forked from blallo/rss-bridge
[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
This commit is contained in:
logmanoriginal
parent
4124c707d4
commit
2500d0df93
1 changed files with 95 additions and 53 deletions
|
|
@ -1,9 +1,9 @@
|
|||
<?php
|
||||
class PinterestBridge extends BridgeAbstract {
|
||||
class PinterestBridge extends FeedExpander {
|
||||
|
||||
const MAINTAINER = 'pauder';
|
||||
const NAME = 'Pinterest Bridge';
|
||||
const URI = 'http://www.pinterest.com/';
|
||||
const URI = 'https://www.pinterest.com';
|
||||
const DESCRIPTION = 'Returns the newest images on a board';
|
||||
|
||||
const PARAMETERS = array(
|
||||
|
|
@ -15,6 +15,13 @@ class PinterestBridge extends BridgeAbstract {
|
|||
'b' => 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){
|
||||
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 $div){
|
||||
private function getUserResults($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 = $json['resourceDataCache'][0]['data']['owner']['image_small_url'];
|
||||
|
||||
foreach($results as $result){
|
||||
$item = array();
|
||||
|
||||
$a = $div->find('a.pinImageWrapper', 0);
|
||||
$img = $a->find('img', 0);
|
||||
$item['uri'] = $result['link'];
|
||||
|
||||
$item['uri'] = $this->getURI() . $a->getAttribute('href');
|
||||
$item['content'] = '<img src="'
|
||||
. htmlentities(str_replace('/236x/', '/736x/', $img->getAttribute('src')))
|
||||
. '" alt="" />';
|
||||
// Some use regular titles, others provide 'advanced' infos, a few
|
||||
// provide even less info. Thus we attempt multiple options.
|
||||
$item['title'] = trim($result['title']);
|
||||
|
||||
$avatar = $div->find('div.creditImg', 0)->find('img', 0);
|
||||
$avatar = $avatar->getAttribute('data-src');
|
||||
$avatar = str_replace("\\", "", $avatar);
|
||||
if($item['title'] === "")
|
||||
$item['title'] = trim($result['rich_summary']['display_name']);
|
||||
|
||||
$username = $div->find('div.creditName', 0);
|
||||
$board = $div->find('div.creditTitle', 0);
|
||||
if($item['title'] === "")
|
||||
$item['title'] = trim($result['description']);
|
||||
|
||||
$item['username'] = $username->innertext;
|
||||
$item['fullname'] = $board->innertext;
|
||||
$item['timestamp'] = strtotime($result['created_at']);
|
||||
$item['username'] = $username;
|
||||
$item['fullname'] = $fullname;
|
||||
$item['avatar'] = $avatar;
|
||||
|
||||
$item['content'] .= '<br /><img align="left" style="margin: 2px 4px;" src="'
|
||||
$item['author'] = $item['username'] . ' (' . $item['fullname'] . ')';
|
||||
$item['content'] = '<img align="left" style="margin: 2px 4px;" src="'
|
||||
. htmlentities($item['avatar'])
|
||||
. '" /> <strong>'
|
||||
. '" /><p><strong>'
|
||||
. $item['username']
|
||||
. '</strong><br />'
|
||||
. $item['fullname'];
|
||||
. '</strong><br>'
|
||||
. $item['fullname']
|
||||
. '</p><br><img src="'
|
||||
. $result['images']['736x']['url']
|
||||
. '" alt="" /><br><p>'
|
||||
. $result['description']
|
||||
. '</p>';
|
||||
|
||||
$item['enclosures'] = array($result['images']['orig']['url']);
|
||||
|
||||
$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!');
|
||||
}
|
||||
|
||||
$json = json_decode($container->innertext, true);
|
||||
private function getSearchResults($html){
|
||||
$json = json_decode($html->find('#jsInit1', 0)->innertext, true);
|
||||
$results = $json['resourceDataCache'][0]['data']['results'];
|
||||
|
||||
foreach($json['itemListElement'] as $element){
|
||||
foreach($results as $result){
|
||||
$item = array();
|
||||
|
||||
$item['uri'] = $element['item']['sharedContent']['author']['url'];
|
||||
$item['title'] = $element['item']['name'];
|
||||
$item['author'] = $element['item']['user']['name'];
|
||||
$item['timestamp'] = strtotime($element['item']['datePublished']);
|
||||
$item['content'] = <<<EOD
|
||||
<a href="{$item['uri']}">
|
||||
<img src="{$element['item']['image']}">
|
||||
</a>
|
||||
<p>{$element['item']['text']}</p>
|
||||
EOD;
|
||||
$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'] . ')';
|
||||
$item['content'] = '<img align="left" style="margin: 2px 4px;" src="'
|
||||
. htmlentities($item['avatar'])
|
||||
. '" /><p><strong>'
|
||||
. $item['username']
|
||||
. '</strong><br>'
|
||||
. $item['fullname']
|
||||
. '</p><br><img src="'
|
||||
. $result['images']['736x']['url']
|
||||
. '" alt="" /><br><p>'
|
||||
. $result['description']
|
||||
. '</p>';
|
||||
|
||||
$item['enclosures'] = array($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');
|
||||
|
|
|
|||
Loading…
Reference in a new issue