<?php
/**
 * Content Analyzer.
 *
 * Analyzes page content via the AI Schema Gen SaaS backend to detect
 * the best schema type, extract entities, and provide recommendations.
 * All AI processing happens server-side — no direct AI provider calls
 * are made from the plugin.
 *
 * @package Schema_AI
 */

if ( ! defined( 'ABSPATH' ) ) {
    exit;
}

class Schema_AI_Content_Analyzer {

    /** @var Schema_AI_Cache */
    private $cache;

    public function __construct() {
        $this->cache = new Schema_AI_Cache();
    }

    /**
     * Analyze page content and detect schema type via the SaaS backend.
     *
     * @param int $post_id WordPress post ID.
     * @return array Analysis result.
     */
    public function analyze_page( $post_id ) {
        $cache_key = 'analysis_' . $post_id . '_' . $this->get_content_hash( $post_id );
        $cached    = $this->cache->get( $cache_key );

        if ( false !== $cached ) {
            return $cached;
        }

        $settings = get_option( 'schema_ai_settings', array() );
        $api_key  = ! empty( $settings['api_key'] ) ? $settings['api_key'] : '';

        if ( empty( $api_key ) ) {
            return $this->fallback_analysis();
        }

        $content  = $this->prepare_content( $post_id );
        $api_base = defined( 'SCHEMA_AI_API_BASE' ) ? SCHEMA_AI_API_BASE : 'https://api.aischemagen.com/api/v1';

        $response = wp_remote_post(
            $api_base . '/analyze',
            array(
                'timeout' => 30,
                'headers' => array(
                    'Content-Type' => 'application/json',
                    'X-API-Key'    => $api_key,
                ),
                'body'    => wp_json_encode( array( 'content' => $content ) ),
            )
        );

        if ( is_wp_error( $response ) || 200 !== wp_remote_retrieve_response_code( $response ) ) {
            return $this->fallback_analysis();
        }

        $body     = json_decode( wp_remote_retrieve_body( $response ), true );
        $analysis = isset( $body['data'] ) ? $body['data'] : null;

        if ( ! is_array( $analysis ) ) {
            return $this->fallback_analysis();
        }

        $this->cache->set( $cache_key, $analysis, WEEK_IN_SECONDS );

        return $analysis;
    }

    /**
     * Prepare content payload to send to the backend.
     */
    public function prepare_content( $post_id ) {
        $post = get_post( $post_id );
        if ( ! $post ) {
            return array();
        }

        return array(
            'url'        => get_permalink( $post_id ),
            'title'      => $post->post_title,
            'content'    => $this->clean_content( $post->post_content ),
            'excerpt'    => get_the_excerpt( $post_id ),
            'post_type'  => $post->post_type,
            'categories' => $this->get_categories( $post_id ),
            'tags'       => $this->get_tags( $post_id ),
            'headings'   => $this->extract_headings( $post->post_content ),
            'images'     => $this->get_images( $post_id ),
            'published'  => $post->post_date,
            'modified'   => $post->post_modified,
        );
    }

    /**
     * Default analysis result when the backend is unreachable.
     */
    private function fallback_analysis() {
        return array(
            'primary_schema'      => 'WebPage',
            'confidence'          => 0.5,
            'alternative_schemas' => array(),
            'detected_entities'   => array(),
            'missing_content'     => array(),
            'reasoning'           => 'Could not reach AI Schema Gen backend.',
        );
    }

    private function clean_content( $content ) {
        $content = do_shortcode( do_blocks( $content ) );
        $content = wp_strip_all_tags( $content );
        return preg_replace( '/\s+/', ' ', trim( $content ) );
    }

    private function get_content_hash( $post_id ) {
        $post = get_post( $post_id );
        return $post ? md5( $post->post_title . '|' . $post->post_content ) : '';
    }

    private function get_categories( $post_id ) {
        $terms = get_the_terms( $post_id, 'category' );
        return ( $terms && ! is_wp_error( $terms ) ) ? wp_list_pluck( $terms, 'name' ) : array();
    }

    private function get_tags( $post_id ) {
        $terms = get_the_terms( $post_id, 'post_tag' );
        return ( $terms && ! is_wp_error( $terms ) ) ? wp_list_pluck( $terms, 'name' ) : array();
    }

    /**
     * Extract headings from HTML content.
     */
    public function extract_headings( $content ) {
        preg_match_all( '/<h([1-6])[^>]*>(.*?)<\/h\1>/i', $content, $matches );
        $headings = array();
        for ( $i = 0; $i < count( $matches[0] ); $i++ ) {
            $headings[] = array(
                'level' => (int) $matches[1][ $i ],
                'text'  => wp_strip_all_tags( $matches[2][ $i ] ),
            );
        }
        return $headings;
    }

    /**
     * Get images from the post.
     */
    public function get_images( $post_id ) {
        $images       = array();
        $thumbnail_id = get_post_thumbnail_id( $post_id );
        if ( $thumbnail_id ) {
            $src = wp_get_attachment_image_src( $thumbnail_id, 'full' );
            if ( $src ) {
                $images[] = array(
                    'url'    => $src[0],
                    'width'  => $src[1],
                    'height' => $src[2],
                );
            }
        }
        return $images;
    }
}