From 2500d0df9329cddfe8c07b49036c42561e784f51 Mon Sep 17 00:00:00 2001 From: logmanoriginal Date: Sun, 9 Apr 2017 23:26:35 +0200 Subject: [PATCH] [PinterestBridge] Fix implementation after DOM changes Due to breaking DOM changes this bridge required re-implementation. With this fix the brige will make use of the JSON data embedded in the returned HTML. The content returned for all contexts is similar with only a few differences due to limitations of the JSON. Feeds returned for a given username and board will by default make use of the provided RSS feed instead of using the custom filter. This bahaviour can be changed by setting the optional parameter '&r=off' (on by default) Notice: The JSON data for userdata and search results is very different, so two functions were implemented to account for that. References #498 --- bridges/PinterestBridge.php | 148 +++++++++++++++++++++++------------- 1 file changed, 95 insertions(+), 53 deletions(-) diff --git a/bridges/PinterestBridge.php b/bridges/PinterestBridge.php index 68989790..7597c43d 100644 --- a/bridges/PinterestBridge.php +++ b/bridges/PinterestBridge.php @@ -1,9 +1,9 @@ array( 'name' => 'board', 'required' => true + ), + 'r' => array( + 'name' => 'Use provided RSS', + 'type' => 'checkbox', + 'required' => false, + 'defaultValue' => 'checked', + 'title' => 'Uncheck to return data via custom filters (more data)' ) ), 'From search' => array( @@ -26,81 +33,116 @@ class PinterestBridge extends BridgeAbstract { ); public function collectData(){ - $html = getSimpleHTMLDOM($this->getURI()); - if(!$html){ - switch($this->queriedContext){ + switch($this->queriedContext){ case 'By username and board': - returnServerError('Username and/or board not found'); + if($this->getInput('r')){ + $this->collectExpandableDatas($this->getURI() . '.rss'); + } else { + $html = getSimpleHTMLDOMCached($this->getURI()); + $this->getUserResults($html); + } + break; case 'From search': - returnServerError('Could not request Pinterest.'); - } + default: + $html = getSimpleHTMLDOMCached($this->getURI()); + $this->getSearchResults($html); } + } - if($this->queriedContext === 'From search'){ - foreach($html->find('div.pinWrapper') as $div){ - $item = array(); + private function getUserResults($html){ + $json = json_decode($html->find('#jsInit1', 0)->innertext, true); + $results = $json['tree']['children'][0]['children'][0]['children'][0]['options']['props']['data']['board_feed']; + $username = $json['resourceDataCache'][0]['data']['owner']['username']; + $fullname = $json['resourceDataCache'][0]['data']['owner']['full_name']; + $avatar = $json['resourceDataCache'][0]['data']['owner']['image_small_url']; - $a = $div->find('a.pinImageWrapper', 0); - $img = $a->find('img', 0); + foreach($results as $result){ + $item = array(); - $item['uri'] = $this->getURI() . $a->getAttribute('href'); - $item['content'] = ''; + $item['uri'] = $result['link']; - $avatar = $div->find('div.creditImg', 0)->find('img', 0); - $avatar = $avatar->getAttribute('data-src'); - $avatar = str_replace("\\", "", $avatar); + // Some use regular titles, others provide 'advanced' infos, a few + // provide even less info. Thus we attempt multiple options. + $item['title'] = trim($result['title']); - $username = $div->find('div.creditName', 0); - $board = $div->find('div.creditTitle', 0); + if($item['title'] === "") + $item['title'] = trim($result['rich_summary']['display_name']); - $item['username'] = $username->innertext; - $item['fullname'] = $board->innertext; - $item['avatar'] = $avatar; + if($item['title'] === "") + $item['title'] = trim($result['description']); - $item['content'] .= '
' + . '" />

' . $item['username'] - . '
' - . $item['fullname']; + . '
' + . $item['fullname'] + . '



' + . $result['description'] + . '

'; - $item['title'] = $img->getAttribute('alt'); - $this->items[] = $item; - } - } elseif($this->queriedContext === 'By username and board'){ - $container = $html->find('SCRIPT[type="application/ld+json"]', 0) - or returnServerError('Unable to find data container!'); + $item['enclosures'] = array($result['images']['orig']['url']); - $json = json_decode($container->innertext, true); + $this->items[] = $item; + } + } - foreach($json['itemListElement'] as $element){ - $item = array(); + private function getSearchResults($html){ + $json = json_decode($html->find('#jsInit1', 0)->innertext, true); + $results = $json['resourceDataCache'][0]['data']['results']; - $item['uri'] = $element['item']['sharedContent']['author']['url']; - $item['title'] = $element['item']['name']; - $item['author'] = $element['item']['user']['name']; - $item['timestamp'] = strtotime($element['item']['datePublished']); - $item['content'] = << - - -

{$element['item']['text']}

-EOD; + foreach($results as $result){ + $item = array(); - $this->items[] = $item; - } + $item['uri'] = self::URI . $result['board']['url']; + + // Some use regular titles, others provide 'advanced' infos, a few + // provide even less info. Thus we attempt multiple options. + $item['title'] = trim($result['title']); + + if($item['title'] === "") + $item['title'] = trim($result['rich_summary']['display_name']); + + if($item['title'] === "") + $item['title'] = trim($result['grid_description']); + + $item['timestamp'] = strtotime($result['created_at']); + $item['username'] = $result['pinner']['username']; + $item['fullname'] = $result['pinner']['full_name']; + $item['avatar'] = $result['pinner']['image_small_url']; + $item['author'] = $item['username'] . ' (' . $item['fullname'] . ')'; + $item['content'] = '

' + . $item['username'] + . '
' + . $item['fullname'] + . '



' + . $result['description'] + . '

'; + + $item['enclosures'] = array($result['images']['orig']['url']); + + $this->items[] = $item; } } public function getURI(){ switch($this->queriedContext){ case 'By username and board': - $uri = self::URI . urlencode($this->getInput('u')) . '/' . urlencode($this->getInput('b')); + $uri = self::URI . '/' . urlencode($this->getInput('u')) . '/' . urlencode($this->getInput('b'));// . '.rss'; break; case 'From search': - $uri = self::URI . 'search/?q=' . urlencode($this->getInput('q')); + $uri = self::URI . '/search/?q=' . urlencode($this->getInput('q')); break; default: return parent::getURI(); } @@ -110,7 +152,7 @@ EOD; public function getName(){ switch($this->queriedContext){ case 'By username and board': - $specific = $this->getInput('u') . '-' . $this->getInput('b'); + $specific = $this->getInput('u') . ' - ' . $this->getInput('b'); break; case 'From search': $specific = $this->getInput('q');