mirror/RSS-Bridge.rss-bridge
1
0
Fork 0
mirror of https://github.com/RSS-Bridge/rss-bridge.git synced 2025-07-24 20:55:11 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Adopt WebDriverAbstract as a solution for active (JavaScript) websites (#3971)

Browse source 
* first working version

---------

Co-authored-by: Dag <me@dvikan.no>
This commit is contained in:
hleskien 2024-02-10 04:42:22 +01:00 • committed by GitHub
parent ff7840d60f
commit 8e8028b786
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 473 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

73
bridges/ScalableCapitalBlogBridge.php Normal file
View file

@ -0,0 +1,73 @@
<?php
use Facebook\WebDriver\WebDriverBy;
use Facebook\WebDriver\WebDriverExpectedCondition;
class ScalableCapitalBlogBridge extends WebDriverAbstract
{
const NAME = 'Scalable Capital Blog';
const URI = 'https://de.scalable.capital/blog';
const DESCRIPTION = 'Alle Artikel';
const MAINTAINER = 'hleskien';
/**
* Adds accept language german to the Chrome Options.
*
* @return Facebook\WebDriver\Chrome\ChromeOptions
*/
protected function getBrowserOptions()
{
$chromeOptions = parent::getBrowserOptions();
$chromeOptions->addArguments(['--accept-lang=de']);
return $chromeOptions;
}
/**
* Puts the content of the first page into the $items array.
*
* @throws Facebook\WebDriver\Exception\NoSuchElementException
* @throws Facebook\WebDriver\Exception\TimeoutException
*/
public function collectData()
{
parent::collectData();
try {
// wait until last item is loaded
$this->getDriver()->wait()->until(WebDriverExpectedCondition::visibilityOfElementLocated(
WebDriverBy::xpath('//div[contains(@class, "articles")]//div[@class="items"]//div[contains(@class, "item")][15]')
));
$this->setIcon($this->getDriver()->findElement(WebDriverBy::xpath('//link[@rel="shortcut icon"]'))->getAttribute('href'));
$items = $this->getDriver()->findElements(WebDriverBy::xpath('//div[contains(@class, "articles")]//div[@class="items"]//div[contains(@class, "item")]'));
foreach ($items as $item) {
$feedItem = new FeedItem();
$feedItem->setEnclosures(['https://de.scalable.capital' . $item->findElement(WebDriverBy::tagName('img'))->getAttribute('src')]);
$heading = $item->findElement(WebDriverBy::tagName('a'));
$feedItem->setTitle($heading->getText());
$feedItem->setURI('https://de.scalable.capital' . $heading->getAttribute('href'));
$feedItem->setContent($item->findElement(WebDriverBy::xpath('.//div[@class="summary"]'))->getText());
$date = $item->findElement(WebDriverBy::xpath('.//div[@class="published-date"]'))->getText();
$feedItem->setTimestamp($this->formatItemTimestamp($date));
$feedItem->setAuthor($item->findElement(WebDriverBy::xpath('.//div[@class="author"]'))->getText());
$this->items[] = $feedItem;
}
} finally {
$this->cleanUp();
}
}
/**
* Converts the given date (dd.mm.yyyy) into a timestamp.
*
* @param $value string
* @return int
*/
protected function formatItemTimestamp($value)
{
$formatter = new IntlDateFormatter('de', IntlDateFormatter::LONG, IntlDateFormatter::NONE);
return $formatter->parse($value);
}
}