cuqmbr/rss-bridge
1
0
Fork 0
mirror of https://github.com/RSS-Bridge/rss-bridge.git synced 2025-04-09 08:28:46 +00:00
Code Issues Packages Projects Releases Wiki Activity

[DuvarOrgBridge] Add Duvar.org bridge for scraping news articles (#4315)

Browse Source 
* Add Duvar.org bridge for scraping news articles

* PR Fixes

* Update DuvarOrgBridge.php to set a default value for the URL suffix

---------

Co-authored-by: Tughan Belbek <Tughan.Belbek@t-hive.io>
This commit is contained in:
Tuğhan Belbek 2024-11-03 20:30:28 +03:00 committed by GitHub
parent bd0fb1da99
commit 8d6d0fa10c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 86 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

86
bridges/DuvarOrgBridge.php Normal file
View File

@ -0,0 +1,86 @@
<?php
class DuvarOrgBridge extends BridgeAbstract
{
const NAME = 'Duvar.org - Haberler';
const MAINTAINER = 'yourname';
const URI = 'https://duvar.org';
const DESCRIPTION = 'Returns the latest articles from Duvar.org - News from Turkey and the world';
const CACHE_TIMEOUT = 3600; // 60min
const PARAMETERS = [[
'postcount' => [
'name' => 'Limit',
'type' => 'number',
'required' => true,
'title' => 'Maximum number of items to return',
'defaultValue' => 20,
],
'urlsuffix' => [
'name' => 'URL Suffix',
'type' => 'list',
'title' => 'Suffix for the URL to scrape a specific section',
'defaultValue' => 'Main',
'values' => [
'Main' => '',
'Balanced' => '/uyumlu',
'Protest' => '/muhalif',
'Center' => '/merkez',
'Alternative' => '/alternatif',
'Global' => '/global',
],
],
]];
public function collectData()
{
$postCount = $this->getInput('postcount');
$urlSuffix = $this->getInput('urlsuffix');
$url = self::URI . $urlSuffix;
$html = getSimpleHTMLDOM($url);
foreach ($html->find('article.news-item') as $data) {
if ($data === null) {
continue;
}
try {
$item = [];
$linkElement = $data->find('h2.news-title a', 0);
$titleElement = $data->find('h2.news-title a', 0);
$timestampElement = $data->find('time.meta-tag.date-tag', 0);
$contentElement = $data->find('div.news-description', 0);
if ($linkElement) {
$item['uri'] = $linkElement->getAttribute('href');
} else {
continue;
}
if ($titleElement) {
$item['title'] = trim($titleElement->plaintext);
} else {
continue;
}
if ($timestampElement) {
$item['timestamp'] = strtotime($timestampElement->plaintext);
} else {
$item['timestamp'] = time();
}
if ($contentElement) {
$item['content'] = trim($contentElement->plaintext);
} else {
$item['content'] = '';
}
$item['uid'] = hash('sha256', $item['title']);
$this->items[] = $item;
if (count($this->items) >= $postCount) {
break;
}
} catch (Exception $e) {
continue;
}
}
}
}