From 96cc1138c0cff274c6516b92c5a81c944c476b90 Mon Sep 17 00:00:00 2001
From: Andreas Kling <kling@serenityos.org>
Date: Sun, 24 May 2020 19:15:16 +0200
Subject: [PATCH] LibWeb: Remove tokenizer's premature character buffering
 optimization

---
 Libraries/LibWeb/Parser/HTMLTokenizer.cpp | 14 ++------------
 Libraries/LibWeb/Parser/HTMLTokenizer.h   |  1 -
 2 files changed, 2 insertions(+), 13 deletions(-)

diff --git a/Libraries/LibWeb/Parser/HTMLTokenizer.cpp b/Libraries/LibWeb/Parser/HTMLTokenizer.cpp
index a5680b7a7e..5c46e36249 100644
--- a/Libraries/LibWeb/Parser/HTMLTokenizer.cpp
+++ b/Libraries/LibWeb/Parser/HTMLTokenizer.cpp
@@ -134,10 +134,9 @@ Optional<HTMLToken> HTMLTokenizer::next_token()
                 }
                 ANYTHING_ELSE
                 {
-                    if (m_current_token.type() != HTMLToken::Type::Character)
-                        create_new_token(HTMLToken::Type::Character);
+                    create_new_token(HTMLToken::Type::Character);
                     m_current_token.m_comment_or_character.data.append(current_input_character.value());
-                    continue;
+                    return m_current_token;
                 }
             }
             END_STATE
@@ -749,7 +748,6 @@ bool HTMLTokenizer::next_few_characters_are(const StringView& string) const
 
 void HTMLTokenizer::create_new_token(HTMLToken::Type type)
 {
-    flush_current_character_or_comment_if_needed();
     m_current_token = {};
     m_current_token.m_type = type;
 }
@@ -761,7 +759,6 @@ HTMLTokenizer::HTMLTokenizer(const StringView& input)
 
 void HTMLTokenizer::will_switch_to([[maybe_unused]] State new_state)
 {
-    flush_current_character_or_comment_if_needed();
 #ifdef TOKENIZER_TRACE
     dbg() << "[" << state_name(m_state) << "] Switch to " << state_name(new_state);
 #endif
@@ -769,16 +766,9 @@ void HTMLTokenizer::will_switch_to([[maybe_unused]] State new_state)
 
 void HTMLTokenizer::will_reconsume_in([[maybe_unused]] State new_state)
 {
-    flush_current_character_or_comment_if_needed();
 #ifdef TOKENIZER_TRACE
     dbg() << "[" << state_name(m_state) << "] Reconsume in " << state_name(new_state);
 #endif
 }
 
-void HTMLTokenizer::flush_current_character_or_comment_if_needed()
-{
-    //if (m_current_token.type() == HTMLToken::Type::Character || m_current_token.type() == HTMLToken::Type::Comment)
-    //        emit_current_token();
-}
-
 }
diff --git a/Libraries/LibWeb/Parser/HTMLTokenizer.h b/Libraries/LibWeb/Parser/HTMLTokenizer.h
index 17b795baa5..b01b0e200e 100644
--- a/Libraries/LibWeb/Parser/HTMLTokenizer.h
+++ b/Libraries/LibWeb/Parser/HTMLTokenizer.h
@@ -147,7 +147,6 @@ private:
 
     void will_switch_to(State);
     void will_reconsume_in(State);
-    void flush_current_character_or_comment_if_needed();
 
     State m_state { State::Data };
     State m_return_state { State::Data };