shihaam/karudhaas
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Pattern fix

Browse Source
This commit is contained in:
Mohamed Jinas
2020-10-01 01:56:48 +05:00
parent 3a2ce441ce
commit 3c6c2a0b21
1 changed files with 21 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
app/Services/Scrapers/DhuvasScraper.php
View File

@@ -27,21 +27,34 @@ class DhuvasScraper
$this->title = $node->text(); $this->title = $node->text();
}); });
$crawler->filter('.elementor-widget-container > p')->each(function ($node) { // Checking for two types of article patterns
$this->content[] = preg_replace("/[a-zA-Z]/","",$node->text()); if ($crawler->filter('.elementor-widget-container > div.normal_text_dv')->count() > 0) {
}); $crawler->filter('.elementor-widget-container > div.normal_text_dv, div.normal_text_dv > p')->each(function ($node) {
if($crawler->filter('span[class*="elementor-icon-list-text elementor-post-info__item elementor-post-info__item--type-author"]')->count() == 1) // Check if any alphabet exist. (To ignore the ads in articles)
{ if (preg_match('/[a-zA-Z]/', $node->text())) {
return;
}
$this->content[] = preg_replace("/[a-zA-Z]/", "", $node->text());
});
} else {
$crawler->filter('.elementor-widget-container > p')->each(function ($node) {
$this->content[] = preg_replace("/[a-zA-Z]/", "", $node->text());
});
}
if ($crawler->filter('span[class*="elementor-icon-list-text elementor-post-info__item elementor-post-info__item--type-author"]')->count() == 1) {
$this->author = $crawler->filter('span[class*="elementor-icon-list-text elementor-post-info__item elementor-post-info__item--type-author"]')->first()->text(); $this->author = $crawler->filter('span[class*="elementor-icon-list-text elementor-post-info__item elementor-post-info__item--type-author"]')->first()->text();
} }
//Remove all the alphabets from string //Remove all the alphabets from string
//preg_replace("/[a-zA-Z]/", "",$string); //preg_replace("/[a-zA-Z]/", "",$string);
return [ return [
'source' => 'Dhuvas', 'source' => 'Dhuvas',
'title' => $this->title, 'title' => $this->title,
'og_title' => $crawler->filter('meta[property*="og:title"]')->first()->attr('content'), 'og_title' => $crawler->filter('meta[property*="og:title"]')->first()->attr('content'),
@@ -49,7 +62,7 @@ class DhuvasScraper
'content' => $this->content, 'content' => $this->content,
'url' => $url, 'url' => $url,
'date' => Carbon::parse($date)->format("Y-m-d H:i:s"), 'date' => Carbon::parse($date)->format("Y-m-d H:i:s"),
'guid' => str_replace("https://dhuvas.mv/","",$url), 'guid' => str_replace("https://dhuvas.mv/", "", $url),
'author' => $this->author, 'author' => $this->author,
'topics' => [ 'topics' => [
[ [