LibWeb: Use Content-Type header to set document encoding

Co-authored-by: Shannon Booth <shannon@serenityos.org>

Userland/Libraries/LibWeb/DOM/DocumentLoading.cpp

@@ -91,9 +91,9 @@ static WebIDL::ExceptionOr<JS::NonnullGCPtr<DOM::Document>> load_html_document(H
     //    causes a load event to be fired.
     else {
         // FIXME: Parse as we receive the document data, instead of waiting for the whole document to be fetched first.
-        auto process_body = JS::create_heap_function(document->heap(), [document, url = navigation_params.response->url().value()](ByteBuffer data) {
-            Platform::EventLoopPlugin::the().deferred_invoke([document = document, data = move(data), url = url] {
-                auto parser = HTML::HTMLParser::create_with_uncertain_encoding(document, data);
+        auto process_body = JS::create_heap_function(document->heap(), [document, url = navigation_params.response->url().value(), mime_type = navigation_params.response->header_list()->extract_mime_type()](ByteBuffer data) {
+            Platform::EventLoopPlugin::the().deferred_invoke([document = document, data = move(data), url = url, mime_type] {
+                auto parser = HTML::HTMLParser::create_with_uncertain_encoding(document, data, mime_type);
                 parser->run(url);
             });
         });

Userland/Libraries/LibWeb/HTML/Parser/HTMLParser.cpp

@@ -4373,11 +4373,11 @@ JS::NonnullGCPtr<HTMLParser> HTMLParser::create_for_scripting(DOM::Document& doc
     return document.heap().allocate_without_realm<HTMLParser>(document);
 }
 
-JS::NonnullGCPtr<HTMLParser> HTMLParser::create_with_uncertain_encoding(DOM::Document& document, ByteBuffer const& input)
+JS::NonnullGCPtr<HTMLParser> HTMLParser::create_with_uncertain_encoding(DOM::Document& document, ByteBuffer const& input, Optional<MimeSniff::MimeType> maybe_mime_type)
 {
     if (document.has_encoding())
         return document.heap().allocate_without_realm<HTMLParser>(document, input, document.encoding().value().to_byte_string());
-    auto encoding = run_encoding_sniffing_algorithm(document, input);
+    auto encoding = run_encoding_sniffing_algorithm(document, input, maybe_mime_type);
     dbgln_if(HTML_PARSER_DEBUG, "The encoding sniffing algorithm returned encoding '{}'", encoding);
     return document.heap().allocate_without_realm<HTMLParser>(document, input, encoding);
 }

Userland/Libraries/LibWeb/HTML/Parser/HTMLParser.h

@@ -12,6 +12,7 @@
 #include <LibWeb/HTML/Parser/HTMLTokenizer.h>
 #include <LibWeb/HTML/Parser/ListOfActiveFormattingElements.h>
 #include <LibWeb/HTML/Parser/StackOfOpenElements.h>
+#include <LibWeb/MimeSniff/MimeType.h>
 
 namespace Web::HTML {
 
@@ -50,7 +51,7 @@ public:
     ~HTMLParser();
 
     static JS::NonnullGCPtr<HTMLParser> create_for_scripting(DOM::Document&);
-    static JS::NonnullGCPtr<HTMLParser> create_with_uncertain_encoding(DOM::Document&, ByteBuffer const& input);
+    static JS::NonnullGCPtr<HTMLParser> create_with_uncertain_encoding(DOM::Document&, ByteBuffer const& input, Optional<MimeSniff::MimeType> maybe_mime_type = {});
     static JS::NonnullGCPtr<HTMLParser> create(DOM::Document&, StringView input, StringView encoding);
 
     void run(HTMLTokenizer::StopAtInsertionPoint = HTMLTokenizer::StopAtInsertionPoint::No);