/**
* blcUtility::normalize_url()
*
* @param string $url
* @params string $base_url (Optional) The base URL is used to convert a relative URL to a fully-qualified one
* @return string A normalized URL or FALSE if the URL is invalid
*/
function normalize_url($url, $base_url = '')
{
//Sometimes links may contain shortcodes. Parse them.
$url = do_shortcode($url);
$parts = @parse_url($url);
if (!$parts) {
return false;
}
//Invalid URL
if (isset($parts['scheme'])) {
//Only HTTP(S) links are checked. Other protocols are not supported.
if ($parts['scheme'] != 'http' && $parts['scheme'] != 'https') {
return false;
}
}
$url = html_entity_decode($url);
$url = preg_replace(array('/([\\?&]PHPSESSID=\\w+)$/i', '/(#[^\\/]*)$/', '/&/', '/^(javascript:.*)/i', '/([\\?&]sid=\\w+)$/i'), array('', '', '&', '', ''), $url);
$url = trim($url);
if ($url == '') {
return false;
}
// turn relative URLs into absolute URLs
if (empty($base_url)) {
$base_url = get_option('siteurl');
}
$url = blcUtility::relative2absolute($base_url, $url);
return $url;
}
/**
* 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;
}
$moduleManager->plugin_activated();
blc_got_unsynched_items();
$blclog->info(sprintf('--- Total: %.3f seconds', microtime(true) - $notification_start));
//Turn off load limiting if it's not available on this server.
$blclog->info('Updating server load limit settings...');
$load = blcUtility::get_server_load();
if (empty($load)) {
$blc_config_manager->options['enable_load_limit'] = false;
$blclog->info('Disable load limit. Cannot retrieve current load average.');
} elseif ($blc_config_manager->options['enable_load_limit'] && !isset($blc_config_manager->options['server_load_limit'])) {
$fifteen_minutes = floatval(end($load));
$default_load_limit = round(max(min($fifteen_minutes * 2, $fifteen_minutes + 2), 4));
$blc_config_manager->options['server_load_limit'] = $default_load_limit;
$blclog->info(sprintf('Set server load limit to %.2f. Current load average is %.2f', $default_load_limit, $fifteen_minutes));
}
//And optimize my DB tables, too (for good measure)
$blclog->info('Optimizing the database...');
$optimize_start = microtime(true);
blcUtility::optimize_database();
$blclog->info(sprintf('--- Total: %.3f seconds', microtime(true) - $optimize_start));
$blclog->info('Completing installation...');
$blc_config_manager->options['installation_complete'] = true;
$blc_config_manager->options['installation_flag_set_on'] = date('c') . ' (' . microtime(true) . ')';
if ($blc_config_manager->save_options()) {
$blclog->info('Configuration saved.');
} else {
$blclog->error('Error saving plugin configuration!');
}
$blclog->info(sprintf('Installation/update completed at %s with %d queries executed.', date_i18n('Y-m-d H:i:s'), $wpdb->num_queries - $queryCnt));
$blclog->info(sprintf('Total time: %.3f seconds', microtime(true) - $activation_start));
$blclog->save();
function build_instance_list_for_email($instances, $max_displayed_links = 5)
{
$result = '';
if (count($instances) > $max_displayed_links) {
$line = sprintf(_n("Here's a list of the first %d broken links:", "Here's a list of the first %d broken links:", $max_displayed_links, 'broken-link-checker'), $max_displayed_links);
} else {
$line = __("Here's a list of the new broken links: ", 'broken-link-checker');
}
$result .= "<p>{$line}</p>";
//Show up to $max_displayed_links broken link instances right in the email.
$displayed = 0;
foreach ($instances as $instance) {
/* @var blcLinkInstance $instance */
$pieces = array(sprintf(__('Link text : %s', 'broken-link-checker'), $instance->ui_get_link_text('email')), sprintf(__('Link URL : <a href="%s">%s</a>', 'broken-link-checker'), htmlentities($instance->get_url()), blcUtility::truncate($instance->get_url(), 70, '')), sprintf(__('Source : %s', 'broken-link-checker'), $instance->ui_get_source('email')));
$link_entry = implode("<br>", $pieces);
$result .= "{$link_entry}<br><br>";
$displayed++;
if ($displayed >= $max_displayed_links) {
break;
}
}
//Add a link to the "Broken Links" tab.
$result .= __("You can see all broken links here:", 'broken-link-checker') . "<br>";
$result .= sprintf('<a href="%1$s">%1$s</a>', admin_url('tools.php?page=view-broken-links'));
return $result;
}
/**
* Get the link URL in ASCII-compatible encoding.
*
* @return string
*/
function get_ascii_url()
{
return blcUtility::idn_to_ascii($this->url);
}
function ui_get_source($container_field = '', $context = 'display')
{
//Display a comment icon.
if ($container_field == 'comment_author_url') {
$image = 'font-awesome/font-awesome-user.png';
} else {
$image = 'font-awesome/font-awesome-comment-alt.png';
}
$image = sprintf('<img src="%s/broken-link-checker/images/%s" class="blc-small-image" title="%3$s" alt="%3$s"> ', WP_PLUGIN_URL, $image, __('Comment', 'broken-link-checker'));
$comment = $this->get_wrapped_object();
//Display a small text sample from the comment
$text_sample = strip_tags($comment->comment_content);
$text_sample = blcUtility::truncate($text_sample, 65);
$html = sprintf('<a href="%s" title="%s"><b>%s</b> — %s</a>', $this->get_edit_url(), esc_attr__('Edit comment'), esc_attr($comment->comment_author), $text_sample);
//Don't show the image in email notifications.
if ($context != 'email') {
$html = $image . $html;
}
return $html;
}
function check($url, $use_get = false)
{
$this->last_headers = '';
$url = $this->clean_url($url);
$result = array('broken' => false);
$log = '';
//Get the BLC configuration. It's used below to set the right timeout values and such.
$conf = blc_get_configuration();
//Init curl.
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $this->urlencodefix($url));
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
//Masquerade as Internet explorer
//$ua = 'Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)';
$ua = 'Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.1.4322; .NET CLR 2.0.50727; .NET CLR 3.0.04506.30)';
curl_setopt($ch, CURLOPT_USERAGENT, $ua);
//Add a semi-plausible referer header to avoid tripping up some bot traps
curl_setopt($ch, CURLOPT_REFERER, home_url());
//Redirects don't work when safe mode or open_basedir is enabled.
if (!blcUtility::is_safe_mode() && !blcUtility::is_open_basedir()) {
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
}
//Set maximum redirects
curl_setopt($ch, CURLOPT_MAXREDIRS, 10);
//Set the timeout
curl_setopt($ch, CURLOPT_TIMEOUT, $conf->options['timeout']);
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $conf->options['timeout']);
//Set the proxy configuration. The user can provide this in wp-config.php
if (defined('WP_PROXY_HOST')) {
curl_setopt($ch, CURLOPT_PROXY, WP_PROXY_HOST);
}
if (defined('WP_PROXY_PORT')) {
curl_setopt($ch, CURLOPT_PROXYPORT, WP_PROXY_PORT);
}
if (defined('WP_PROXY_USERNAME')) {
$auth = WP_PROXY_USERNAME;
if (defined('WP_PROXY_PASSWORD')) {
$auth .= ':' . WP_PROXY_PASSWORD;
}
curl_setopt($ch, CURLOPT_PROXYUSERPWD, $auth);
}
//Make CURL return a valid result even if it gets a 404 or other error.
curl_setopt($ch, CURLOPT_FAILONERROR, false);
$nobody = !$use_get;
//Whether to send a HEAD request (the default) or a GET request
$parts = @parse_url($url);
if ($parts['scheme'] == 'https') {
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
//Required to make HTTPS URLs work.
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, false);
$nobody = false;
//Can't use HEAD with HTTPS.
}
if ($nobody) {
//If possible, use HEAD requests for speed.
curl_setopt($ch, CURLOPT_NOBODY, true);
} else {
//If we must use GET at least limit the amount of downloaded data.
curl_setopt($ch, CURLOPT_HTTPHEADER, array('Range: bytes=0-2048'));
//2 KB
}
//Register a callback function which will process the HTTP header(s).
//It can be called multiple times if the remote server performs a redirect.
curl_setopt($ch, CURLOPT_HEADERFUNCTION, array(&$this, 'read_header'));
//Execute the request
$start_time = microtime_float();
curl_exec($ch);
$measured_request_duration = microtime_float() - $start_time;
$info = curl_getinfo($ch);
//Store the results
$result['http_code'] = intval($info['http_code']);
$result['final_url'] = $info['url'];
$result['request_duration'] = $info['total_time'];
$result['redirect_count'] = $info['redirect_count'];
//CURL doesn't return a request duration when a timeout happens, so we measure it ourselves.
//It is useful to see how long the plugin waited for the server to respond before assuming it timed out.
if (empty($result['request_duration'])) {
$result['request_duration'] = $measured_request_duration;
}
//Determine if the link counts as "broken"
if ($result['http_code'] == 0) {
$result['broken'] = true;
$error_code = curl_errno($ch);
$log .= sprintf("%s [Error #%d]\n", curl_error($ch), $error_code);
//We only handle a couple of CURL error codes; most are highly esoteric.
//libcurl "CURLE_" constants can't be used here because some of them have
//different names or values in PHP.
switch ($error_code) {
case 6:
//CURLE_COULDNT_RESOLVE_HOST
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['status_text'] = __('Server Not Found', 'broken-link-checker');
break;
case 28:
//CURLE_OPERATION_TIMEDOUT
$result['timeout'] = true;
break;
case 7:
//CURLE_COULDNT_CONNECT
//More often than not, this error code indicates that the connection attempt
//timed out. This heuristic tries to distinguish between connections that fail
//due to timeouts and those that fail due to other causes.
if ($result['request_duration'] >= 0.9 * $conf->options['timeout']) {
$result['timeout'] = true;
} else {
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['status_text'] = __('Connection Failed', 'broken-link-checker');
}
break;
default:
$result['status_code'] = BLC_LINK_STATUS_WARNING;
$result['status_text'] = __('Unknown Error', 'broken-link-checker');
}
} else {
$result['broken'] = $this->is_error_code($result['http_code']);
}
curl_close($ch);
if ($nobody && $result['broken']) {
//The site in question might be expecting GET instead of HEAD, so lets retry the request
//using the GET verb.
return $this->check($url, true);
//Note : normally a server that doesn't allow HEAD requests on a specific resource *should*
//return "405 Method Not Allowed". Unfortunately, there are sites that return 404 or
//another, even more general, error code instead. So just checking for 405 wouldn't be enough.
}
//When safe_mode or open_basedir is enabled CURL will be forbidden from following redirects,
//so redirect_count will be 0 for all URLs. As a workaround, set it to 1 when the HTTP
//response codes indicates a redirect but redirect_count is zero.
//Note to self : Extracting the Location header might also be helpful.
if ($result['redirect_count'] == 0 && in_array($result['http_code'], array(301, 302, 303, 307))) {
$result['redirect_count'] = 1;
}
//Build the log from HTTP code and headers.
$log .= '=== ';
if ($result['http_code']) {
$log .= sprintf(__('HTTP code : %d', 'broken-link-checker'), $result['http_code']);
} else {
$log .= __('(No response)', 'broken-link-checker');
}
$log .= " ===\n\n";
$log .= $this->last_headers;
if (!empty($result['broken']) && !empty($result['timeout'])) {
$log .= "\n(" . __("Most likely the connection timed out or the domain doesn't exist.", 'broken-link-checker') . ')';
}
$result['log'] = $log;
//The hash should contain info about all pieces of data that pertain to determining if the
//link is working.
$result['result_hash'] = implode('|', array($result['http_code'], !empty($result['broken']) ? 'broken' : '0', !empty($result['timeout']) ? 'timeout' : '0', md5($result['final_url'])));
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;
}
/**
* Convert an internationalized domain name (or URL) from ASCII-compatible encoding to UTF8.
*
* @param string $url
* @return string
*/
static function idn_to_utf8($url)
{
$idn = blcUtility::get_idna_converter();
if ($idn != null) {
$url = $idn->decode($url);
}
return $url;
}
function check($url, $use_get = false)
{
$this->last_headers = '';
$url = $this->clean_url($url);
$result = array('broken' => false);
$log = '';
//Get the BLC configuration. It's used below to set the right timeout values and such.
$conf = blc_get_configuration();
//Init curl.
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $this->urlencodefix($url));
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
//Masquerade as Internet explorer
//$ua = 'Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)';
$ua = 'Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.1.4322; .NET CLR 2.0.50727; .NET CLR 3.0.04506.30)';
curl_setopt($ch, CURLOPT_USERAGENT, $ua);
//Add a semi-plausible referer header to avoid tripping up some bot traps
curl_setopt($ch, CURLOPT_REFERER, get_option('home'));
//Redirects don't work when safe mode or open_basedir is enabled.
if (!blcUtility::is_safe_mode() && !blcUtility::is_open_basedir()) {
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
}
//Set maximum redirects
curl_setopt($ch, CURLOPT_MAXREDIRS, 10);
//Set the timeout
curl_setopt($ch, CURLOPT_TIMEOUT, $conf->options['timeout']);
//Set the proxy configuration. The user can provide this in wp-config.php
if (defined('WP_PROXY_HOST')) {
curl_setopt($ch, CURLOPT_PROXY, WP_PROXY_HOST);
}
if (defined('WP_PROXY_PORT')) {
curl_setopt($ch, CURLOPT_PROXYPORT, WP_PROXY_PORT);
}
if (defined('WP_PROXY_USERNAME')) {
$auth = WP_PROXY_USERNAME;
if (defined('WP_PROXY_PASSWORD')) {
$auth .= ':' . WP_PROXY_PASSWORD;
}
curl_setopt($ch, CURLOPT_PROXYUSERPWD, $auth);
}
//Make CURL return a valid result even if it gets a 404 or other error.
curl_setopt($ch, CURLOPT_FAILONERROR, false);
$nobody = !$use_get;
//Whether to send a HEAD request (the default) or a GET request
$parts = @parse_url($url);
if ($parts['scheme'] == 'https') {
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 0);
//Required to make HTTPS URLs work.
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, 0);
//Likewise.
$nobody = false;
//Can't use HEAD with HTTPS.
}
if ($nobody) {
//If possible, use HEAD requests for speed.
curl_setopt($ch, CURLOPT_NOBODY, true);
} else {
//If we must use GET at least limit the amount of downloaded data.
curl_setopt($ch, CURLOPT_HTTPHEADER, array('Range: bytes=0-2048'));
//2 KB
}
//Register a callback function which will process the HTTP header(s).
//It can be called multiple times if the remote server performs a redirect.
curl_setopt($ch, CURLOPT_HEADERFUNCTION, array(&$this, 'read_header'));
//Execute the request
curl_exec($ch);
$info = curl_getinfo($ch);
curl_close($ch);
//Store the results
$result['http_code'] = intval($info['http_code']);
$result['timeout'] = $result['http_code'] == 0;
//If the code is 0 then it's probably a timeout
$result['final_url'] = $info['url'];
$result['request_duration'] = $info['total_time'];
$result['redirect_count'] = $info['redirect_count'];
//Determine if the link counts as "broken"
$result['broken'] = $this->is_error_code($result['http_code']) || $result['timeout'];
if ($nobody && $result['broken']) {
//The site in question might be expecting GET instead of HEAD, so lets retry the request
//using the GET verb.
return $this->check($url, true);
//Note : normally a server that doesn't allow HEAD requests on a specific resource *should*
//return "405 Method Not Allowed". Unfortunately, there are sites that return 404 or
//another, even more general, error code instead. So just checking for 405 wouldn't be enough.
}
//When safe_mode or open_basedir is enabled CURL will be forbidden from following redirects,
//so redirect_count will be 0 for all URLs. As a workaround, set it to 1 when the HTTP
//response codes indicates a redirect but redirect_count is zero.
//Note to self : Extracting the Location header might also be helpful.
if ($result['redirect_count'] == 0 && in_array($result['http_code'], array(301, 302, 303, 307))) {
$result['redirect_count'] = 1;
}
//Build the log from HTTP code and headers.
//TODO: Put some kind of a color-coded error explanation at the top of the log, not a cryptic HTTP code.
$log .= '=== ';
if ($result['http_code']) {
$log .= sprintf(__('HTTP code : %d', 'broken-link-checker'), $result['http_code']);
} else {
$log .= __('(No response)', 'broken-link-checker');
}
$log .= " ===\n\n";
$log .= $this->last_headers;
if ($result['broken'] && $result['timeout']) {
$log .= "\n(" . __("Most likely the connection timed out or the domain doesn't exist.", 'broken-link-checker') . ')';
}
$result['log'] = $log;
//The hash should contain info about all pieces of data that pertain to determining if the
//link is working.
$result['result_hash'] = implode('|', array($result['http_code'], $result['broken'] ? 'broken' : '0', $result['timeout'] ? 'timeout' : '0', md5($result['final_url'])));
return $result;
}
function ui_get_source_comment($container, $container_field = '')
{
//Display a comment icon.
if ($container_field == 'comment_author_url') {
$image = 'font-awesome/font-awesome-user.png';
} else {
$image = 'font-awesome/font-awesome-comment-alt.png';
}
$comment = $container->get_wrapped_object();
//Display a small text sample from the comment
$text_sample = strip_tags($comment->comment_content);
$text_sample = blcUtility::truncate($text_sample, 65);
return array('image' => $image, 'text_sample' => $text_sample, 'comment_author' => esc_attr($comment->comment_author), 'comment_id' => esc_attr($comment->comment_ID), 'comment_status' => wp_get_comment_status($comment->comment_ID), 'container_post_title' => get_the_title($comment->comment_post_ID), 'container_post_status' => get_post_status($comment->comment_post_ID), 'container_post_ID' => $comment->comment_post_ID);
}
function check($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;
}
/**
* 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;
}
function send_email_notifications()
{
global $wpdb;
//Find links that have been detected as broken since the last sent notification.
$last_notification = date('Y-m-d H:i:s', $this->conf->options['last_notification_sent']);
$where = $wpdb->prepare('( first_failure >= %s )', $last_notification);
$links = blc_get_links(array('s_filter' => 'broken', 'where_expr' => $where, 'load_instances' => true, 'max_results' => 0));
if (empty($links)) {
return;
}
$cnt = count($links);
//Prepare email message
$subject = sprintf(__("[%s] Broken links detected", 'broken-link-checker'), html_entity_decode(get_option('blogname'), ENT_QUOTES));
$body = sprintf(_n("Broken Link Checker has detected %d new broken link on your site.", "Broken Link Checker has detected %d new broken links on your site.", $cnt, 'broken-link-checker'), $cnt);
$body .= "<br>";
$max_displayed_links = 5;
if ($cnt > $max_displayed_links) {
$line = sprintf(_n("Here's a list of the first %d broken links:", "Here's a list of the first %d broken links:", $max_displayed_links, 'broken-link-checker'), $max_displayed_links);
} else {
$line = __("Here's a list of the new broken links: ", 'broken-link-checker');
}
$body .= "<p>{$line}</p>";
//Show up to $max_displayed_links broken link instances right in the email.
$displayed = 0;
foreach ($links as $link) {
$instances = $link->get_instances();
foreach ($instances as $instance) {
$pieces = array(sprintf(__('Link text : %s', 'broken-link-checker'), $instance->ui_get_link_text('email')), sprintf(__('Link URL : <a href="%s">%s</a>', 'broken-link-checker'), htmlentities($link->url), blcUtility::truncate($link->url, 70, '')), sprintf(__('Source : %s', 'broken-link-checker'), $instance->ui_get_source('email')));
$link_entry = implode("<br>", $pieces);
$body .= "{$link_entry}<br><br>";
$displayed++;
if ($displayed >= $max_displayed_links) {
break 2;
//Exit both foreach loops
}
}
}
//Add a link to the "Broken Links" tab.
$body .= __("You can see all broken links here:", 'broken-link-checker') . "<br>";
$link_page = admin_url('tools.php?page=view-broken-links');
$body .= sprintf('<a href="%1$s">%1$s</a>', $link_page);
//Need to override the default 'text/plain' content type to send a HTML email.
add_filter('wp_mail_content_type', array(&$this, 'override_mail_content_type'));
//Send the notification
$rez = wp_mail(get_option('admin_email'), $subject, $body);
if ($rez) {
$this->conf->options['last_notification_sent'] = time();
$this->conf->save_options();
}
//Remove the override so that it doesn't interfere with other plugins that might
//want to send normal plaintext emails.
remove_filter('wp_mail_content_type', array(&$this, 'override_mail_content_type'));
}
?>
</ul>
</div>
</div>
<?php
}
}
?>
<!-- Admin Menu Editor Pro ad -->
<?php
if ($show_ame_ad) {
//Display an ad for Admin Menu Editor.
//We're A/B testing a bunch of different ad copies.
$ame_copy_variants = array(array('a', "Add, delete, hide, or move any admin menu item."), array('b', "Organize your admin menu the way you want it."), array('c', "Hide, move or customize admin menus. Perfect for client sites."));
$ad_copy_index = intval(blcUtility::constrained_hash(get_site_url(), 0, count($ame_copy_variants)));
$ad_copy = $ame_copy_variants[$ad_copy_index];
$ad_url = sprintf('http://w-shadow.com/admin-menu-editor-pro/?utm_source=broken_link_checker&utm_medium=text_link&utm_campaign=Plugins&utm_content=%s', urlencode('ad_copy_') . $ad_copy[0]);
?>
<div class="postbox" id="advertising">
<h3 class="hndle"><?php
_e('More plugins by Janis Elsts', 'broken-link-checker');
?>
</h3>
<div class="inside">
<p class="ws-ame-ad-copy"><?php
echo $ad_copy[1];
?>
</p>
<p class="ws-ame-ad-link">
<a href="<?php
/**
* 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;
}
/**
* @param blcLink $link
* @param blcLinkInstance[] $instances
*/
function column_status($link, $instances)
{
printf('<table class="mini-status" title="%s">', esc_attr(__('Show more info about this link', 'broken-link-checker')));
$status = $link->analyse_status();
printf('<tr class="link-status-row link-status-%s">
<td>
<span class="http-code">%s</span> <span class="status-text">%s</span>
</td>
</tr>', $status['code'], empty($link->http_code) ? '' : $link->http_code, $status['text']);
//Last checked...
if ($link->last_check != 0) {
$last_check = _x('Checked', 'checked how long ago', 'broken-link-checker') . ' ';
$last_check .= blcUtility::fuzzy_delta(time() - $link->last_check, 'ago');
printf('<tr class="link-last-checked"><td>%s</td></tr>', $last_check);
}
//Broken for...
if ($link->broken) {
$delta = time() - $link->first_failure;
$broken_for = blcUtility::fuzzy_delta($delta);
printf('<tr class="link-broken-for"><td>%s %s</td></tr>', __('Broken for', 'broken-link-checker'), $broken_for);
}
echo '</table>';
}
