<?php
namespace Elementor\Modules\Agents\Components\Readability;
use Elementor\Modules\Agents\Components\Readability\Extractors\Block_Extractor;
use Elementor\Modules\Agents\Components\Readability\Extractors\Classic_Extractor;
use Elementor\Modules\Agents\Components\Readability\Extractors\Elementor_Extractor;
use Elementor\Modules\Agents\Components\Readability\Extractors\Extractor_Interface;
use Elementor\Modules\Agents\Components\Readability\Extractors\Rendered_Extractor;
if ( ! defined( 'ABSPATH' ) ) {
exit;
}
/**
* Orchestrates the content-extractor chain.
*
* Extractors are walked in ascending priority order. The first extractor
* whose `can_handle()` returns true AND whose result contains at least
* {@see Content_Extractor::MIN_CONTENT_CHARS} non-whitespace characters is
* used. Extraction falls through to the next extractor on an empty or
* below-threshold result.
*
* Third-party extractors may be injected via the
* `elementor/agents/markdown/extractors` filter.
*
* Scope can be narrowed to Elementor-built content only via the
* `elementor/agents/markdown/coverage` filter (values: `all`, `elementor_only`).
*/
class Content_Extractor {
/**
* Minimum number of non-whitespace characters an extraction must contain
* before it is accepted as a valid result.
*/
const MIN_CONTENT_CHARS = 20;
/**
* @var Extractor_Interface[]
*/
private array $extractors = [];
/**
* Constructor — registers the built-in extractors and applies the
* third-party injection filter.
*/
public function __construct() {
$defaults = [
new Elementor_Extractor(),
new Block_Extractor(),
new Classic_Extractor(),
new Rendered_Extractor(),
];
/**
* Filter the list of content extractors.
*
* Third-party code may add, remove, or reorder extractors.
* Each entry must implement Extractor_Interface.
*
* @param Extractor_Interface[] $extractors
*/
$extractors = apply_filters( 'elementor/agents/markdown/extractors', $defaults );
$extractors = array_values( array_filter(
$extractors,
static fn( $extractor ) => $extractor instanceof Extractor_Interface
) );
usort( $extractors, static fn( Extractor_Interface $a, Extractor_Interface $b ) => $a->get_priority() <=> $b->get_priority() );
$this->extractors = $extractors;
}
/**
* Run the extractor chain once and return both the winning extractor ID
* and the extracted body markdown.
*
* @param \WP_Post $post
* @return array{id: string, body: string}
*/
public function extract_with_id( \WP_Post $post ): array {
foreach ( $this->get_active_extractors( $post ) as $extractor ) {
if ( ! $extractor->can_handle( $post ) ) {
continue;
}
$result = $extractor->extract( $post );
if ( $this->is_sufficient( $result ) ) {
return [
'id' => $extractor->get_id(),
'body' => $result,
];
}
}
return [
'id' => '',
'body' => '',
];
}
/**
* Extract body markdown from a post using the extractor chain.
*
* Returns only body content — no frontmatter.
*
* @param \WP_Post $post
* @return string Body markdown, or '' when no extractor produces content.
*/
public function extract( \WP_Post $post ): string {
return $this->extract_with_id( $post )['body'];
}
/**
* Return the ID of the extractor that would produce content for the post.
*
* When both the extractor ID and body are needed, prefer `extract_with_id()`.
*
* @param \WP_Post $post
* @return string Extractor ID, or '' when no extractor produces content.
*/
public function get_extractor_id( \WP_Post $post ): string {
return $this->extract_with_id( $post )['id'];
}
// -------------------------------------------------------------------------
// Private helpers
// -------------------------------------------------------------------------
/**
* Return the list of extractors filtered by the coverage setting.
*
* When the `elementor/agents/markdown/coverage` filter returns
* `elementor_only`, only the Elementor_Extractor is included.
*
* @param \WP_Post $post
* @return Extractor_Interface[]
*/
private function get_active_extractors( \WP_Post $post ): array {
/**
* Filter the extraction coverage mode.
*
* Accepted values:
* 'all' – use the full extractor chain (default).
* 'elementor_only' – skip non-Elementor extractors.
*
* @param string $coverage
* @param \WP_Post $post
*/
$coverage = (string) apply_filters( 'elementor/agents/markdown/coverage', 'all', $post );
if ( 'elementor_only' === $coverage ) {
return array_filter(
$this->extractors,
static fn( Extractor_Interface $e ) => 'elementor' === $e->get_id()
);
}
return $this->extractors;
}
/**
* Return true when the extracted string is non-empty and meets the minimum
* non-whitespace character threshold.
*
* @param string $result
* @return bool
*/
private function is_sufficient( string $result ): bool {
if ( '' === $result ) {
return false;
}
return strlen( preg_replace( '/\s+/', '', $result ) ) >= self::MIN_CONTENT_CHARS;
}
}