<?php
/**
 * Link-density signal check.
 *
 * @package ZeroSpamForms
 */

namespace TwoOTwoCode\ZeroSpamForms\Engine\Checks;

defined( 'ABSPATH' ) || exit;

use TwoOTwoCode\ZeroSpamForms\Engine\CheckResult;
use TwoOTwoCode\ZeroSpamForms\Engine\ReasonCodeRegistry;
use TwoOTwoCode\ZeroSpamForms\Engine\SettingsSnapshot;
use TwoOTwoCode\ZeroSpamForms\Engine\SpamCheckInterface;
use TwoOTwoCode\ZeroSpamForms\Engine\SubmissionContext;

/**
 * Flags submissions with unusually dense links.
 */
final class LinkDensityCheck implements SpamCheckInterface {
	/**
	 * Minimum links before density can be suspicious.
	 */
	private const MINIMUM_LINKS = 2;

	/**
	 * Get check name.
	 */
	public function get_name(): string {
		return 'link_density';
	}

	/**
	 * Whether the check is enabled.
	 *
	 * @param SettingsSnapshot $settings Immutable settings snapshot.
	 */
	public function is_enabled( SettingsSnapshot $settings ): bool {
		return $settings->is_check_enabled( 'link_density_check_enabled' );
	}

	/**
	 * Evaluate link density.
	 *
	 * @param SubmissionContext $context  Normalized submission context.
	 * @param SettingsSnapshot  $settings Immutable settings snapshot.
	 */
	public function evaluate( SubmissionContext $context, SettingsSnapshot $settings ): CheckResult {
		$text = implode( ' ', $context->get_field_values() );

		if ( '' === trim( $text ) ) {
			return CheckResult::pass();
		}

		$link_count = $this->count_links( $text );

		if ( $link_count < self::MINIMUM_LINKS ) {
			return CheckResult::pass();
		}

		$token_count = $this->count_tokens( $text );
		$density     = $link_count / max( 1, $token_count );

		if ( $density > $settings->get_link_density_threshold() ) {
			return CheckResult::suspicious( ReasonCodeRegistry::EXCESSIVE_LINKS );
		}

		return CheckResult::pass();
	}

	/**
	 * Count URLs.
	 *
	 * @param string $text Text to inspect.
	 */
	private function count_links( string $text ): int {
		preg_match_all( '/(?:https?:\/\/|www\.)[^\s<>"\']+/i', $text, $matches );

		return count( $matches[0] );
	}

	/**
	 * Count whitespace-delimited tokens.
	 *
	 * @param string $text Text to inspect.
	 */
	private function count_tokens( string $text ): int {
		$tokens = preg_split( '/\s+/', trim( $text ) );

		if ( ! is_array( $tokens ) ) {
			return 0;
		}

		return count(
			array_filter(
				$tokens,
				static function ( string $token ): bool {
					return '' !== $token;
				}
			)
		);
	}
}