' . \esc_html( $error_message ) . '
'; \printf( /* translators: 1: php.ini, 2: post_max_size, 3: upload_max_filesize */ \esc_html__( 'File is empty. Please upload something more substantial. This error could also be caused by uploads being disabled in your %1$s file or by %2$s being defined as smaller than %3$s in %1$s.', 'activitypub' ), 'php.ini', 'post_max_size', 'upload_max_filesize' ); echo '

'; return false; } $file_info = \wp_check_filetype( \sanitize_file_name( $_FILES['import']['name'] ), array( 'zip' => 'application/zip' ) ); if ( 'application/zip' !== $file_info['type'] ) { echo '

' . \esc_html( $error_message ) . '
'; \esc_html_e( 'The uploaded file must be a ZIP archive. Please try again with the correct file format.', 'activitypub' ); echo '

'; return false; } $overrides = array( 'test_form' => false, 'test_type' => false, ); $upload = \wp_handle_upload( $_FILES['import'], $overrides ); if ( isset( $upload['error'] ) ) { echo '

' . \esc_html( $error_message ) . '
'; echo \esc_html( $upload['error'] ) . '

'; return false; } // Construct the attachment array. $attachment = array( 'post_title' => \wp_basename( $upload['file'] ), 'post_content' => $upload['url'], 'post_mime_type' => $upload['type'], 'guid' => $upload['url'], 'context' => 'import', 'post_status' => 'private', ); // Save the data. self::$import_id = \wp_insert_attachment( $attachment, $upload['file'] ); // Schedule a cleanup for one day from now in case of failed import or missing wp_import_cleanup() call. \wp_schedule_single_event( \time() + DAY_IN_SECONDS, 'importer_scheduled_cleanup', array( self::$import_id ) ); return true; } /** * Import options. */ public static function import_options() { $author = 0; if ( isset( self::$outbox['orderedItems'][0] ) ) { $users = \get_users( array( 'fields' => 'ID', // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_query 'meta_query' => array( array( 'key' => $GLOBALS['wpdb']->get_blog_prefix() . 'activitypub_also_known_as', 'value' => self::$outbox['orderedItems'][0]['actor'], 'compare' => 'LIKE', ), ), ) ); if ( ! empty( $users ) ) { $author = $users[0]; } } ?> wp_content_dir() . 'import/'; self::$archive = $import_folder . \basename( \basename( $file, '.txt' ), '.zip' ); // Clean up working directory. if ( $wp_filesystem->is_dir( self::$archive ) ) { $wp_filesystem->delete( self::$archive, true ); } // Unzip package to working directory. \unzip_file( $file, self::$archive ); self::maybe_unwrap_archive(); if ( ! $wp_filesystem->exists( self::$archive . '/outbox.json' ) ) { echo '

' . \esc_html( $error_message ) . '
'; echo \esc_html__( 'The archive does not contain an Outbox file, please try again.', 'activitypub' ) . '

'; return; } self::$outbox = \json_decode( $wp_filesystem->get_contents( self::$archive . '/outbox.json' ), true ); \wp_suspend_cache_invalidation(); \wp_defer_term_counting( true ); \wp_defer_comment_counting( true ); /** * Fires when the Mastodon import starts. */ \do_action( 'import_start' ); $result = self::import_posts(); \wp_suspend_cache_invalidation( false ); \wp_defer_term_counting( false ); \wp_defer_comment_counting( false ); $wp_filesystem->delete( $import_folder, true ); \wp_import_cleanup( self::$import_id ); if ( \is_wp_error( $result ) ) { echo '

' . \esc_html( $error_message ) . '
'; echo \esc_html( $result->get_error_message() ) . '

'; } else { echo '

'; /* translators: Home URL */ \printf( \wp_kses_post( \__( 'All done. Have fun!', 'activitypub' ) ), \esc_url( \admin_url() ) ); echo '

'; } /** * Fires when the Mastodon import ends. */ \do_action( 'import_end' ); } /** * Process posts. * * Uses a multi-pass approach: * 1. Categorize posts into regular posts and self-replies. * 2. Import regular posts (root posts and external replies) as WordPress posts. * 3. Import self-replies as comments on their parent posts. * * @return true|\WP_Error True on success, WP_Error on failure. */ public static function import_posts() { $skipped = array(); $imported = 0; // Pass 1: Categorize posts. $posts_to_import = array(); $self_replies = array(); foreach ( self::$outbox['orderedItems'] as $post ) { // Skip boosts. if ( 'Announce' === $post['type'] ) { continue; } if ( ! is_activity_public( $post ) ) { continue; } if ( self::is_self_reply( $post ) ) { $self_replies[] = $post; } else { // Root posts and external replies are imported as WordPress posts. $posts_to_import[] = $post; } } // Pass 2: Import regular posts as WordPress posts. $source_to_post_id = array(); foreach ( $posts_to_import as $post ) { $created = false; $result = self::import_as_post( $post, $created ); if ( \is_wp_error( $result ) ) { return $result; } // Map it either way: an already-imported post is still the parent for its replies. $source_to_post_id[ $post['object']['id'] ] = $result; if ( $created ) { ++$imported; } else { $skipped[] = $post['object']['id']; } } // Pass 3: Import self-replies as comments (sorted by date for correct threading). \usort( $self_replies, static function ( $a, $b ) { return \strtotime( $a['published'] ) <=> \strtotime( $b['published'] ); } ); $source_to_comment_id = array(); $comments_skipped = array(); $comments_imported = 0; foreach ( $self_replies as $post ) { $result = self::import_as_comment( $post, $source_to_post_id, $source_to_comment_id ); if ( $result ) { ++$comments_imported; } else { $comments_skipped[] = $post['object']['id']; } } // Output results. if ( ! empty( $skipped ) ) { echo '

' . \esc_html__( 'Skipped posts:', 'activitypub' ) . '
'; echo \wp_kses( \implode( '
', $skipped ), array( 'br' => array() ) ); echo '

'; } if ( ! empty( $comments_skipped ) ) { echo '

' . \esc_html__( 'Skipped comments:', 'activitypub' ) . '
'; echo \wp_kses( \implode( '
', $comments_skipped ), array( 'br' => array() ) ); echo '

'; } /* translators: %s: Number of posts */ echo '

' . \esc_html( \sprintf( \_n( 'Imported %s post.', 'Imported %s posts.', $imported, 'activitypub' ), \number_format_i18n( $imported ) ) ) . '

'; if ( $comments_imported > 0 ) { /* translators: %s: Number of comments */ echo '

' . \esc_html( \sprintf( \_n( 'Imported %s comment from self-reply threads.', 'Imported %s comments from self-reply threads.', $comments_imported, 'activitypub' ), \number_format_i18n( $comments_imported ) ) ) . '

'; } return true; } /** * Check if a post is a self-reply (thread continuation). * * A self-reply is when a user replies to their own post, creating a thread. * * @param array $post The Mastodon activity. * * @return bool True if replying to own post. */ private static function is_self_reply( $post ) { if ( empty( $post['object']['inReplyTo'] ) ) { return false; } /* * Compare base URLs (actor URL should be a prefix of inReplyTo for self-replies). * * Example: * - actor: https://mastodon.social/users/example * - inReplyTo: https://mastodon.social/users/example/statuses/123 * * Adding a trailing slash ensures we don't match partial usernames * (e.g., "example" shouldn't match "example2"). */ return \str_starts_with( $post['object']['inReplyTo'], \rtrim( $post['actor'], '/' ) . '/' ); } /** * Import a single activity as a WordPress post. * * @param array $post The Mastodon activity. * @param bool $created Set to false when the post was already imported and the * return value is the existing id rather than a new one. * * @return int|\WP_Error Post ID, either newly created or the one already imported, * or WP_Error on failure. */ private static function import_as_post( $post, &$created ) { /* * An imported post is handled the same as a federated one: `Sanitize::content()` * normalizes and sanitizes the content. The * import runs as a user with `unfiltered_html`, so kses filters are not installed * for this request and `wp_insert_post()` would otherwise store whatever the archive * contained; these posts are published publicly, so they are the wider exposure. */ $post_data = array( 'post_author' => self::$author, 'post_date' => $post['published'], // Slashed like the federated path: wp_insert_post() unslashes what it is given. 'post_excerpt' => \wp_slash( \is_string( $post['object']['summary'] ?? null ) ? \wp_strip_all_tags( $post['object']['summary'] ) : '' ), 'post_content' => \wp_slash( Sanitize::content( $post['object']['content'] ?? '' ) ), 'post_status' => 'publish', 'post_type' => 'post', 'meta_input' => array( '_source_id' => $post['object']['id'] ), 'tags_input' => \array_map( static function ( $tag ) { if ( 'Hashtag' === $tag['type'] ) { return \ltrim( $tag['name'], '#' ); } return ''; }, $post['object']['tag'] ?? array() ), ); /** * Filter the post data before inserting it into the database. * * @param array $post_data The post data to be inserted. * @param array $post The Mastodon Create activity. */ $post_data = \apply_filters( 'activitypub_import_mastodon_post_data', $post_data, $post ); /* * Match on the archive's own id first. Falling back to `post_exists()` alone keys * de-duplication on an exact `post_content` match, so any change to what we store * (a new sanitizer, for one) makes everything imported by an older version stop * matching and come back as a duplicate. */ $post_exists = self::get_post_by_source_id( $post['object']['id'] ?? '', $post_data['post_type'] ); if ( ! $post_exists ) { $post_exists = \post_exists( '', $post_data['post_content'], $post_data['post_date'], $post_data['post_type'] ); } /** * Filter ID of the existing post corresponding to post currently importing. * * Return 0 to force the post to be imported. Filter the ID to be something else * to override which existing post is mapped to the imported post. * * @see post_exists() * * @param int $post_exists Post ID, or 0 if post did not exist. * @param array $post_data The post array to be inserted. */ $post_exists = \apply_filters( 'wp_import_existing_post', $post_exists, $post_data ); if ( $post_exists ) { /* * Report it as already imported, but hand the id back: pass 3 maps self-replies * onto their parent through this return value, and a reply whose parent we * skipped would otherwise find no parent and be dropped. */ $created = false; return $post_exists; } $created = true; $post_id = \wp_insert_post( $post_data, true ); if ( \is_wp_error( $post_id ) ) { return $post_id; } \set_post_format( $post_id, 'status' ); // Process attachments if enabled. if ( self::$fetch_attachments && ! empty( $post['object']['attachment'] ) ) { // Prepend archive path to attachment URLs for local files. $attachments = \array_map( array( self::class, 'prepend_archive_path' ), $post['object']['attachment'] ); Attachments::import( $attachments, $post_id, self::$author ); } return $post_id; } /** * Find a post already imported under an archive object id. * * @since 9.3.0 * * @param string $source_id The archive object id. * @param string $post_type The post type to look in. * * @return int The post ID, or 0 when the object has not been imported yet. */ private static function get_post_by_source_id( $source_id, $post_type ) { if ( ! \is_string( $source_id ) || '' === $source_id ) { return 0; } /* * `suppress_filters` is left at the get_posts() default of true on purpose: a * de-duplication probe is the last query a third-party `posts_where` should be * able to rewrite, since hiding the existing post brings the duplicates back. */ /* * Every status, trash included. `post_exists()` below takes no status argument, so * it always matched a trashed import on content; `'any'` would be narrower, because * it drops the statuses flagged `exclude_from_search`. */ $posts = \get_posts( array( 'post_type' => $post_type, 'post_status' => \get_post_stati(), 'numberposts' => 1, 'fields' => 'ids', 'meta_key' => '_source_id', // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_key 'meta_value' => $source_id, // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_value ) ); return $posts ? (int) $posts[0] : 0; } /** * Import a self-reply as a comment on its parent post. * * @param array $post The Mastodon activity. * @param array $source_to_post_id Mapping of source IDs to WordPress post IDs. * @param array $source_to_comment_id Mapping of source IDs to WordPress comment IDs (passed by reference). * * @return int|false Comment ID on success, false if parent not found or skipped. */ private static function import_as_comment( $post, $source_to_post_id, &$source_to_comment_id ) { $in_reply_to = $post['object']['inReplyTo']; // Find parent - could be a post or another comment. $parent_post_id = null; $parent_comment_id = 0; if ( isset( $source_to_post_id[ $in_reply_to ] ) ) { // Replying to a root post or external reply. $parent_post_id = $source_to_post_id[ $in_reply_to ]; } elseif ( isset( $source_to_comment_id[ $in_reply_to ] ) ) { // Replying to another comment (nested thread). $parent_comment_id = $source_to_comment_id[ $in_reply_to ]; $parent_comment = \get_comment( $parent_comment_id ); if ( $parent_comment ) { $parent_post_id = $parent_comment->comment_post_ID; } } // If we couldn't find the parent, skip this comment. if ( ! $parent_post_id ) { return false; } // Check for duplicate. $existing_comments = \get_comments( array( 'post_id' => $parent_post_id, 'meta_key' => 'source_id', // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_key 'meta_value' => $post['object']['id'], // phpcs:ignore WordPress.DB.SlowDBQuery.slow_db_query_meta_value 'number' => 1, ) ); if ( ! empty( $existing_comments ) ) { // Already imported, add to mapping and skip. $source_to_comment_id[ $post['object']['id'] ] = $existing_comments[0]->comment_ID; return false; } $comment_data = array( 'comment_post_ID' => $parent_post_id, 'comment_parent' => $parent_comment_id, 'comment_author' => \wp_slash( \get_the_author_meta( 'display_name', self::$author ) ), /* * Sanitize the archive's content explicitly. The import runs as a user with * `unfiltered_html`, so `kses_init()` installs none of the kses filters for * this request: neither `wp_insert_comment()` nor the `pre_comment_*` chain * would touch this value. Core then prints `comment_content` unescaped, on * the front end and in the Dashboard "Activity" widget. * * The comment allowlist, not the post one, so this field is cleaned the same * way Collection\Interactions cleans a live-federated reply. */ // Slashed like the federated path: wp_insert_comment() unslashes what it is given. 'comment_content' => \wp_slash( Sanitize::comment_content( $post['object']['content'] ?? '' ) ), 'comment_date' => $post['published'], 'user_id' => self::$author, 'comment_approved' => 1, ); $comment_id = \wp_insert_comment( $comment_data ); if ( $comment_id ) { \update_comment_meta( $comment_id, 'source_id', $post['object']['id'] ); $source_to_comment_id[ $post['object']['id'] ] = $comment_id; } return $comment_id; } /** * Header. */ public static function header() { echo '

'; echo '

' . \esc_html__( 'Import from Mastodon (Beta)', 'activitypub' ) . '

'; } /** * Footer. */ public static function footer() { echo '

'; } /** * Intro. */ public static function greet() { echo '

'; echo '

' . \wp_kses( \sprintf( /* translators: %s: URL to Mastodon export documentation */ \__( 'This importer allows you to bring your Mastodon posts into your WordPress site. For a smooth import experience, check out the Mastodon documentation.', 'activitypub' ), 'https://docs.joinmastodon.org/user/moving/#export' ), array( 'a' => array( 'href' => array(), 'target' => array(), ), ) ) . '

'; echo '

' . \esc_html__( 'Here’s how to get started:', 'activitypub' ) . '

'; echo '

' . \wp_kses( \__( 'Log in to your Mastodon account and go to Preferences > Import and Export.', 'activitypub' ), array( 'strong' => array() ) ) . '
' . \esc_html__( 'Request a new archive of your data and wait for the email notification.', 'activitypub' ) . '
' . \wp_kses( \__( 'Download the archive file (it will be a .zip file).', 'activitypub' ), array( 'code' => array() ) ) . '
' . \esc_html__( 'Upload that file below to begin the import process.', 'activitypub' ) . '

'; \wp_import_upload_form( 'admin.php?import=mastodon&step=1' ); echo '

'; } /** * Prepend archive path to local attachment URLs. * * @param array $attachment The attachment array. * * @return array The attachment array with updated URL. */ private static function prepend_archive_path( $attachment ) { if ( ! empty( $attachment['url'] ) && ! \preg_match( '#^https?://#i', $attachment['url'] ) ) { $attachment['url'] = self::$archive . $attachment['url']; } return $attachment; } /** * Detect and unwrap single nested directory in archive. * * Some Mastodon exports wrap all files in a root folder. This method * detects this pattern and updates the archive path to point inside it. */ private static function maybe_unwrap_archive() { global $wp_filesystem; $files = $wp_filesystem->dirlist( self::$archive ); // Check if there's exactly one directory at root level. if ( \count( $files ) !== 1 ) { return; } $first = \reset( $files ); if ( 'd' !== $first['type'] ) { return; } // Check if outbox.json exists inside the nested directory. $nested_path = self::$archive . '/' . $first['name']; if ( $wp_filesystem->exists( $nested_path . '/outbox.json' ) ) { self::$archive = $nested_path; } } }