Fixing #62625

ThomasLandauer · ThomasLandauer · commit 0c5e8f20c74c · 2025-12-06T16:47:47.000+01:00
diff --git a/Crawler.php b/Crawler.php
@@ -1110,7 +1110,11 @@ private function parseXhtml(string $htmlContent, string $charset = 'UTF-8'): \DO
 
         $internalErrors = libxml_use_internal_errors(true);
 
-        $document = \Dom\HTMLDocument::createFromString($htmlContent, \Dom\HTML_NO_DEFAULT_NS, $charset);
+        try {
+            $document = \Dom\HTMLDocument::createFromString($htmlContent, \Dom\HTML_NO_DEFAULT_NS, $charset);
+        } catch (\ValueError) {
+            $document = \Dom\HTMLDocument::createFromString($htmlContent, \Dom\HTML_NO_DEFAULT_NS);
+        }
 
         libxml_use_internal_errors($internalErrors);
 
diff --git a/Tests/CrawlerTest.php b/Tests/CrawlerTest.php
@@ -1335,7 +1335,7 @@ public function testAddXmlContentWithErrors()
     #[Group('legacy')]
     public function testHtml5ParserNotSameAsNativeParserForSpecificHtml()
     {
-        // Html who create a bug specific to the DOM extension (see https://github.com/symfony/symfony/issues/28596)
+        // HTML that creates a bug specific to the DOM extension (see https://github.com/symfony/symfony/issues/28596)
         $html = '<!DOCTYPE html><html><body><h1><p>Foo</p></h1></body></html>';
 
         $html5Crawler = new Crawler(null, null, null, true);
@@ -1396,6 +1396,14 @@ public function testAlpineJs()
         $this->assertCount(3, $crawler->filterXPath('//div'));
     }
 
+    public function testInvalidCharset()
+    {
+        $email = "Content-Type: text/html; charset=foobar\n\n<!DOCTYPE html><html><body>One Two Three</body></html>";
+        $crawler = $this->createCrawler($email);
+        // Not really needed anymore, since the test would already have crashed with: ValueError: Dom\HTMLDocument::createFromString(): Argument #3 ($overrideEncoding) must be a valid document encoding
+        $this->assertSame('Content-Type: text/html; charset=foobar One Two Three', $crawler->text());
+    }
+
     protected function createTestCrawler($uri = null)
     {
         $html = $this->getDoctype().'
