LesLibres.org
/
Rss-Bridge
2
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

[FolhaDeSaoPaulo]: Add new Bridge (#1426) 

* [FolhaDeSaoPaulo]: Add new Bridge
This commit is contained in:
somini 2020-02-04 16:19:39 +00:00 committed by GitHub
parent f52eb43f8c
commit 275662b8d4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 51 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

51
bridges/FolhaDeSaoPauloBridge.php Normal file
View File

@ -0,0 +1,51 @@
<?php
class FolhaDeSaoPauloBridge extends FeedExpander {
const MAINTAINER = 'somini';
const NAME = 'Folha de São Paulo';
const URI = 'https://www1.folha.uol.com.br';
const DESCRIPTION = 'Returns the newest posts from Folha de São Paulo (full text)';
const PARAMETERS = array(
array(
'feed' => array(
'name' => 'Feed sub-URL',
'type' => 'text',
'title' => 'Select the sub-feed (see https://www1.folha.uol.com.br/feed/)',
'exampleValue' => 'emcimadahora/rss091.xml',
)
)
);
protected function parseItem($item){
$item = parent::parseItem($item);
$articleHTMLContent = getSimpleHTMLDOMCached($item['uri']);
if($articleHTMLContent) {
foreach ($articleHTMLContent->find('div.c-news__body .is-hidden') as $toRemove) {
$toRemove->innertext = '';
}
$item_content = $articleHTMLContent->find('div.c-news__body', 0);
if ($item_content) {
$text = $item_content->innertext;
$text = strip_tags($text, '<p><b><a><blockquote><img><em>');
$item['content'] = $text;
}
} else {
Debug::log('???: ' . $item['uri']);
}
return $item;
}
public function collectData(){
$feed_input = $this->getInput('feed');
if (substr($feed_input, 0, strlen(self::URI)) === self::URI) {
Debug::log('Input:: ' . $feed_input);
$feed_url = $feed_input;
} else {
/* TODO: prepend `/` if missing */
$feed_url = self::URI . '/' . $this->getInput('feed');
}
Debug::log('URL: ' . $feed_url);
$this->collectExpandableDatas($feed_url);
}
}