blallo/rss-bridge
1
0
Fork 1
Code Issues Pull requests Releases Wiki Activity

[Numerama] update bridge post extraction (issue #158)

Browse source
This commit is contained in:
Mitsukarenai 2015-10-12 17:13:27 +02:00
parent f11d6fc2ad
commit 01d6859f77
1 changed files with 12 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

13
bridges/NumeramaBridge.php
View file

@ -2,12 +2,12 @@
/**
* RssBridgeNumerama
* Returns the 5 newest posts from http://www.numerama.com (full text)
* 2014-05-25
*
* @name Numerama
* @homepage http://www.numerama.com/
* @description Returns the 5 newest posts from Numerama (full text)
* @maintainer mitsukarenai
* @update 2015-10-12
*/
class NumeramaBridge extends BridgeAbstract{
@ -18,13 +18,16 @@ class NumeramaBridge extends BridgeAbstract{
$string = str_replace(']]>', '', $string);
return $string;
}
function NumeramaExtractContent($url) {
function NumeramaExtractContent($url)
{
$html2 = file_get_html($url);
$text = $html2->find('h2.intro', 0)->innertext;
$text = $text.$html2->find('div.content', 0)->innertext;
$text = strip_tags($text, '<p><b><a><blockquote><img><em><ul><ol>');
$text = $html2->find('section[class=related-article]', 0)->innertext = ''; // remove related articles block
$text = '<img alt="" style="max-width:300px;" src="'.$html2->find('meta[property=og:image]', 0)->getAttribute('content').'">'; // add post picture
$text = $text.$html2->find('article[class=post-content]', 0)->innertext; // extract the post
return $text;
}
$html = file_get_html('http://www.numerama.com/rss/news.rss') or $this->returnError('Could not request Numerama.', 404);
$limit = 0;