boyska/rss-bridge
1
0
Fork 0
forked from blallo/rss-bridge
Code Releases Activity

[GooglePlusPostBridge] Fix bridge implementation

Browse source 
This bridge was broken due to DOM changes. This commit fixes
most of the broken code. Hashtags do no longer work because
they are no longer supported/provided.

The timing might be off as the source only provides a rough
relative value like '1 hour' or '1 year'.

Closes #485
This commit is contained in:
logmanoriginal 2017-03-18 21:06:48 +01:00
parent 59025d96bc
commit bd92392921
1 changed files with 22 additions and 18 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

40
bridges/GooglePlusPostBridge.php
View file

@ -30,25 +30,27 @@ class GooglePlusPostBridge extends BridgeAbstract{
) or returnServerError('No results for this query.'); ) or returnServerError('No results for this query.');
// get title, url, ... there is a lot of intresting stuff in meta // get title, url, ... there is a lot of intresting stuff in meta
$this->_title = $html->find('meta[property]', 0)->getAttribute('content'); $this->_title = $html->find('meta[property=og:title]', 0)->getAttribute('content');
$this->_url = $html->find('meta[itemprop=url]', 0)->getAttribute('content'); $this->_url = $html->find('meta[property=og:url]', 0)->getAttribute('content');
// div[jsmodel=XNmfOc] // I don't even know where to start with this discusting html...
foreach($html->find('div.yt') as $post){ foreach($html->find('div[jsname=WsjYwc]') as $post){
$item = array(); $item = array();
// $item['content'] = $post->find('div.Al', 0)->innertext;
$item['author'] = $item['fullname'] = $post->find('header.lea h3 a', 0)->innertext; $item['author'] = $item['fullname'] = $post->find('div div div div a', 0)->innertext;
$item['id'] = $post->getAttribute('id'); $item['id'] = $post->find('div div div', 0)->getAttribute('id');
$item['title'] = $item['fullname'] = $post->find('header.lea', 0)->plaintext; $item['avatar'] = $post->find('div img', 0)->src;
$item['avatar'] = $post->find('div.ys img', 0)->src; $item['uri'] = self::URI . $post->find('div div div a', 1)->href;
$item['uri'] = self::URI . $post->find('a.o-U-s', 0)->href; $item['timestamp'] = strtotime(
$item['timestamp'] = strtotime($post->find('a.o-U-s', 0)->plaintext); '+' . preg_replace('/[^0-9A-Za-z]/',
'',
$post->find('div div div a span', 1)->getAttribute('aria-label')));
// hashtag to treat : https://plus.google.com/explore/tag // hashtag to treat : https://plus.google.com/explore/tag
$hashtags = array(); // $hashtags = array();
foreach($post->find('a.d-s') as $hashtag){ // foreach($post->find('a.d-s') as $hashtag){
$hashtags[trim($hashtag->plaintext)] = self::URI . $hashtag->href; // $hashtags[trim($hashtag->plaintext)] = self::URI . $hashtag->href;
} // }
$item['content'] = ''; $item['content'] = '';
@ -63,7 +65,10 @@ class GooglePlusPostBridge extends BridgeAbstract{
. $item['avatar'] . $item['avatar']
. '" /></a></div>'; . '" /></a></div>';
$content = $post->find('div.Al', 0); $content = $post->find('div div[id^=body] div div', 0);
// extract plaintext
$item['content_simple'] = $content->plaintext;
$item['title'] = substr($item['content_simple'], 0, 72) . '...';
// XXX ugly but I don't have any idea how to do a better stuff, // XXX ugly but I don't have any idea how to do a better stuff,
// str_replace on link doesn't work as expected and ask too many checks // str_replace on link doesn't work as expected and ask too many checks
@ -86,8 +91,7 @@ class GooglePlusPostBridge extends BridgeAbstract{
$item['content'] .= '<div style="margin-top: -1.5em">' . $content . '</div>'; $item['content'] .= '<div style="margin-top: -1.5em">' . $content . '</div>';
$item['content'] = trim(strip_tags($item['content'], '<a><p><div><img>')); $item['content'] = trim(strip_tags($item['content'], '<a><p><div><img>'));
// extract plaintext
$item['content_simple'] = $post->find('div.Al', 0)->plaintext;
$this->items[] = $item; $this->items[] = $item;
} }
} }