<?php
/**
 * Comment extractor (opt-in).
 *
 * @package Unbreak
 */

namespace Unbreak\Extract;

use Unbreak\Support\Url;

defined( 'ABSPATH' ) || exit;

/**
 * Links in approved comments: comment_content and comment_author_url.
 */
final class CommentExtractor {

	public const SOURCE_TYPE = 'comment';

	/**
	 * Extract links from one comment. Returns null when the comment should not be indexed.
	 *
	 * @param int  $comment_id     Comment ID.
	 * @param bool $include_images Include images.
	 * @return array<int,array{url:string,element:string,anchor_text:string,context:string,source_key:string}>|null
	 */
	public static function extract( int $comment_id, bool $include_images = true ): ?array {
		$comment = get_comment( $comment_id );
		if ( ! $comment instanceof \WP_Comment || '1' !== (string) $comment->comment_approved ) {
			return null;
		}
		$base  = (string) get_permalink( (int) $comment->comment_post_ID );
		$base  = '' !== $base ? $base : home_url( '/' );
		$links = HtmlParser::extract( (string) $comment->comment_content, $base, 'comment_content', $include_images );

		if ( '' !== (string) $comment->comment_author_url ) {
			$url = Url::normalize( (string) $comment->comment_author_url, $base );
			if ( null !== $url ) {
				$links[] = array(
					'url'         => $url,
					'element'     => 'a',
					'anchor_text' => HtmlParser::truncate( (string) $comment->comment_author, 255 ),
					'context'     => __( 'Comment author website', 'unbreak' ),
					'source_key'  => 'comment_author_url',
				);
			}
		}
		return $links;
	}
}