cuqmbr/rss-bridge
1
0
Fork 0
mirror of https://github.com/RSS-Bridge/rss-bridge.git synced 2025-04-09 16:38:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

[NYTBridge] Fix article parsing (#2106)

Browse Source 
Co-authored-by: podiki <podiki@users.noreply.github.com>
This commit is contained in:
Corentin Garcia 2021-06-30 12:14:25 +02:00 committed by GitHub
parent 9399ebb2c6
commit 0de2db853f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 17 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
bridges/NYTBridge.php
View File

@ -4,22 +4,33 @@ class NYTBridge extends FeedExpander {
const MAINTAINER = 'IceWreck'; const MAINTAINER = 'IceWreck';
const NAME = 'New York Times Bridge'; const NAME = 'New York Times Bridge';
const URI = 'https://www.nytimes.com/'; const URI = 'https://www.nytimes.com/';
const CACHE_TIMEOUT = 3600; const CACHE_TIMEOUT = 900; // 15 minutes
const DESCRIPTION = 'RSS feed for the New York Times'; const DESCRIPTION = 'RSS feed for the New York Times';
public function collectData(){ public function collectData(){
$this->collectExpandableDatas('https://rss.nytimes.com/services/xml/rss/nyt/HomePage.xml', 15); $this->collectExpandableDatas('https://rss.nytimes.com/services/xml/rss/nyt/HomePage.xml', 40);
} }
protected function parseItem($newsItem){ protected function parseItem($newsItem){
$item = parent::parseItem($newsItem); $item = parent::parseItem($newsItem);
$article = '';
// $articlePage gets the entire page's contents // $articlePage gets the entire page's contents
$articlePage = getSimpleHTMLDOM($newsItem->link); $articlePage = getSimpleHTMLDOM($newsItem->link);
// handle subtitle
$subtitle = $articlePage->find('p.css-w6ymp8', 0);
if ($subtitle != null) {
$article .= '<strong>' . $subtitle->plaintext . '</strong>';
}
// figure contain's the main article image // figure contain's the main article image
$article = $articlePage->find('figure', 0); $article .= $articlePage->find('figure', 0) . '<br />';
// p > css-exrw3m has the actual article
foreach($articlePage->find('p.css-exrw3m') as $element) // section.meteredContent has the actual article
$article = $article . $element; foreach($articlePage->find('section.meteredContent p') as $element)
$article .= '' . $element . '';
$item['content'] = $article; $item['content'] = $article;
return $item; return $item;
} }