/**
* html_entity_decode with recursion
*
* @param $string string|array
* @param $flags integer default is ENT_COMPAT | ENT_HTML401
* @param $encoding string default is ini_get('default_charset')
* @return string|array
*/
function htmlEntityDecode($string, $flags = null, $encoding = null)
{
if (!isset($flags)) {
$flags = ENT_COMPAT | ENT_HTML401;
}
if (!isset($encoding)) {
$encoding = ini_get('default_charset');
}
if (is_array($string)) {
foreach ($string as $key => $value) {
$string[$key] = htmlEntityDecode($value, $flags, $encoding);
}
} else {
$string = html_entity_decode($string, $flags, $encoding);
}
return $string;
}
/**
* Examine a page at {@link $Url} for title, description & images.
*
* Be sure to check the resultant array for any Exceptions that occurred while retrieving the page.
*
* @param string $url The url to examine.
* @param integer $timeout How long to allow for this request.
* Default Garden.SocketTimeout or 1, 0 to never timeout. Default is 0.
* @param bool $sendCookies Whether or not to send browser cookies with the request.
* @return array Returns an array containing Url, Title, Description, Images (array) and Exception
* (if there were problems retrieving the page).
*/
function fetchPageInfo($url, $timeout = 3, $sendCookies = false)
{
$PageInfo = array('Url' => $url, 'Title' => '', 'Description' => '', 'Images' => array(), 'Exception' => false);
try {
// Make sure the URL is valid.
$urlParts = parse_url($url);
if ($urlParts === false || !in_array(val('scheme', $urlParts), array('http', 'https'))) {
throw new Exception('Invalid URL.', 400);
}
if (!defined('HDOM_TYPE_ELEMENT')) {
require_once PATH_LIBRARY . '/vendors/simplehtmldom/simple_html_dom.php';
}
$Request = new ProxyRequest();
$PageHtml = $Request->Request(array('URL' => $url, 'Timeout' => $timeout, 'Cookies' => $sendCookies));
if (!$Request->status()) {
throw new Exception('Couldn\'t connect to host.', 400);
}
$Dom = str_get_html($PageHtml);
if (!$Dom) {
throw new Exception('Failed to load page for parsing.');
}
// FIRST PASS: Look for open graph title, desc, images
$PageInfo['Title'] = domGetContent($Dom, 'meta[property=og:title]');
Trace('Getting og:description');
$PageInfo['Description'] = domGetContent($Dom, 'meta[property=og:description]');
foreach ($Dom->find('meta[property=og:image]') as $Image) {
if (isset($Image->content)) {
$PageInfo['Images'][] = $Image->content;
}
}
// SECOND PASS: Look in the page for title, desc, images
if ($PageInfo['Title'] == '') {
$PageInfo['Title'] = $Dom->find('title', 0)->plaintext;
}
if ($PageInfo['Description'] == '') {
Trace('Getting meta description');
$PageInfo['Description'] = domGetContent($Dom, 'meta[name=description]');
}
// THIRD PASS: Look in the page contents
if ($PageInfo['Description'] == '') {
foreach ($Dom->find('p') as $element) {
Trace('Looking at p for description.');
if (strlen($element->plaintext) > 150) {
$PageInfo['Description'] = $element->plaintext;
break;
}
}
if (strlen($PageInfo['Description']) > 400) {
$PageInfo['Description'] = SliceParagraph($PageInfo['Description'], 400);
}
}
// Final: Still nothing? remove limitations
if ($PageInfo['Description'] == '') {
foreach ($Dom->find('p') as $element) {
Trace('Looking at p for description (no restrictions)');
if (trim($element->plaintext) != '') {
$PageInfo['Description'] = $element->plaintext;
break;
}
}
}
// Page Images
if (count($PageInfo['Images']) == 0) {
$Images = domGetImages($Dom, $url);
$PageInfo['Images'] = array_values($Images);
}
$PageInfo['Title'] = htmlEntityDecode($PageInfo['Title']);
$PageInfo['Description'] = htmlEntityDecode($PageInfo['Description']);
} catch (Exception $ex) {
$PageInfo['Exception'] = $ex->getMessage();
}
return $PageInfo;
}
/**
* Render the entire head module.
*/
public function toString()
{
// Add the canonical Url if necessary.
if (method_exists($this->_Sender, 'CanonicalUrl') && !c('Garden.Modules.NoCanonicalUrl', false)) {
$CanonicalUrl = $this->_Sender->canonicalUrl();
if (!isUrl($CanonicalUrl)) {
$CanonicalUrl = Gdn::router()->ReverseRoute($CanonicalUrl);
}
$this->_Sender->canonicalUrl($CanonicalUrl);
// $CurrentUrl = url('', true);
// if ($CurrentUrl != $CanonicalUrl) {
$this->addTag('link', array('rel' => 'canonical', 'href' => $CanonicalUrl));
// }
}
// Include facebook open-graph meta information.
if ($FbAppID = c('Plugins.Facebook.ApplicationID')) {
$this->addTag('meta', array('property' => 'fb:app_id', 'content' => $FbAppID));
}
$SiteName = c('Garden.Title', '');
if ($SiteName != '') {
$this->addTag('meta', array('property' => 'og:site_name', 'content' => $SiteName));
}
$Title = htmlEntityDecode(Gdn_Format::text($this->title('', true)));
if ($Title != '') {
$this->addTag('meta', array('name' => 'twitter:title', 'property' => 'og:title', 'content' => $Title));
}
if (isset($CanonicalUrl)) {
$this->addTag('meta', array('property' => 'og:url', 'content' => $CanonicalUrl));
}
if ($Description = trim(Gdn_Format::reduceWhiteSpaces($this->_Sender->Description()))) {
$this->addTag('meta', array('name' => 'description', 'property' => 'og:description', 'content' => $Description));
}
$hasRelevantImage = false;
// Default to the site logo if there were no images provided by the controller.
if (count($this->_Sender->Image()) == 0) {
$Logo = c('Garden.ShareImage', c('Garden.Logo', ''));
if ($Logo != '') {
// Fix the logo path.
if (stringBeginsWith($Logo, 'uploads/')) {
$Logo = substr($Logo, strlen('uploads/'));
}
$Logo = Gdn_Upload::url($Logo);
$this->addTag('meta', array('property' => 'og:image', 'content' => $Logo));
}
} else {
foreach ($this->_Sender->Image() as $Img) {
$this->addTag('meta', array('name' => 'twitter:image', 'property' => 'og:image', 'content' => $Img));
$hasRelevantImage = true;
}
}
// For the moment at least, only discussions are supported.
if ($Title && val('DiscussionID', $this->_Sender)) {
if ($hasRelevantImage) {
$twitterCardType = 'summary_large_image';
} else {
$twitterCardType = 'summary';
}
// Let's force a description for the image card since it makes sense to see a card with only an image and a title.
if (!$Description && $twitterCardType === 'summary_large_image') {
$Description = '...';
}
// Card && Title && Description are required
if ($twitterCardType && $Description) {
$this->addTag('meta', array('name' => 'twitter:description', 'content' => $Description));
$this->addTag('meta', array('name' => 'twitter:card', 'content' => $twitterCardType));
}
}
$this->fireEvent('BeforeToString');
$Tags = $this->_Tags;
// Make sure that css loads before js (for jquery)
usort($this->_Tags, array('HeadModule', 'TagCmp'));
// "link" comes before "script"
$Tags2 = $this->_Tags;
// Start with the title.
$Head = '<title>' . Gdn_Format::text($this->title()) . "</title>\n";
$TagStrings = array();
// Loop through each tag.
foreach ($this->_Tags as $Index => $Attributes) {
$Tag = $Attributes[self::TAG_KEY];
// Inline the content of the tag, if necessary.
if (val('_hint', $Attributes) == 'inline') {
$Path = val('_path', $Attributes);
if ($Path && !stringBeginsWith($Path, 'http')) {
$Attributes[self::CONTENT_KEY] = file_get_contents($Path);
if (isset($Attributes['src'])) {
$Attributes['_src'] = $Attributes['src'];
unset($Attributes['src']);
}
if (isset($Attributes['href'])) {
$Attributes['_href'] = $Attributes['href'];
unset($Attributes['href']);
}
}
}
// If we set an IE conditional AND a "Not IE" condition, we will need to make a second pass.
do {
// Reset tag string
$TagString = '';
// IE conditional? Validates condition.
$IESpecific = isset($Attributes['_ie']) && preg_match('/((l|g)t(e)? )?IE [0-9\\.]/', $Attributes['_ie']);
// Only allow $NotIE if we're not doing a conditional this loop.
$NotIE = !$IESpecific && isset($Attributes['_notie']);
// Open IE conditional tag
if ($IESpecific) {
$TagString .= '<!--[if ' . $Attributes['_ie'] . ']>';
}
if ($NotIE) {
$TagString .= '<!--[if !IE]> -->';
}
// Build tag
$TagString .= ' <' . $Tag . Attribute($Attributes, '_');
if (array_key_exists(self::CONTENT_KEY, $Attributes)) {
$TagString .= '>' . $Attributes[self::CONTENT_KEY] . '</' . $Tag . '>';
} elseif ($Tag == 'script') {
$TagString .= '></script>';
} else {
$TagString .= ' />';
}
// Close IE conditional tag
if ($IESpecific) {
$TagString .= '<![endif]-->';
}
if ($NotIE) {
$TagString .= '<!-- <![endif]-->';
}
// Cleanup (prevent infinite loop)
if ($IESpecific) {
unset($Attributes['_ie']);
}
$TagStrings[] = $TagString;
} while ($IESpecific && isset($Attributes['_notie']));
// We need a second pass
}
//endforeach
$Head .= implode("\n", array_unique($TagStrings));
foreach ($this->_Strings as $String) {
$Head .= $String;
$Head .= "\n";
}
return $Head;
}
/**
* Examine a page at {@link $Url} for title, description & images.
*
* Be sure to check the resultant array for any Exceptions that occurred while retrieving the page.
*
* @param string $url The url to examine.
* @param integer $timeout How long to allow for this request.
* Default Garden.SocketTimeout or 1, 0 to never timeout. Default is 0.
* @param bool $sendCookies Whether or not to send browser cookies with the request.
* @return array Returns an array containing Url, Title, Description, Images (array) and Exception
* (if there were problems retrieving the page).
*/
function fetchPageInfo($url, $timeout = 3, $sendCookies = false)
{
$PageInfo = array('Url' => $url, 'Title' => '', 'Description' => '', 'Images' => array(), 'Exception' => false);
try {
// Make sure the URL is valid.
$urlParts = parse_url($url);
if ($urlParts === false || !in_array(val('scheme', $urlParts), array('http', 'https'))) {
throw new Exception('Invalid URL.', 400);
}
$Request = new ProxyRequest();
$PageHtml = $Request->Request(array('URL' => $url, 'Timeout' => $timeout, 'Cookies' => $sendCookies, 'Redirects' => true));
if (!$Request->status()) {
throw new Exception('Couldn\'t connect to host.', 400);
}
$Dom = pQuery::parseStr($PageHtml);
if (!$Dom) {
throw new Exception('Failed to load page for parsing.');
}
// FIRST PASS: Look for open graph title, desc, images
$PageInfo['Title'] = domGetContent($Dom, 'meta[property="og:title"]');
Trace('Getting og:description');
$PageInfo['Description'] = domGetContent($Dom, 'meta[property="og:description"]');
foreach ($Dom->query('meta[property="og:image"]') as $Image) {
if ($Image->attr('content')) {
$PageInfo['Images'][] = $Image->attr('content');
}
}
// SECOND PASS: Look in the page for title, desc, images
if ($PageInfo['Title'] == '') {
$PageInfo['Title'] = $Dom->query('title')->text();
}
if ($PageInfo['Description'] == '') {
Trace('Getting meta description');
$PageInfo['Description'] = domGetContent($Dom, 'meta[name="description"]');
}
// THIRD PASS: Look in the page contents
if ($PageInfo['Description'] == '') {
foreach ($Dom->query('p') as $element) {
Trace('Looking at p for description.');
if (strlen($element->plaintext) > 150) {
$PageInfo['Description'] = $element->text();
break;
}
}
if (strlen($PageInfo['Description']) > 400) {
$PageInfo['Description'] = SliceParagraph($PageInfo['Description'], 400);
}
}
// Final: Still nothing? remove limitations
if ($PageInfo['Description'] == '') {
foreach ($Dom->query('p') as $element) {
Trace('Looking at p for description (no restrictions)');
if (trim($element->text()) != '') {
$PageInfo['Description'] = $element->text();
break;
}
}
}
// Page Images
if (count($PageInfo['Images']) == 0) {
$Images = domGetImages($Dom, $url);
$PageInfo['Images'] = array_values($Images);
}
$PageInfo['Title'] = htmlEntityDecode($PageInfo['Title']);
$PageInfo['Description'] = htmlEntityDecode($PageInfo['Description']);
} catch (Exception $ex) {
$PageInfo['Exception'] = $ex->getMessage();
}
return $PageInfo;
}
