aggregator.parser.inc

<?php

/**
 * @file
 * Parser functions for the aggregator module.
 */

/**
 * Implements hook_aggregator_parse_info().
 */
function aggregator_aggregator_parse_info() {
  return array(
    'title' => t('Default parser'),
    'description' => t('Parses RSS, Atom and RDF feeds.'),
  );
}

/**
 * Implements hook_aggregator_parse().
 */
function aggregator_aggregator_parse($feed) {
  global $channel, $image;

  // Filter the input data.
  if (aggregator_parse_feed($feed->source_string, $feed)) {
    $modified = empty($feed->http_headers['last-modified']) ? 0 : strtotime($feed->http_headers['last-modified']);

    // Prepare the channel data.
    foreach ($channel as $key => $value) {
      $channel[$key] = trim($value);
    }

    // Prepare the image data (if any).
    foreach ($image as $key => $value) {
      $image[$key] = trim($value);
    }
    $etag = empty($feed->http_headers['etag']) ? '' : $feed->http_headers['etag'];

    // Add parsed data to the feed object.
    $feed->link = !empty($channel['link']) ? $channel['link'] : '';
    $feed->description = !empty($channel['description']) ? $channel['description'] : '';
    $feed->image = !empty($image['url']) ? $image['url'] : '';
    $feed->etag = $etag;
    $feed->modified = $modified;

    // Clear the cache.
    cache_clear_all();
    return TRUE;
  }
  return FALSE;
}

/**
 * Parses a feed and stores its items.
 *
 * @param $data
 *   The feed data.
 * @param $feed
 *   An object describing the feed to be parsed.
 *
 * @return
 *   FALSE on error, TRUE otherwise.
 */
function aggregator_parse_feed(&$data, $feed) {
  global $items, $image, $channel;

  // Unset the global variables before we use them.
  unset($GLOBALS['element'], $GLOBALS['item'], $GLOBALS['tag']);
  $items = array();
  $image = array();
  $channel = array();

  // Parse the data.
  $xml_parser = drupal_xml_parser_create($data);
  xml_set_element_handler($xml_parser, 'aggregator_element_start', 'aggregator_element_end');
  xml_set_character_data_handler($xml_parser, 'aggregator_element_data');
  if (!xml_parse($xml_parser, $data, 1)) {
    watchdog('aggregator', 'The feed from %site seems to be broken, due to an error "%error" on line %line.', array(
      '%site' => $feed->title,
      '%error' => xml_error_string(xml_get_error_code($xml_parser)),
      '%line' => xml_get_current_line_number($xml_parser),
    ), WATCHDOG_WARNING);
    drupal_set_message(t('The feed from %site seems to be broken, because of error "%error" on line %line.', array(
      '%site' => $feed->title,
      '%error' => xml_error_string(xml_get_error_code($xml_parser)),
      '%line' => xml_get_current_line_number($xml_parser),
    )), 'error');
    return FALSE;
  }
  xml_parser_free($xml_parser);

  // We reverse the array such that we store the first item last, and the last
  // item first. In the database, the newest item should be at the top.
  $items = array_reverse($items);

  // Initialize items array.
  $feed->items = array();
  foreach ($items as $item) {

    // Prepare the item:
    foreach ($item as $key => $value) {
      $item[$key] = trim($value);
    }

    // Resolve the item's title. If no title is found, we use up to 40
    // characters of the description ending at a word boundary, but not
    // splitting potential entities.
    if (!empty($item['title'])) {
      $item['title'] = $item['title'];
    }
    elseif (!empty($item['description'])) {
      $item['title'] = preg_replace('/^(.*)[^\\w;&].*?$/', "\\1", truncate_utf8($item['description'], 40));
    }
    else {
      $item['title'] = '';
    }

    // Resolve the items link.
    if (!empty($item['link'])) {
      $item['link'] = $item['link'];
    }
    else {
      $item['link'] = $feed->link;
    }

    // Atom feeds have an ID tag instead of a GUID tag.
    if (!isset($item['guid'])) {
      $item['guid'] = isset($item['id']) ? $item['id'] : '';
    }

    // Atom feeds have a content and/or summary tag instead of a description tag.
    if (!empty($item['content:encoded'])) {
      $item['description'] = $item['content:encoded'];
    }
    elseif (!empty($item['summary'])) {
      $item['description'] = $item['summary'];
    }
    elseif (!empty($item['content'])) {
      $item['description'] = $item['content'];
    }

    // Try to resolve and parse the item's publication date.
    $date = '';
    foreach (array(
      'pubdate',
      'dc:date',
      'dcterms:issued',
      'dcterms:created',
      'dcterms:modified',
      'issued',
      'created',
      'modified',
      'published',
      'updated',
    ) as $key) {
      if (!empty($item[$key])) {
        $date = $item[$key];
        break;
      }
    }
    $item['timestamp'] = strtotime($date);
    if ($item['timestamp'] === FALSE) {
      $item['timestamp'] = aggregator_parse_w3cdtf($date);

      // Aggregator_parse_w3cdtf() returns FALSE on failure.
    }

    // Resolve dc:creator tag as the item author if author tag is not set.
    if (empty($item['author']) && !empty($item['dc:creator'])) {
      $item['author'] = $item['dc:creator'];
    }
    $item += array(
      'author' => '',
      'description' => '',
    );

    // Store on $feed object. This is where processors will look for parsed items.
    $feed->items[] = $item;
  }
  return TRUE;
}

/**
 * Performs an action when an opening tag is encountered.
 *
 * Callback function used by xml_parse() within aggregator_parse_feed().
 */
function aggregator_element_start($parser, $name, $attributes) {
  global $item, $element, $tag, $items, $channel;
  $name = strtolower($name);
  switch ($name) {
    case 'image':
    case 'textinput':
    case 'summary':
    case 'tagline':
    case 'subtitle':
    case 'logo':
    case 'info':
      $element = $name;
      break;
    case 'id':
    case 'content':
      if ($element != 'item') {
        $element = $name;
      }
    case 'link':

      // According to RFC 4287, link elements in Atom feeds without a 'rel'
      // attribute should be interpreted as though the relation type is
      // "alternate".
      if (!empty($attributes['HREF']) && (empty($attributes['REL']) || $attributes['REL'] == 'alternate')) {
        if ($element == 'item') {
          $items[$item]['link'] = $attributes['HREF'];
        }
        else {
          $channel['link'] = $attributes['HREF'];
        }
      }
      break;
    case 'item':
      $element = $name;
      $item += 1;
      break;
    case 'entry':
      $element = 'item';
      $item += 1;
      break;
  }
  $tag = $name;
}

/**
 * Performs an action when a closing tag is encountered.
 *
 * Callback function used by xml_parse() within aggregator_parse_feed().
 */
function aggregator_element_end($parser, $name) {
  global $element;
  switch ($name) {
    case 'image':
    case 'textinput':
    case 'item':
    case 'entry':
    case 'info':
      $element = '';
      break;
    case 'id':
    case 'content':
      if ($element == $name) {
        $element = '';
      }
  }
}

/**
 * Performs an action when data is encountered.
 *
 * Callback function used by xml_parse() within aggregator_parse_feed().
 */
function aggregator_element_data($parser, $data) {
  global $channel, $element, $items, $item, $image, $tag;
  $items += array(
    $item => array(),
  );
  switch ($element) {
    case 'item':
      $items[$item] += array(
        $tag => '',
      );
      $items[$item][$tag] .= $data;
      break;
    case 'image':
    case 'logo':
      $image += array(
        $tag => '',
      );
      $image[$tag] .= $data;
      break;
    case 'link':
      if ($data) {
        $items[$item] += array(
          $tag => '',
        );
        $items[$item][$tag] .= $data;
      }
      break;
    case 'content':
      $items[$item] += array(
        'content' => '',
      );
      $items[$item]['content'] .= $data;
      break;
    case 'summary':
      $items[$item] += array(
        'summary' => '',
      );
      $items[$item]['summary'] .= $data;
      break;
    case 'tagline':
    case 'subtitle':
      $channel += array(
        'description' => '',
      );
      $channel['description'] .= $data;
      break;
    case 'info':
    case 'id':
    case 'textinput':

      // The sub-element is not supported. However, we must recognize
      // it or its contents will end up in the item array.
      break;
    default:
      $channel += array(
        $tag => '',
      );
      $channel[$tag] .= $data;
  }
}

/**
 * Parses 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 $date_str
 *   A string with a potentially W3C DTF date.
 *
 * @return
 *   A timestamp if parsed successfully or FALSE if not.
 */
function aggregator_parse_w3cdtf($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;
  }
}