Update Cleaner.php

liorphp · web-flow · commit 667b83aad2e0 · 2025-02-24T23:28:18.000+02:00
diff --git a/src/PHPHtmlParser/Dom/Cleaner.php b/src/PHPHtmlParser/Dom/Cleaner.php
@@ -10,6 +10,9 @@
 
 class Cleaner implements CleanerInterface
 {
+    mb_regex_encoding('UTF-8');
+    mb_regex_set_options('m');
+
     /**
      * Cleans the html of any none-html information.
      *
@@ -74,7 +77,7 @@ public function clean(string $str, Options $options, string $defaultCharset): st
 
         // strip out <script> tags
         if ($options->isRemoveScripts()) {
-            $str = \mb_eregi_replace("<\s*script[^>]*[^/]>(.*?)<\s*/\s*script\s*>", '', $str);
+            $str = preg_replace("/<\s*script[^>]*[^\/]>(.*?)<\s*\/\s*script\s*>/is", '', $str);
             if ($str === false) {
                 throw new LogicalException('mb_eregi_replace returned false instead of a string. Error when attempting to remove scripts 1.');
             }

Original file line number	Diff line number	Diff line change
`@@ -10,6 +10,9 @@`
`10`	`10`
`11`	`11`	`class Cleaner implements CleanerInterface`
`12`	`12`	`{`
	`13`	`+ mb_regex_encoding('UTF-8');`
	`14`	`+ mb_regex_set_options('m');`
	`15`	`+`
`13`	`16`	`/**`
`14`	`17`	`* Cleans the html of any none-html information.`
`15`	`18`	`*`
`@@ -74,7 +77,7 @@ public function clean(string $str, Options $options, string $defaultCharset): st`
`74`	`77`
`75`	`78`	`// strip out <script> tags`
`76`	`79`	`if ($options->isRemoveScripts()) {`
`77`		`- $str = \mb_eregi_replace("<\sscript[^>][^/]>(.?)<\s/\sscript\s>", '', $str);`
	`80`	`+ $str = preg_replace("/<\sscript[^>][^\/]>(.?)<\s\/\sscript\s>/is", '', $str);`
`78`	`81`	`if ($str === false) {`
`79`	`82`	`throw new LogicalException('mb_eregi_replace returned false instead of a string. Error when attempting to remove scripts 1.');`
`80`	`83`	`}`