<?php
/**
 * Sanitizers shared by settings and per-object SEO fields.
 *
 * @package ElevenRank
 * @since   1.0.0
 */

namespace ElevenRank;

defined( 'ABSPATH' ) || exit;

/**
 * Sanitizers for values core's helpers would damage.
 *
 * @since 1.0.0
 */
final class Sanitize {

	/**
	 * Sanitizes a title or description template such as "%title% %sep% %sitename%".
	 *
	 * `sanitize_text_field()` must not be used for templates: it removes anything that
	 * looks like a percent-encoded octet, so `%date%` and `%category%` would lose their
	 * first characters.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value Raw value.
	 * @return string Plain text on one line, without tags.
	 */
	public static function template( $value ): string {
		if ( ! is_scalar( $value ) ) {
			return '';
		}

		$text      = wp_strip_all_tags( (string) $value );
		$collapsed = preg_replace( '/\s+/u', ' ', $text );

		// A null result means the text is not valid UTF-8; fall back to a byte-wise pass.
		if ( null === $collapsed ) {
			$collapsed = preg_replace( '/\s+/', ' ', $text );
		}

		return trim( (string) $collapsed );
	}

	/**
	 * Sanitizes an http(s) URL. Anything else becomes an empty string.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value Raw value.
	 * @return string
	 */
	public static function url( $value ): string {
		if ( ! is_scalar( $value ) ) {
			return '';
		}

		return esc_url_raw( trim( (string) $value ), array( 'http', 'https' ) );
	}

	/**
	 * Sanitizes a template and cuts it to a maximum length.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value      Raw value.
	 * @param int   $max_length Maximum number of characters; 0 for no limit.
	 * @return string
	 */
	public static function text( $value, int $max_length = 0 ): string {
		$text = self::template( $value );

		return $max_length > 0 && mb_strlen( $text ) > $max_length ? mb_substr( $text, 0, $max_length ) : $text;
	}

	/**
	 * Sanitizes a list of focus keywords: plain text, unique, at most five.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value Raw value.
	 * @return string[]
	 */
	public static function keywords( $value ): array {
		if ( ! is_array( $value ) ) {
			return array();
		}

		$keywords = array();
		$seen     = array();

		foreach ( $value as $item ) {
			$keyword = self::text( $item, 100 );
			$folded  = function_exists( 'mb_strtolower' ) ? mb_strtolower( $keyword ) : strtolower( $keyword );

			if ( '' === $keyword || isset( $seen[ $folded ] ) ) {
				continue;
			}

			$seen[ $folded ] = true;
			$keywords[]      = $keyword;

			if ( count( $keywords ) >= 5 ) {
				break;
			}
		}

		return $keywords;
	}

	/**
	 * Sanitizes the per-object robots settings.
	 *
	 * The five on/off directives are kept only when explicitly set, so "not set" (use the
	 * site default) stays different from "false" (force index). The three limits must
	 * be valid values.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value Raw value.
	 * @return array<string, bool|string>
	 */
	public static function robots( $value ): array {
		if ( ! is_array( $value ) ) {
			return array();
		}

		$clean = array();

		foreach ( array( 'noindex', 'nofollow', 'noarchive', 'nosnippet', 'noimageindex' ) as $flag ) {
			if ( isset( $value[ $flag ] ) && '' !== $value[ $flag ] ) {
				$clean[ $flag ] = rest_sanitize_boolean( $value[ $flag ] );
			}
		}

		$limits = array(
			'max_snippet'       => 'max-snippet',
			'max_image_preview' => 'max-image-preview',
			'max_video_preview' => 'max-video-preview',
		);

		foreach ( $limits as $key => $directive ) {
			$limit = isset( $value[ $key ] ) ? self::robots_limit( $directive, $value[ $key ] ) : '';

			if ( '' !== $limit ) {
				$clean[ $key ] = $limit;
			}
		}

		return $clean;
	}

	/**
	 * Validates a max-snippet / max-image-preview / max-video-preview value.
	 *
	 * @since 1.0.0
	 *
	 * @param string $directive Directive name such as `max-snippet`.
	 * @param mixed  $value     Raw value.
	 * @return string The value, or an empty string when it is not valid.
	 */
	public static function robots_limit( string $directive, $value ): string {
		if ( ! is_scalar( $value ) ) {
			return '';
		}

		$value = strtolower( trim( (string) $value ) );

		if ( 'max-image-preview' === $directive ) {
			return in_array( $value, array( 'none', 'standard', 'large' ), true ) ? $value : '';
		}

		return 1 === preg_match( '/^(-1|\d{1,6})$/', $value ) ? $value : '';
	}

	/**
	 * Cleans the text of a robots.txt file.
	 *
	 * Tags are removed, line endings become "\n", control characters (other than tab and
	 * newline) are dropped, trailing whitespace is trimmed from every line and the text is
	 * cut to a maximum length. The result is plain text; it is sent as `text/plain`, never as HTML.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value      Raw value.
	 * @param int   $max_length Maximum number of characters; 0 means no limit.
	 * @return string
	 */
	public static function robots_txt( $value, int $max_length = 10000 ): string {
		if ( ! is_scalar( $value ) ) {
			return '';
		}

		$text = wp_strip_all_tags( (string) $value, false );
		$text = str_replace( array( "\r\n", "\r" ), "\n", $text );
		$text = (string) preg_replace( '/[^\P{C}\t\n]/u', '', $text );
		$text = (string) preg_replace( '/[ \t]+$/m', '', $text );
		$text = trim( $text, "\n" );

		if ( $max_length > 0 && mb_strlen( $text ) > $max_length ) {
			$text = rtrim( mb_substr( $text, 0, $max_length ) );
		}

		return $text;
	}

	/**
	 * Turns a list into unique positive integers.
	 *
	 * Accepts an array, or text with ids separated by commas, spaces or line breaks.
	 *
	 * @since 1.0.0
	 *
	 * @param mixed $value Raw value.
	 * @param int   $max   Maximum number of ids to keep.
	 * @return int[]
	 */
	public static function ids( $value, int $max = 1000 ): array {
		if ( is_string( $value ) ) {
			$value = preg_split( '/[\s,;]+/', $value, -1, PREG_SPLIT_NO_EMPTY );
		}

		if ( ! is_array( $value ) ) {
			return array();
		}

		$ids = array();

		foreach ( $value as $item ) {
			if ( ! is_scalar( $item ) || 1 !== preg_match( '/^\d{1,10}$/', trim( (string) $item ) ) ) {
				continue;
			}

			$id = (int) $item;

			if ( $id > 0 ) {
				$ids[ $id ] = $id;
			}

			if ( count( $ids ) >= $max ) {
				break;
			}
		}

		return array_values( $ids );
	}
}