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

Update AvasScraper.php

Browse Source
This commit is contained in:
Mohamed Jinas
2020-09-24 13:43:00 +05:00
parent 4c72042aeb
commit 94fda48b90
1 changed files with 17 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
app/Services/Scrapers/AvasScraper.php
View File

@@ -1,5 +1,4 @@
<?php
namespace App\Services\Scrapers;
use Goutte\Client;
@@ -11,8 +10,9 @@ class AvasScraper
protected $content;
protected $author;
protected $image = "images/noimg.jpg";
protected $topics = [];
public function __construct()
{
$this->client = new Client;
@@ -25,18 +25,20 @@ class AvasScraper
$title = $crawler->filter('h1')->first()->text();
$image = ($crawler->filter('figure img')->first()->attr('src')) ? $crawler->filter('figure img')->first()->attr('src') : "/images/noimg.jpg";
if($crawler->filter('figure img')->count() == 1)
{
$this->image = $crawler->filter('figure img')->first()->attr('src');
}
$crawler->filter('.post_content p')->each(function ($node) {
$this->content[] = preg_replace("/[a-zA-Z]/", "", $node->text());
$this->content[] = preg_replace("/[a-zA-Z]/","",$node->text());
});
$crawler->filter('div[class*="border-t border-grey-light border-dotted mt-7 py-3"] a')->each(function ($node) {
//Removing the show more tags button
if ($node->text() == "+") {
if($node->text() == "+")
{
return;
}
$this->topics[] = [
@@ -45,26 +47,27 @@ class AvasScraper
];
});
if ($crawler->filter('div[class*="font-waheed text-grey ml-3 pl-3 text-lg border-l border-grey border-dotted"] a')->count() == 1) {
if($crawler->filter('div[class*="font-waheed text-grey ml-3 pl-3 text-lg border-l border-grey border-dotted"] a')->count() == 1)
{
$this->author = $crawler->filter('div[class*="font-waheed text-grey ml-3 pl-3 text-lg border-l border-grey border-dotted"] a')->first()->text();
}
//Remove all the alphabets from string
//preg_replace("/[a-zA-Z]/", "",$string);
return [
return [
'source' => 'Avas',
'title' => $title,
'og_title' => $crawler->filter('meta[property*="og:title"]')->first()->attr('content'),
'image' => $image,
'image' => $this->image,
'content' => $this->content,
'url' => $url,
'date' => Carbon::parse($crawler->filter('timeago')->first()->attr('datetime'))->format("Y-m-d H:i:s"),
'guid' => str_replace("https://avas.mv/", "", $url),
'guid' => str_replace("https://avas.mv/","",$url),
'author' => $this->author,
'topics' => $this->topics
];
}
}
}