/**
* Extract embedded elements from a HTML string.
*
* Returns an array of IFrame elements found in the input string.
* Elements without a 'src' attribute are skipped.
*
* Each array item has the same basic structure as the array items
* returned by blcUtility::extract_tags(), plus an additional 'embed_code' key
* that contains the full HTML code for the entire <ifram> tag.
*
* @uses blcUtility::extract_tags() This function is a simple wrapper around extract_tags()
*
* @param string $html
* @return array
*/
function extract_embeds($html)
{
$results = array();
//remove all <code></code> blocks first
$html = preg_replace('/<code[^>]*>.+?<\\/code>/si', ' ', $html);
//Find likely-looking <object> elements
$iframes = blcUtility::extract_tags($html, 'iframe', false, true);
foreach ($iframes as $embed) {
if (empty($embed['attributes']['src'])) {
continue;
}
$embed['embed_code'] = $embed['full_tag'];
$results[] = $embed;
}
return $results;
}
/**
* Extract embedded elements from a HTML string.
*
* This function returns an array of <embed> elements found in the input
* string. Embeds without a 'src' attribute are skipped.
*
* Each array item has the same basic structure as the array items
* returned by blcUtility::extract_tags(), plus an additional 'embed_code' key
* that contains the HTML code for the element. If the embed element is wrapped
* in an <object>, the 'embed_code' key contains the full HTML for the entire
* <object> + <embed> structure.
*
* @uses blcUtility::extract_tags() This function is a simple wrapper around extract_tags()
*
* @param string $html
* @return array
*/
function extract_embeds($html)
{
$results = array();
//remove all <code></code> blocks first
$html = preg_replace('/<code[^>]*>.+?<\\/code>/si', ' ', $html);
//Find likely-looking <object> elements
$objects = blcUtility::extract_tags($html, 'object', false, true);
foreach ($objects as $candidate) {
//Find the <embed> tag
$embed = blcUtility::extract_tags($candidate['full_tag'], 'embed', true);
if (empty($embed)) {
continue;
}
$embed = reset($embed);
//Take the first (and only) found <embed> element
if (empty($embed['attributes']['src'])) {
continue;
}
$embed['embed_code'] = $candidate['full_tag'];
$results[] = $embed;
//Remove the element so it doesn't come up when we search for plain <embed> elements.
$html = str_replace($candidate['full_tag'], ' ', $html);
}
//Find <embed> elements not wrapped in an <object> element.
$embeds = blcUtility::extract_tags($html, 'embed', false, true);
foreach ($embeds as $embed) {
if (!empty($embed['attributes']['src'])) {
$embed['embed_code'] = $embed['full_tag'];
$results[] = $embed;
}
}
return $results;
}
function check($url)
{
$result = array('final_url' => $url, 'redirect_count' => 0, 'timeout' => false, 'broken' => false, 'log' => "<em>(Using YouTube API)</em>\n\n", 'result_hash' => '');
//Extract the video ID from the URL
$components = @parse_url($url);
parse_str($components['query'], $query);
$video_id = $query['v'];
//Fetch video data from the YouTube API
$api_url = 'http://gdata.youtube.com/feeds/api/videos/' . $video_id;
$conf =& blc_get_configuration();
$args = array('timeout' => $conf->options['timeout']);
$start = microtime_float();
$response = wp_remote_get($api_url, $args);
$result['request_duration'] = microtime_float() - $start;
//Placeholders for video restriction data
$state_name = $state_reason = '';
//Got anything?
if (is_wp_error($response)) {
$result['log'] .= "Error.\n" . $response->get_error_message();
//WP doesn't make it easy to distinguish between different internal errors.
$result['broken'] = true;
$result['http_code'] = 0;
} else {
$result['http_code'] = intval($response['response']['code']);
switch ($result['http_code']) {
case 404:
//Not found
$result['log'] .= __('Video Not Found', 'broken-link-checker');
$result['broken'] = true;
$result['http_code'] = 0;
$result['status_text'] = __('Video Not Found', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_ERROR;
break;
case 403:
//Forbidden. Usually means that the video has been removed. Body contains details.
$result['log'] .= $response['body'];
$result['broken'] = true;
$result['http_code'] = 0;
$result['status_text'] = __('Video Removed', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_ERROR;
break;
case 400:
//Bad request. Usually means that the video ID is incorrect. Body contains details.
$result['log'] .= $response['body'];
$result['broken'] = true;
$result['http_code'] = 0;
$result['status_text'] = __('Invalid Video ID', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_WARNING;
break;
case 200:
//Video exists, but may be restricted. Check for <yt:state> tags.
//See http://code.google.com/apis/youtube/2.0/reference.html#youtube_data_api_tag_yt:state
//Can we count on an XML parser being installed? No, probably not.
//Back to our makeshift tag "parser" we go.
$state = blcUtility::extract_tags($response['body'], 'yt:state', false);
if (empty($state)) {
//Phew, no restrictions.
$result['log'] .= __("Video OK", 'broken-link-checker');
$result['status_text'] = __('OK', 'link status', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_OK;
$result['http_code'] = 0;
} else {
//Get the state name and code and append them to the log
$state = reset($state);
$state_name = $state['attributes']['name'];
$state_reason = isset($state['attributes']['reasonCode']) ? $state['attributes']['reasonCode'] : '';
$result['result_hash'] = 'youtube_api|' . $state_name . '|' . $state_reason;
$result['log'] .= sprintf(__('Video status : %s%s', 'broken-link-checker'), $state_name, $state_reason ? ' [' . $state_reason . ']' : '');
//A couple of restricted states are not that bad
$state_ok = $state_name == 'processing' || $state_name == 'restricted' && $state_reason == 'limitedSyndication';
if ($state_ok) {
$result['broken'] = false;
$result['status_text'] = __('OK', 'link status', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_OK;
$result['http_code'] = 0;
} else {
$result['broken'] = true;
$result['status_text'] = __('Video Restricted', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['http_code'] = 0;
}
}
//Add the video title to the log, purely for information.
//http://code.google.com/apis/youtube/2.0/reference.html#youtube_data_api_tag_media:title
$title = blcUtility::extract_tags($response['body'], 'media:title', false);
if (!empty($title)) {
$result['log'] .= "\n\nTitle : \"" . $title[0]['contents'] . '"';
}
break;
default:
$result['log'] .= $result['http_code'] . $response['response']['message'];
$result['log'] .= "\n" . __('Unknown YouTube API response received.');
break;
}
}
//The hash should contain info about all pieces of data that pertain to determining if the
//link is working.
$result['result_hash'] = implode('|', array('youtube', $result['http_code'], $result['broken'] ? 'broken' : '0', $result['timeout'] ? 'timeout' : '0', $state_name, $state_reason));
return $result;
}
/**
* Extract <embed> elements from a HTML string.
*
* This function returns an array of <embed> elements found in the input
* string. Only <embed>'s that are inside <object>'s are considered. Embeds
* without a 'src' attribute are skipped.
*
* Each array item has the same basic structure as the array items
* returned by blcUtility::extract_tags(), plus an additional 'wrapper' key
* that contains similarly structured info about the wrapping <object> tag.
*
* @uses blcUtility::extract_tags() This function is a simple wrapper around extract_tags()
*
* @param string $html
* @return array
*/
function extract_embeds($html)
{
$results = array();
//remove all <code></code> blocks first
$content = preg_replace('/<code[^>]*>.+?<\\/code>/si', ' ', $content);
//Find likely-looking <object> elements
$objects = blcUtility::extract_tags($html, 'object', false, true);
foreach ($objects as $candidate) {
//Find the <embed> tag
$embed = blcUtility::extract_tags($candidate['full_tag'], 'embed', false);
if (empty($embed)) {
continue;
}
$embed = reset($embed);
//Take the first (and only) found <embed> element
if (empty($embed['attributes']['src'])) {
continue;
}
$embed['wrapper'] = $candidate;
$results[] = $embed;
}
return $results;
}
/**
* Check a YouTube API response that contains a single playlist.
*
* @param array $response
* @param array $result
* @return array
*/
protected function check_playlist($response, $result)
{
switch ($result['http_code']) {
case 404:
//Not found
$result['log'] .= __('Playlist Not Found', 'broken-link-checker');
$result['broken'] = true;
$result['http_code'] = 0;
$result['status_text'] = __('Playlist Not Found', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_ERROR;
break;
case 403:
//Forbidden. We're unlikely to see this code for playlists, but lets allow it.
$result['log'] .= $response['body'];
$result['broken'] = true;
$result['status_text'] = __('Playlist Restricted', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_ERROR;
break;
case 400:
//Bad request. Probably indicates a client error (invalid API request). Body contains details.
$result['log'] .= $response['body'];
$result['broken'] = true;
$result['status_text'] = __('Invalid Playlist', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_WARNING;
break;
case 200:
//The playlist exists, but some of the videos may be restricted.
//Check for <yt:state> tags.
$video_states = blcUtility::extract_tags($response['body'], 'yt:state', false);
if (empty($video_states)) {
//No restrictions. Does the playlist have any entries?
$entries = blcUtility::extract_tags($response['body'], 'entry', false);
if (!empty($entries)) {
//All is well.
$result['log'] .= __("Playlist OK", 'broken-link-checker');
$result['status_text'] = __('OK', 'link status', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_OK;
$result['http_code'] = 0;
} else {
//An empty playlist. It is possible that all of the videos
//have been deleted. Treat it as a warning.
$result['log'] .= __("This playlist has no entries or all entries have been deleted.", 'broken-link-checker');
$result['status_text'] = __('Empty Playlist', 'link status', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['http_code'] = 0;
$result['broken'] = true;
}
} else {
//Treat the playlist as broken if at least one video is inaccessible.
foreach ($video_states as $state) {
$state_name = $state['attributes']['name'];
$state_reason = isset($state['attributes']['reasonCode']) ? $state['attributes']['reasonCode'] : '';
if (!$this->is_state_ok($state_name, $state_reason)) {
$result['log'] .= sprintf(__('Video status : %s%s', 'broken-link-checker'), $state_name, $state_reason ? ' [' . $state_reason . ']' : '');
$result['state_name'] = $state_name;
$result['state_reason'] = $state_reason;
$result['broken'] = true;
$result['status_text'] = __('Video Restricted', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['http_code'] = 0;
break;
}
}
if (!$result['broken']) {
$result['status_text'] = __('OK', 'link status', 'broken-link-checker');
$result['status_code'] = BLC_LINK_STATUS_OK;
$result['http_code'] = 0;
}
}
//Add the playlist title to the log, purely for information.
$title = blcUtility::extract_tags($response['body'], 'title', false);
if (!empty($title)) {
$result['log'] .= "\n\nPlaylist title : \"" . $title[0]['contents'] . '"';
}
break;
default:
$result['log'] .= $result['http_code'] . $response['response']['message'];
$result['log'] .= "\n" . __('Unknown YouTube API response received.');
break;
}
return $result;
}
/**
* Apply a callback function to all HTML links found in a string and return the results.
*
* The link data array will contain at least these keys :
* 'href' - the URL of the link (with htmlentitydecode() already applied).
* '#raw' - the raw link code, e.g. the entire '<a href="...">...</a>' tag of a HTML link.
* '#offset' - the offset within $content at which the first character of the link tag was found.
* '#link_text' - the link's anchor text, if any. May contain HTML tags.
*
* Any attributes of the link tag will also be included in the returned array as attr_name => attr_value
* pairs. This function will also automatically decode any HTML entities found in attribute values.
*
* @see blcParser::map()
*
* @param string $content A text string to parse for links.
* @param callback $callback Callback function to apply to all found links.
* @param mixed $extra If the optional $extra param. is supplied, it will be passed as the second parameter to the function $callback.
* @return array An array of all detected links after applying $callback to each of them.
*/
function map($content, $callback, $extra = null)
{
$results = array();
//Find all links
$links = blcUtility::extract_tags($content, 'a', false, true);
//Iterate over the links and apply $callback to each
foreach ($links as $link) {
//Massage the found link into a form required for the callback function
$param = $link['attributes'];
$param = array_merge($param, array('#raw' => $link['full_tag'], '#offset' => $link['offset'], '#link_text' => $link['contents'], 'href' => isset($link['attributes']['href']) ? $link['attributes']['href'] : ''));
//Prepare arguments for the callback
$params = array($param);
if (isset($extra)) {
$params[] = $extra;
}
//Execute & store :)
$results[] = call_user_func_array($callback, $params);
}
return $results;
}
