common_syndication_parser.inc

<?php

/**
 * @file
 *   Downloading and parsing functions for Common Syndication Parser.
 *   Pillaged from FeedAPI common syndication parser.
 *
 * @todo Restructure. OO could work wonders here.
 * @todo Write unit tests.
 * @todo Keep in Feeds project or host on Drupal?
 */

/**
 * Parse the feed into a data structure.
 *
 * @param string $string
 *   The feed object (contains the URL or the parsed XML structure).
 *
 * @return array|false
 *   The structured datas extracted from the feed or FALSE in case of failures.
 */
function common_syndication_parser_parse($string) {
  // SimpleXML can only deal with XML declaration at the start of the document,
  // so remove any surrounding whitespace.
  $string = trim($string);

  @ $xml = simplexml_load_string($string, NULL, LIBXML_NOERROR | LIBXML_NOWARNING | LIBXML_NOCDATA);

  // Got a malformed XML.
  if ($xml === FALSE || is_null($xml)) {
    return FALSE;
  }
  $feed_type = _parser_common_syndication_feed_format_detect($xml);
  if ($feed_type ==  "atom1.0") {
    return _parser_common_syndication_atom10_parse($xml);
  }
  if ($feed_type == "RSS2.0" || $feed_type == "RSS0.91" || $feed_type == "RSS0.92") {
    return _parser_common_syndication_RSS20_parse($xml);
  }
  if ($feed_type == "RDF") {
    return _parser_common_syndication_RDF10_parse($xml);
  }
  return FALSE;
}

/**
 * Determine the feed format of a SimpleXML parsed object structure.
 *
 * @param SimpleXMLElement $xml
 *   SimpleXML-preprocessed feed.
 *
 * @return string|false
 *   The feed format short description or FALSE if not compatible.
 */
function _parser_common_syndication_feed_format_detect($xml) {
  if (!is_object($xml)) {
    return FALSE;
  }
  $attr = $xml->attributes();
  $type = strtolower($xml->getName());
  if (isset($xml->entry) && $type == "feed") {
    return "atom1.0";
  }
  if ($type == "rss" && $attr["version"] == "2.0") {
    return "RSS2.0";
  }
  if ($type == "rdf" && isset($xml->channel)) {
    return "RDF";
  }
  if ($type == "rss" && $attr["version"] == "0.91") {
    return "RSS0.91";
  }
  if ($type == "rss" && $attr["version"] == "0.92") {
    return "RSS0.92";
  }
  return FALSE;
}

/**
 * Parse atom feeds.
 */
function _parser_common_syndication_atom10_parse($feed_XML) {
  $parsed_source = array();

  $ns = array(
    "georss" => "http://www.georss.org/georss",
  );

  $base = $feed_XML->xpath("@base");
  $base = (string) array_shift($base);
  if (!valid_url($base, TRUE)) {
    $base = FALSE;
  }

  // Detect the title
  $parsed_source['title'] = isset($feed_XML->title) ? _parser_common_syndication_title("{$feed_XML->title}") : "";
  // Detect the description
  $parsed_source['description'] = isset($feed_XML->subtitle) ? "{$feed_XML->subtitle}" : "";

  $parsed_source['link'] = _parser_common_syndication_link($feed_XML->link);
  if (valid_url($parsed_source['link']) && !valid_url($parsed_source['link'], TRUE) && !empty($base)) {
    $parsed_source['link'] = $base . $parsed_source['link'];
  }

  $parsed_source['items'] = array();

  foreach ($feed_XML->entry as $news) {

    $original_url = NULL;
    $guid = !empty($news->id) ? "{$news->id}" : NULL;
    if (valid_url($guid, TRUE)) {
      $original_url = $guid;
    }

    $georss = (array)$news->children($ns["georss"]);
    $geoname = '';
    if (isset($georss['featureName'])) {
      $geoname = "{$georss['featureName']}";
    }

    $latlon =
    $lat =
    $lon = NULL;
    if (isset($georss['point'])) {
      $latlon = explode(' ', $georss['point']);
      $lat = "{$latlon[0]}";
      $lon = "{$latlon[1]}";
      if (!$geoname) {
        $geoname = "{$lat} {$lon}";
      }
    }

    $additional_taxonomies = array();
    if (isset($news->category)) {
      $additional_taxonomies['ATOM Categories'] = array();
      $additional_taxonomies['ATOM Domains'] = array();
      foreach ($news->category as $category) {
        if (isset($category['scheme'])) {
          $domain = "{$category['scheme']}";
          if (!empty($domain)) {
              if (!isset($additional_taxonomies['ATOM Domains'][$domain])) {
                $additional_taxonomies['ATOM Domains'][$domain] = array();
              }
              $additional_taxonomies['ATOM Domains'][$domain][] = count($additional_taxonomies['ATOM Categories']) - 1;
          }
        }
        $additional_taxonomies['ATOM Categories'][] = "{$category['term']}";
      }
    }

    $title = "{$news->title}";

    $body = '';
    if (!empty($news->content)) {
      foreach ($news->content->children() as $child)  {
        $body .= $child->asXML();
      }
      $body .= "{$news->content}";
    }
    elseif (!empty($news->summary)) {
      foreach ($news->summary->children() as $child)  {
        $body .= $child->asXML();
      }
      $body .= "{$news->summary}";
    }

    if (!empty($news->content['src'])) {
      // some src elements in some valid atom feeds contained no urls at all
      if (valid_url("{$news->content['src']}", TRUE)) {
        $original_url = "{$news->content['src']}";
      }
    }

    $original_author = '';
    if (!empty($news->source->author->name)) {
      $original_author = "{$news->source->author->name}";
    }
    elseif (!empty($news->author->name)) {
      $original_author = "{$news->author->name}";
    }
    elseif (!empty($feed_XML->author->name)) {
      $original_author = "{$feed_XML->author->name}";
    }

    $original_url = _parser_common_syndication_link($news->link);

    $item = array();
    $item['title'] = _parser_common_syndication_title($title, $body);
    $item['description'] = $body;
    $item['author_name'] = $original_author;

    // Fall back to updated for timestamp if both published and issued are
    // empty.
    if (isset($news->published)) {
      $item['timestamp'] = _parser_common_syndication_parse_date("{$news->published}");
    }
    elseif (isset($news->issued)) {
       $item['timestamp'] = _parser_common_syndication_parse_date("{$news->issued}");
    }
    elseif (isset($news->updated)) {
      $item['timestamp'] = _parser_common_syndication_parse_date("{$news->updated}");
    }

    $item['url'] = trim($original_url);
    if (valid_url($item['url']) && !valid_url($item['url'], TRUE) && !empty($base)) {
      $item['url'] = $base . $item['url'];
    }
    // Fall back on URL if GUID is empty.
    if (!empty($guid)) {
      $item['guid'] = $guid;
    }
    else {
      $item['guid'] = $item['url'];
    }
    $item['geolocations'] = array();
    if ($lat && $lon) {
      $item['geolocations'] = array(
        array(
          'name' => $geoname,
          'lat' => $lat,
          'lon' => $lon,
        ),
      );
    }
    $item['tags'] = isset($additional_taxonomies['ATOM Categories']) ? $additional_taxonomies['ATOM Categories'] : array();
    $item['domains'] = isset($additional_taxonomies['ATOM Domains']) ? $additional_taxonomies['ATOM Domains'] : array();
    $parsed_source['items'][] = $item;
  }
  return $parsed_source;
}

/**
 * Parse RDF Site Summary (RSS) 1.0 feeds in RDF/XML format.
 *
 * @see http://web.resource.org/rss/1.0/
 */
function _parser_common_syndication_RDF10_parse($feed_XML) {
  // Declare some canonical standard prefixes for well-known namespaces:
  static $canonical_namespaces = array(
    'rdf'      => 'http://www.w3.org/1999/02/22-rdf-syntax-ns#',
    'rdfs'     => 'http://www.w3.org/2000/01/rdf-schema#',
    'xsi'      => 'http://www.w3.org/2001/XMLSchema-instance#',
    'xsd'      => 'http://www.w3.org/2001/XMLSchema#',
    'owl'      => 'http://www.w3.org/2002/07/owl#',
    'dc'       => 'http://purl.org/dc/elements/1.1/',
    'dcterms'  => 'http://purl.org/dc/terms/',
    'dcmitype' => 'http://purl.org/dc/dcmitype/',
    'foaf'     => 'http://xmlns.com/foaf/0.1/',
    'rss'      => 'http://purl.org/rss/1.0/',
  );

  // Get all namespaces declared in the feed element.
  $namespaces = $feed_XML->getNamespaces(TRUE);

  // Process the <rss:channel> resource containing feed metadata:
  foreach ($feed_XML->children($canonical_namespaces['rss'])->channel as $rss_channel) {
    $parsed_source = array(
      'title'       => _parser_common_syndication_title((string) $rss_channel->title),
      'description' => (string) $rss_channel->description,
      'link'        => (string) $rss_channel->link,
      'items'       => array(),
    );
    break;
  }

  // Process each <rss:item> resource contained in the feed:
  foreach ($feed_XML->children($canonical_namespaces['rss'])->item as $rss_item) {

    // Extract all available RDF statements from the feed item's RDF/XML
    // tags, allowing for both the item's attributes and child elements to
    // contain RDF properties:
    $rdf_data = array();
    foreach ($namespaces as $ns => $ns_uri) {
      // Note that we attempt to normalize the found property name
      // namespaces to well-known 'standard' prefixes where possible, as the
      // feed may in principle use any arbitrary prefixes and we should
      // still be able to correctly handle it.
      foreach ($rss_item->attributes($ns_uri) as $attr_name => $attr_value) {
        $ns_prefix = ($ns_prefix = array_search($ns_uri, $canonical_namespaces)) ? $ns_prefix : $ns;
        $rdf_data[$ns_prefix . ':' . $attr_name][] = (string) $attr_value;
      }
      foreach ($rss_item->children($ns_uri) as $rss_property) {
        $ns_prefix = ($ns_prefix = array_search($ns_uri, $canonical_namespaces)) ? $ns_prefix : $ns;
        $rdf_data[$ns_prefix . ':' . $rss_property->getName()][] = (string) $rss_property;
      }
    }

    // Declaratively define mappings that determine how to construct the result object.
    $item = _parser_common_syndication_RDF10_item($rdf_data, array(
      'title'       => array('rss:title', 'dc:title'),
      'description' => array('rss:description', 'dc:description', 'content:encoded'),
      'url'         => array('rss:link', 'rdf:about'),
      'author_name' => array('dc:creator', 'dc:publisher'),
      'guid'        => 'rdf:about',
      'timestamp'   => 'dc:date',
      'tags'        => 'dc:subject'
    ));

    // Special handling for the title:
    $item['title'] = _parser_common_syndication_title($item['title'], $item['description']);

    // Parse any date/time values into Unix timestamps:
    $item['timestamp'] = _parser_common_syndication_parse_date($item['timestamp']);

    // If no GUID found, use the URL of the feed.
    if (empty($item['guid'])) {
      $item['guid'] = $item['url'];
    }

    // Add every found RDF property to the feed item.
    $item['rdf'] = array();
    foreach ($rdf_data as $rdf_property => $rdf_value) {
      // looks nicer in the mapper UI
      // @todo Revisit, not used with feedapi mapper anymore.
      $rdf_property = str_replace(':', '_', $rdf_property);
      $item['rdf'][$rdf_property] = $rdf_value;
    }

    $parsed_source['items'][] = $item;
  }

  return $parsed_source;
}

function _parser_common_syndication_RDF10_property($rdf_data, $rdf_properties = array()) {
  $rdf_properties = is_array($rdf_properties) ? $rdf_properties : array_slice(func_get_args(), 1);
  foreach ($rdf_properties as $rdf_property) {
    if ($rdf_property && !empty($rdf_data[$rdf_property])) {
      // remove empty strings
      return array_filter($rdf_data[$rdf_property], 'strlen');
    }
  }
}

function _parser_common_syndication_RDF10_item($rdf_data, $mappings) {
  foreach ($mappings as $k => $v) {
    $values = _parser_common_syndication_RDF10_property($rdf_data, $v);
    $mappings[$k] = !is_array($values) || count($values) > 1 ? $values : reset($values);
  }
  return $mappings;
}

/**
 * Parse RSS2.0 feeds.
 */
function _parser_common_syndication_RSS20_parse($feed_XML) {

  $ns = array(
    "content" => "http://purl.org/rss/1.0/modules/content/",
     "dc" => "http://purl.org/dc/elements/1.1/",
     "georss" => "http://www.georss.org/georss",
  );

  $parsed_source = array();
  // Detect the title.
  $parsed_source['title'] = isset($feed_XML->channel->title) ? _parser_common_syndication_title("{$feed_XML->channel->title}") : "";
  // Detect the description.
  $parsed_source['description'] = isset($feed_XML->channel->description) ? "{$feed_XML->channel->description}" : "";
  // Detect the link.
  $parsed_source['link'] = isset($feed_XML->channel->link) ? "{$feed_XML->channel->link}" : "";
  $parsed_source['items'] = array();

  foreach ($feed_XML->xpath('//item') as $news) {
    $title = $body = $original_author = $original_url = $guid = '';

    // Get optional source url.
    $source_url = (string) $news->source['url'];

    $category = $news->xpath('category');
    // Get children for current namespace.
    $content = (array)$news->children($ns["content"]);
    $dc      = (array)$news->children($ns["dc"]);
    $georss  = (array)$news->children($ns["georss"]);
    $news = (array) $news;
    $news['category'] = $category;

    if (isset($news['title'])) {
      $title = "{$news['title']}";
    }

    if (isset($news['description'])) {
      $body = "{$news['description']}";
    }
    // Some sources use content:encoded as description i.e.
    // PostNuke PageSetter module.
    if (isset($news['encoded'])) {  // content:encoded for PHP < 5.1.2.
      if (strlen($body) < strlen("{$news['encoded']}")) {
        $body = "{$news['encoded']}";
      }
    }
    if (isset($content['encoded'])) { // content:encoded for PHP >= 5.1.2.
      if (strlen($body) < strlen("{$content['encoded']}")) {
        $body = "{$content['encoded']}";
      }
    }
    if (!isset($body)) {
      $body = "{$news['title']}";
    }

    if (!empty($news['author'])) {
      $original_author = "{$news['author']}";
    }
    elseif (!empty($dc["creator"])) {
      $original_author = (string)$dc["creator"];
    }

    if (!empty($news['link'])) {
      $original_url = "{$news['link']}";
      $guid = $original_url;
    }

    if (!empty($news['guid'])) {
      $guid = "{$news['guid']}";
    }

    if (!empty($georss['featureName'])) {
      $geoname = "{$georss['featureName']}";
    }

    $lat =
    $lon =
    $latlon =
    $geoname = NULL;
    if (!empty($georss['point'])) {
      $latlon = explode(' ', $georss['point']);
      $lat = "{$latlon[0]}";
      $lon = "{$latlon[1]}";
      if (!$geoname) {
        $geoname = "$lat $lon";
      }
    }

    $additional_taxonomies = array();
    $additional_taxonomies['RSS Categories'] = array();
    $additional_taxonomies['RSS Domains'] = array();
    if (isset($news['category'])) {
      foreach ($news['category'] as $category) {
        $additional_taxonomies['RSS Categories'][] = "{$category}";
        if (isset($category['domain'])) {
          $domain = "{$category['domain']}";
          if (!empty($domain)) {
              if (!isset($additional_taxonomies['RSS Domains'][$domain])) {
                $additional_taxonomies['RSS Domains'][$domain] = array();
              }
              $additional_taxonomies['RSS Domains'][$domain][] = count($additional_taxonomies['RSS Categories']) - 1;
          }
        }
      }
    }

    $item = array();
    $item['title'] = _parser_common_syndication_title($title, $body);
    $item['description'] = $body;
    $item['author_name'] = $original_author;
    if (!empty($news['pubDate'])) {
      $item['timestamp'] = _parser_common_syndication_parse_date($news['pubDate']);
    }
    elseif (!empty($dc['date'])) {
      $item['timestamp'] = _parser_common_syndication_parse_date($dc['date']);
    }
    else {
      $item['timestamp'] = time();
    }
    $item['url'] = trim($original_url);
    $item['guid'] = $guid;
    if (!empty($news['source'])) {
      $item['source:title'] = $news['source'];
    }
    else {
      $item['source:title'] = NULL;
    }
    $item['source:url'] = trim($source_url);

    $item['geolocations'] = array();
    if (isset($geoname, $lat, $lon)) {
      $item['geolocations'] = array(
        array(
          'name' => $geoname,
          'lat' => $lat,
          'lon' => $lon,
        ),
      );
    }

    $item['domains'] = $additional_taxonomies['RSS Domains'];
    $item['tags'] = $additional_taxonomies['RSS Categories'];
    $parsed_source['items'][] = $item;
  }
  return $parsed_source;
}

/**
 * Parse a date comes from a feed.
 *
 * @param string $date_str
 *   The date string in various formats.
 *
 * @return int
 *   The timestamp of the string or the current time if can't be parsed.
 */
function _parser_common_syndication_parse_date($date_str) {
  // PHP < 5.3 doesn't like the GMT- notation for parsing timezones.
  $date_str = str_replace('GMT-', '-', $date_str);
  $date_str = str_replace('GMT+', '+', $date_str);
  $parsed_date = strtotime($date_str);

  if ($parsed_date === FALSE || $parsed_date == -1) {
    $parsed_date = _parser_common_syndication_parse_w3cdtf($date_str);
  }

  if (($parsed_date === FALSE || $parsed_date == -1)) {
    // PHP does not support the UT timezone. Fake it. The system that generated
    // this, Google Groups, probably meant UTC.
    $date_str = strtolower(trim($date_str));
    $last_three = substr($date_str, strlen($date_str) - 3, 3);

    if ($last_three == ' ut') {
      $parsed_date = strtotime($date_str . 'c');
    }
  }

  return $parsed_date === FALSE ? time() : $parsed_date;
}

/**
 * Parse the W3C date/time format, a subset of ISO 8601.
 *
 * PHP date parsing functions do not handle this format.
 * See http://www.w3.org/TR/NOTE-datetime for more information.
 * Originally from MagpieRSS (http://magpierss.sourceforge.net/).
 *
 * @param string $date_str
 *   A potentially W3C DTF date.
 *
 * @return int|false
 *   A timestamp if parsed successfully or FALSE if not.
 */
function _parser_common_syndication_parse_w3cdtf($date_str) {
  if (preg_match('/(\d{4})-(\d{2})-(\d{2})T(\d{2}):(\d{2})(:(\d{2}))?(?:([-+])(\d{2}):?(\d{2})|(Z))?/', $date_str, $match)) {
    list($year, $month, $day, $hours, $minutes, $seconds) = array($match[1], $match[2], $match[3], $match[4], $match[5], $match[6]);
    // Calculate the epoch for current date assuming GMT.
    $epoch = gmmktime($hours, $minutes, $seconds, $month, $day, $year);
    if ($match[10] != 'Z') { // Z is zulu time, aka GMT
      list($tz_mod, $tz_hour, $tz_min) = array($match[8], $match[9], $match[10]);
      // Zero out the variables.
      if (!$tz_hour) {
        $tz_hour = 0;
      }
      if (!$tz_min) {
        $tz_min = 0;
      }
      $offset_secs = (($tz_hour * 60) + $tz_min) * 60;
      // Is timezone ahead of GMT?  If yes, subtract offset.
      if ($tz_mod == '+') {
        $offset_secs *= -1;
      }
      $epoch += $offset_secs;
    }
    return $epoch;
  }
  else {
    return FALSE;
  }
}

/**
 * Extract the link that points to the original content (back to site or
 * original article)
 *
 * @param array $links
 *   Array of SimpleXML objects
 *
 * @return string
 *   An URL if found. An empty string otherwise.
 */
function _parser_common_syndication_link($links) {
  $to_link = '';
  if (count($links) > 0) {
    foreach ($links as $link) {
      $link = $link->attributes();
      $to_link = isset($link["href"]) ? "{$link["href"]}" : "";
      if (isset($link["rel"])) {
        if ("{$link["rel"]}" == 'alternate') {
          break;
        }
      }
    }
  }
  return $to_link;
}

/**
 * Prepare raw data to be a title
 */
function _parser_common_syndication_title($title, $body = FALSE) {
  if (empty($title) && !empty($body)) {
    // Explode to words and use the first 3 words.
    $words = preg_split('/[\s,]+/', strip_tags($body));
    $title = implode(' ', array_slice($words, 0, 3));
  }
  return $title;
}