate_title_detected' => $duplicate,
            'page_title' => wp_strip_all_tags($title),
            'removed_duplicated_heading' => $removed,
            'first_content_heading_before_normalization' => $first_before ? $first_before['text'] : '',
            'first_content_heading_after_normalization' => $first_after ? $first_after['text'] : '',
            'theme_title_visible' => $theme_title_visible,
            'theme_title_should_be_hidden' => $manual_action,
            'custom_hero_mode' => $mode === 'custom_hero',
            'downgraded_h1_count' => $downgraded,
            'changed' => $content !== $original,
            'message' => $removed !== ''
                ? 'Duplicate page title heading was removed to prevent repeated H1 output.'
                : ($manual_action ? 'This page uses a custom hero H1. Hide the automatic theme page title to avoid duplicate headings.' : ''),
            'post_type' => $post_type,
            'checked_at' => function_exists('current_time') ? current_time('mysql') : gmdate('Y-m-d H:i:s'),
        ];

        return ['content' => $content, 'report' => $report];
    }

    public static function normalize_content($post_id, $content, $context = []) {
        $result = self::normalize($post_id, $content, $context);
        return $result['content'];
    }

    public static function record_report($post_id, $report) {
        $post_id = absint($post_id);
        if (!$post_id || !is_array($report)) return false;
        return update_post_meta($post_id, self::META_REPORT, $report);
    }

    public static function titles_match($page_title, $heading) {
        $title = self::normalize_title($page_title);
        $heading = self::normalize_title($heading);
        if ($title === '' || $heading === '') return false;
        if ($title === $heading) return true;

        $aliases = [
            'frequently asked questions' => 'faq',
            'frequently asked question' => 'faq',
            'questions and answers' => 'faq',
        ];
        $title = $aliases[$title] ?? $title;
        $heading = $aliases[$heading] ?? $heading;
        if ($title === $heading) return true;

        similar_text($title, $heading, $percent);
        if ($percent >= 86) return true;

        $shorter = strlen($title) <= strlen($heading) ? $title : $heading;
        $longer = $shorter === $title ? $heading : $title;
        if (strlen($shorter) >= 5 && strpos($longer, $shorter) !== false) {
            $extra = trim(str_replace($shorter, '', $longer));
            if ($extra === '' || preg_match('/^(?:page|guide|overview|wordpress|website)$/', $extra)) return true;
        }
        return false;
    }

    protected static function normalize_title($text) {
        $text = html_entity_decode(wp_strip_all_tags((string)$text), ENT_QUOTES | ENT_HTML5, 'UTF-8');
        $text = self::lower($text);
        $text = str_replace(['&', '+'], ' and ', $text);
        $text = preg_replace('/[^\p{L}\p{N}]+/u', ' ', $text);
        $text = trim(preg_replace('/\s+/u', ' ', $text));

        $brands = ['ai webmaster studio', 'ai webmaster foundation', 'ai webmaster', 'aiwos cloud', 'aiwos'];
        $site = function_exists('get_bloginfo') ? self::lower(trim(wp_strip_all_tags((string)get_bloginfo('name')))) : '';
        if ($site !== '') $brands[] = preg_replace('/[^\p{L}\p{N}]+/u', ' ', $site);
        $brands = array_unique($brands);
        usort($brands, function($left, $right) {
            return strlen($right) <=> strlen($left);
        });
        foreach ($brands as $brand) {
            $brand = trim($brand);
            if ($brand === '') continue;
            $text = preg_replace('/^(?:' . preg_quote($brand, '/') . ')\s+/u', '', $text);
            $text = preg_replace('/\s+(?:' . preg_quote($brand, '/') . ')$/u', '', $text);
        }
        return trim(preg_replace('/\s+/u', ' ', $text));
    }

    protected static function first_heading($content) {
        $pattern = '/(?:<!--\s*wp:heading\b[^>]*-->\s*)?<h([1-3])\b([^>]*)>(.*?)<\/h\1>\s*(?:<!--\s*\/wp:heading\s*-->)?/is';
        if (!preg_match($pattern, (string)$content, $match, PREG_OFFSET_CAPTURE)) return null;
        return [
            'level' => intval($match[1][0]),
            'attributes' => $match[2][0],
            'inner_html' => $match[3][0],
            'text' => trim(wp_strip_all_tags($match[3][0])),
            'offset' => intval($match[0][1]),
            'length' => strlen($match[0][0]),
        ];
    }

    protected static function replace_heading_level_at($content, $heading, $level) {
        $replacement = '<h' . intval($level) . $heading['attributes'] . '>' . $heading['inner_html'] . '</h' . intval($level) . '>';
        return substr_replace($content, $replacement, $heading['offset'], $heading['length']);
    }

    protected static function downgrade_body_h1s($content, &$count) {
        $content = preg_replace_callback(
            '/(<!--\s*wp:heading\b.*?-->\s*)<h1\b([^>]*)>(.*?)<\/h1>(\s*<!--\s*\/wp:heading\s*-->)/is',
            function($match) use (&$count) {
                $count++;
                $opening = preg_replace('/"level"\s*:\s*1/', '"level":2', $match[1]);
                return $opening . '<h2' . $match[2] . '>' . $match[3] . '</h2>' . $match[4];
            },
            (string)$content
        );

        return preg_replace_callback('/<h1\b([^>]*)>(.*?)<\/h1>/is', function($match) use (&$count) {
            $count++;
            return '<h2' . $match[1] . '>' . $match[2] . '</h2>';
        }, $content);
    }

    protected static function contains_h1($content) {
        return preg_match('/<h1\b/i', (string)$content) === 1;
    }

    protected static function theme_title_is_visible($post_id, $context) {
        if (array_key_exists('theme_title_visible', $context)) return (bool)$context['theme_title_visible'];
        if (!empty($context['theme_title_hidden'])) return false;

        $visible = true;
        if ($post_id) {
            if (class_exists('AIWEST_Canvas') && AIWEST_Canvas::is_canvas_page($post_id)) {
                return get_post_meta($post_id, '_aiwest_custom_hero_mode', true) !== 'yes';
            }
            $hidden_values = ['1', 'yes', 'true', 'on', 'disabled', 'hide', 'hidden'];
            foreach (['_astra_disable_title', '_disable_page_title', 'hide_page_title', '_hide_title', 'site-post-title'] as $key) {
                $value = self::lower((string)get_post_meta($post_id, $key, true));
                if (in_array($value, $hidden_values, true)) {
                    $visible = false;
                    break;
                }
            }
            $elementor = get_post_meta($post_id, '_elementor_page_settings', true);
            if (is_array($elementor) && !empty($elementor['hide_title']) && in_array(self::lower((string)$elementor['hide_title']), $hidden_values, true)) {
                $visible = false;
            }
        }

        return (bool)apply_filters('aiwest_theme_displays_page_title', $visible, $post_id, $context);
    }

    protected static function lower($text) {
        return function_exists('mb_strtolower') ? mb_strtolower((string)$text, 'UTF-8') : strtolower((string)$text);
    }
}

function aiwest_normalize_duplicate_page_title_heading($post_id, $content, $context = []) {
    return AIWEST_Page_Title_Heading_Normalizer::normalize($post_id, $content, $context);
}

AIWEST_Page_Title_Heading_Normalizer::init();