<?php
/**
 * Turns a parsed Markdown file into a prepared item and then into a post.
 *
 * @package Draftdrop
 */

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

/**
 * Import logic: validation (prepare) and post creation (create).
 */
class Draftdrop_Importer {

	const MAX_IMAGE_BYTES = 8388608; // 8 MB.
	const HASH_META       = '_draftdrop_hash';

	/**
	 * Validate and normalise one file. Never writes to the database.
	 *
	 * @param string $raw      File contents.
	 * @param string $filename Original file name (display only).
	 * @return array Prepared item.
	 */
	public static function prepare( $raw, $filename ) {
		$raw  = (string) $raw;
		$item = array(
			'filename'         => (string) $filename,
			'errors'           => array(),
			'warnings'         => array(),
			'notes'            => array(),
			'title'            => '',
			'slug'             => '',
			'status'           => 'draft',
			'excerpt'          => '',
			'categories'       => array(),
			'tags'             => array(),
			'seo_title'        => '',
			'meta_description' => '',
			'focus_keyword'    => '',
			'featured_image'   => '',
			'image_alt'        => '',
			'content'          => '',
			'hash'             => sha1( $raw ),
			'duplicate'        => 0,
		);

		$valid_utf8 = function_exists( 'wp_is_valid_utf8' ) ? wp_is_valid_utf8( $raw ) : seems_utf8( $raw ); // phpcs:ignore WordPress.WP.DeprecatedFunctions.seems_utf8Found -- only used on WordPress older than 6.9.
		if ( ! $valid_utf8 ) {
			$item['errors'][] = __( 'The file is not valid UTF-8.', 'draftdrop' );
			return $item;
		}

		$parsed = Draftdrop_Frontmatter::parse( $raw );
		$meta   = $parsed['meta'];

		if ( in_array( 'no_front_matter', $parsed['errors'], true ) ) {
			$item['errors'][] = __( 'No front matter found. The file must start with a line of three dashes.', 'draftdrop' );
			return $item;
		}
		if ( in_array( 'bad_line', $parsed['errors'], true ) ) {
			$item['warnings'][] = __( 'Some front matter lines could not be read and were ignored.', 'draftdrop' );
		}

		$item['title'] = self::text( isset( $meta['title'] ) ? $meta['title'] : '' );
		if ( '' === $item['title'] ) {
			$item['errors'][] = __( 'The title is missing.', 'draftdrop' );
		}

		$slug_source  = isset( $meta['slug'] ) ? self::text( $meta['slug'] ) : '';
		$item['slug'] = sanitize_title( '' !== $slug_source ? $slug_source : $item['title'] );
		if ( '' === $item['slug'] && '' !== $item['title'] ) {
			$item['errors'][] = __( 'A valid slug could not be created.', 'draftdrop' );
		}

		$status = isset( $meta['status'] ) ? strtolower( self::text( $meta['status'] ) ) : 'draft';
		if ( '' === $status ) {
			$status = 'draft';
		}
		if ( ! in_array( $status, array( 'draft', 'pending' ), true ) ) {
			/* translators: %s: status given in the file. */
			$item['warnings'][] = sprintf( __( 'Status "%s" is not allowed in this version. Saved as draft.', 'draftdrop' ), $status );
			$status             = 'draft';
		}
		$item['status'] = $status;

		$item['excerpt'] = isset( $meta['excerpt'] ) ? sanitize_textarea_field( is_array( $meta['excerpt'] ) ? implode( ' ', $meta['excerpt'] ) : (string) $meta['excerpt'] ) : '';

		$cats               = isset( $meta['categories'] ) ? $meta['categories'] : ( isset( $meta['category'] ) ? $meta['category'] : array() );
		$item['categories'] = self::terms( $cats, false );
		$item['tags']       = self::terms( isset( $meta['tags'] ) ? $meta['tags'] : array(), true );

		$item['seo_title']        = self::text( isset( $meta['seo_title'] ) ? $meta['seo_title'] : '' );
		$item['meta_description'] = self::text( isset( $meta['meta_description'] ) ? $meta['meta_description'] : '' );
		$item['focus_keyword']    = self::text( isset( $meta['focus_keyword'] ) ? $meta['focus_keyword'] : '' );
		$item['image_alt']        = self::text( isset( $meta['image_alt'] ) ? $meta['image_alt'] : '' );

		if ( '' === $item['meta_description'] ) {
			$item['warnings'][] = __( 'No meta description.', 'draftdrop' );
		} elseif ( self::length( $item['meta_description'] ) > 160 ) {
			/* translators: %d: number of characters. */
			$item['warnings'][] = sprintf( __( 'Meta description is %d characters (about 160 recommended).', 'draftdrop' ), self::length( $item['meta_description'] ) );
		}
		if ( '' !== $item['seo_title'] && self::length( $item['seo_title'] ) > 60 ) {
			/* translators: %d: number of characters. */
			$item['warnings'][] = sprintf( __( 'SEO title is %d characters (about 60 recommended).', 'draftdrop' ), self::length( $item['seo_title'] ) );
		}
		if ( '' === $item['focus_keyword'] ) {
			$item['warnings'][] = __( 'No focus keyphrase.', 'draftdrop' );
		}

		$image = isset( $meta['featured_image'] ) ? esc_url_raw( self::text( $meta['featured_image'] ) ) : '';
		if ( '' !== $image ) {
			if ( wp_http_validate_url( $image ) ) {
				$item['featured_image'] = $image;
			} else {
				$item['warnings'][] = __( 'The featured image URL is not valid and will be skipped.', 'draftdrop' );
			}
		}

		// Body: Markdown to HTML, sanitize, then to blocks.
		$markdown  = ltrim( $parsed['body'], "\n" );
		$parsedown = new Draftdrop_Parsedown();
		$html      = $parsedown->text( $markdown );
		$cleaned   = Draftdrop_Sanitizer::clean( $html );
		if ( ! empty( $cleaned['removed'] ) ) {
			/* translators: %s: list of HTML tags, for example "script, iframe". */
			$item['warnings'][] = sprintf( __( 'Unsafe HTML was removed: %s.', 'draftdrop' ), implode( ', ', $cleaned['removed'] ) );
		}
		$html = wp_kses_post( $cleaned['html'] );

		if ( preg_match( '#^\s*<h1[^>]*>.*?</h1>\s*#is', $html ) ) {
			$html            = preg_replace( '#^\s*<h1[^>]*>.*?</h1>\s*#is', '', $html, 1 );
			$item['notes'][] = __( 'Leading H1 removed (the title comes from front matter).', 'draftdrop' );
		}
		if ( '' === trim( wp_strip_all_tags( $html ) ) && false === stripos( $html, '<img' ) ) {
			$item['errors'][] = __( 'The article body is empty.', 'draftdrop' );
		}

		// Sites that edit posts in the Classic editor get plain HTML; block comments would only add noise there.
		$use_blocks = function_exists( 'use_block_editor_for_post_type' ) ? use_block_editor_for_post_type( 'post' ) : true;
		$use_blocks = (bool) apply_filters( 'draftdrop_use_blocks', $use_blocks );
		if ( $use_blocks ) {
			$item['content'] = Draftdrop_Blocks::convert( $html );
		} else {
			$item['content'] = $html;
			$item['notes'][] = __( 'Classic editor detected: the article is saved as plain HTML (no blocks).', 'draftdrop' );
		}

		if ( false !== stripos( $html, '<img' ) ) {
			$item['notes'][] = __( 'Images inside the text keep their original (remote) URLs.', 'draftdrop' );
		}

		// Duplicate detection.
		if ( empty( $item['errors'] ) ) {
			$item['duplicate'] = self::find_duplicate( $item['slug'], $item['hash'] );
			if ( $item['duplicate'] ) {
				$item['errors'][] = __( 'Already imported or a post with this slug exists. Skipped.', 'draftdrop' );
			}
		}

		return $item;
	}

	/**
	 * Create the post for a prepared item. Call only for items without errors.
	 *
	 * @param array $item Prepared item.
	 * @return array{post_id:int, notes:string[], error:string}
	 */
	public static function create( array $item ) {
		$result = array(
			'post_id' => 0,
			'notes'   => array(),
			'error'   => '',
		);

		if ( ! current_user_can( 'edit_posts' ) ) {
			$result['error'] = __( 'You are not allowed to create posts.', 'draftdrop' );
			return $result;
		}

		$meta = array( self::HASH_META => $item['hash'] );
		// Written at insert time so the SEO plugin indexes the values immediately.
		if ( '' !== $item['seo_title'] ) {
			$meta['_yoast_wpseo_title'] = $item['seo_title'];
		}
		if ( '' !== $item['meta_description'] ) {
			$meta['_yoast_wpseo_metadesc'] = $item['meta_description'];
		}
		if ( '' !== $item['focus_keyword'] ) {
			$meta['_yoast_wpseo_focuskw'] = $item['focus_keyword'];
		}

		$postarr = array(
			'post_type'    => 'post',
			'post_status'  => $item['status'],
			'post_title'   => $item['title'],
			'post_name'    => $item['slug'],
			'post_excerpt' => $item['excerpt'],
			'post_content' => $item['content'],
			'post_author'  => get_current_user_id(),
			'meta_input'   => $meta,
		);

		$cat_ids = self::resolve_categories( $item['categories'], $result['notes'] );
		if ( ! empty( $cat_ids ) ) {
			$postarr['post_category'] = $cat_ids;
		}
		if ( ! empty( $item['tags'] ) ) {
			$postarr['tags_input'] = $item['tags'];
		}

		$post_id = wp_insert_post( wp_slash( $postarr ), true );
		if ( is_wp_error( $post_id ) ) {
			$result['error'] = $post_id->get_error_message();
			return $result;
		}
		$result['post_id'] = (int) $post_id;

		if ( '' !== $item['featured_image'] ) {
			$image = self::sideload_featured_image( $post_id, $item['featured_image'], $item['image_alt'], $item['title'] );
			if ( is_wp_error( $image ) ) {
				/* translators: %s: error message. */
				$result['notes'][] = sprintf( __( 'Featured image skipped: %s', 'draftdrop' ), $image->get_error_message() );
			}
		}

		return $result;
	}

	/**
	 * Look up an existing post with the same content hash or slug.
	 *
	 * @param string $slug Post slug.
	 * @param string $hash Content hash.
	 * @return int Post ID or 0.
	 */
	private static function find_duplicate( $slug, $hash ) {
		$by_hash = get_posts(
			array(
				'post_type'      => 'post',
				'post_status'    => 'any',
				'meta_key'       => self::HASH_META, // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_key
				'meta_value'     => $hash, // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_value
				'posts_per_page' => 1,
				'fields'         => 'ids',
				'no_found_rows'  => true,
			)
		);
		if ( ! empty( $by_hash ) ) {
			return (int) $by_hash[0];
		}

		$by_slug = get_posts(
			array(
				'post_type'      => 'post',
				'post_status'    => 'any',
				'name'           => $slug,
				'posts_per_page' => 1,
				'fields'         => 'ids',
				'no_found_rows'  => true,
			)
		);
		return empty( $by_slug ) ? 0 : (int) $by_slug[0];
	}

	/**
	 * Resolve category names to IDs, creating missing ones when allowed.
	 *
	 * @param string[] $names Category names or slugs.
	 * @param string[] $notes Notes (by reference).
	 * @return int[]
	 */
	private static function resolve_categories( array $names, array &$notes ) {
		$ids = array();
		foreach ( $names as $name ) {
			$term = get_term_by( 'name', $name, 'category' );
			if ( ! $term ) {
				$term = get_term_by( 'slug', sanitize_title( $name ), 'category' );
			}
			if ( $term ) {
				$ids[] = (int) $term->term_id;
				continue;
			}
			if ( current_user_can( 'manage_categories' ) ) {
				$created = wp_insert_term( $name, 'category' );
				if ( ! is_wp_error( $created ) ) {
					$ids[] = (int) $created['term_id'];
					/* translators: %s: category name. */
					$notes[] = sprintf( __( 'Category "%s" was created.', 'draftdrop' ), $name );
					continue;
				}
			}
			/* translators: %s: category name. */
			$notes[] = sprintf( __( 'Category "%s" does not exist and was not assigned.', 'draftdrop' ), $name );
		}
		return array_values( array_unique( $ids ) );
	}

	/**
	 * Download an image and set it as the featured image.
	 *
	 * @param int    $post_id Post ID.
	 * @param string $url     Image URL (http/https, validated).
	 * @param string $alt     Alt text.
	 * @param string $title   Attachment title.
	 * @return true|WP_Error
	 */
	private static function sideload_featured_image( $post_id, $url, $alt, $title ) {
		if ( ! current_user_can( 'upload_files' ) ) {
			return new WP_Error( 'draftdrop_cap', __( 'You are not allowed to upload files.', 'draftdrop' ) );
		}

		require_once ABSPATH . 'wp-admin/includes/file.php';
		require_once ABSPATH . 'wp-admin/includes/media.php';
		require_once ABSPATH . 'wp-admin/includes/image.php';

		$tmp = download_url( $url, 30 ); // Uses wp_safe_remote_get(): private and loopback hosts are refused.
		if ( is_wp_error( $tmp ) ) {
			return $tmp;
		}

		$size = filesize( $tmp );
		if ( false === $size || $size > self::MAX_IMAGE_BYTES ) {
			wp_delete_file( $tmp );
			return new WP_Error( 'draftdrop_size', __( 'The image is larger than 8 MB.', 'draftdrop' ) );
		}

		$info       = wp_getimagesize( $tmp );
		$extensions = array(
			'image/jpeg' => 'jpg',
			'image/png'  => 'png',
			'image/gif'  => 'gif',
			'image/webp' => 'webp',
		);
		if ( ! is_array( $info ) || empty( $info['mime'] ) || ! isset( $extensions[ $info['mime'] ] ) ) {
			wp_delete_file( $tmp );
			return new WP_Error( 'draftdrop_type', __( 'Only JPEG, PNG, GIF and WebP images are accepted.', 'draftdrop' ) );
		}

		$base = sanitize_file_name( pathinfo( (string) wp_parse_url( $url, PHP_URL_PATH ), PATHINFO_FILENAME ) );
		if ( '' === $base ) {
			$base = 'draftdrop-image';
		}
		$file_array = array(
			'name'     => $base . '.' . $extensions[ $info['mime'] ],
			'tmp_name' => $tmp,
		);

		$attachment_id = media_handle_sideload( $file_array, $post_id, $title );
		if ( is_wp_error( $attachment_id ) ) {
			wp_delete_file( $tmp );
			return $attachment_id;
		}

		if ( '' !== $alt ) {
			update_post_meta( $attachment_id, '_wp_attachment_image_alt', $alt );
		}
		set_post_thumbnail( $post_id, $attachment_id );
		return true;
	}

	/**
	 * Sanitize a scalar (or join a list) into plain text.
	 *
	 * @param mixed $value Value.
	 * @return string
	 */
	private static function text( $value ) {
		if ( is_array( $value ) ) {
			$value = implode( ', ', $value );
		}
		return sanitize_text_field( (string) $value );
	}

	/**
	 * Normalise a list of terms.
	 *
	 * @param mixed $value        String or list.
	 * @param bool  $split_commas Split a plain string on commas.
	 * @return string[]
	 */
	private static function terms( $value, $split_commas ) {
		if ( is_string( $value ) ) {
			$value = $split_commas ? explode( ',', $value ) : array( $value );
		}
		if ( ! is_array( $value ) ) {
			return array();
		}
		$out = array();
		foreach ( $value as $term ) {
			$term = sanitize_text_field( (string) $term );
			if ( '' !== $term ) {
				$out[] = $term;
			}
		}
		return array_values( array_unique( $out ) );
	}

	/**
	 * Character length that is safe for multibyte text.
	 *
	 * @param string $s String.
	 * @return int
	 */
	private static function length( $s ) {
		return function_exists( 'mb_strlen' ) ? mb_strlen( $s, 'UTF-8' ) : strlen( $s );
	}
}