add parser arriah.ru

2022-08-18 01:49:10 +03:00
parent cb7c116251
commit a13b193d5b
3 changed files with 221 additions and 0 deletions
--- a/arriah.php
+++ b/arriah.php
@@ -0,0 +1,127 @@
+<?php
+
+
+
+include_once '/vhosts/beta/wp-load.php';
+
+arriah_parser();
+
+
+function arriah_parser(){
+
+    global $wpdb;
+   
+   
+    $feed = 'https://new.arriah.ru/rss.php';
+
+
+    $arrContextOptions=array(
+        "ssl"=>array(
+            "verify_peer"=>false,
+            "verify_peer_name"=>false,
+        ),
+    );  
+   
+    $xml = file_get_contents($feed, false, stream_context_create($arrContextOptions));   
+
+    $result = array();
+    $res = array();
+
+    $xml = simplexml_load_string($xml);
+    $n_old = 0;
+    $nn = 0;
+
+    foreach ($xml->xpath('//item') as $item){
+
+        $link = (string)$item->link;
+        $link = trim($link, '/');
+        $f_end = strrpos($link,'/');
+        
+        if ($f_end){
+            $code = substr($link, $f_end+1);
+        }
+        else
+            continue; 
+
+
+        if (isset($find_news['ID'])){
+            $n_old++;
+            continue;
+        }
+        else {
+            
+            $date = (string)$item->pubDate;
+            $mysqldate = date('Y-m-d H:i:s',strtotime($date));
+
+            $title = (string)$item->title;
+            $title = wp_strip_all_tags($title);
+            $text = (string)$item->description;
+            $text = str_replace('src="/local/', 'src="https://new.arriah.ru/local/', $text);
+            //$text = replace_br($text);
+            //$text = delete_p($text);
+            $text = str_replace('<br>', '<br><br>', $text);
+            $slug = express_slug(vij_slug($title));
+     
+            $sql = "SELECT `ID` from `wp_posts` WHERE `post_name` = '$slug' and `post_date` = '$mysqldate'";
+            $result = $wpdb->get_results($sql);
+            
+            if ($result) {
+                continue;
+            }
+                        
+
+            $my_post = [
+                'post_title'    => $title,
+                'post_name' => $slug,
+                'post_type' => 'post',
+                'post_content'  => $text,
+                'post_date' => $mysqldate,
+                'post_status'   => 'publish',
+                'post_author'   => 1,
+                'comment_status' => 'closed',
+                'post_category' => array( 37 )
+            ];
+            
+            wp_insert_post( $my_post); 
+            
+        }  
+
+    }
+
+  return True;
+
+}
+
+
+function file_get_contents_curl( $url ) {
+
+    $ch = curl_init();
+  
+    curl_setopt( $ch, CURLOPT_AUTOREFERER, TRUE );
+    curl_setopt( $ch, CURLOPT_HEADER, 0 );
+    curl_setopt( $ch, CURLOPT_RETURNTRANSFER, 1 );
+    curl_setopt( $ch, CURLOPT_URL, $url );
+    curl_setopt( $ch, CURLOPT_FOLLOWLOCATION, TRUE );
+  
+    $data = curl_exec( $ch );
+    curl_close( $ch );
+  
+    return $data;
+  
+}
+
+
+
+function replace_br($html) {
+       
+    $html = preg_replace('#(?:<br\s*/?>\s*?){2,}#', '</p><p>', $html);
+    return "<p>$html</p>";
+
+}
+
+function delete_p($html){
+
+    $pattern = "/<p[^>]*><\\/p[^>]*>/"; 
+    return preg_replace($pattern, '', $html); 
+
+}