From 9677d8eeac2dc2bfd96d634b178e05a28aaff27b Mon Sep 17 00:00:00 2001
From: Idan Horowitz <idan.horowitz@gmail.com>
Date: Fri, 17 Nov 2023 15:14:07 +0200
Subject: [PATCH] LibWeb: Reject improperly encoded XML documents as not
 well-formed

---
 Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp b/Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp
index bcc22a0df9..be680b8d50 100644
--- a/Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp
+++ b/Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp
@@ -159,6 +159,9 @@ bool build_xml_document(DOM::Document& document, ByteBuffer const& data, Optiona
         decoder = TextCodec::decoder_for(encoding);
     }
     VERIFY(decoder.has_value());
+    // Well-formed XML documents contain only properly encoded characters
+    if (!decoder->validate(data))
+        return false;
     auto source = decoder->to_utf8(data).release_value_but_fixme_should_propagate_errors();
     XML::Parser parser(source, { .resolve_external_resource = resolve_xml_resource });
     XMLDocumentBuilder builder { document };