includes/importers/helpers/wxr_importer/class-themeisle-ob-wxr-importer.php
<?php
/**
* Author: Andrei Baicus <andrei@themeisle.com>
* Created on: 12/07/2018
*
* @package themeisle-onboarding
* @phpcs:disable Squiz.Commenting.FunctionComment.Missing
* @phpcs:disable WordPress.NamingConventions.ValidVariableName.UsedPropertyNotSnakeCase
* todo: clean up for both phpcs rules.
*/
/**
* Class Themeisle_OB_WXR_Importer
*/
class Themeisle_OB_WXR_Importer extends WP_Importer {
/**
* Maximum supported WXR version
*/
const MAX_WXR_VERSION = 1.2;
/**
* Regular expression for checking if a post references an attachment
*
* Note: This is a quick, weak check just to exclude text-only posts. More
* vigorous checking is done later to verify.
*/
const REGEX_HAS_ATTACHMENT_REFS = '!
(
# Match anything with an image or attachment class
class=[\'"].*?\b(wp-image-\d+|attachment-[\w\-]+)\b
|
# Match anything that looks like an upload URL
src=[\'"][^\'"]*(
[0-9]{4}/[0-9]{2}/[^\'"]+\.(jpg|jpeg|png|gif)
|
content/uploads[^\'"]+
)[\'"]
)!ix';
/**
* Enable debug.
*
* @var bool
*/
private $debug = false;
/**
* Version of WXR we're importing.
*
* Defaults to 1.0 for compatibility. Typically overridden by a
* `<wp:wxr_version>` tag at the start of the file.
*
* @var string
*/
protected $version = '1.0';
/**
* Categories.
*
* @var array
*/
protected $categories = array();
/**
* Tags
*
* @var array
*/
protected $tags = array();
/**
* Base URL
*
* @var string
*/
protected $base_url = '';
/**
* Full URL
*
* @var string
*/
protected $full_url = '';
/**
* Missing menu items.
*
* @var array
*/
protected $missing_menu_items = array();
/**
* Mapping array
*
* @var array
*/
public $mapping = array();
/**
* Items that require remapping.
*
* @var array
*/
protected $requires_remapping = array();
/**
* Exists
*
* @var array
*/
protected $exists = array();
/**
* Override user slug.
*
* @var array
*/
protected $user_slug_override = array();
/**
* URL Remapping.
*
* @var array
*/
protected $url_remap = array();
/**
* Featured images.
*
* @var array
*/
protected $featured_images = array();
/**
* Options.
*
* @var array
*/
protected $options = array();
/**
* Constructor
*
* @param array $options options.
*
* @var bool $prefill_existing_posts Should we prefill `post_exists` calls? (True prefills and uses more
* memory, false checks once per imported post and takes longer. Default is true.)
* @var bool $prefill_existing_comments Should we prefill `comment_exists` calls? (True prefills and uses more
* memory, false checks once per imported comment and takes longer. Default is true.)
* @var bool $prefill_existing_terms Should we prefill `term_exists` calls? (True prefills and uses more
* memory, false checks once per imported term and takes longer. Default is true.)
* @var bool $update_attachment_guids Should attachment GUIDs be updated to the new URL? (True updates the
* GUID, which keeps compatibility with v1, false doesn't update, and allows deduplication and reimporting.
* Default is false.)
* @var bool $fetch_attachments Fetch attachments from the remote server. (True fetches and creates
* attachment posts, false skips attachments. Default is false.)
* @var bool $aggressive_url_search Should we search/replace for URLs aggressively? (True searches all
* posts' content for old URLs and replaces, false checks for `<img class="wp-image-*">` only. Default is
* false.)
* @var int $default_author User ID to use if author is missing or invalid. (Default is null, which
* leaves posts unassigned.)
* }
*/
public function __construct( $options = array() ) {
// Initialize some important variables
$empty_types = array(
'post' => array(),
'term' => array(),
'user' => array(),
);
$this->mapping = $empty_types;
$this->mapping['user_slug'] = array();
$this->mapping['term_id'] = array();
$this->requires_remapping = $empty_types;
$this->exists = $empty_types;
$this->options = wp_parse_args(
$options,
array(
'prefill_existing_posts' => true,
'prefill_existing_terms' => true,
'update_attachment_guids' => true,
'fetch_attachments' => true,
'aggressive_url_search' => false,
'default_author' => null,
)
);
}
/**
* Get a stream reader for the file.
*
* @param string $file Path to the XML file.
*
* @return XMLReader|WP_Error Reader instance on success, error otherwise.
*/
protected function get_reader( $file ) {
$old_value = null;
$reader = new XMLReader();
$status = $reader->open( $file );
if ( ! $status ) {
return new WP_Error( 'wxr_importer.cannot_parse', 'Could not open the file for parsing.' );
}
return $reader;
}
/**
* The main controller for the actual import stage.
*
* @param string $file Path to the WXR file for importing
*
* @return XMLReader | WP_Error | array
*/
public function import( $file ) {
add_filter( 'import_post_meta_key', array( $this, 'is_valid_meta_key' ) );
add_filter( 'http_request_timeout', array( &$this, 'bump_request_timeout' ) );
$result = $this->import_start( $file );
if ( is_wp_error( $result ) ) {
return $result;
}
$reader = $this->get_reader( $file );
if ( is_wp_error( $reader ) ) {
return $reader;
}
$this->version = '1.0';
$this->base_url = '';
while ( $reader->read() ) {
if ( $reader->nodeType !== XMLReader::ELEMENT ) {
continue;
}
switch ( $reader->name ) {
case 'wp:wxr_version':
$this->version = $reader->readString();
$reader->next();
break;
case 'wp:base_site_url':
$this->base_url = $reader->readString();
$reader->next();
break;
case 'wp:base_blog_url':
$this->full_url = $reader->readString();
$reader->next();
break;
case 'item':
$node = $reader->expand();
$parsed = $this->parse_post_node( $node );
if ( is_wp_error( $parsed ) ) {
$reader->next();
break;
}
$this->process_post( $parsed['data'], $parsed['meta'], $parsed['comments'], $parsed['terms'] );
$reader->next();
break;
case 'wp:author':
$reader->next();
break;
case 'wp:category':
$node = $reader->expand();
$parsed = $this->parse_term_node( $node, 'category' );
if ( is_wp_error( $parsed ) ) {
$reader->next();
break;
}
$this->process_term( $parsed['data'], $parsed['meta'] );
$reader->next();
break;
case 'wp:tag':
$node = $reader->expand();
$parsed = $this->parse_term_node( $node, 'tag' );
if ( is_wp_error( $parsed ) ) {
$reader->next();
break;
}
$this->process_term( $parsed['data'], $parsed['meta'] );
$reader->next();
break;
case 'wp:term':
$node = $reader->expand();
$parsed = $this->parse_term_node( $node );
if ( is_wp_error( $parsed ) ) {
$reader->next();
break;
}
$this->process_term( $parsed['data'], $parsed['meta'] );
$reader->next();
break;
default:
break;
}
}
// Post Processing and remapping.
$this->post_process();
if ( $this->options['aggressive_url_search'] ) {
$this->replace_attachment_urls_in_content();
}
$this->remap_featured_images();
$this->import_end();
}
/**
* Parses the WXR file and prepares us for the task of processing parsed data
*
* @param string $file Path to the WXR file for importing
*
* @return WP_Error
*/
protected function import_start( $file ) {
if ( ! is_file( $file ) ) {
return new WP_Error( 'wxr_importer.file_missing', 'The file does not exist, please try again.' );
}
// Suspend bunches of stuff in WP core
wp_defer_term_counting( true );
wp_defer_comment_counting( true );
wp_suspend_cache_invalidation( true );
// Prefill exists calls if told to
if ( $this->options['prefill_existing_posts'] ) {
$this->prefill_existing_posts();
}
if ( $this->options['prefill_existing_terms'] ) {
$this->prefill_existing_terms();
}
/**
* Begin the import.
*/
do_action( 'ti_ob_content_import_start' );
}
/**
* Performs post-import cleanup of files and the cache
*/
protected function import_end() {
// Re-enable stuff in core
wp_suspend_cache_invalidation( false );
wp_cache_flush();
foreach ( get_taxonomies() as $tax ) {
delete_option( "{$tax}_children" );
_get_term_hierarchy( $tax );
}
wp_defer_term_counting( false );
wp_defer_comment_counting( false );
/**
* Complete the import.
*/
do_action( 'ti_ob_content_import_end' );
}
/**
* Set the user mapping.
*
* @param array $mapping List of map arrays (containing `old_slug`, `old_id`, `new_id`)
*/
public function set_user_mapping( $mapping ) {
foreach ( $mapping as $map ) {
if ( empty( $map['old_slug'] ) || empty( $map['old_id'] ) || empty( $map['new_id'] ) ) {
continue;
}
$old_slug = $map['old_slug'];
$old_id = $map['old_id'];
$new_id = $map['new_id'];
$this->mapping['user'][ $old_id ] = $new_id;
$this->mapping['user_slug'][ $old_slug ] = $new_id;
}
}
/**
* Set the user slug overrides.
*
* Allows overriding the slug in the import with a custom/renamed version.
*
* @param string[] $overrides Map of old slug to new slug.
*/
public function set_user_slug_overrides( $overrides ) {
foreach ( $overrides as $original => $renamed ) {
$this->user_slug_override[ $original ] = $renamed;
}
}
/**
* Parse a post node into post data.
*
* @param DOMNode $node Parent node of post data (typically `item`).
*
* @return array|WP_Error Post data array on success, error otherwise.
*/
protected function parse_post_node( $node ) {
$data = array();
$meta = array();
$comments = array();
$terms = array();
foreach ( $node->childNodes as $child ) {
// We only care about child elements
if ( $child->nodeType !== XML_ELEMENT_NODE ) {
continue;
}
switch ( $child->tagName ) {
case 'wp:post_type':
$data['post_type'] = $child->textContent;
break;
case 'title':
$data['post_title'] = $child->textContent;
break;
case 'guid':
$data['guid'] = $child->textContent;
break;
case 'dc:creator':
$data['post_author'] = $child->textContent;
break;
case 'content:encoded':
$data['post_content'] = $child->textContent;
break;
case 'excerpt:encoded':
$data['post_excerpt'] = $child->textContent;
break;
case 'wp:post_id':
$data['post_id'] = $child->textContent;
break;
case 'wp:post_date':
$data['post_date'] = $child->textContent;
break;
case 'wp:post_date_gmt':
$data['post_date_gmt'] = $child->textContent;
break;
case 'wp:comment_status':
break;
case 'wp:ping_status':
$data['ping_status'] = $child->textContent;
break;
case 'wp:post_name':
$data['post_name'] = $child->textContent;
break;
case 'wp:status':
$data['post_status'] = $child->textContent;
if ( $data['post_status'] === 'auto-draft' ) {
// Bail now
return new WP_Error(
'wxr_importer.post.cannot_import_draft',
'Cannot import auto-draft posts',
$data
);
}
break;
case 'wp:post_parent':
$data['post_parent'] = $child->textContent;
break;
case 'wp:menu_order':
$data['menu_order'] = $child->textContent;
break;
case 'wp:post_password':
$data['post_password'] = $child->textContent;
break;
case 'wp:is_sticky':
$data['is_sticky'] = $child->textContent;
break;
case 'wp:attachment_url':
$data['attachment_url'] = $child->textContent;
break;
case 'wp:postmeta':
$meta_item = $this->parse_meta_node( $child );
if ( ! empty( $meta_item ) ) {
$meta[] = $meta_item;
}
break;
case 'wp:comment':
break;
case 'category':
$term_item = $this->parse_category_node( $child );
if ( ! empty( $term_item ) ) {
$terms[] = $term_item;
}
break;
}
}
return compact( 'data', 'meta', 'comments', 'terms' );
}
/**
* Create new posts based on import information
*
* @param $data
* @param $meta
* @param $comments
* @param $terms
*
* @return bool
*
* Posts marked as having a parent which doesn't exist will become top level items.
* Doesn't create a new post if: the post type doesn't exist, the given post ID
* is already noted as imported or a post with the same title and date already exists.
* Note that new/updated terms, comments and meta are imported for the last of the above.
*/
protected function process_post( $data, $meta, $comments, $terms ) {
/**
* Pre-process post data.
*
* @param array $data Post data. (Return empty to skip.)
* @param array $meta Meta data.
* @param array $comments Comments on the post.
* @param array $terms Terms on the post.
*/
$data = apply_filters( 'wxr_importer.pre_process.post', $data, $meta, $comments, $terms );
if ( empty( $data ) ) {
return false;
}
$original_id = isset( $data['post_id'] ) ? (int) $data['post_id'] : 0;
$parent_id = isset( $data['post_parent'] ) ? (int) $data['post_parent'] : 0;
// Have we already processed this?
if ( isset( $this->mapping['post'][ $original_id ] ) ) {
return;
}
$post_type_object = get_post_type_object( $data['post_type'] );
// Is this type even valid?
if ( ! $post_type_object ) {
return false;
}
$post_exists = $this->post_exists( $data );
if ( $post_exists ) {
/**
* Post processing already imported.
*
* @param array $data Raw data imported for the post.
*/
do_action( 'wxr_importer.process_already_imported.post', $data );
return false;
}
// Map the parent post, or mark it as one we need to fix
$requires_remapping = false;
if ( $parent_id ) {
if ( isset( $this->mapping['post'][ $parent_id ] ) ) {
$data['post_parent'] = $this->mapping['post'][ $parent_id ];
} else {
$meta[] = array(
'key' => '_wxr_import_parent',
'value' => $parent_id,
);
$requires_remapping = true;
$data['post_parent'] = 0;
}
}
$data['post_author'] = (int) get_current_user_id();
// Does the post look like it contains attachment images?
if ( preg_match( self::REGEX_HAS_ATTACHMENT_REFS, $data['post_content'] ) ) {
$meta[] = array(
'key' => '_wxr_import_has_attachment_refs',
'value' => true,
);
$requires_remapping = true;
}
// Whitelist to just the keys we allow
$postdata = array(
'import_id' => $data['post_id'],
);
$allowed = array(
'post_author' => true,
'post_date' => true,
'post_date_gmt' => true,
'post_content' => true,
'post_excerpt' => true,
'post_title' => true,
'post_status' => true,
'post_name' => true,
'comment_status' => true,
'ping_status' => true,
'guid' => true,
'post_parent' => true,
'menu_order' => true,
'post_type' => true,
'post_password' => true,
);
foreach ( $data as $key => $value ) {
if ( ! isset( $allowed[ $key ] ) ) {
continue;
}
$postdata[ $key ] = $data[ $key ];
}
$postdata = apply_filters( 'wp_import_post_data_processed', $postdata, $data );
if ( 'attachment' === $postdata['post_type'] ) {
if ( ! $this->options['fetch_attachments'] ) {
/**
* Post processing skipped.
*
* @param array $data Raw data imported for the post.
* @param array $meta Raw meta data, already processed by {@see process_post_meta}.
*/
do_action( 'wxr_importer.process_skipped.post', $data, $meta );
return false;
}
$remote_url = ! empty( $data['attachment_url'] ) ? $data['attachment_url'] : $data['guid'];
$post_id = $this->process_attachment( $postdata, $meta, $remote_url );
} else {
$post_id = wp_insert_post( $postdata, true );
do_action( 'wp_import_insert_post', $post_id, $original_id, $postdata, $data );
}
if ( is_wp_error( $post_id ) ) {
/**
* Post processing failed.
*
* @param \WP_Error $post_id Error object.
* @param array $data Raw data imported for the post.
* @param array $meta Raw meta data, already processed by {@see process_post_meta}.
* @param array $comments Raw comment data, already processed by {@see process_comments}.
* @param array $terms Raw term data, already processed.
*/
do_action( 'wxr_importer.process_failed.post', $post_id, $data, $meta, $comments, $terms );
return false;
}
// Ensure stickiness is handled correctly too
if ( $data['is_sticky'] === '1' ) {
stick_post( $post_id );
}
// map pre-import ID to local ID
$this->mapping['post'][ $original_id ] = (int) $post_id;
if ( $requires_remapping ) {
$this->requires_remapping['post'][ $post_id ] = true;
}
$this->mark_post_exists( $data, $post_id );
// Handle the terms too
$terms = apply_filters( 'wp_import_post_terms', $terms, $post_id, $data );
if ( ! empty( $terms ) ) {
$term_ids = array();
foreach ( $terms as $term ) {
$taxonomy = $term['taxonomy'];
$key = sha1( $taxonomy . ':' . $term['slug'] );
if ( isset( $this->mapping['term'][ $key ] ) ) {
$term_ids[ $taxonomy ][] = (int) $this->mapping['term'][ $key ];
} else {
$meta[] = array(
'key' => '_wxr_import_term',
'value' => $term,
);
}
}
foreach ( $term_ids as $tax => $ids ) {
$tt_ids = wp_set_post_terms( $post_id, $ids, $tax );
do_action( 'wp_import_set_post_terms', $tt_ids, $ids, $tax, $post_id, $data );
}
}
$this->process_post_meta( $meta, $post_id, $data );
if ( 'nav_menu_item' === $data['post_type'] ) {
$this->process_menu_item_meta( $post_id, $data, $meta );
}
/**
* Post processing completed.
*
* @param int $post_id New post ID.
* @param array $data Raw data imported for the post.
* @param array $meta Raw meta data, already processed by {@see process_post_meta}.
* @param array $comments Raw comment data, already processed by {@see process_comments}.
* @param array $terms Raw term data, already processed.
*/
do_action( 'wxr_importer.processed.post', $post_id, $data, $meta, $comments, $terms );
}
/**
* Attempt to create a new menu item from import data
*
* Fails for draft, orphaned menu items and those without an associated nav_menu
* or an invalid nav_menu term. If the post type or term object which the menu item
* represents doesn't exist then the menu item will not be imported (waits until the
* end of the import to retry again before discarding).
*
* @param int $post_id item id.
*/
protected function process_menu_item_meta( $post_id, $data, $meta ) {
$item_type = get_post_meta( $post_id, '_menu_item_type', true );
$original_object_id = get_post_meta( $post_id, '_menu_item_object_id', true );
$object_id = null;
$requires_remapping = false;
switch ( $item_type ) {
case 'taxonomy':
if ( isset( $this->mapping['term_id'][ $original_object_id ] ) ) {
$object_id = $this->mapping['term_id'][ $original_object_id ];
} else {
add_post_meta( $post_id, '_wxr_import_menu_item', wp_slash( $original_object_id ) );
$requires_remapping = true;
}
break;
case 'post_type':
if ( isset( $this->mapping['post'][ $original_object_id ] ) ) {
$object_id = $this->mapping['post'][ $original_object_id ];
} else {
add_post_meta( $post_id, '_wxr_import_menu_item', wp_slash( $original_object_id ) );
$requires_remapping = true;
}
break;
case 'custom':
$custom_url = get_post_meta( $post_id, '_menu_item_url', true );
$home_url = get_home_url();
$new_url = str_replace( $this->full_url, $home_url, $custom_url );
update_post_meta( $post_id, '_menu_item_url', $new_url );
$object_id = $post_id;
break;
default:
// associated object is missing or not imported yet, we'll retry later
$this->missing_menu_items[] = null;
break;
}
if ( $requires_remapping ) {
$this->requires_remapping['post'][ $post_id ] = true;
}
if ( empty( $object_id ) ) {
// Nothing needed here.
return;
}
update_post_meta( $post_id, '_menu_item_object_id', wp_slash( $object_id ) );
}
/**
* If fetching attachments is enabled then attempt to create a new attachment
*
* @param array $post Attachment post details from WXR
*
* @return int|\WP_Error Post ID on success, \WP_Error otherwise
*/
protected function process_attachment( $post, $meta, $remote_url ) {
// try to use _wp_attached file for upload folder placement to ensure the same location as the export site
// e.g. location is 2003/05/image.jpg but the attachment post_date is 2010/09, see media_handle_upload()
$post['upload_date'] = $post['post_date'];
foreach ( $meta as $meta_item ) {
if ( $meta_item['key'] !== '_wp_attached_file' ) {
continue;
}
if ( preg_match( '%^[0-9]{4}/[0-9]{2}%', $meta_item['value'], $matches ) ) {
$post['upload_date'] = $matches[0];
}
break;
}
// if the URL is absolute, but does not contain address, then upload it assuming base_site_url
if ( preg_match( '|^/[\w\W]+$|', $remote_url ) ) {
$remote_url = rtrim( $this->base_url, '/' ) . $remote_url;
}
$upload = $this->fetch_remote_file( $remote_url, $post );
if ( is_wp_error( $upload ) ) {
return $upload;
}
$info = wp_check_filetype( $upload['file'] );
if ( ! $info ) {
return new \WP_Error( 'attachment_processing_error', 'Invalid file type' );
}
$post['post_mime_type'] = $info['type'];
// WP really likes using the GUID for display. Allow updating it.
// See https://core.trac.wordpress.org/ticket/33386
if ( $this->options['update_attachment_guids'] ) {
$post['guid'] = $upload['url'];
}
// as per wp-admin/includes/upload.php
$post_id = wp_insert_attachment( $post, $upload['file'] );
if ( is_wp_error( $post_id ) ) {
return $post_id;
}
$attachment_metadata = wp_generate_attachment_metadata( $post_id, $upload['file'] );
wp_update_attachment_metadata( $post_id, $attachment_metadata );
// Map this image URL later if we need to
$this->url_remap[ $remote_url ] = $upload['url'];
// If we have a HTTPS URL, ensure the HTTP URL gets replaced too
if ( substr( $remote_url, 0, 8 ) === 'https://' ) {
$insecure_url = 'http' . substr( $remote_url, 5 );
$this->url_remap[ $insecure_url ] = $upload['url'];
}
return $post_id;
}
/**
* Parse a meta node into meta data.
*
* @param \DOMElement $node Parent node of meta data (typically `wp:postmeta` or `wp:commentmeta`).
*
* @return array|null Meta data array on success, or null on error.
*/
protected function parse_meta_node( $node ) {
foreach ( $node->childNodes as $child ) {
// We only care about child elements
if ( $child->nodeType !== XML_ELEMENT_NODE ) {
continue;
}
switch ( $child->tagName ) {
case 'wp:meta_key':
$key = $child->textContent;
break;
case 'wp:meta_value':
$value = $child->textContent;
break;
}
}
if ( empty( $key ) || empty( $value ) ) {
return null;
}
return compact( 'key', 'value' );
}
/**
* Process and import post meta items.
*
* @param array $meta List of meta data arrays
* @param int $post_id Post to associate with
* @param array $post Post data
*
* @return int|\WP_Error Number of meta items imported on success, error otherwise.
*/
protected function process_post_meta( $meta, $post_id, $post ) {
if ( empty( $meta ) ) {
return true;
}
foreach ( $meta as $meta_item ) {
/**
* Pre-process post meta data.
*
* @param array $meta_item Meta data. (Return empty to skip.)
* @param int $post_id Post the meta is attached to.
*/
$meta_item = apply_filters( 'wxr_importer.pre_process.post_meta', $meta_item, $post_id );
if ( empty( $meta_item ) ) {
return false;
}
$key = apply_filters( 'import_post_meta_key', $meta_item['key'], $post_id, $post );
$value = false;
if ( '_edit_last' === $key ) {
$value = intval( $meta_item['value'] );
if ( ! isset( $this->mapping['user'][ $value ] ) ) {
// Skip!
continue;
}
$value = $this->mapping['user'][ $value ];
}
if ( $key ) {
// export gets meta straight from the DB so could have a serialized string
if ( ! $value ) {
$value = maybe_unserialize( $meta_item['value'] );
}
if ( $key === '_elementor_data' ) {
require_once 'class-themeisle-ob-elementor-meta-handler.php';
$meta_handler = new Themeisle_OB_Elementor_Meta_Handler( $value, $this->full_url );
$meta_handler->filter_meta();
}
add_post_meta( $post_id, $key, $value );
do_action( 'import_post_meta', $post_id, $key, $value );
// if the post has a featured image, take note of this in case of remap
if ( '_thumbnail_id' === $key ) {
$this->featured_images[ $post_id ] = (int) $value;
}
}
}
return true;
}
protected function parse_category_node( $node ) {
$data = array(
// Default taxonomy to "category", since this is a `<category>` tag
'taxonomy' => 'category',
);
if ( $node->hasAttribute( 'domain' ) ) {
$data['taxonomy'] = $node->getAttribute( 'domain' );
}
if ( $node->hasAttribute( 'nicename' ) ) {
$data['slug'] = $node->getAttribute( 'nicename' );
}
$data['name'] = $node->textContent;
if ( empty( $data['slug'] ) ) {
return null;
}
// Just for extra compatibility
if ( $data['taxonomy'] === 'tag' ) {
$data['taxonomy'] = 'post_tag';
}
return $data;
}
protected function parse_term_node( $node, $type = 'term' ) {
$data = array();
$meta = array();
$tag_name = array(
'id' => 'wp:term_id',
'taxonomy' => 'wp:term_taxonomy',
'slug' => 'wp:term_slug',
'parent' => 'wp:term_parent',
'name' => 'wp:term_name',
'description' => 'wp:term_description',
);
$taxonomy = null;
// Special casing!
switch ( $type ) {
case 'category':
$tag_name['slug'] = 'wp:category_nicename';
$tag_name['parent'] = 'wp:category_parent';
$tag_name['name'] = 'wp:cat_name';
$tag_name['description'] = 'wp:category_description';
$tag_name['taxonomy'] = null;
$data['taxonomy'] = 'category';
break;
case 'tag':
$tag_name['slug'] = 'wp:tag_slug';
$tag_name['parent'] = null;
$tag_name['name'] = 'wp:tag_name';
$tag_name['description'] = 'wp:tag_description';
$tag_name['taxonomy'] = null;
$data['taxonomy'] = 'post_tag';
break;
}
foreach ( $node->childNodes as $child ) {
// We only care about child elements
if ( $child->nodeType !== XML_ELEMENT_NODE ) {
continue;
}
$key = array_search( $child->tagName, $tag_name );
if ( $key ) {
$data[ $key ] = $child->textContent;
}
}
if ( empty( $data['taxonomy'] ) ) {
return null;
}
// Compatibility with WXR 1.0
if ( $data['taxonomy'] === 'tag' ) {
$data['taxonomy'] = 'post_tag';
}
return compact( 'data', 'meta' );
}
protected function process_term( $data, $meta ) {
/**
* Pre-process term data.
*
* @param array $data Term data. (Return empty to skip.)
* @param array $meta Meta data.
*/
$data = apply_filters( 'wxr_importer.pre_process.term', $data, $meta );
if ( empty( $data ) ) {
return false;
}
$original_id = isset( $data['id'] ) ? (int) $data['id'] : 0;
$mapping_key = sha1( $data['taxonomy'] . ':' . $data['slug'] );
$existing = $this->term_exists( $data );
if ( $existing ) {
/**
* Term processing already imported.
*
* @param array $data Raw data imported for the term.
*/
do_action( 'wxr_importer.process_already_imported.term', $data );
$this->mapping['term'][ $mapping_key ] = $existing;
$this->mapping['term_id'][ $original_id ] = $existing;
return false;
}
// WP really likes to repeat itself in export files
if ( isset( $this->mapping['term'][ $mapping_key ] ) ) {
return false;
}
$termdata = array();
$allowed = array(
'slug' => true,
'description' => true,
);
foreach ( $data as $key => $value ) {
if ( ! isset( $allowed[ $key ] ) ) {
continue;
}
$termdata[ $key ] = $data[ $key ];
}
$result = wp_insert_term( $data['name'], $data['taxonomy'], $termdata );
if ( is_wp_error( $result ) ) {
do_action( 'wp_import_insert_term_failed', $result, $data );
/**
* Term processing failed.
*
* @param \WP_Error $result Error object.
* @param array $data Raw data imported for the term.
* @param array $meta Meta data supplied for the term.
*/
do_action( 'wxr_importer.process_failed.term', $result, $data, $meta );
return false;
}
$term_id = $result['term_id'];
$this->mapping['term'][ $mapping_key ] = $term_id;
$this->mapping['term_id'][ $original_id ] = $term_id;
do_action( 'wp_import_insert_term', $term_id, $data );
/**
* Term processing completed.
*
* @param int $term_id New term ID.
* @param array $data Raw data imported for the term.
*/
do_action( 'wxr_importer.processed.term', $term_id, $data );
}
/**
* Attempt to download a remote file attachment
*
* @param string $url URL of item to fetch
* @param array $post Attachment details
*
* @return array|WP_Error Local file location details on success, \WP_Error otherwise
*/
protected function fetch_remote_file( $url, $post ) {
// extract the file name and extension from the url
$file_name = basename( $url );
// get placeholder file in the upload dir with a unique, sanitized filename
$upload = wp_upload_bits( $file_name, 0, '', $post['upload_date'] );
if ( $upload['error'] ) {
return new WP_Error( 'upload_dir_error', $upload['error'] );
}
// fetch the remote url and write it to the placeholder file
$response = wp_remote_get(
$url,
array(
'stream' => true,
'filename' => $upload['file'],
)
);
// request failed
if ( is_wp_error( $response ) ) {
unlink( $upload['file'] );
return $response;
}
$code = (int) wp_remote_retrieve_response_code( $response );
// make sure the fetch was successful
if ( $code !== 200 ) {
unlink( $upload['file'] );
return new WP_Error(
'import_file_error',
sprintf(
'Remote server returned %1$d %2$s for %3$s',
$code,
get_status_header_desc( $code ),
$url
)
);
}
$filesize = filesize( $upload['file'] );
$headers = wp_remote_retrieve_headers( $response );
if ( isset( $headers['content-length'] ) && $filesize !== (int) $headers['content-length'] ) {
unlink( $upload['file'] );
return new WP_Error( 'import_file_error', 'Remote file is incorrect size' );
}
if ( 0 === $filesize ) {
unlink( $upload['file'] );
return new WP_Error( 'import_file_error', 'Zero size file downloaded' );
}
$max_size = (int) $this->max_attachment_size();
if ( ! empty( $max_size ) && $filesize > $max_size ) {
unlink( $upload['file'] );
$message = sprintf( 'Remote file is too large, limit is %s' );
return new WP_Error( 'import_file_error', $message );
}
return $upload;
}
protected function post_process() {
if ( ! empty( $this->requires_remapping['post'] ) ) {
$this->post_process_posts( $this->requires_remapping['post'] );
}
}
protected function post_process_posts( $todo ) {
foreach ( $todo as $post_id => $_ ) {
$data = array();
$parent_id = get_post_meta( $post_id, '_wxr_import_parent', true );
if ( ! empty( $parent_id ) ) {
// Have we imported the parent now?
if ( isset( $this->mapping['post'][ $parent_id ] ) ) {
$data['post_parent'] = $this->mapping['post'][ $parent_id ];
}
}
$author_slug = get_post_meta( $post_id, '_wxr_import_user_slug', true );
if ( ! empty( $author_slug ) ) {
// Have we imported the user now?
if ( isset( $this->mapping['user_slug'][ $author_slug ] ) ) {
$data['post_author'] = $this->mapping['user_slug'][ $author_slug ];
}
}
$has_attachments = get_post_meta( $post_id, '_wxr_import_has_attachment_refs', true );
if ( ! empty( $has_attachments ) ) {
$post = get_post( $post_id );
$content = $post->post_content;
// Replace all the URLs we've got
$new_content = str_replace( array_keys( $this->url_remap ), $this->url_remap, $content );
if ( $new_content !== $content ) {
$data['post_content'] = $new_content;
}
}
if ( get_post_type( $post_id ) === 'nav_menu_item' ) {
$this->post_process_menu_item( $post_id );
}
// Do we have updates to make?
if ( empty( $data ) ) {
continue;
}
// Run the update
$data['ID'] = $post_id;
$result = wp_update_post( $data, true );
if ( is_wp_error( $result ) ) {
continue;
}
// Clear out our temporary meta keys
delete_post_meta( $post_id, '_wxr_import_parent' );
delete_post_meta( $post_id, '_wxr_import_user_slug' );
delete_post_meta( $post_id, '_wxr_import_has_attachment_refs' );
}
}
protected function post_process_menu_item( $post_id ) {
$menu_object_id = get_post_meta( $post_id, '_wxr_import_menu_item', true );
if ( empty( $menu_object_id ) ) {
// No processing needed!
return;
}
$menu_item_type = get_post_meta( $post_id, '_menu_item_type', true );
switch ( $menu_item_type ) {
case 'taxonomy':
if ( isset( $this->mapping['term_id'][ $menu_object_id ] ) ) {
$menu_object = $this->mapping['term_id'][ $menu_object_id ];
}
break;
case 'post_type':
if ( isset( $this->mapping['post'][ $menu_object_id ] ) ) {
$menu_object = $this->mapping['post'][ $menu_object_id ];
}
break;
default:
// Cannot handle this.
return;
}
if ( ! empty( $menu_object ) ) {
update_post_meta( $post_id, '_menu_item_object_id', wp_slash( $menu_object ) );
}
delete_post_meta( $post_id, '_wxr_import_menu_item' );
}
/**
* Use stored mapping information to update old attachment URLs
*/
protected function replace_attachment_urls_in_content() {
global $wpdb;
// make sure we do the longest urls first, in case one is a substring of another
uksort( $this->url_remap, array( $this, 'cmpr_strlen' ) );
foreach ( $this->url_remap as $from_url => $to_url ) {
// remap urls in post_content
$query = $wpdb->prepare( "UPDATE {$wpdb->posts} SET post_content = REPLACE(post_content, %s, %s)", $from_url, $to_url );
$wpdb->query( $query );
// remap enclosure urls
$query = $wpdb->prepare( "UPDATE {$wpdb->postmeta} SET meta_value = REPLACE(meta_value, %s, %s) WHERE meta_key='enclosure'", $from_url, $to_url );
$wpdb->query( $query );
}
}
/**
* Update _thumbnail_id meta to new, imported attachment IDs
*/
function remap_featured_images() {
// cycle through posts that have a featured image
foreach ( $this->featured_images as $post_id => $value ) {
if ( isset( $this->mapping['post'][ $value ] ) ) {
$new_id = $this->mapping['post'][ $value ];
// only update if there's a difference
if ( $new_id !== $value ) {
update_post_meta( $post_id, '_thumbnail_id', $new_id );
}
}
}
}
/**
* Decide if the given meta key maps to information we will want to import
*
* @param string $key The meta key to check
*
* @return string|bool The key if we do want to import, false if not
*/
public function is_valid_meta_key( $key ) {
// skip attachment metadata since we'll regenerate it from scratch
// skip _edit_lock as not relevant for import
if ( in_array( $key, array( '_wp_attached_file', '_wp_attachment_metadata', '_edit_lock' ) ) ) {
return false;
}
return $key;
}
/**
* Decide what the maximum file size for downloaded attachments is.
* Default is 0 (unlimited), can be filtered via import_attachment_size_limit
*
* @return int Maximum attachment file size to import
*/
protected function max_attachment_size() {
return apply_filters( 'import_attachment_size_limit', 0 );
}
/**
* Added to http_request_timeout filter to force timeout at 60 seconds during import
*
* @access protected
* @return int 60
*/
function bump_request_timeout( $val ) {
return 60;
}
/**
* Return the difference in length between two strings
*/
function cmpr_strlen( $a, $b ) {
return strlen( $b ) - strlen( $a );
}
/**
* Prefill existing post data.
*
* This preloads all GUIDs into memory, allowing us to avoid hitting the
* database when we need to check for existence. With larger imports, this
* becomes prohibitively slow to perform SELECT queries on each.
*
* By preloading all this data into memory, it's a constant-time lookup in
* PHP instead. However, this does use a lot more memory, so for sites doing
* small imports onto a large site, it may be a better tradeoff to use
* on-the-fly checking instead.
*/
protected function prefill_existing_posts() {
global $wpdb;
$posts = $wpdb->get_results( "SELECT ID, guid FROM {$wpdb->posts}" );
foreach ( $posts as $item ) {
$this->exists['post'][ $item->guid ] = $item->ID;
}
}
/**
* Does the post exist?
*
* @param array $data Post data to check against.
*
* @return int|bool Existing post ID if it exists, false otherwise.
*/
protected function post_exists( $data ) {
// Constant-time lookup if we prefilled
$exists_key = $data['guid'];
if ( $this->options['prefill_existing_posts'] ) {
return isset( $this->exists['post'][ $exists_key ] ) ? $this->exists['post'][ $exists_key ] : false;
}
// No prefilling, but might have already handled it
if ( isset( $this->exists['post'][ $exists_key ] ) ) {
return $this->exists['post'][ $exists_key ];
}
// Still nothing, try post_exists, and cache it
$exists = post_exists( $data['post_title'], $data['post_content'], $data['post_date'] );
$this->exists['post'][ $exists_key ] = $exists;
return $exists;
}
/**
* Mark the post as existing.
*
* @param array $data Post data to mark as existing.
* @param int $post_id Post ID.
*/
protected function mark_post_exists( $data, $post_id ) {
$exists_key = $data['guid'];
$this->exists['post'][ $exists_key ] = $post_id;
}
/**
* Prefill existing term data.
*
* @see self::prefill_existing_posts() for justification of why this exists.
*/
protected function prefill_existing_terms() {
global $wpdb;
$query = "SELECT t.term_id, tt.taxonomy, t.slug FROM {$wpdb->terms} AS t";
$query .= " JOIN {$wpdb->term_taxonomy} AS tt ON t.term_id = tt.term_id";
$terms = $wpdb->get_results( $query );
foreach ( $terms as $item ) {
$exists_key = sha1( $item->taxonomy . ':' . $item->slug );
$this->exists['term'][ $exists_key ] = $item->term_id;
}
}
/**
* Does the term exist?
*
* @param array $data Term data to check against.
*
* @return int|bool Existing term ID if it exists, false otherwise.
*/
protected function term_exists( $data ) {
$exists_key = sha1( $data['taxonomy'] . ':' . $data['slug'] );
// Constant-time lookup if we prefilled
if ( $this->options['prefill_existing_terms'] ) {
return isset( $this->exists['term'][ $exists_key ] ) ? $this->exists['term'][ $exists_key ] : false;
}
// No prefilling, but might have already handled it
if ( isset( $this->exists['term'][ $exists_key ] ) ) {
return $this->exists['term'][ $exists_key ];
}
// Still nothing, try comment_exists, and cache it
$exists = term_exists( $data['slug'], $data['taxonomy'] );
if ( is_array( $exists ) ) {
$exists = $exists['term_id'];
}
$this->exists['term'][ $exists_key ] = $exists;
return $exists;
}
/**
* Mark the term as existing.
*
* @param array $data Term data to mark as existing.
* @param int $term_id Term ID.
*/
protected function mark_term_exists( $data, $term_id ) {
$exists_key = sha1( $data['taxonomy'] . ':' . $data['slug'] );
$this->exists['term'][ $exists_key ] = $term_id;
}
}