LibWeb: Make document.write() work while document is parsing

This necessitated making HTMLParser ref-counted, and having it register
itself with Document when created. That makes it possible for scripts to
add new input at the current parser insertion point.

There is now a reference cycle between Document and HTMLParser. This
cycle is explicitly broken by calling Document::detach_parser() at the
end of HTMLParser::run().

This is a huge progression on ACID3, from 31% to 49%! :^)

Tests/LibWeb/test-web.cpp

@@ -150,10 +150,10 @@ JS_DEFINE_NATIVE_FUNCTION(TestWebGlobalObject::wait_for_page_to_load)
     loader.load_sync(
         request,
         [&](auto data, auto&, auto) {
-            Web::HTML::HTMLParser parser(document, data, "utf-8");
+            auto parser = Web::HTML::HTMLParser::create(document, data, "utf-8");
             // Now parse the HTML page.
-            parser.run(next_page_to_load.value());
-            g_page_view->set_document(&parser.document());
+            parser->run(next_page_to_load.value());
+            g_page_view->set_document(&parser->document());
             // Note: Unhandled exceptions are just dropped here.
 
             // Run the "after" hooks

Userland/Libraries/LibWeb/DOM/Document.cpp

@@ -256,7 +256,7 @@ ExceptionOr<Document*> Document::open(String const&, String const&)
     set_quirks_mode(QuirksMode::No);
 
     // 16. Create a new HTML parser and associate it with document. This is a script-created parser (meaning that it can be closed by the document.open() and document.close() methods, and that the tokenizer will wait for an explicit call to document.close() before emitting an end-of-file token). The encoding confidence is irrelevant.
-    m_parser = make<HTML::HTMLParser>(*this);
+    m_parser = HTML::HTMLParser::create_for_scripting(*this);
 
     // 17. Set the insertion point to point at just before the end of the input stream (which at this point will be empty).
     m_parser->tokenizer().update_insertion_point();
@@ -1324,4 +1324,14 @@ bool Document::has_focus() const
     return true;
 }
 
+void Document::set_parser(Badge<HTML::HTMLParser>, HTML::HTMLParser& parser)
+{
+    m_parser = parser;
+}
+
+void Document::detach_parser(Badge<HTML::HTMLParser>)
+{
+    m_parser = nullptr;
+}
+
 }

Userland/Libraries/LibWeb/DOM/Document.h

@@ -313,6 +313,9 @@ public:
 
     bool has_focus() const;
 
+    void set_parser(Badge<HTML::HTMLParser>, HTML::HTMLParser&);
+    void detach_parser(Badge<HTML::HTMLParser>);
+
 private:
     explicit Document(const AK::URL&);
 
@@ -358,7 +361,7 @@ private:
     RefPtr<Core::Timer> m_style_update_timer;
     RefPtr<Core::Timer> m_layout_update_timer;
 
-    OwnPtr<HTML::HTMLParser> m_parser;
+    RefPtr<HTML::HTMLParser> m_parser;
     bool m_active_parser_was_aborted { false };
 
     String m_source;

Userland/Libraries/LibWeb/HTML/DOMParser.cpp

@@ -34,11 +34,11 @@ NonnullRefPtr<DOM::Document> DOMParser::parse_from_string(String const& string,
         // 2. Create an HTML parser parser, associated with document.
         // 3. Place string into the input stream for parser. The encoding confidence is irrelevant.
         // FIXME: We don't have the concept of encoding confidence yet.
-        HTMLParser parser(document, string, "UTF-8");
+        auto parser = HTMLParser::create(document, string, "UTF-8");
 
         // 4. Start parser and let it run until it has consumed all the characters just inserted into the input stream.
         // FIXME: This is to match the default URL. Instead, pass in this's relevant global object's associated Document's URL.
-        parser.run("about:blank");
+        parser->run("about:blank");
     } else {
         // -> Otherwise
         // FIXME: 1. Create an XML parser parse, associated with document, and with XML scripting support disabled.

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

@@ -121,8 +121,8 @@ static bool is_html_integration_point(DOM::Element const& element)
 RefPtr<DOM::Document> parse_html_document(StringView data, const AK::URL& url, const String& encoding)
 {
     auto document = DOM::Document::create(url);
-    HTMLParser parser(document, data, encoding);
-    parser.run(url);
+    auto parser = HTMLParser::create(document, data, encoding);
+    parser->run(url);
     return document;
 }
 
@@ -131,6 +131,7 @@ HTMLParser::HTMLParser(DOM::Document& document, StringView input, const String&
     , m_document(document)
 {
     m_tokenizer.set_parser({}, *this);
+    m_document->set_parser({}, *this);
     m_document->set_should_invalidate_styles_on_attribute_changes(false);
     auto standardized_encoding = TextCodec::get_standardized_encoding(encoding);
     VERIFY(standardized_encoding.has_value());
@@ -140,6 +141,7 @@ HTMLParser::HTMLParser(DOM::Document& document, StringView input, const String&
 HTMLParser::HTMLParser(DOM::Document& document)
     : m_document(document)
 {
+    m_document->set_parser({}, *this);
     m_tokenizer.set_parser({}, *this);
 }
 
@@ -201,6 +203,7 @@ void HTMLParser::run(const AK::URL& url)
     m_document->set_source(m_tokenizer.source());
     run();
     the_end();
+    m_document->detach_parser({});
 }
 
 // https://html.spec.whatwg.org/multipage/parsing.html#the-end
@@ -3180,44 +3183,44 @@ DOM::Document& HTMLParser::document()
 NonnullRefPtrVector<DOM::Node> HTMLParser::parse_html_fragment(DOM::Element& context_element, StringView markup)
 {
     auto temp_document = DOM::Document::create();
-    HTMLParser parser(*temp_document, markup, "utf-8");
-    parser.m_context_element = context_element;
-    parser.m_parsing_fragment = true;
-    parser.document().set_quirks_mode(context_element.document().mode());
+    auto parser = HTMLParser::create(*temp_document, markup, "utf-8");
+    parser->m_context_element = context_element;
+    parser->m_parsing_fragment = true;
+    parser->document().set_quirks_mode(context_element.document().mode());
 
     if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
-        parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
+        parser->m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
     } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
-        parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
+        parser->m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
     } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
-        parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
+        parser->m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
     } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
         if (context_element.document().is_scripting_enabled())
-            parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
+            parser->m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
     } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
-        parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
+        parser->m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
     }
 
     auto root = create_element(context_element.document(), HTML::TagNames::html, Namespace::HTML);
-    parser.document().append_child(root);
-    parser.m_stack_of_open_elements.push(root);
+    parser->document().append_child(root);
+    parser->m_stack_of_open_elements.push(root);
 
     if (context_element.local_name() == HTML::TagNames::template_) {
-        parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
+        parser->m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
     }
 
     // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
 
-    parser.reset_the_insertion_mode_appropriately();
+    parser->reset_the_insertion_mode_appropriately();
 
     for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
         if (is<HTMLFormElement>(*form_candidate)) {
-            parser.m_form_element = verify_cast<HTMLFormElement>(*form_candidate);
+            parser->m_form_element = verify_cast<HTMLFormElement>(*form_candidate);
             break;
         }
     }
 
-    parser.run(context_element.document().url());
+    parser->run(context_element.document().url());
 
     NonnullRefPtrVector<DOM::Node> children;
     while (RefPtr<DOM::Node> child = root->first_child()) {
@@ -3228,13 +3231,23 @@ NonnullRefPtrVector<DOM::Node> HTMLParser::parse_html_fragment(DOM::Element& con
     return children;
 }
 
-NonnullOwnPtr<HTMLParser> HTMLParser::create_with_uncertain_encoding(DOM::Document& document, const ByteBuffer& input)
+NonnullRefPtr<HTMLParser> HTMLParser::create_for_scripting(DOM::Document& document)
+{
+    return adopt_ref(*new HTMLParser(document));
+}
+
+NonnullRefPtr<HTMLParser> HTMLParser::create_with_uncertain_encoding(DOM::Document& document, const ByteBuffer& input)
 {
     if (document.has_encoding())
-        return make<HTMLParser>(document, input, document.encoding().value());
+        return adopt_ref(*new HTMLParser(document, input, document.encoding().value()));
     auto encoding = run_encoding_sniffing_algorithm(document, input);
     dbgln("The encoding sniffing algorithm returned encoding '{}'", encoding);
-    return make<HTMLParser>(document, input, encoding);
+    return adopt_ref(*new HTMLParser(document, input, encoding));
+}
+
+NonnullRefPtr<HTMLParser> HTMLParser::create(DOM::Document& document, StringView input, String const& encoding)
+{
+    return adopt_ref(*new HTMLParser(document, input, encoding));
 }
 
 // https://html.spec.whatwg.org/multipage/parsing.html#html-fragment-serialisation-algorithm

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

@@ -41,15 +41,15 @@ namespace Web::HTML {
 
 RefPtr<DOM::Document> parse_html_document(StringView, const AK::URL&, const String& encoding);
 
-class HTMLParser {
+class HTMLParser : public RefCounted<HTMLParser> {
     friend class HTMLTokenizer;
 
 public:
-    HTMLParser(DOM::Document&, StringView input, const String& encoding);
-    HTMLParser(DOM::Document&);
     ~HTMLParser();
 
-    static NonnullOwnPtr<HTMLParser> create_with_uncertain_encoding(DOM::Document&, const ByteBuffer& input);
+    static NonnullRefPtr<HTMLParser> create_for_scripting(DOM::Document&);
+    static NonnullRefPtr<HTMLParser> create_with_uncertain_encoding(DOM::Document&, ByteBuffer const& input);
+    static NonnullRefPtr<HTMLParser> create(DOM::Document&, StringView input, String const& encoding);
 
     void run();
     void run(const AK::URL&);
@@ -76,6 +76,9 @@ public:
     size_t script_nesting_level() const { return m_script_nesting_level; }
 
 private:
+    HTMLParser(DOM::Document&, StringView input, const String& encoding);
+    HTMLParser(DOM::Document&);
+
     const char* insertion_mode_name() const;
 
     DOM::QuirksMode which_quirks_mode(const HTMLToken&) const;

Userland/Libraries/LibWeb/Loader/FrameLoader.cpp

@@ -46,8 +46,8 @@ static bool build_markdown_document(DOM::Document& document, const ByteBuffer& d
     if (!markdown_document)
         return false;
 
-    HTML::HTMLParser parser(document, markdown_document->render_to_html(), "utf-8");
-    parser.run(document.url());
+    auto parser = HTML::HTMLParser::create(document, markdown_document->render_to_html(), "utf-8");
+    parser->run(document.url());
     return true;
 }
 
@@ -116,8 +116,8 @@ static bool build_gemini_document(DOM::Document& document, const ByteBuffer& dat
     dbgln_if(GEMINI_DEBUG, "Gemini data:\n\"\"\"{}\"\"\"", gemini_data);
     dbgln_if(GEMINI_DEBUG, "Converted to HTML:\n\"\"\"{}\"\"\"", html_data);
 
-    HTML::HTMLParser parser(document, html_data, "utf-8");
-    parser.run(document.url());
+    auto parser = HTML::HTMLParser::create(document, html_data, "utf-8");
+    parser->run(document.url());
     return true;
 }
 
@@ -226,9 +226,9 @@ bool FrameLoader::load(const AK::URL& url, Type type)
 void FrameLoader::load_html(StringView html, const AK::URL& url)
 {
     auto document = DOM::Document::create(url);
-    HTML::HTMLParser parser(document, html, "utf-8");
-    parser.run(url);
-    browsing_context().set_active_document(&parser.document());
+    auto parser = HTML::HTMLParser::create(document, html, "utf-8");
+    parser->run(url);
+    browsing_context().set_active_document(&parser->document());
 }
 
 // FIXME: Use an actual templating engine (our own one when it's built, preferably