mirror of
https://github.com/RSS-Bridge/rss-bridge.git
synced 2024-11-26 11:26:31 +03:00
052844f5e1
simplehtmldom 1.9 introduced new functions to recursively remove nodes from the DOM. This allows removing elements without the need to re-load the document by using $html->load($html->save()), which is very inefficient. Find more information about remove() at https://simplehtmldom.sourceforge.io/docs/1.9/api/simple_html_dom_node/remove/
63 lines
2 KiB
PHP
63 lines
2 KiB
PHP
<?php
|
|
class EconomistBridge extends BridgeAbstract {
|
|
const NAME = 'The Economist: Latest Updates';
|
|
const URI = 'https://www.economist.com';
|
|
const DESCRIPTION = 'Fetches the latest updates from the Economist.';
|
|
const MAINTAINER = 'thefranke';
|
|
const CACHE_TIMEOUT = 3600; // 1h
|
|
|
|
public function getIcon() {
|
|
return 'https://www.economist.com/sites/default/files/econfinal_favicon.ico';
|
|
}
|
|
|
|
public function collectData() {
|
|
$html = getSimpleHTMLDOM(self::URI . '/latest/')
|
|
or returnServerError('Could not fetch latest updates form The Economist.');
|
|
|
|
foreach($html->find('article') as $element) {
|
|
|
|
$a = $element->find('a', 0);
|
|
$href = self::URI . $a->href;
|
|
$full = getSimpleHTMLDOMCached($href);
|
|
$article = $full->find('article', 0);
|
|
|
|
$header = $article->find('h1', 0);
|
|
$author = $article->find('span[itemprop="author"]', 0);
|
|
$time = $article->find('time[itemprop="dateCreated"]', 0);
|
|
$content = $article->find('div[itemprop="description"]', 0);
|
|
|
|
// Remove newsletter subscription box
|
|
$newsletter = $content->find('div[class="newsletter-form__message"]', 0);
|
|
if ($newsletter)
|
|
$newsletter->remove();
|
|
|
|
$newsletterForm = $content->find('form', 0);
|
|
if ($newsletterForm)
|
|
$newsletterForm->remove();
|
|
|
|
// Remove next and previous article URLs at the bottom
|
|
$nextprev = $content->find('div[class="blog-post__next-previous-wrapper"]', 0);
|
|
if ($nextprev)
|
|
$nextprev->remove();
|
|
|
|
$section = [ $article->find('h3[itemprop="articleSection"]', 0)->plaintext ];
|
|
|
|
$item = array();
|
|
$item['title'] = $header->find('span', 0)->innertext . ': '
|
|
. $header->find('span', 1)->innertext;
|
|
|
|
$item['uri'] = $href;
|
|
$item['timestamp'] = strtotime($time->datetime);
|
|
$item['author'] = $author->innertext;
|
|
$item['categories'] = $section;
|
|
|
|
$item['content'] = '<img style="max-width: 100%" src="'
|
|
. $a->find('img', 0)->src . '">' . $content->innertext;
|
|
|
|
$this->items[] = $item;
|
|
|
|
if (count($this->items) >= 10)
|
|
break;
|
|
}
|
|
}
|
|
}
|