syndicatedpost.class.php

<?php
require_once dirname(__FILE__) . '/feedtime.class.php';
require_once dirname(__FILE__) . '/syndicatedpostterm.class.php';
require_once dirname(__FILE__) . '/syndicatedpostxpathquery.class.php';

/**
 * class SyndicatedPost: FeedWordPress uses to manage the conversion of
 * incoming items from the feed parser into posts for the WordPress
 * database. It contains several internal management methods primarily
 * of interest to someone working on the FeedWordPress source, as well
 * as some utility methods for extracting useful data from many
 * different feed formats, which may be useful to FeedWordPress users
 * who make use of feed data in PHP add-ons and filters.
 *
 * @version 2017.1018
 */
class SyndicatedPost {
	/** @var MagpieRSS|null  MagpieRSS representation. */
	var $item = null;
	/**  @var SimplePie_Item|null  SimplePie_Item representation. */
	var $entry = null;

	var $link = null;
	var $feed = null;
	var $feedmeta = null;

	var $xmlns = array();

	var $post  = array();

	var $named = array();
	var $preset_terms = array();
	var $feed_terms = array();

	var $_freshness = null;
	var $_wp_id = null;
	var $_wp_post = null;

	/**
	 * SyndicatedPost constructor: Given a feed item and the source from
	 * which it was taken, prepare a post that can be inserted into the
	 * WordPress database on request, or updated in place if it has already
	 * been syndicated.
	 *
	 * @param array $item The item syndicated from the feed.
	 * @param SyndicatedLink $source The feed it was syndicated from.
	 */
	public function __construct( $item, $source ) {
		if ( empty( $item ) and empty( $source ) )
			return;

		if ( is_array( $item )
		and isset( $item['simplepie'] )
		and isset( $item['magpie'] ) ) :
			$this->entry = $item['simplepie'];
			$this->item = $item['magpie'];
			$item = $item['magpie'];
		elseif ( is_a( $item, 'SimplePie_Item' ) ) :
			$this->entry = $item;

			// convert to Magpie for compat purposes
			$mpie = new MagpieFromSimplePie( $source->simplepie, $this->entry );
			$this->item = $mpie->get_item();

			// done with conversion object
			$mpie = NULL; unset( $mpie );
		else :
			$this->item = $item;
		endif;

		$this->link = $source;
		$this->feed = $source->magpie;
		$this->feedmeta = $source->settings;

		FeedWordPress::diagnostic( 'feed_items', 'Considering item [' . $this->guid() . '] "' . $this->entry->get_title().'"');

		# Dealing with namespaces can get so fucking fucked.
		$this->xmlns['forward'] = $source->magpie->_XMLNS_FAMILIAR;
		$this->xmlns['reverse'] = array();
		foreach ( $this->xmlns['forward'] as $url => $ns ) :
			if ( ! isset( $this->xmlns['reverse'][ $ns ] ) ) :
				$this->xmlns['reverse'][ $ns ] = array();
			endif;
			$this->xmlns['reverse'][ $ns ][] = $url;
		endforeach;

		// Fucking SimplePie.
		$this->xmlns['reverse']['rss'][] = '';

		// Trigger global syndicated_item filter.
		$changed = apply_filters( 'syndicated_item', $this->item, $this );
		$this->item = $changed;

		// Allow for feed-specific syndicated_item filters.
		$changed = apply_filters(
			"syndicated_item_" . $source->uri(),
			$this->item,
			$this
		);
		$this->item = $changed;

		# Filters can halt further processing by returning NULL
		if ( is_null( $this->item ) ) :
			$this->post = NULL;
		else :

			# Note that nothing is run through esc_sql() here.
			# That's deliberate. The escaping is done at the point
			# of insertion, not here, to avoid double-escaping and
			# to avoid screwing with syndicated_post filters

			$this->post['post_title'] = apply_filters(
				'syndicated_item_title',
				$this->entry->get_title(),
				$this
			);

			$this->named['author'] = apply_filters(
				'syndicated_item_author',
				$this->author(),
				$this
			);
			// This just gives us an alphanumeric name for the author.
			// We look up (or create) the numeric ID for the author
			// in SyndicatedPost::add().

			$this->post['post_content'] = apply_filters(
				'syndicated_item_content',
				$this->content(),
				$this
			);

			$excerpt = apply_filters( 'syndicated_item_excerpt', $this->excerpt(), $this );

			if ( !empty( $excerpt ) ):
				$this->post['post_excerpt'] = $excerpt;
			endif;

			// Dealing with timestamps in WordPress is so fucking fucked.
			$offset = (int) get_option( 'gmt_offset' ) * 60 * 60;
			$post_date_gmt     = $this->published( array( 'default' => -1 ) );
			$post_modified_gmt = $this->updated( array( 'default' => -1 ) );

			$this->post['post_date_gmt']     = gmdate( 'Y-m-d H:i:s', $post_date_gmt );
			$this->post['post_date']         = gmdate( 'Y-m-d H:i:s', $post_date_gmt + $offset );
			$this->post['post_modified_gmt'] = gmdate( 'Y-m-d H:i:s', $post_modified_gmt );
			$this->post['post_modified']     = gmdate( 'Y-m-d H:i:s', $post_modified_gmt + $offset );

			// Use feed-level preferences or the global default.
			$this->post['post_status']    = $this->link->syndicated_status( 'post', 'publish' );
			$this->post['comment_status'] = $this->link->syndicated_status( 'comment', 'closed' );
			$this->post['ping_status']    = $this->link->syndicated_status( 'ping', 'closed' );

			// Unique ID (hopefully a unique tag: URI); failing that, the permalink
			$this->post['guid'] = apply_filters( 'syndicated_item_guid', $this->guid(), $this );

			// User-supplied custom settings to apply to each post.
			// Do first so that FWP-generated custom settings will
			// overwrite if necessary; thus preventing any munging.
			$postMetaIn = $this->link->postmeta( array( "parsed" => true ) );
			$postMetaOut = array();

			foreach ( $postMetaIn as $key => $meta ) :
				$postMetaOut[ $key ] = $meta->do_substitutions( $this );
			endforeach;

			foreach ( $postMetaOut as $key => $values ) :
				if ( is_null( $values ) ) { // have chosen to replace value with empty string
					$values = [ '' ];
				}
				$this->post['meta'][ $key ] = array();
				foreach ( $values as $value ) :
					$this->post['meta'][ $key ][] = apply_filters( "syndicated_post_meta_{$key}", $value, $this );
				endforeach;
			endforeach;

			// RSS 2.0 / Atom 1.0 enclosure support
			$enclosures = $this->entry->get_enclosures();
			if ( is_array( $enclosures ) ) : foreach ( $enclosures as $enclosure ) :
				$this->post['meta']['enclosure'][] =
					apply_filters( 'syndicated_item_enclosure_url',    $enclosure->get_link(),   $this ) . "\n".
					apply_filters( 'syndicated_item_enclosure_length', $enclosure->get_length(), $this ) . "\n".
					apply_filters( 'syndicated_item_enclosure_type',   $enclosure->get_type(),   $this );
			endforeach; endif;

			// In case you want to point back to the blog this was
			// syndicated from.

			$sourcemeta['syndication_source'] = apply_filters(
				'syndicated_item_source_title',
				$this->link->name(),
				$this
			);
			$sourcemeta['syndication_source_uri'] = apply_filters(
				'syndicated_item_source_link',
				$this->link->homepage(),
				$this
			);
			$sourcemeta['syndication_source_id'] = apply_filters(
				'syndicated_item_source_id',
				$this->link->guid(),
				$this
			);

			// Make use of atom:source data, if present in an aggregated feed
			$entry_source = $this->source();
			if ( ! is_null( $entry_source ) ) :
				foreach ( $entry_source as $what => $value ) :
					if ( ! is_null( $value ) ) :
						if ( 'title' == $what ) : $key = 'syndication_source';
						elseif ( 'feed' == $what ) : $key = 'syndication_feed';
						else : $key = "syndication_source_{$what}";
						endif;

						$sourcemeta["{$key}_original"] = apply_filters(
							'syndicated_item_original_source_' . $what,
							$value,
							$this
						);
					endif;
				endforeach;
			endif;

			foreach ( $sourcemeta as $meta_key => $value ) :
				if ( !is_null( $value ) ) :
					$this->post['meta'][ $meta_key ] = $value;
				endif;
			endforeach;

			// Store information on human-readable and machine-readable comment URIs

			// Human-readable comment URI
			$commentLink = apply_filters( 'syndicated_item_comments', $this->comment_link(), $this );
			if ( ! is_null( $commentLink) ) : $this->post['meta']['rss:comments'] = $commentLink; endif;

			// Machine-readable content feed URI
			$commentFeed = apply_filters( 'syndicated_item_commentrss', $this->comment_feed(), $this );
			if ( ! is_null( $commentFeed ) ) :	$this->post['meta']['wfw:commentRSS'] = $commentFeed; endif;
			// Yeah, yeah, now I know that it's supposed to be
			// wfw:commentRss. Oh well. Path dependence, sucka.

			// Store information to identify the feed that this came from
			if ( isset( $this->feedmeta['link/uri'] ) ) :
				$this->post['meta']['syndication_feed'] = $this->feedmeta['link/uri'];
			endif;
			if ( isset( $this->feedmeta['link/id'] ) ) :
				$this->post['meta']['syndication_feed_id'] = $this->feedmeta['link/id'];
			endif;

			if ( isset( $this->item['source_link_self'] ) ) :
				$this->post['meta']['syndication_feed_original'] = $this->item['source_link_self'];
			endif;

			// In case you want to know the external permalink...
			$this->post['meta']['syndication_permalink'] = apply_filters( 'syndicated_item_link', $this->permalink() );

			// Store a hash of the post content for checking whether something needs to be updated
			$this->post['meta']['syndication_item_hash'] = $this->update_hash();

			// Categories, Tags, and other Terms: from settings assignments (global settings, subscription settings),
			// and from feed assignments (item metadata, post content)
			$this->preset_terms = apply_filters( 'syndicated_item_preset_terms', $this->get_terms_from_settings(), $this );
			$this->feed_terms   = apply_filters( 'syndicated_item_feed_terms', $this->get_terms_from_feeds(), $this );

			$this->post['post_type'] = apply_filters(
				'syndicated_post_type',
				$this->link->setting( 'syndicated post type', 'syndicated_post_type', 'post' ),
				$this
			);
		endif;
	} /* SyndicatedPost::__construct() */

	#####################################
	#### EXTRACT DATA FROM FEED ITEM ####
	#####################################

	function substitution_function ($name) {
		$ret = NULL;

		switch ($name) :
		// Allowed PHP string functions
		case 'trim':
		case 'ltrim':
		case 'rtrim':
		case 'strtoupper':
		case 'strtolower':
		case 'urlencode':
		case 'urldecode':
			$ret = $name;
		endswitch;
		return $ret;
	}

	/**
	 * SyndicatedPost::query uses an XPath-like syntax to query arbitrary
	 * elements within the syndicated item.
	 *
	 * @param string $path
	 * @returns array of string values representing contents of matching
	 * elements or attributes
	 */
	 public function query ($path) {
		$xq = new SyndicatedPostXPathQuery(array("path" => $path));

		$feedChannel = array_merge(
			$this->get_feed_root_element(),
			$this->get_feed_channel_elements()
		);

		$matches = $xq->match(array(
		"type" => $this->link->simplepie->get_type(),
		"xmlns" => $this->xmlns,
		"map" => array(
			"/" => array($this->entry->data),
			"item" => array($this->entry->data),
			"feed" => $feedChannel,
			"channel" => $feedChannel
		),
		"context" => $this->entry->data,
		"parent" => $feedChannel,
		));

		return $matches;
	} /* SyndicatedPost::query() */

	function get_feed_root_element () {
		$matches = array();
		foreach ($this->link->simplepie->data['child'] as $ns => $root) :
			foreach ($root as $element => $data) :
				$matches = array_merge($matches, $data);
			endforeach;
		endforeach;
		return $matches;
	} /* SyndicatedPost::get_feed_root_element() */

	function get_feed_channel_elements () {
		$rss = array(
				SIMPLEPIE_NAMESPACE_RSS_090,
				SIMPLEPIE_NAMESPACE_RSS_10,
				'http://backend.userland.com/RSS2',
				SIMPLEPIE_NAMESPACE_RSS_20,
		);

		$matches = array();
		foreach ($rss as $ns) :
			$data = $this->link->simplepie->get_feed_tags($ns, 'channel');
			if ( !is_null($data)) :
				$matches = array_merge($matches, $data);
			endif;
		endforeach;
		return $matches;
	} /* SyndicatedPost::get_feed_channel_elements() */

	public function get_categories ($params = array()) {
		return $this->entry->get_categories();
	}

	public function title ($params = array()) {
		return $this->entry->get_title();
	} /* SyndicatedPost::title () */

	public function content ($params = array())
	{
		$params = wp_parse_args($params, array(
		"full only" => false,
		));

		$content = NULL;

		// FIXME: This is one of the main places in the code still using
		// the outmoded SimplePie - to - Magpie construction. We could
		// replace using SimplePie_Item::get_tags() here. (Or if really
		// ambitious we could attempt to just use
		// SimplePie_Item::get_content() with content-only set to TRUE
		// and some sanitization in effect. -CJ 1jul14

		// atom:content, standard method of providing full HTML content
		// in Atom feeds.
		if (isset($this->item['atom_content'])) :
			$content = $this->item['atom_content'];
		elseif (isset($this->item['atom']['atom_content'])) :
			$content = $this->item['atom']['atom_content'];

		// Some exotics: back in the day, before widespread convergence
		// on content:encoding, some RSS feeds took advantage of XML
		// namespaces to use an inline xhtml:body or xhtml:div element
		// for full-content HTML. (E.g. Sam Ruby's feed, IIRC.)
		elseif (isset($this->item['xhtml']['body'])) :
			$content = $this->item['xhtml']['body'];
		elseif (isset($this->item['xhtml']['div'])) :
			$content = $this->item['xhtml']['div'];

		// content:encoded, most common method of providing full HTML in
		// RSS 2.0 feeds.
		elseif (isset($this->item['content']['encoded']) and $this->item['content']['encoded']):
			$content = $this->item['content']['encoded'];

		// Fall back on elements that sometimes may contain full HTML
		// but sometimes not.
		elseif ( ! $params['full only']) :

			// description element is sometimes used for full HTML
			// sometimes for summary text in RSS. (By the letter of
			// the standard, it should just be for summary text.)
			if (isset($this->item['description'])) :
				$content = $this->item['description'];
			endif;

		endif;

		return $content;
	} /* SyndicatedPost::content() */

	public function excerpt () {
		# Identify and sanitize excerpt: atom:summary, or rss:description
		$excerpt = $this->entry->get_description();

		# Many RSS feeds use rss:description, inadvisably, to
		# carry the entire post (typically with escaped HTML).
		# If that's what happened, we don't want the full
		# content for the excerpt.
		$content = $this->content();

		// Ignore whitespace, case, and tag cruft.
		$theExcerpt = preg_replace('/\s+/', '', strtolower(strip_tags(html_entity_decode($excerpt))));
		$theContent = preg_replace('/\s+/', '', strtolower(strip_tags(html_entity_decode($content))));
		if ( empty($excerpt) or $theExcerpt == $theContent ) :
			# If content is available, generate an excerpt.
			if ( strlen(trim($content)) > 0 ) :
				$excerpt = strip_tags($content);
				if (strlen($excerpt) > 255) :
					if (is_object($this->link) and is_object($this->link->simplepie)) :
						$encoding = $this->link->simplepie->get_encoding();
					else :
						$encoding = get_option('blog_charset', 'utf8');
					endif;
					$excerpt = mb_substr($excerpt,0,252,$encoding).'...';
				endif;
			endif;
		endif;

		return $excerpt;
	} /* SyndicatedPost::excerpt() */

	/**
	 * SyndicatedPost::permalink: returns the permalink for the post, as provided by the
	 * source feed.
	 *
	 * @return string The URL of the original permalink for this syndicated post
	 */
	public function permalink () {
		// Handles explicit <link> elements and also RSS 2.0 cases with
		// <guid isPermaLink="true">, etc. Hooray!
		$permalink = $this->entry->get_link();
		return $permalink;
	} /* SyndicatedPost::permalink () */

	public function created ($params = array()) {
		$unfiltered = false; // $default = NULL; // seems to be unused on this function (gwyneth 20230916)
		extract($params);

		$date = '';
		if (isset($this->item['dc']['created'])) :
			$date = $this->item['dc']['created'];
		elseif (isset($this->item['dcterms']['created'])) :
			$date = $this->item['dcterms']['created'];
		elseif (isset($this->item['created'])): // Atom 0.3
			$date = $this->item['created'];
		endif;

		$time = new FeedTime($date);
		$tstamp = $time->timestamp();
		if ( ! $unfiltered) :
			apply_filters('syndicated_item_created', $tstamp, $this);
		endif;
		return $tstamp;
	} /* SyndicatedPost::created() */

	public function published ($params = array(), $default = NULL) {
		$fallback = true; $unfiltered = false;
		if ( !is_array($params)) : // Old style
			$fallback = $params;
		else : // New style
			extract($params);
		endif;

		$date = '';
		$tstamp = null;

		# RSS is a fucking mess. Figure out whether we have a date in
		# <dc:date>, <issued>, <pubDate>, etc., and get it into Unix
		# epoch format for reformatting. If we can't find anything,
		# we'll use the last-updated time.
		if (isset($this->item['dc']['date'])):				// Dublin Core
			$date = $this->item['dc']['date'];
		elseif (isset($this->item['dcterms']['issued'])) :		// Dublin Core extensions
			$date = $this->item['dcterms']['issued'];
		elseif (isset($this->item['published'])) : 			// Atom 1.0
			$date = $this->item['published'];
		elseif (isset($this->item['issued'])): 				// Atom 0.3
			$date = $this->item['issued'];
		elseif (isset($this->item['pubdate'])):				// RSS 2.0
			$date = $this->item['pubdate'];
		endif;

		if (strlen($date) > 0) :
			$time = new FeedTime($date);
			$tstamp = $time->timestamp();
		elseif ($fallback) :						// Fall back to <updated> / <modified> if present
			$tstamp = $this->updated(/*fallback=*/ false, /*default=*/ $default);
		endif;

		# If everything failed, then default to the current time.
		if (is_null($tstamp)) :
			if (-1 == $default) :
				$tstamp = time();
			else :
				$tstamp = $default;
			endif;
		endif;

		if ( ! $unfiltered) :
			$tstamp = apply_filters('syndicated_item_published', $tstamp, $this);
		endif;
		return $tstamp;
	} /* SyndicatedPost::published() */

	public function updated ($params = array(), $default = -1) {
		$fallback = true; $unfiltered = false;
		if ( !is_array($params)) : // Old style
			$fallback = $params;
		else : // New style
			extract($params);
		endif;

		$date = '';
		$tstamp = null;

		# As far as I know, only dcterms and Atom have reliable ways to
		# specify when something was *modified* last. If neither is
		# available, then we'll try to get the time of publication.
		if (isset($this->item['dc']['modified'])) : 			// Not really correct
			$date = $this->item['dc']['modified'];
		elseif (isset($this->item['dcterms']['modified'])) :		// Dublin Core extensions
			$date = $this->item['dcterms']['modified'];
		elseif (isset($this->item['modified'])):			// Atom 0.3
			$date = $this->item['modified'];
		elseif (isset($this->item['updated'])):			// Atom 1.0
			$date = $this->item['updated'];
		endif;

		if (strlen($date) > 0) :
			$time = new FeedTime($date);
			$tstamp = $time->timestamp();
		elseif ($fallback) :						// Fall back to issued / dc:date
			$tstamp = $this->published(/*fallback=*/ false, /*default=*/ $default);
		endif;

		# If everything failed, then default to the current time.
		if (is_null($tstamp)) :
			if (-1 == $default) :
				$tstamp = time();
			else :
				$tstamp = $default;
			endif;
		endif;

		if ( ! $unfiltered) :
			$tstamp = apply_filters('syndicated_item_updated', $tstamp, $this);
		endif;
		return $tstamp;
	} /* SyndicatedPost::updated() */

	var $_hashes = array();
	function stored_hashes ($id = NULL) {
		if (is_null($id)) :
			$id = $this->wp_id();
		endif;

		if ( !isset($this->_hashes[$id])) :
			$this->_hashes[$id] = get_post_custom_values(
				'syndication_item_hash', $id
			);
			if (is_null($this->_hashes[$id])) :
				$this->_hashes[$id] = array();
			endif;
		endif;
		return $this->_hashes[$id];
	}

	function update_hash ($hashed = true) {
		// Basis for tracking possible changes to item.
		$hash = array(
			"title" => $this->entry->get_title(),
			"link" => $this->permalink(),
			"content" => $this->content(),
			"excerpt" => $this->excerpt(),
		);

		if ($hashed) :
			$hash = md5(serialize($hash));
		endif;

		return $hash;
	} /* SyndicatedPost::update_hash() */

	/**
	 * SyndicatedPost::normalize_guid_prefix(): generates a normalized URL
	 * prefix (including scheme, authority, full path, and the beginning of
	 * a query string) for creating guids that conform to WordPress's
	 * internal constraints on the URL space for valid guids. To create a
	 * normalized guid, just concatenate a valid URL query parameter value
	 * to the returned URL.
	 *
	 * @return string The URL prefix generated.
	 *
	 * @uses trailingslashit()
	 * @uses home_url()
	 * @uses apply_filters()
	 */
	static function normalize_guid_prefix () {
		$scheme = 'http';
		$url = trailingslashit(home_url(/*path=*/ '', $scheme));
		return apply_filters('syndicated_item_guid_normalized_prefix', $url . '?guid=', $scheme);
	} /* SyndicatedPost::normalize_guid_prefix() */

	static function normalize_guid ($guid) {
		$guid = trim($guid);
		if (preg_match('/^[0-9a-z]{32}$/i', $guid)) : // MD5
			$guid = SyndicatedPost::normalize_guid_prefix().strtolower($guid);
		elseif ((strlen(esc_url($guid)) == 0) or (esc_url($guid) != $guid)) :
			$guid = SyndicatedPost::normalize_guid_prefix().md5($guid);
		endif;
		$guid = trim($guid);

		return $guid;
	} /* SyndicatedPost::normalize_guid() */

	static function alternative_guid_prefix () {
		$scheme = 'https';
		$url = trailingslashit(home_url(/*path=*/ '', $scheme));
		return apply_filters('syndicated_item_guid_normalized_prefix', $url . '?guid=', $scheme);
	}
	static function alternative_guid ($guid) {
		$guid = trim($guid);
		if (preg_match('/^[0-9a-z]{32}$/i', $guid)) : // MD5
			$guid = SyndicatedPost::alternative_guid_prefix().strtolower($guid);
		elseif ((strlen(esc_url($guid)) == 0) or (esc_url($guid) != $guid)) :
			$guid = SyndicatedPost::alternative_guid_prefix().md5($guid);
		endif;
		$guid = trim($guid);

		return $guid;
	} /* SyndicatedPost::normalize_guid() */

	public function guid () {
		$guid = null;
		if (isset($this->item['id'])):						// Atom 0.3 / 1.0
			$guid = $this->item['id'];
		elseif (isset($this->item['atom']['id'])) :		// Namespaced Atom
			$guid = $this->item['atom']['id'];
		elseif (isset($this->item['guid'])) :				// RSS 2.0
			$guid = $this->item['guid'];
		elseif (isset($this->item['dc']['identifier'])) :	// yeah, right
			$guid = $this->item['dc']['identifier'];
		endif;

		// Un-set or too long to use as-is. Generate a tag: URI.
		if (is_null($guid) or strlen($guid) > 250) :
			// In case we need to check this again
			$original_guid = $guid;

			// The feed does not seem to have provided us with a
			// usable unique identifier, so we'll have to cobble
			// together a tag: URI that might work for us. The base
			// of the URI will be the host name of the feed source ...
			$bits = parse_url($this->link->uri());
			$guid = 'tag:'.$bits['host'];

			// Some ill-mannered feeds (for example, certain feeds
			// coming from Google Calendar) have extraordinarily long
			// guids -- so long that they exceed the 255 character
			// width of the WordPress guid field. But if the string
			// gets clipped by MySQL, uniqueness tests will fail
			// forever after and the post will be endlessly
			// reduplicated. So, instead, Guids Of A Certain Length
			// are hashed down into a nice, manageable tag: URI.
			if ( !is_null($original_guid)) :
				$guid .= ',2010-12-03:id.'.md5($original_guid);

			// If we have a date of creation, then we can use that
			// to uniquely identify the item. (On the other hand, if
			// the feed producer was consicentious enough to
			// generate dates of creation, she probably also was
			// conscientious enough to generate unique identifiers.)
			elseif ( !is_null($this->created())) :
				$guid .= '://post.'.date('YmdHis', $this->created());

			// Otherwise, use both the URI of the item, *and* the
			// item's title. We have to use both because titles are
			// often not unique, and sometimes links aren't unique
			// either (e.g. Bitch (S)HITLIST, Mozilla Dot Org news,
			// some podcasts). But it's rare to have *both* the same
			// title *and* the same link for two different items. So
			// this is about the best we can do.
			else :
				$link = $this->permalink();
				if (is_null($link)) : $link = $this->link->uri(); endif;
				$guid .= '://'.md5($link.'/'.$this->title());
			endif;
		endif;
		return $guid;
	} /* SyndicatedPost::guid() */

	public function author () {
		$author = array ();

		$aa = $this->entry->get_authors();
		if (is_countable($aa) and count($aa) > 0) :
			$a = reset($aa);

			$author = array(
			'name' => $a->get_name(),
			'email' => $a->get_email(),
			'uri' => $a->get_link(),
			);
		endif;

		if (FEEDWORDPRESS_COMPATIBILITY) :
			// Search through the MagpieRSS elements: Atom, Dublin Core, RSS
			if (isset($this->item['author_name'])):
				$author['name'] = $this->item['author_name'];
			elseif (isset($this->item['dc']['creator'])):
				$author['name'] = $this->item['dc']['creator'];
			elseif (isset($this->item['dc']['contributor'])):
				$author['name'] = $this->item['dc']['contributor'];
			elseif (isset($this->feed->channel['dc']['creator'])) :
				$author['name'] = $this->feed->channel['dc']['creator'];
			elseif (isset($this->feed->channel['dc']['contributor'])) :
				$author['name'] = $this->feed->channel['dc']['contributor'];
			elseif (isset($this->feed->channel['author_name'])) :
				$author['name'] = $this->feed->channel['author_name'];
			elseif ($this->feed->is_rss() and isset($this->item['author'])) :
				// The author element in RSS is allegedly an
				// e-mail address, but lots of people don't use
				// it that way. So let's make of it what we can.
				$author = parse_email_with_realname($this->item['author']);

				if ( !isset($author['name'])) :
					if (isset($author['email'])) :
						$author['name'] = $author['email'];
					else :
						$author['name'] = $this->feed->channel['title'];
					endif;
				endif;
			endif;
		endif;

		if ( !isset($author['name']) or is_null($author['name'])) :
			// Nothing found. Try some crappy defaults.
			if ($this->link->name()) :
				$author['name'] = $this->link->name();
			else :
				$url = parse_url($this->link->uri());
				$author['name'] = $url['host'];
			endif;
		endif;

		if (FEEDWORDPRESS_COMPATIBILITY) :
			if (isset($this->item['author_email'])):
				$author['email'] = $this->item['author_email'];
			elseif (isset($this->feed->channel['author_email'])) :
				$author['email'] = $this->feed->channel['author_email'];
			endif;

			if (isset($this->item['author_uri'])):
				$author['uri'] = $this->item['author_uri'];
			elseif (isset($this->item['author_url'])):
				$author['uri'] = $this->item['author_url'];
			elseif (isset($this->feed->channel['author_uri'])) :
				$author['uri'] = $this->item['author_uri'];
			elseif (isset($this->feed->channel['author_url'])) :
				$author['uri'] = $this->item['author_url'];
			elseif (isset($this->feed->channel['link'])) :
				$author['uri'] = $this->feed->channel['link'];
			endif;
		endif;

		return $author;
	} /* SyndicatedPost::author() */

	/**
	 * SyndicatedPost::get_terms_from_settings(): Return an array of terms to associate with the incoming
	 * post based on the Categories, Tags, and other terms associated with each new post by the user's
	 * settings (global and feed-specific).
	 *
	 * @since 2016.0331
	 * @return array of lists, each element has the taxonomy for a key ('category', 'post_tag', etc.),
	 * 		and a list of term codes (either alphanumeric names, or ID numbers encoded in a format that
	 * 		SyndicatedLink::category_ids() can understand) within that taxonomy
	 *
	 */
	public function get_terms_from_settings () {
		// Categories: start with default categories, if any.
		$cats = array();
		if ('no' != $this->link->setting('add/category', NULL, 'yes')) :
			$fc = get_option("feedwordpress_syndication_cats");
			if ($fc) :
				$cats = array_merge($cats, explode("\n", $fc));
			endif;
		endif;

		$fc = $this->link->setting('cats',NULL, array());
		if (is_array($fc)) :
			$cats = array_merge($cats, $fc);
		endif;
		$preset_terms['category'] = $cats;

		// Tags: start with default tags, if any
		$tags = array();
		if ('no' != $this->link->setting('add/post_tag', NULL, 'yes')) :
			$ft = get_option("feedwordpress_syndication_tags", NULL);
			$tags = (is_null($ft) ? array() : explode(FEEDWORDPRESS_CAT_SEPARATOR, $ft));
		endif;

		$ft = $this->link->setting('tags', NULL, array());
		if (is_array($ft)) :
			$tags = array_merge($tags, $ft);
		endif;
		$preset_terms['post_tag'] = $tags;

		$taxonomies = $this->link->taxonomies();
		$feedTerms = $this->link->setting('terms', NULL, array());
		$globalTerms = get_option('feedwordpress_syndication_terms', array());
		$specials = array('category' => 'cats', 'post_tag' => 'tags');

		foreach ($taxonomies as $tax) :
			// category and tag settings have already previously been handled
			// but if this is from another taxonomy, then...
			if ( !isset($specials[$tax])) :
				$terms = array();

				// See if we should get the globals
				if ('no' != $this->link->setting("add/$tax", NULL, 'yes')) :
					if (isset($globalTerms[$tax])) :
						$terms = $globalTerms[$tax];
					endif;
				endif;

				// Now merge in the locals
				if (isset($feedTerms[$tax])) :
					$terms = array_merge($terms, $feedTerms[$tax]);
				endif;

				// That's all, folks.
				$preset_terms[$tax] = $terms;
			endif;
		endforeach;

		return $preset_terms;
	} /* SyndicatedPost::get_terms_from_settings () */

	/**
	 * SyndicatedPost::get_terms_from_feeds(): Return an array of terms to associate with the incoming
	 * post based on the contents of the subscribed feed (atom:category and rss:category elements, dc:subject
	 * elements, tags embedded using microformats in the post content, etc.)
	 *
	 * @since 2016.0331
	 * @return array of lists, each element has the taxonomy for a key ('category', 'post_tag', etc.),
	 * 		and a list of alphanumeric term names
	 */
	public function get_terms_from_feeds () {
		// Now add categories from the post, if we have 'em
		$cats = array();
		$post_cats = $this->entry->get_categories();
		if (is_array($post_cats)) : foreach ($post_cats as $cat) :
			$cat_name = $cat->get_term();
			if ( ! $cat_name) : $cat_name = $cat->get_label(); endif;

			if ($this->link->setting('cat_split', NULL, NULL)) :
				$pcre = "\007".$this->feedmeta['cat_split']."\007";
				$cats = array_merge(
					$cats,
					preg_split(
						$pcre,
						$cat_name,
						-1 /*=no limit*/,
						PREG_SPLIT_NO_EMPTY
					)
				);
			else :
				$cats[] = $cat_name;
			endif;
		endforeach; endif;

		$feed_terms['category'] = apply_filters('syndicated_item_categories', $cats, $this);

		// Scan post for /a[@rel='tag'] and use as tags if present
		$tags = $this->inline_tags();
		$feed_terms['post_tag'] = apply_filters('syndicated_item_tags', $tags, $this);

		return $feed_terms;
	} /* SyndicatedPost::get_terms_from_feeds () */

	/**
	 * SyndicatedPost::inline_tags: Return a list of all the tags embedded
	 * in post content using the a[@rel="tag"] microformat.
	 *
	 * @since 2010.0630
	 * @return array of string values containing the name of each tag
	 */
	function inline_tags () {
		$tags = array();
		$content = $this->content();
		$pattern = FeedWordPressHTML::tagWithAttributeRegex('a', 'rel', 'tag');
		preg_match_all($pattern, $content, $refs, PREG_SET_ORDER);
		if (is_countable($refs) and count($refs) > 0) :
			foreach ($refs as $ref) :
				$tag = FeedWordPressHTML::tagWithAttributeMatch($ref);
				$tags[] = $tag['content'];
			endforeach;
		endif;
		return $tags;
	}

	/**
	 * SyndicatedPost::isTaggedAs: Test whether a feed item is
	 * tagged / categorized with a given string. Case and leading and
	 * trailing whitespace are ignored.
	 *
	 * @param string $tag Tag to check for
	 *
	 * @return bool Whether or not at least one of the categories / tags on
	 *	$this->item is set to $tag (modulo case and leading and trailing
	 * 	whitespace)
	 */
	function isTaggedAs ($tag) {
		$desiredTag = strtolower(trim($tag)); // Normalize case and whitespace

		// Check to see if this is tagged with $tag
		$currentCategory = 'category';
		$currentCategoryNumber = 1;

		// If we have the new MagpieRSS, the number of category elements
		// on this item is stored under index "category#".
		if (isset($this->item['category#'])) :
			$numberOfCategories = (int) $this->item['category#'];

		// We REALLY shouldn't have the old and busted MagpieRSS, but in
		// case we do, it doesn't support multiple categories, but there
		// might still be a single value under the "category" index.
		elseif (isset($this->item['category'])) :
			$numberOfCategories = 1;

		// No standard category or tag elements on this feed item.
		else :
			$numberOfCategories = 0;

		endif;

		$isSoTagged = false; // Innocent until proven guilty

		// Loop through category elements; if there are multiple
		// elements, they are indexed as category, category#2,
		// category#3, ... category#N
		while ($currentCategoryNumber <= $numberOfCategories) :
			if ($desiredTag == strtolower(trim($this->item[$currentCategory]))) :
				$isSoTagged = true; // Got it!
				break;
			endif;

			$currentCategoryNumber += 1;
			$currentCategory = 'category#'.$currentCategoryNumber;
		endwhile;

		return $isSoTagged;
	} /* SyndicatedPost::isTaggedAs() */

	/**
	 * SyndicatedPost::enclosures: returns an array with any enclosures
	 * that may be attached to this syndicated item.
	 *
	 * @param string $type If you only want enclosures that match a certain
	 *	MIME type or group of MIME types, you can limit the enclosures
	 *	that will be returned to only those with a MIME type which
	 *	matches this regular expression.
	 * @return array
	 */
	function enclosures( $type = '/.*/' ) {
		$enclosures = array();

		if ( isset( $this->item['enclosure#'] ) ) :
			// Loop through enclosure, enclosure#2, enclosure#3, ....
			for ( $i = 1; $i <= $this->item['enclosure#']; $i++ ) :
				$eid = ( ( $i > 1 ) ? "#{$i}" : "" );	// this was set as #{$id} — was that a typo? (gwyneth 20230919)

				// Does it match the type we want?
				if ( preg_match($type, $this->item["enclosure{$eid}@type"] ) ) :
					$enclosures[] = array(
						"url"    => $this->item["enclosure{$eid}@url"],
						"type"   => $this->item["enclosure{$eid}@type"],
						"length" => $this->item["enclosure{$eid}@length"],
					);
				endif;
			endfor;
		endif;
		return $enclosures;
	} /* SyndicatedPost::enclosures() */

	function source ($what = NULL) {
		$ret = NULL;
		$source = $this->entry->get_source();
		if ($source) :
			$ret = array();
			$ret['title'] = $source->get_title();
			$ret['uri'] = $source->get_link();
			$ret['feed'] = $source->get_link(0, 'self');

			if ($id_tags = $source->get_source_tags(SIMPLEPIE_NAMESPACE_ATOM_10, 'id')) :
				$ret['id'] = $id_tags[0]['data'];
			elseif ($id_tags = $source->get_source_tags(SIMPLEPIE_NAMESPACE_ATOM_03, 'id')) :
				$ret['id'] = $id_tags[0]['data'];
			elseif ($id_tags = $source->get_source_tags(SIMPLEPIE_NAMESPACE_RSS_20, 'guid')) :
				$ret['id'] = $id_tags[0]['data'];
			elseif ($id_tags = $source->get_source_tags(SIMPLEPIE_NAMESPACE_RSS_10, 'guid')) :
				$ret['id'] = $id_tags[0]['data'];
			elseif ($id_tags = $source->get_source_tags(SIMPLEPIE_NAMESPACE_RSS_090, 'guid')) :
				$ret['id'] = $id_tags[0]['data'];
			endif;
		endif;

		if ( !is_null($what) and is_scalar($what)) :
			$ret = $ret[$what];
		endif;
		return $ret;
	}

	function comment_link () {
		$url = null;

		// RSS 2.0 has a standard <comments> element:
		// "<comments> is an optional sub-element of <item>. If present,
		// it is the url of the comments page for the item."
		// <http://cyber.law.harvard.edu/rss/rss.html#ltcommentsgtSubelementOfLtitemgt>
		if (isset($this->item['comments'])) :
			$url = $this->item['comments'];
		endif;

		// The convention in Atom feeds is to use a standard <link>
		// element with @rel="replies" and @type="text/html".
		// Unfortunately, SimplePie_Item::get_links() allows us to filter
		// by the value of @rel, but not by the value of @type. *sigh*

		// Try Atom 1.0 first
		$linkElements = $this->entry->get_item_tags(SIMPLEPIE_NAMESPACE_ATOM_10, 'link');

		// Fall back and try Atom 0.3
		if (is_null($linkElements)) : $linkElements =  $this->entry->get_item_tags(SIMPLEPIE_NAMESPACE_ATOM_03, 'link'); endif;

		// Now loop through the elements, screening by @rel and @type
		if (is_array($linkElements)) : foreach ($linkElements as $link) :
			$rel = (isset($link['attribs']['']['rel']) ? $link['attribs']['']['rel'] : 'alternate');
			$type = (isset($link['attribs']['']['type']) ? $link['attribs']['']['type'] : NULL);
			$href = (isset($link['attribs']['']['href']) ? $link['attribs']['']['href'] : NULL);

			if (strtolower($rel)=='replies' and $type=='text/html' and !is_null($href)) :
				$url = $href;
			endif;
		endforeach; endif;

		return $url;
	}

	function comment_feed () {
		$feed = null;

		// Well Formed Web comment feeds extension for RSS 2.0
		// <http://www.sellsbrothers.com/spout/default.aspx?content=archive.htm#exposingRssComments>
		//
		// N.B.: Correct capitalization is wfw:commentRss, but
		// wfw:commentRSS is common in the wild (partly due to a typo in
		// the original spec). In any case, our item array is normalized
		// to all lowercase anyways.
		if (isset($this->item['wfw']['commentrss'])) :
			$feed = $this->item['wfw']['commentrss'];
		endif;

		// In Atom 1.0, the convention is to use a standard link element
		// with @rel="replies". Sometimes this is also used to pass a
		// link to the human-readable comments page, so we also need to
		// check link/@type for a feed MIME type.
		//
		// Which is why I'm not using the SimplePie_Item::get_links()
		// method here, incidentally: it doesn't allow you to filter by
		// @type. *sigh*
		if (isset($this->item['link_replies'])) :
			// There may be multiple <link rel="replies"> elements; feeds have a feed MIME type
			$N = isset($this->item['link_replies#']) ? $this->item['link_replies#'] : 1;
			for ($i = 1; $i <= $N; $i++) :
				$currentElement = 'link_replies'.(($i > 1) ? '#'.$i : '');
				if (isset($this->item[$currentElement.'@type'])
				and preg_match("\007application/(atom|rss|rdf)\+xml\007i", $this->item[$currentElement.'@type'])) :
					$feed = $this->item[$currentElement];
				endif;
			endfor;
		endif;
		return $feed;
	} /* SyndicatedPost::comment_feed() */

	##################################
	#### BUILT-IN CONTENT FILTERS ####
	##################################

	var $uri_attrs = array (
		array('a', 'href'),
		array('applet', 'codebase'),
		array('area', 'href'),
		array('blockquote', 'cite'),
		array('body', 'background'),
		array('del', 'cite'),
		array('form', 'action'),
		array('frame', 'longdesc'),
		array('frame', 'src'),
		array('iframe', 'longdesc'),
		array('iframe', 'src'),
		array('head', 'profile'),
		array('img', 'longdesc'),
		array('img', 'src'),
		array('img', 'usemap'),
		array('input', 'src'),
		array('input', 'usemap'),
		array('ins', 'cite'),
		array('link', 'href'),
		array('object', 'classid'),
		array('object', 'codebase'),
		array('object', 'data'),
		array('object', 'usemap'),
		array('q', 'cite'),
		array('script', 'src')
	); /* var SyndicatedPost::$uri_attrs */

	var $_base = null;

	function resolve_single_relative_uri ($refs) {
		$tag = FeedWordPressHTML::attributeMatch($refs);
		$url = SimplePie_Misc::absolutize_url($tag['value'], $this->_base);

		return $tag['prefix'] . $url . $tag['suffix'];
	} /* function SyndicatedPost::resolve_single_relative_uri() */

	static function resolve_relative_uris ($content, $obj) {
		$set = $obj->link->setting('resolve relative', 'resolve_relative', 'yes');
		if ($set and $set != 'no') :
			// Fallback: if we don't have anything better, use the
			// item link from the feed
			$obj->_base = $obj->permalink(); // Reset the base for resolving relative URIs

			// What we should do here, properly, is to use
			// SimplePie_Item::get_base() -- but that method is
			// currently broken. Or getting down and dirty in the
			// SimplePie representation of the content tags and
			// grabbing the xml_base member for the content element.
			// Maybe someday...

			foreach ($obj->uri_attrs as $pair) :
				list($tag, $attr) = $pair;
				$pattern = FeedWordPressHTML::attributeRegex($tag, $attr);

				// FIXME: Encountered issue while testing an extremely long (= 88827 characters) item
				// Relying on preg_replace_callback() here can cause a PHP seg fault on my development
				// server. preg_match_all() causes a similar problem. Apparently this is a PCRE issue
				// Cf. discussion of similar issue <https://bugs.php.net/bug.php?id=65009>
				$content = preg_replace_callback (
					$pattern,
					array($obj, 'resolve_single_relative_uri'),
					$content
				);

			endforeach;
		endif;

		return $content;
	} /* function SyndicatedPost::resolve_relative_uris () */

	var $strip_attrs = array (
		array('[a-z]+', 'target'),
//		array('[a-z]+', 'style'),
//		array('[a-z]+', 'on[a-z]+'),
	);

	function strip_attribute_from_tag ($refs) {
		$tag = FeedWordPressHTML::attributeMatch($refs);
		return $tag['before_attribute'].$tag['after_attribute'];
	}

	static function sanitize_content ($content, $obj) {
		# This kind of sucks. I intend to replace it with
		# lib_filter sometime soon.
		foreach ($obj->strip_attrs as $pair):
			list($tag,$attr) = $pair;
			$pattern = FeedWordPressHTML::attributeRegex($tag, $attr);

			$content = preg_replace_callback (
				$pattern,
				array($obj, 'strip_attribute_from_tag'),
				$content
			);
		endforeach;
		return $content;
	} /* SyndicatedPost::sanitize() */

	#####################
	#### POST STATUS ####
	#####################

	/**
	 * SyndicatedPost::filtered: check whether or not this post has been
	 * screened out by a registered filter.
	 *
	 * @return bool TRUE iff post has been filtered out by a previous filter
	 */
	function filtered () {
		return is_null($this->post);
	} /* SyndicatedPost::filtered() */

	/**
	 * SyndicatedPost::freshness: check whether post is a new post to be
	 * inserted, a previously syndicated post that needs to be updated to
	 * match the latest revision, or a previously syndicated post that is
	 * still up-to-date.
	 *
	 * @return int A status code representing the freshness of the post
	 *	-1 = post already syndicated; has a revision that needs to be stored, but not updated to
	 *	0 = post already syndicated; no update needed
	 *	1 = post already syndicated, but needs to be updated to latest
	 *	2 = post has not yet been syndicated; needs to be created
	 */
	function freshness ($format = 'number') {

		if ($this->filtered()) : // This should never happen.
			FeedWordPressDiagnostic::critical_bug('SyndicatedPost', $this, __LINE__, __FILE__);
		endif;

		if (is_null($this->_freshness)) : // Not yet checked and cached.
			$guid = $this->post['guid'];
			$eguid = esc_sql($this->post['guid']);

			$q = new WP_Query(array(
				'fields' => '_synfresh', // id, guid, post_modified_gmt
				'ignore_sticky_posts' => true,
				'guid' => $eguid,	// it's always better to use the escaped version, right? (gwyneth 20230915)
			));

			$old_post = NULL;
			if ($q->have_posts()) :
				while ($q->have_posts()) : $q->the_post();
					if (get_post_type($q->post->ID) == $this->post['post_type']):
						$old_post = $q->post;
					endif;
				endwhile;
			endif;

			if (is_null($old_post)) : // No post with this guid
				FeedWordPress::diagnostic('feed_items:freshness', 'Item ['.$guid.'] "'.$this->entry->get_title().'" is a NEW POST.');
				$this->_wp_id = NULL;
				$this->_freshness = 2; // New content
			else :
				// Presume there is nothing new until we find
				// something new.
				$updated = false;
				$live = false;

				// Pull the list of existing revisions to get
				// timestamps.
				$revisions = wp_get_post_revisions($old_post->ID);
				foreach ($revisions as $rev) :
					$revisions_ts[] = mysql2date('G', $rev->post_modified_gmt);
				endforeach;

				$revisions_ts[] = mysql2date('G', $old_post->post_modified_gmt);
				$last_rev_ts = end($revisions_ts);
				$updated_ts = $this->updated(/*fallback=*/ true, /*default=*/ NULL);

				// If we have an explicit updated timestamp,
				// check that against existing stamps.
				if ( !is_null($updated_ts)) :
					$updated = !in_array($updated_ts, $revisions_ts);

					// If this a newer revision, make it go
					// live. If an older one, just record
					// the contents.
					$live = ($updated and ($updated_ts > $last_rev_ts));
				endif;

				// This is a revision we haven't seen before, judging by the date.

				$updatedReason = NULL;
				if ($updated) :
					$updatedReason = preg_replace(
						"/\s+/", " ",
						'has been marked with a new timestamp ('
						.date('Y-m-d H:i:s', $updated_ts)
						." > "
						.date('Y-m-d H:i:s', $last_rev_ts)
						.')'
					);

				// The date does not indicate a new revision, so
				// let's check the hash.
				else :
					// Or the hash...
					$hash = $this->update_hash();
					$seen = $this->stored_hashes($old_post->ID);
					if (is_countable($seen) and count($seen) > 0) :
						$updated = !in_array($hash, $seen); // Not seen yet?
					else :
						$updated = true; // Can't find syndication meta-data
					endif;

					if ($updated and FeedWordPressDiagnostic::is_on('feed_items:freshness:reasons')) :
						// In the absence of definitive
						// timestamp information, we
						// just have to assume that a
						// hash we haven't seen before
						// is a newer version.
						$live = true;

						$updatedReason = ' has a not-yet-seen update hash: '
						.MyPHP::val($hash)
						.' not in {'
						.implode(", ", array_map(array('FeedWordPress', 'val'), $seen))
						.'}. Basis: '
						.MyPHP::val(array_keys($this->update_hash(false)));
					endif;
				endif;

				$frozen = false;
				if ($updated) : // Ignore if the post is frozen
					$frozen = ('yes' == $this->link->setting('freeze updates', 'freeze_updates', NULL));
					if ( ! $frozen) :
						$frozen_value = get_post_meta($old_post->ID, '_syndication_freeze_updates', /*single=*/ true);
						$frozen = ( !is_null($frozen_value) and ('yes' == $frozen_value));

						if ($frozen) :
							$updatedReason = ' IS BLOCKED FROM BEING UPDATED BY A UPDATE LOCK ON THIS POST, EVEN THOUGH IT '.$updatedReason;
						endif;
					else :
						$updatedReason = ' IS BLOCKED FROM BEING UPDATED BY A FEEDWORDPRESS UPDATE LOCK, EVEN THOUGH IT '.$updatedReason;
					endif;
				endif;
				$live = ($live and ! $frozen);

				if ($updated) :
					FeedWordPress::diagnostic('feed_items:freshness', 'Item ['.$guid.'] "'.$this->entry->get_title().'" is an update of an existing post.');
					if ( !is_null($updatedReason)) :
						$updatedReason = preg_replace('/\s+/', ' ', $updatedReason);
						FeedWordPress::diagnostic('feed_items:freshness:reasons', 'Item ['.$guid.'] "'.$this->entry->get_title().'" '.$updatedReason);
					endif;

					$this->_freshness = apply_filters('syndicated_item_freshness', ($live ? 1 : -1), $updated, $frozen, $updated_ts, $last_rev_ts, $this);

					$this->_wp_id = $old_post->ID;
					$this->_wp_post = $old_post;

					// We want this to keep a running list of all the
					// processed update hashes.
					$this->post['meta']['syndication_item_hash'] = array_merge(
						$this->stored_hashes(),
						array($this->update_hash())
					);
				else :
					FeedWordPress::diagnostic('feed_items:freshness', 'Item ['.$guid.'] "'.$this->entry->get_title().'" is a duplicate of an existing post.');
					$this->_freshness = 0; // Same old, same old
					$this->_wp_id = $old_post->ID;
				endif;
			endif;
		endif;

		switch ($format) :
		case 'status' :
			switch ($this->_freshness) :
			case -1:
				$ret = 'stored';
				break;
			case 0:
				$ret = NULL;
				break;
			case 1:
				$ret = 'updated';
				break;
			case 2:
			default:
				$ret = 'new';
				break;
			endswitch;
			break;
		case 'number' :
		default :
			$ret = $this->_freshness;
		endswitch;


		return $ret;
	} /* SyndicatedPost::freshness () */

	function has_fresh_content () {
		return ( ! $this->filtered() and $this->freshness() != 0 );
	} /* SyndicatedPost::has_fresh_content () */

	function this_revision_needs_original_post ($freshness = NULL) {
		if (is_null($freshness)) :
			$freshness = $this->freshness();
		endif;
		return ( $freshness >= 2 );
	}

	function this_revision_is_current ($freshness = NULL) {
		if (is_null($freshness)) :
			$freshness = $this->freshness();
		endif;
		return ( $freshness >= 1 );
	} /* SyndicatedPost::this_revision_is_current () */

	function fresh_content_is_update () {
		return ($this->freshness() < 2);
	} /* SyndicatedPost::fresh_content_is_update () */

	function fresh_storage_diagnostic () {
		$ret = NULL;
		switch ($this->freshness()) :
		case -1 :
			$ret = 'Storing alternate revision of existing post # '.$this->wp_id().', "'.$this->post['post_title'].'"';
			break;
		case 1 :
			$ret = 'Updating existing post # '.$this->wp_id().', "'.$this->post['post_title'].'"';
			break;
		case 2 :
		default :
			$ret = 'Inserting new post "'.$this->post['post_title'].'"';
			break;
		endswitch;
		return $ret;
	} /* SyndicatedPost::fresh_storage_diagnostic() */

	function fresh_storage_hook () {
		$ret = NULL;
		switch ($this->freshness()) :
		case -1 :
		case 1 :
			$ret = 'update_syndicated_item';
			break;
		case 2 :
		default :
			$ret = 'post_syndicated_item';
			break;
		endswitch;
		return $ret;
	} /* SyndicatedPost::fresh_storage_hook () */

	#################################################
	#### INTERNAL STORAGE AND MANAGEMENT METHODS ####
	#################################################

	function wp_id () {
		if ($this->filtered()) : // This should never happen.
			FeedWordPressDiagnostic::critical_bug('SyndicatedPost', $this, __LINE__, __FILE__);
		endif;

		if (is_null($this->_wp_id) and is_null($this->_freshness)) :
			$fresh = $this->freshness(); // sets WP DB id in the process
		endif;
		return $this->_wp_id;
	}

	/**
	 * SyndicatedPost::secure_author_id(). Look up, or create, a numeric ID
	 * for the author of the incoming post.
	 *
	 * side effect: int|NULL stored in $this->post['post_author']
	 * side effect: IF no valid author is found, NULL stored in $this->post
	 * side effect: diagnostic output in case item is rejected with NULL author
	 *
	 * @used-by SyndicatedPost::store
	 *
	 * @uses SyndicatedPost::post
	 * @uses SyndicatedPost::author_id
	 * @uses SyndicatedLink::setting
	 * @uses FeedWordPress::diagnostic
	 */
	protected function secure_author_id () {
		# -- Look up, or create, numeric ID for author
		$this->post['post_author'] = $this->author_id (
			$this->link->setting('unfamiliar author', 'unfamiliar_author', 'create')
		);

		if (is_null($this->post['post_author'])) :
			FeedWordPress::diagnostic('feed_items:rejected', 'Filtered out item ['.$this->guid().'] without syndication: no author available');
			$this->post = NULL;
		endif;
	} /* SyndicatedPost::secure_author_id() */

	/**
	 * SyndicatedPost::secure_term_ids(). Look up, or create, numeric IDs
	 * for the terms (categories, tags, etc.) assigned to the incoming post,
	 * whether by global settings, feed settings, or by the tags on the feed.
	 *
	 * side effect: array of term ids stored in $this->post['tax_input']
	 * side effect: IF settings or filters determine post should be filtered out,
	 * 	NULL stored in $this->post
	 *
	 * @used-by SyndicatedPost::store
	 *
	 * @uses apply_filters
	 * @uses SyndicatedLink::setting
	 * @uses SyndicatedPost::category_ids
	 * @uses SyndicatedPost::preset_terms
	 * @uses SyndicatedPost::post
	 */
	protected function secure_term_ids () {
		$mapping = apply_filters('syndicated_post_terms_mapping', array(
			'category' => array('abbr' => 'cats', 'unfamiliar' => 'category', 'domain' => array('category', 'post_tag')),
			'post_tag' => array('abbr' => 'tags', 'unfamiliar' => 'post_tag', 'domain' => array('post_tag')),
		), $this);

		$termSet = array(); $valid = null;
		foreach ($this->feed_terms as $what => $anTerms) :
			// Default to using the inclusive procedures (for cats) rather than exclusive (for inline tags)
			$taxes = (isset($mapping[$what]) ? $mapping[$what] : $mapping['category']);
			$unfamiliar = $taxes['unfamiliar'];

			if ( !is_null($this->post)) : // Not filtered out yet
				# -- Look up, or create, numeric ID for categories
				$taxonomies = $this->link->setting("match/".$taxes['abbr'], 'match_'.$taxes['abbr'], $taxes['domain']);

				// Eliminate dummy variables
				$taxonomies = array_filter( $taxonomies, 'remove_dummy_zero' );

				// Allow FWP add-on filters to control the taxonomies we use to search for a term
				$taxonomies = apply_filters("syndicated_post_terms_match", $taxonomies, $what, $this);
				$taxonomies = apply_filters("syndicated_post_terms_match_{$what}", $taxonomies, $this);

				// Allow FWP add-on filters to control with greater precision what happens on unmatched
				$unmatched = apply_filters("syndicated_post_terms_unfamiliar",
					$this->link->setting(
						"unfamiliar {$unfamiliar}",
						"unfamiliar_{$unfamiliar}",
						'create:'.$unfamiliar
					),
					$what,
					$this
				);

				$terms = $this->category_ids (
					$anTerms,
					$unmatched,
					/*taxonomies=*/ $taxonomies,
					array(
					  'singleton' => false, // I don't like surprises
					  'filters' => true,
					)
				);

				if (is_null($terms) or is_null($termSet)) :
					// filtered out -- no matches
				else :
					$valid = true;

					// filter mode off, or at least one match
					foreach ($terms as $tax => $term_ids) :
						if ( !isset($termSet[$tax])) :
							$termSet[$tax] = array();
						endif;
						$termSet[$tax] = array_merge($termSet[$tax], $term_ids);
					endforeach;
				endif;
			endif;
		endforeach;

		if (is_null($valid)) : // Plonked
			$this->post = NULL;
		else : // We can proceed
			$this->post['tax_input'] = array();
			foreach ($termSet as $tax => $term_ids) :
				if ( !isset($this->post['tax_input'][$tax])) :
					$this->post['tax_input'][$tax] = array();
				endif;
				$this->post['tax_input'][$tax] = array_merge(
					$this->post['tax_input'][$tax],
					$term_ids
				);
			endforeach;

			// Now let's add on the feed and global presets
			foreach ($this->preset_terms as $tax => $term_ids) :
				if ( !isset($this->post['tax_input'][$tax])) :
					$this->post['tax_input'][$tax] = array();
				endif;

				$this->post['tax_input'][$tax] = array_merge (
					$this->post['tax_input'][$tax],
					$this->category_ids (
					/*terms=*/ $term_ids,
					/*unfamiliar=*/ 'create:'.$tax, // These are presets; for those added in a tagbox editor, the tag may not yet exist
					/*taxonomies=*/ array($tax),
					array(
					  'singleton' => true,
					))
				);
			endforeach;
		endif;
	} /* SyndicatedPost::secure_term_ids() */

	/**
	 * SyndicatedPost::store
	 *
	 * @uses SyndicatedPost::secure_author_id
	 */
	public function store () {
		if ($this->filtered()) : // This should never happen.
			FeedWordPressDiagnostic::critical_bug('SyndicatedPost', $this, __LINE__, __FILE__);
		endif;

		$freshness = $this->freshness();		// see above: this sets WP DB id in the process (gwyneth 20230919)
		if ($this->has_fresh_content()) :
			$this->secure_author_id();
		endif;

		if ($this->has_fresh_content()) : // Was this filtered during author_id lookup?
			$this->secure_term_ids();
		endif;

		// We have to check again in case the post has been filtered
		// during the category/tags/taxonomy terms lookup
		if ($this->has_fresh_content()) :
			// Filter some individual fields

			// If there already is a post slug (from syndication or by manual
			// editing) don't cause WP to overwrite it by sending in a NULL
			// post_name. Props Chris Fritz 2012-11-28.
			$post_name = (is_null($this->_wp_post) ? NULL : $this->_wp_post->post_name);

			// Allow filters to set post slug. Props niska.
			$post_name = apply_filters('syndicated_post_slug', $post_name, $this);
			if ( !empty($post_name)) :
				$this->post['post_name'] = $post_name;
			endif;

			$this->post = apply_filters('syndicated_post', $this->post, $this);

			// Allow for feed-specific syndicated_post filters.
			$this->post = apply_filters(
				"syndicated_post_".$this->link->uri(),
				$this->post,
				$this
			);
		endif;

		// Hook in early to make sure these get inserted if at all possible
		add_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'add_rss_meta'),
			/*priority=*/ -10000, /* very early */
			/*arguments=*/ 3
		);

		$ret = false;
		if ($this->has_fresh_content()) :
			$diag = $this->fresh_storage_diagnostic();
			if ( !is_null($diag)) :
				FeedWordPress::diagnostic('syndicated_posts', $diag);
			endif;

			$this->insert_post(/*update=*/ $this->fresh_content_is_update(), $this->freshness());

			$hook = $this->fresh_storage_hook();
			if ( !is_null($hook)) :
				do_action($hook, $this->wp_id(), $this);
			endif;

			$ret = $this->freshness('status');
		endif;

		// If this is a legit, non-filtered post, tag it as found on the
		// feed regardless of fresh or stale status
		if ( ! $this->filtered()) :
			$key = '_feedwordpress_retire_me_' . $this->link->id;
			delete_post_meta($this->wp_id(), $key);

			$status = get_post_field('post_status', $this->wp_id());
			if ('fwpretired'==$status and $this->link->is_non_incremental()) :
				FeedWordPress::diagnostic('syndicated_posts', "Un-retiring previously retired post # ".$this->wp_id()." due to re-appearance on non-incremental feed.");
				set_post_field('post_status', $this->post['post_status'], $this->wp_id());
				wp_transition_post_status($this->post['post_status'], $status, $this->post);
			elseif ('fwpzapped'==$status) :
				// Set this new revision up to be
				// blanked on the next update.
				add_post_meta($this->wp_id(), '_feedwordpress_zapped_blank_me', 2, /*single=*/ true);
			endif;
		endif;

		// Remove add_rss_meta hook
		remove_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'add_rss_meta'),
			/*priority=*/ -10000, /* very early */
			/*arguments=*/ 3
		);

		return $ret;
	} /* function SyndicatedPost::store () */

	function insert_post ($update = false, $freshness = 2) {
		$dbpost = $this->normalize_post(/*new=*/ true);

		$ret = null;

		if ( !is_null($dbpost)) :
			$dbpost['post_pingback'] = false; // Tell WP 2.1 and 2.2 not to process for pingbacks

			// This is a ridiculous fucking kludge necessitated by WordPress 2.6 munging authorship meta-data
			add_action('_wp_put_post_revision', array($this, 'fix_revision_meta'));

			// Always run syndicated content through kses, using the
			// rules WordPress applies to users without `unfiltered_html`,
			// no matter who (if anyone) triggered this update.
			$kses_forced = ! has_filter('content_save_pre', 'wp_filter_post_kses');
			if ($kses_forced) :
				kses_init_filters();
			endif;

			// Let kses keep <video>, <audio>, and iframes from a short
			// list of trusted video hosts in syndicated post content.
			$allow_embeds = apply_filters('feedwordpress_allow_syndicated_embeds', true, $this);
			if ($allow_embeds) :
				add_filter('content_save_pre', array(__CLASS__, 'kses_embeds_open'), 9);
				add_filter('content_save_pre', array(__CLASS__, 'kses_embeds_close'), 11);
			endif;

			// kses allows position:fixed/sticky in style attributes,
			// which lets feed content overlay the whole page.
			add_filter('content_save_pre', array(__CLASS__, 'strip_fixed_positioning'), 12);
			add_filter('excerpt_save_pre', array(__CLASS__, 'strip_fixed_positioning'), 12);

			if ($update and function_exists('get_post_field')) :
				// Don't munge status fields that the user may
				// have reset manually
				$doNotMunge = array('post_status', 'comment_status', 'ping_status');

				foreach ($doNotMunge as $field) :
					$dbpost[$field] = get_post_field($field, $this->wp_id());
				endforeach;
			endif;

			// WP3's wp_insert_post scans current_user_can() for the
			// tax_input, with no apparent way to override. Ugh.
			add_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'add_terms'),
			/*priority=*/ -10001, /* very early */
			/*arguments=*/ 3
			);

			// WP3 appears to override whatever you give it for
			// post_modified. Ugh.
			add_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'fix_post_modified_ts'),
			/*priority=*/ -10000, /* very early */
			/*arguments=*/ 3
			);

			if ($update) :
				$this->post['ID'] = $this->wp_id();
				$dbpost['ID'] = $this->post['ID'];
			endif;

			// O.K., is this a new post? If so, we need to create
			// the basic post record before we do anything else.
			if ($this->this_revision_needs_original_post()) :
				// *sigh*, for handling inconsistent slash expectations < 3.6
				$sdbpost = $this->db_sanitize_post($dbpost);

				// Go ahead and insert the first post record to
				// anchor the revision history.

				$this->_wp_id = wp_insert_post($sdbpost, /*return wp_error=*/ true);

				$dbpost['ID'] = $this->_wp_id;
			endif;

			// Sanity check: if the attempt to insert post
			// returned an error, then feeding that error
			// object in to _wp_put_post_revision() would
			// cause a fatal error. Better to break out.
			if ( !is_wp_error($this->_wp_id)) :
				// Now that we've made sure the original exists, insert
				// this version here as a revision.
				$revision_id = _wp_put_post_revision($dbpost, /*autosave=*/ false);

				if ( ! $this->this_revision_needs_original_post()) :

					if ($this->this_revision_is_current()) :

						wp_restore_post_revision($revision_id);

					else :

						// If we do not activate this revision, then the
						// add_rss_meta will not be called, which is
						// more or less as it should be, but that means
						// we have to actively record this revision's
						// update hash from here.
						$postId = $this->post['ID'];
						$key = 'syndication_item_hash';
						$hash = $this->update_hash();
						FeedWordPress::diagnostic('syndicated_posts:meta_data', "Adding post meta-datum to post [$postId]: [$key] = ".FeedWordPress::val($hash, /*no newlines=*/ true));
						add_post_meta(	$postId, $key, $hash, /*unique=*/ false );
					endif;
				endif;
			endif;

			remove_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'add_terms'),
			/*priority=*/ -10001, /* very early */
			/*arguments=*/ 3
			);

			remove_action(
			/*hook=*/ 'transition_post_status',
			/*callback=*/ array($this, 'fix_post_modified_ts'),
			/*priority=*/ -10000, /* very early */
			/*arguments=*/ 3
			);

			// Turn off ridiculous fucking kludge #1
			remove_action('_wp_put_post_revision', array($this, 'fix_revision_meta'));

			// Restore normal content filtering for the current user.
			if ($allow_embeds) :
				remove_filter('content_save_pre', array(__CLASS__, 'kses_embeds_open'), 9);
				remove_filter('content_save_pre', array(__CLASS__, 'kses_embeds_close'), 11);
			endif;
			remove_filter('content_save_pre', array(__CLASS__, 'strip_fixed_positioning'), 12);
			remove_filter('excerpt_save_pre', array(__CLASS__, 'strip_fixed_positioning'), 12);
			if ($kses_forced) :
				kses_init();
			endif;

			$this->validate_post_id($dbpost, $update, array(__CLASS__, __FUNCTION__));

			$ret = $this->_wp_id;
		endif;
		return $ret;
	} /* function SyndicatedPost::insert_post () */

	/**
	 * SyndicatedPost::kses_allow_embeds() -- wp_kses_allowed_html filter that
	 * adds <video>, <audio>, their <source>/<track> children, and a minimal
	 * <iframe> to the 'post' context. Only hooked in (by kses_embeds_open())
	 * for the duration of kses's pass over syndicated post_content.
	 *
	 * Iframe attributes are deliberately minimal: never srcdoc (runs in our
	 * origin), allow (delegates browser permissions), sandbox, name or style.
	 * Iframe hosts are checked afterward by restrict_iframe_hosts().
	 */
	static function kses_allow_embeds ($allowed, $context) {
		if ('post' !== $context) :
			return $allowed;
		endif;

		$allowed['video'] = array_merge(isset($allowed['video']) ? $allowed['video'] : array(), array(
			'controls' => true, 'width' => true, 'height' => true, 'preload' => true,
			'muted' => true, 'loop' => true, 'playsinline' => true, 'src' => true, 'poster' => true,
		));
		$allowed['audio'] = array_merge(isset($allowed['audio']) ? $allowed['audio'] : array(), array(
			'controls' => true, 'preload' => true, 'muted' => true, 'loop' => true, 'src' => true,
		));
		$allowed['source'] = array('src' => true, 'type' => true);
		$allowed['track'] = array('src' => true, 'kind' => true, 'srclang' => true, 'label' => true, 'default' => true);
		$allowed['iframe'] = array(
			'src' => true, 'width' => true, 'height' => true,
			'title' => true, 'allowfullscreen' => true, 'loading' => true,
		);
		return $allowed;
	} /* SyndicatedPost::kses_allow_embeds () */

	/**
	 * SyndicatedPost::kses_embeds_open() -- content_save_pre @ 9, just
	 * before kses (@ 10): widen the allowlist for this one kses pass.
	 */
	static function kses_embeds_open ($content) {
		add_filter('wp_kses_allowed_html', array(__CLASS__, 'kses_allow_embeds'), 10, 2);
		return $content;
	} /* SyndicatedPost::kses_embeds_open () */

	/**
	 * SyndicatedPost::kses_embeds_close() -- content_save_pre @ 11, just
	 * after kses: close the widened allowlist, then host-check iframes.
	 */
	static function kses_embeds_close ($content) {
		remove_filter('wp_kses_allowed_html', array(__CLASS__, 'kses_allow_embeds'), 10);
		return self::restrict_iframe_hosts($content);
	} /* SyndicatedPost::kses_embeds_close () */

	/**
	 * SyndicatedPost::restrict_iframe_hosts() -- keep only https iframes on
	 * an exact host + path allowlist; remove all others. Fails closed: if
	 * the HTML parser can't account for every iframe, strip them all.
	 *
	 * @param string $content Slashed post content (as on content_save_pre).
	 * @return string Slashed post content.
	 */
	static function restrict_iframe_hosts ($content) {
		if (false === stripos($content, '<iframe')) :
			return $content;
		endif;

		$youtube = '#^/embed/[A-Za-z0-9_-]+/?$#';
		$hosts = apply_filters('feedwordpress_allowed_iframe_hosts', array(
			'www.youtube.com' => $youtube,
			'youtube.com' => $youtube,
			'www.youtube-nocookie.com' => $youtube,
			'youtube-nocookie.com' => $youtube,
			'player.vimeo.com' => '#^/video/[0-9]+/?$#',
			'open.spotify.com' => '#^/embed/(track|album|playlist|episode|show|artist)/[A-Za-z0-9]+/?$#',
		));

		$html = wp_unslash($content);
		$p = new WP_HTML_Tag_Processor($html);
		$bad = false; $seen = 0;
		while ($p->next_tag('IFRAME')) :
			if ($p->is_tag_closer()) :
				continue;
			endif;
			$seen++;

			$parts = wp_parse_url((string) $p->get_attribute('src')); // entity-decoded
			$host = (is_array($parts) and isset($parts['host'])) ? strtolower($parts['host']) : '';
			$ok = ($host and isset($hosts[$host], $parts['scheme'], $parts['path'])
				and 'https' === strtolower($parts['scheme'])
				and ! isset($parts['user']) and ! isset($parts['pass']) and ! isset($parts['port'])
				and preg_match($hosts[$host], $parts['path']));

			if ($ok) :
				$p->set_attribute('allow', 'encrypted-media; picture-in-picture; fullscreen');
				$p->set_attribute('referrerpolicy', 'strict-origin-when-cross-origin');
			else :
				$p->remove_attribute('src');
				$p->set_attribute('data-fwp-blocked-iframe', '1');
				$bad = true;
			endif;
		endwhile;

		// paused_at_incomplete_token() is WP 6.5+; on 6.2-6.4 the count
		// comparison below is what catches an unclosed or unparsed iframe.
		$visited_all = ! (method_exists($p, 'paused_at_incomplete_token') and $p->paused_at_incomplete_token());
		$html = $p->get_updated_html();

		if ( ! $visited_all or preg_match_all('#<iframe\b#i', $html) !== $seen) :
			return wp_slash(preg_replace('#</?iframe\b[^>]*>?#i', '', $html));
		endif;

		if ($bad) :
			$html = preg_replace('#<iframe\b[^>]*\bdata-fwp-blocked-iframe="1"[^>]*>.*?</iframe>#is', '', $html);
			$html = preg_replace('#<iframe\b[^>]*\bdata-fwp-blocked-iframe="1"[^>]*>#is', '', $html);
		endif;
		return wp_slash($html);
	} /* SyndicatedPost::restrict_iframe_hosts () */

	/**
	 * SyndicatedPost::strip_fixed_positioning() -- content_save_pre and
	 * excerpt_save_pre @ 12, after kses: remove any CSS `position`
	 * declaration from style attributes unless its value is static,
	 * relative or absolute. That removes fixed and sticky (incl.
	 * -webkit-sticky), and also var(...) values, which kses allows and
	 * which could otherwise resolve to fixed via a custom property.
	 *
	 * @param string $content Slashed content (as on *_save_pre).
	 * @return string Slashed content.
	 */
	static function strip_fixed_positioning ($content) {
		if (false === stripos($content, 'position')) :
			return $content;
		endif;

		$html = wp_unslash($content);
		$p = new WP_HTML_Tag_Processor($html);
		while ($p->next_tag()) :
			$style = $p->get_attribute('style'); // entity-decoded
			if ( ! is_string($style) or false === stripos($style, 'position')) :
				continue;
			endif;

			$kept = array();
			foreach (explode(';', $style) as $decl) :
				if (preg_match('/^\s*position\s*:(.*)$/is', $decl, $m)
				and ! preg_match('/^\s*(static|relative|absolute)\s*(!\s*important\s*)?$/i', $m[1])) :
					continue;
				endif;
				$kept[] = $decl;
			endforeach;

			$style = trim(implode(';', $kept), " \t\n\r;");
			if ('' === $style) :
				$p->remove_attribute('style');
			else :
				$p->set_attribute('style', $style);
			endif;
		endwhile;
		$html = $p->get_updated_html();

		// Fail closed: an unterminated tag at the end of the content is
		// never visited by the parser (and browsers drop it anyway), but
		// text appended later, e.g. boilerplate, could complete it. Drop it.
		$html = preg_replace('#<[a-z][^>]*$#i', '', $html);

		return wp_slash($html);
	} /* SyndicatedPost::strip_fixed_positioning () */

	/**
	 * SyndicatedPost::insert_new(). Uses the data collected in this post object to insert
	 * a new post into the wp_posts table.
	 *
	 * @uses SyndicatedPost::insert_post
	 */
	function insert_new () {
		$this->insert_post(/*update=*/ false, 1);
	} /* SyndicatedPost::insert_new() */

	/**
	 * SyndicatedPost::insert_new(). Uses the data collected in this post object to update
	 * an existing post in the wp_posts table.
	 *
	 * @uses SyndicatedPost::insert_post
	 */
	function update_existing () {
		$this->insert_post(/*update=*/ true, 2);
	} /* SyndicatedPost::update_existing() */

	/**
	 * SyndicatedPost::normalize_post()
	 *
	 * @param bool $new If true, this post is to be inserted anew. If false, it is an update of an existing post. (Unused)
	 * @return array A normalized representation of the post ready to be inserted into the database or sent to the WordPress API functions
	 */
	function normalize_post( $new = true ) {
		$out = $this->post;

		$fullPost = $out['post_title'].$out['post_content'];
		$fullPost .= (isset($out['post_excerpt']) ? $out['post_excerpt'] : '');
		if (strlen($fullPost) < 1) :
			// FIXME: Option for filtering out empty posts
		endif;
		if (strlen($out['post_title'])==0) :
			$offset = (int) get_option('gmt_offset') * 60 * 60;
			if (isset($this->post['meta']['syndication_source'])) :
				$source_title = $this->post['meta']['syndication_source'];
			else :
				$feed_url = parse_url($this->post['meta']['syndication_feed']);
				$source_title = $feed_url['host'];
			endif;

			$out['post_title'] = $source_title
				.' '.gmdate('Y-m-d H:i:s', $this->published() + $offset);
			// FIXME: Option for what to fill a blank title with...
		endif;

		// Normalize the guid if necessary.
		$out['guid'] = SyndicatedPost::normalize_guid($out['guid']);

		return $out;
	}

	public function db_sanitize_post_check_encoding ($out) {
		// Check encoding recursively: every string field needs to be checked
		// for character encoding issues. This is a bit problematic because we
		// *should* be using DB_CHARSET, but DB_CHARSET sometimes has values
		// that work for MySQL but not for PHP mb_check_encoding. So instead
		// we must rely on WordPress setting blog_charset and hope that the user
		// has got their database encoding set up to roughly match
		$charset = get_option('blog_charset', 'utf8');

		foreach ($out as $key => $value) :
			if (is_string($value)) :

				if ( !function_exists('mb_check_encoding') or mb_check_encoding($value, $charset)) :
					$out[$key] = $value;
				else :
					$fromCharset = mb_detect_encoding($value, mb_detect_order(), /*strict=*/ true);
					$out[$key] = mb_convert_encoding($value, $charset, $fromCharset);
				endif;

			elseif (is_array($value)) :
				$out[$key] = $this->db_sanitize_post_check_encoding($value);

			else :
				$out[$key] = $value;
			endif;

		endforeach;

		return $out;
	} /* SyndicatedPost::db_sanitize_post_check_encoding () */

	function db_sanitize_post ($out) {
		global $wp_db_version;

		$out = $this->db_sanitize_post_check_encoding($out);

		// < 3.6. Core API, including `wp_insert_post()`, expects
		// properly slashed data. If `wp_slash()` exists, then
		// this is after the big change-over in how data slashing
		// was handled.
		if ( !function_exists('wp_slash')) :

			foreach ($out as $key => $value) :
				if (is_string($value)) :
					$out[$key] = esc_sql($value);
				else :
					$out[$key] = $value;
				endif;
			endforeach;

		// For revisions [@23416,@23554), core API expects
		// unslashed data. Cf. <https://core.trac.wordpress.org/browser/trunk/wp-includes/post.php?rev=23416>
		// 	NOOP for those revisions.

		// In revisions @23554 to present, `wp_insert_post()`
		// expects slashed data once again.
		// Cf. <https://core.trac.wordpress.org/changeset/23554/trunk/wp-includes/post.php?contextall=1>
		// But at least now we can use the wp_slash API function to do that.
		// Hooray.

		elseif ($wp_db_version >= 23524) :

			$out = wp_slash($out);

		endif;

		return $out;
	}

	/**
	 * SyndicatedPost::validate_post_id()
	 *
	 * @param array $dbpost An array representing the post we attempted to insert or update
	 * @param mixed $ns A string or array representing the namespace (class, method) whence this method was called.
	 */
	function validate_post_id ($dbpost, $is_update, $ns) {
		if (is_array($ns)) : $ns = implode('::', $ns);
		else : $ns = (string) $ns; endif;

		// This should never happen.
		if ( !is_numeric($this->_wp_id) or ($this->_wp_id == 0)) :
			$verb = ($is_update ? 'update existing' : 'insert new');
			$guid = $this->guid();
			$url = $this->permalink();
			$feed = $this->link->uri(array('add_params' => true));

			// wp_insert_post failed. Diagnostics, or barf up a critical bug
			// notice if we are in debug mode.
			$mesg = "Failed to $verb item [$guid]. WordPress API returned no valid post ID.\n"
				."\t\tID = ".serialize($this->_wp_id)."\n"
				."\t\tURL = ".MyPHP::val($url)
				."\t\tFeed = ".MyPHP::val($feed);

			FeedWordPress::diagnostic('updated_feeds:errors', "WordPress API error: $mesg");
			FeedWordPress::diagnostic('feed_items:rejected', $mesg);

			$mesg = <<<EOM
The WordPress API returned an invalid post ID
			   when FeedWordPress tried to $verb item $guid
			   [URL: $url]
			   from the feed at $feed

$ns::_wp_id
EOM;
			FeedWordPressDiagnostic::noncritical_bug(
				/*message=*/ $mesg,
				/*var =*/ array(
					"\$this->_wp_id" => $this->_wp_id,
					"\$dbpost" => $dbpost,
				),
				/*line # =*/ __LINE__, /*filename=*/ __FILE__
			);
		endif;
	} /* SyndicatedPost::validate_post_id() */

	/**
	 * SyndicatedPost::fix_revision_meta() - Ensures that we get the meta
	 * data (authorship, guid, etc.) that we want when storing revisions of
	 * a syndicated post.
	 *
	 * In their infinite wisdom, the WordPress coders seem to have made it
	 * completely impossible for a plugin that uses wp_insert_post() to set
	 * certain meta-data (such as the author) when you store an old revision
	 * of an updated post. Instead, it uses the WordPress defaults (= cur.
	 * active user ID if the process is running with a user logged in, or
	 * = #0 if there is no user logged in). This results in bogus authorship
	 * data for revisions that are syndicated from off the feed, unless we
	 * use a ridiculous kludge like this to end-run the munging of meta-data
	 * by _wp_put_post_revision.
	 *
	 * @param int $revision_id The revision ID to fix up meta-data
	 */
	function fix_revision_meta( $revision_id ) {
		$post_author = (int) $this->post['post_author'];	// is this a global? And if it is, where is it defined? (gwyneth 20230919)

		$revision_id = (int) $revision_id;

		// Let's fix the author.
		set_post_field( 'post_author', $this->post['post_author'], $revision_id );

		// Let's fix the GUID to a dummy URL with the update hash.
		set_post_field( 'guid', 'http://feedwordpress.radgeek.com/?rev=' . $this->update_hash(), $revision_id );

		// Let's fire an event for add-ons and filters
		do_action( 'syndicated_post_fix_revision_meta', $revision_id, $this );

	} /* SyndicatedPost::fix_revision_meta () */

	/**
	 * SyndicatedPost::add_terms() -- if FeedWordPress is processing an
	 * automatic update, that generally means that wp_insert_post() is being
	 * called under the user credentials of whoever is viewing the blog at
	 * the time -- which usually means no user at all. But wp_insert_post()
	 * checks current_user_can() before assigning any of the terms in a
	 * post's tax_input structure -- which is unfortunate, since
	 * current_user_can() always returns FALSE when there is no current user
	 * logged in. Meaning that automatic updates get no terms assigned.
	 *
	 * So, wp_insert_post() is not going to do the term assignments for us.
	 * If you want something done right....
	 *
	 * @param string $new_status Unused action parameter.
	 * @param string $old_status Unused action parameter.
	 * @param object $post The database record for the post just inserted.
	 */
	function add_terms ($new_status, $old_status, $post) {

		if ($new_status!='inherit') : // Bail if we are creating a revision.
			if ( is_array($this->post) and isset($this->post['tax_input']) and is_array($this->post['tax_input']) ) :
				foreach ($this->post['tax_input'] as $taxonomy => $terms) :
					if (is_array($terms)) :
						$terms = array_filter($terms); // strip out empties
					endif;

					$res = wp_set_post_terms(
						/*post_id=*/ $post->ID,
						/*terms=*/ $terms,
						/*taxonomy=*/ $taxonomy
					);

					FeedWordPress::diagnostic(
						'syndicated_posts:categories',
						'Category: post('.json_encode($post->ID).') '.$taxonomy
						.' := '
						.json_encode($terms)
						.' / result: '
						.json_encode($res)
					);

				endforeach;
			endif;
		endif;

	} /* SyndicatedPost::add_terms () */

	/**
	 * SyndicatedPost::fix_post_modified_ts() -- We would like to set
	 * post_modified and post_modified_gmt to reflect the value of
	 * <atom:updated> or equivalent elements on the feed. Unfortunately,
	 * wp_insert_post() refuses to acknowledge explicitly-set post_modified
	 * fields and overwrites them, either with the post_date (if new) or the
	 * current timestamp (if updated).
	 *
	 * So, wp_insert_post() is not going to do the last-modified assignments
	 * for us. If you want something done right....
	 *
	 * @param string $new_status Unused action parameter.
	 * @param string $old_status Unused action parameter.
	 * @param object $post The database record for the post just inserted.
	 */
	function fix_post_modified_ts ($new_status, $old_status, $post) {
		global $wpdb;
		if ($new_status!='inherit') : // Bail if we are creating a revision.
			$wpdb->update( $wpdb->posts, /*data=*/ array(
			'post_modified' => $this->post['post_modified'],
			'post_modified_gmt' => $this->post['post_modified_gmt'],
			), /*where=*/ array('ID' => $post->ID) );
		endif;
	} /* SyndicatedPost::fix_post_modified_ts () */

	/**
	 * SyndicatedPost::add_rss_meta: adds interesting meta-data to each entry
	 * using the space for custom keys. The set of keys and values to add is
	 * specified by the keys and values of $post['meta']. This is used to
	 * store anything that the WordPress user might want to access from a
	 * template concerning the post's original source that isn't provided
	 * for by standard WP meta-data (i.e., any interesting data about the
	 * syndicated post other than author, title, timestamp, categories, and
	 * guid). It's also used to hook into WordPress's support for
	 * enclosures.
	 *
	 * @param string $new_status Unused action parameter.
	 * @param string $old_status Unused action parameter.
	 * @param object $post The database record for the post just inserted.
	 */
	function add_rss_meta($new_status, $old_status, $post) {
		global $wpdb;
		if ($new_status!='inherit') : // Bail if we are creating a revision.
			FeedWordPress::diagnostic('syndicated_posts:meta_data', 'Adding post meta-data: {'.implode(", ", array_keys($this->post['meta'])).'}');

			if ( is_array($this->post) and isset($this->post['meta']) and is_array($this->post['meta']) ) :
				$postId = $post->ID;

				// Aggregated posts should NOT send out pingbacks.
				// WordPress 2.1-2.2 claim you can tell them not to
				// using $post_pingback, but they don't listen, so we
				// make sure here.
				$result = $wpdb->query("
				DELETE FROM $wpdb->postmeta
				WHERE post_id='$postId' AND meta_key='_pingme'
				");

				foreach ( $this->post['meta'] as $key => $values ) :
					$eKey = esc_sql($key);

					// If this is an update, clear out the old
					// values to avoid duplication.
					$result = $wpdb->query("
					DELETE FROM $wpdb->postmeta
					WHERE post_id='$postId' AND meta_key='$eKey'
					");

					// Allow for either a single value or an array
					if ( !is_array($values)) $values = array($values);
					foreach ( $values as $value ) :
					FeedWordPress::diagnostic('syndicated_posts:meta_data', "Adding post meta-datum to post [$postId]: [$key] = ".MyPHP::val($value, /*no newlines=*/ true));
						add_post_meta($postId, $key, $value, /*unique=*/ false);
					endforeach;
				endforeach;

				if ( $result === false)
				{
					error_log( "delete query failed: " . $wpdb->last_error );
				}
			endif;
		endif;
	} /* SyndicatedPost::add_rss_meta () */

	/**
	 * SyndicatedPost::author_id (): get the ID for an author name from
	 * the feed. Create the author if necessary.
	 *
	 * @param string $unfamiliar_author
	 *
	 * @return NULL|int The numeric ID of the author to attribute the post to
	 *	NULL if the post should be filtered out.
	 */
	function author_id ($unfamiliar_author = 'create') {
		global $wpdb;

		$a = $this->named['author'];

		$source = $this->source();
		$forbidden = apply_filters('feedwordpress_forbidden_author_names',
			array('admin', 'administrator', 'www', 'root'));

		// Prepare the list of candidates to try for author name: name from
		// feed, original source title (if any), immediate source title live
		// from feed, subscription title, prettied version of feed homepage URL,
		// prettied version of feed URL, or, failing all, use "unknown author"
		// as last resort

		$candidates = array();
		$candidates[] = $a['name'];
		if ( !is_null($source)) : $candidates[] = $source['title']; endif;
		$candidates[] = $this->link->name(/*fromFeed=*/ true);
		$candidates[] = $this->link->name(/*fromFeed=*/ false);
		if (strlen($this->link->homepage()) > 0) : $candidates[] = feedwordpress_display_url($this->link->homepage()); endif;
		$candidates[] = feedwordpress_display_url($this->link->uri());
		$candidates[] = 'unknown author';

		// Pick the first one that works from the list, screening against empty
		// or forbidden names.

		$author = NULL;
		foreach ($candidates as $candidate) {
			if ( !is_null($candidate)
				and (strlen(trim($candidate)) > 0)
				and !in_array(strtolower(trim($candidate)), $forbidden)) :
					$author = $candidate;
					break;
			endif;
		}

		$email = (isset($a['email']) ? $a['email'] : NULL);
		$authorUrl = (isset($a['uri']) ? $a['uri'] : NULL);


		$hostUrl = $this->link->homepage();
		if (is_null($hostUrl) or (strlen($hostUrl) < 0)) :
			$hostUrl = $this->link->uri();
		endif;

		$match_author_by_email = !('yes' == get_option("feedwordpress_do_not_match_author_by_email"));
		if ($match_author_by_email and !FeedWordPress::is_null_email($email)) :
			$test_email = $email;
		else :
			$test_email = NULL;
		endif;

		// Never can be too careful...
		$login = sanitize_user($author, /*strict=*/ true);

		// Possible for, e.g., foreign script author names
		if (strlen($login) < 1) :
			// No usable characters in author name for a login.
			// (Sometimes results from, e.g., foreign scripts.)
			//
			// We just need *something* in Western alphanumerics,
			// so let's try the domain name.
			//
			// Uniqueness will be guaranteed below if necessary.

			$url = parse_url($hostUrl);

			$login = sanitize_user($url['host'], /*strict=*/ true);
			if (strlen($login) < 1) :
				// This isn't working. Frak it.
				$login = 'syndicated';
			endif;
		endif;

		$login = apply_filters('pre_user_login', $login);

		$nice_author = sanitize_title($author);
		$nice_author = apply_filters('pre_user_nicename', $nice_author);

		$reg_author = esc_sql(preg_quote($author ?: "(unknown)"));
		$author = esc_sql($author);
		$email = esc_sql($email);
		$test_email = esc_sql($test_email);
		$authorUrl = esc_sql($authorUrl);

		// Check for an existing author rule....
		if (isset($this->link->settings['map authors']['name']['*'])) :
			$author_rule = $this->link->settings['map authors']['name']['*'];
		elseif (isset($this->link->settings['map authors']['name'][strtolower(trim($author))])) :
			$author_rule = $this->link->settings['map authors']['name'][strtolower(trim($author))];
		else :
			$author_rule = NULL;
		endif;

		// User name is mapped to a particular author. If that author ID exists, use it.
		// Note: get_userdata() is a pluggable function, so it might not be available yet;
		// check if it exists before using it, or WP will throw a fatal error. (gwyneth 20220223)
		if (is_numeric($author_rule) and function_exists('get_userdata') and get_userdata((int) $author_rule)) :
			$id = (int) $author_rule;

		// User name is filtered out
		elseif ('filter' == $author_rule) :
			$id = NULL;

		else :
			// Check the database for an existing author record that might fit

			// First try the user core data table.
			$id = $wpdb->get_var(
			"SELECT ID FROM $wpdb->users
			WHERE TRIM(LCASE(display_name)) = TRIM(LCASE('$author'))
			OR TRIM(LCASE(user_login)) = TRIM(LCASE('$author'))
			OR (
				LENGTH(TRIM(LCASE(user_email))) > 0
				AND TRIM(LCASE(user_email)) = TRIM(LCASE('$test_email'))
			)");

			// If that fails, look for aliases in the user meta data table
			if (is_null($id)) :
				$id = $wpdb->get_var(
				"SELECT user_id FROM $wpdb->usermeta
				WHERE
					(meta_key = 'description' AND TRIM(LCASE(meta_value)) = TRIM(LCASE('$author')))
					OR (
						meta_key = 'description'
						AND TRIM(LCASE(meta_value))
						RLIKE CONCAT(
							'(^|\\n)a\\.?k\\.?a\\.?( |\\t)*:?( |\\t)*',
							TRIM(LCASE('$reg_author')),
							'( |\\t|\\r)*(\\n|\$)'
						)
					)
				");
			endif;

			// ... if you don't find one, then do what you need to do
			if (is_null($id)) :
				if ($unfamiliar_author === 'create') :
					$userdata = array();

					#-- we need *something* for the email here or WordPress
					#-- is liable to pitch a fit. So, make something up if
					#-- necessary. (Ugh.)
					if (strlen($email) == 0 or FeedWordPress::is_null_email($email)) :
						$url = parse_url($hostUrl);
						$email = $nice_author.'@'.$url['host'];
					endif;

					#-- user table data
					$userdata['ID'] = NULL; // new user
					$userdata['user_login'] = $login;
					$userdata['user_nicename'] = $nice_author;
					$userdata['user_pass'] = substr(md5(uniqid(microtime())), 0, 6); // just something random to lock it up
					$userdata['user_email'] = $email;
					$userdata['user_url'] = $authorUrl;
					$userdata['nickname'] = $author;

					$parts = preg_split('/\s+/', trim($author), 2);
					if (isset($parts[0])) : $userdata['first_name'] = $parts[0]; endif;
					if (isset($parts[1])) : $userdata['last_name'] = $parts[1]; endif;

					$userdata['display_name'] = $author;
					$userdata['role'] = 'contributor';

					#-- loop. Keep trying to add the user until you get it
					#-- right. Or until PHP crashes, I guess.
					$insanity = 0;
					do {
						$id = wp_insert_user($userdata);
						if (is_wp_error($id)) :
							$codes = $id->get_error_code();
							switch ($codes) :
							case 'empty_user_login' :
							case 'existing_user_login' :
							case 'invalid_username' :
								// Add a random disambiguator
								$userdata['user_login'] .= substr(md5(uniqid(microtime())), 0, 6);
								break;
							case 'user_login_too_long' :
								// Limit length to 53 characters; if we end up needing a random disambiguator,
								// we should still have space to add it.
								$userdata['user_login'] = mb_substr( $userdata['user_login'], 0, 53 );
								break;
							case 'user_nicename_too_long' :
								// Add a limited 50 characters user_nicename based on user_login
                                				$userdata['user_nicename'] = mb_substr( $userdata['user_login'], 0, 50 );
								break;
							case 'existing_user_email' :
								// Disassemble email for username, host
								$parts = explode('@', $userdata['user_email'], 2);

								// Add a random disambiguator as a gmail-style username extension
								$parts[0] .= '+'.substr(md5(uniqid(microtime())), 0, 6);

								// Reassemble
								$userdata['user_email'] = $parts[0].'@'.$parts[1];
								break;
							case 'user_url_too_long' :
								// WordPress limits user_url to 100 characters. Long
								// URLs are common for pingback authors in comment
								// feeds. Fall back to the site's root URL, or none.
								$url = wp_parse_url($userdata['user_url']);
								$root = (isset($url['scheme'], $url['host']) ? $url['scheme'].'://'.$url['host'].'/' : '');
								$userdata['user_url'] = (mb_strlen($root) <= 100 ? $root : '');
								break;
							default :
								if ( $insanity > 10 ) :
									// Try some settings that are unlikely to cause complaint...
									$url = parse_url($hostUrl);

									$userdata['user_login'] = substr(md5(uniqid(microtime())), 0, 6);
									$userdata['user_nicename'] = $userdata['user_login'];
									$userdata['user_email'] = 'noreply@' . $url['host'];
									$userdata['user_url'] = '';
								endif;
							endswitch;
						endif;
						$insanity = $insanity + 1;

					// Stop doing the same thing and expecting a different result
					} while (is_wp_error($id) and $insanity <= 50);

					// $id should now contain the numeric ID of a newly minted
					// user account. Let's mark them as having been generated
					// by FeedWordPress in the usermeta table, as per the
					// suggestion of @boonebgorges, in case we need to process,
					// winnow, filter, or merge syndicated author accounts, &c.
					if ( !is_wp_error($id)) :
						add_user_meta($id, 'feedwordpress_generated', 1);
					else :
						$id = null;
					endif;
				// also see comment above regarding get_userdata(). (gwyneth 20220223)
				elseif ( is_numeric( $unfamiliar_author ) and function_exists( 'get_userdata ') and get_userdata((int) $unfamiliar_author ) ) :
					$id = (int) $unfamiliar_author;
				elseif ( $unfamiliar_author === 'default' ) :
					$id = 1;
				endif;
			endif;
		endif;

		if ($id) :
			$this->link->settings['map authors']['name'][strtolower(trim($author))] = $id;

			// Multisite: Check whether the author has been recorded
			// on *this* blog before. If not, put her down as a
			// Contributor for *this* blog.
			$user = new WP_User((int) $id);
			if (empty($user->roles)) :
				$user->add_role('contributor');
			endif;
		endif;
		return $id;
	} /* function SyndicatedPost::author_id () */

	/**
	 * category_ids: look up (and create) category ids from a list of
	 * categories
	 *
	 * @param array $cats
	 * @param string $unfamiliar_category
	 * @param array|null $taxonomies
	 * @return array
	 */
	function category_ids ($cats, $unfamiliar_category = 'create', $taxonomies = NULL, $params = array()) {
		return $this->link->category_ids($this, $cats, $unfamiliar_category, $taxonomies, $params);
	} /* SyndicatedPost::category_ids () */

} /* class SyndicatedPost */