From 3c6c2a0b2149b59e774cf409f0909ad616a54221 Mon Sep 17 00:00:00 2001
From: Mohamed Jinas <j@live.mv>
Date: Thu, 1 Oct 2020 01:56:48 +0500
Subject: [PATCH] Pattern fix

---
 app/Services/Scrapers/DhuvasScraper.php | 29 ++++++++++++++++++-------
 1 file changed, 21 insertions(+), 8 deletions(-)

diff --git a/app/Services/Scrapers/DhuvasScraper.php b/app/Services/Scrapers/DhuvasScraper.php
index 2428a76..1a0d74a 100644
--- a/app/Services/Scrapers/DhuvasScraper.php
+++ b/app/Services/Scrapers/DhuvasScraper.php
@@ -27,21 +27,34 @@ class DhuvasScraper
             $this->title = $node->text();
         });
 
-        $crawler->filter('.elementor-widget-container > p')->each(function ($node) {
-            $this->content[] = preg_replace("/[a-zA-Z]/","",$node->text());
-        });
+        // Checking for two types of article patterns
+        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();
         }
 
 
-     
+
 
         //Remove all the alphabets from string
         //preg_replace("/[a-zA-Z]/", "",$string);
-       return [
+        return [
             'source'    => 'Dhuvas',
             'title'      => $this->title,
             'og_title'   => $crawler->filter('meta[property*="og:title"]')->first()->attr('content'),
@@ -49,7 +62,7 @@ class DhuvasScraper
             'content'    => $this->content,
             'url'        => $url,
             'date'       => Carbon::parse($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,
             'topics'       =>  [
                 [