LibWeb: Remove StringBuilder from HTMLToken::m_comment_or_character

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

@@ -317,7 +317,7 @@ void HTMLDocumentParser::handle_initial(HTMLToken& token)
     }
 
     if (token.is_comment()) {
-        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
+        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data));
         document().append_child(move(comment));
         return;
     }
@@ -347,7 +347,7 @@ void HTMLDocumentParser::handle_before_html(HTMLToken& token)
     }
 
     if (token.is_comment()) {
-        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
+        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data));
         document().append_child(move(comment));
         return;
     }
@@ -520,7 +520,7 @@ AnythingElse:
 
 void HTMLDocumentParser::insert_comment(HTMLToken& token)
 {
-    auto data = token.m_comment_or_character.data.to_string();
+    auto data = token.m_comment_or_character.data;
     auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
     adjusted_insertion_location.parent->insert_before(adopt_ref(*new DOM::Comment(document(), data)), adjusted_insertion_location.insert_before_sibling);
 }
@@ -832,7 +832,7 @@ void HTMLDocumentParser::handle_after_body(HTMLToken& token)
     }
 
     if (token.is_comment()) {
-        auto data = token.m_comment_or_character.data.to_string();
+        auto data = token.m_comment_or_character.data;
         auto& insertion_location = m_stack_of_open_elements.first();
         insertion_location.append_child(adopt_ref(*new DOM::Comment(document(), data)));
         return;
@@ -870,7 +870,7 @@ void HTMLDocumentParser::handle_after_body(HTMLToken& token)
 void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
 {
     if (token.is_comment()) {
-        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
+        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data));
         document().append_child(move(comment));
         return;
     }
@@ -2751,7 +2751,7 @@ void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
 void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
 {
     if (token.is_comment()) {
-        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
+        auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data));
         document().append_child(move(comment));
         return;
     }

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

@@ -53,7 +53,7 @@ String HTMLToken::to_string() const
 
     if (type() == HTMLToken::Type::Comment || type() == HTMLToken::Type::Character) {
         builder.append(" { data: '");
-        builder.append(m_comment_or_character.data.to_string());
+        builder.append(m_comment_or_character.data);
         builder.append("' }");
     }
 

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

@@ -8,7 +8,6 @@
 
 #include <AK/FlyString.h>
 #include <AK/String.h>
-#include <AK/StringBuilder.h>
 #include <AK/Types.h>
 #include <AK/Utf8View.h>
 #include <AK/Vector.h>
@@ -34,7 +33,10 @@ public:
     {
         HTMLToken token;
         token.m_type = Type::Character;
-        token.m_comment_or_character.data.append(code_point);
+        StringBuilder builder;
+        // FIXME: This narrows code_point to char, should this be append_code_point() instead?
+        builder.append(code_point);
+        token.m_comment_or_character.data = builder.to_string();
         return token;
     }
 
@@ -56,7 +58,7 @@ public:
     u32 code_point() const
     {
         VERIFY(is_character());
-        Utf8View view(m_comment_or_character.data.string_view());
+        Utf8View view(m_comment_or_character.data);
         VERIFY(view.length() == 1);
         return *view.begin();
     }
@@ -209,7 +211,7 @@ private:
     // Type::Comment
     // Type::Character
     struct {
-        StringBuilder data;
+        String data;
     } m_comment_or_character;
 
     Position m_start_position;

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

@@ -74,17 +74,18 @@ namespace Web::HTML {
         goto new_state;                                                 \
     } while (0)
 
-#define FLUSH_CODEPOINTS_CONSUMED_AS_A_CHARACTER_REFERENCE                                           \
-    do {                                                                                             \
-        for (auto code_point : m_temporary_buffer) {                                                 \
-            if (consumed_as_part_of_an_attribute()) {                                                \
-                m_current_builder.append_code_point(code_point);                                     \
-            } else {                                                                                 \
-                create_new_token(HTMLToken::Type::Character);                                        \
-                m_current_token.m_comment_or_character.data.append_code_point(code_point);           \
-                m_queued_tokens.enqueue(m_current_token);                                            \
-            }                                                                                        \
-        }                                                                                            \
+#define FLUSH_CODEPOINTS_CONSUMED_AS_A_CHARACTER_REFERENCE                               \
+    do {                                                                                 \
+        for (auto code_point : m_temporary_buffer) {                                     \
+            if (consumed_as_part_of_an_attribute()) {                                    \
+                m_current_builder.append_code_point(code_point);                         \
+            } else {                                                                     \
+                create_new_token(HTMLToken::Type::Character);                            \
+                m_current_builder.append_code_point(code_point);                         \
+                m_current_token.m_comment_or_character.data = consume_current_builder(); \
+                m_queued_tokens.enqueue(m_current_token);                                \
+            }                                                                            \
+        }                                                                                \
     } while (0)
 
 #define DONT_CONSUME_NEXT_INPUT_CHARACTER \
@@ -139,12 +140,13 @@ namespace Web::HTML {
         return m_queued_tokens.dequeue();         \
     } while (0)
 
-#define EMIT_CHARACTER(code_point)                                                 \
-    do {                                                                           \
-        create_new_token(HTMLToken::Type::Character);                              \
-        m_current_token.m_comment_or_character.data.append_code_point(code_point); \
-        m_queued_tokens.enqueue(m_current_token);                                  \
-        return m_queued_tokens.dequeue();                                          \
+#define EMIT_CHARACTER(code_point)                                               \
+    do {                                                                         \
+        create_new_token(HTMLToken::Type::Character);                            \
+        m_current_builder.append_code_point(code_point);                         \
+        m_current_token.m_comment_or_character.data = consume_current_builder(); \
+        m_queued_tokens.enqueue(m_current_token);                                \
+        return m_queued_tokens.dequeue();                                        \
     } while (0)
 
 #define EMIT_CURRENT_CHARACTER \
@@ -402,6 +404,7 @@ _StartOfFunction:
             {
                 ON('>')
                 {
+                    m_current_token.m_comment_or_character.data = consume_current_builder();
                     SWITCH_TO_AND_EMIT_CURRENT_TOKEN(Data);
                 }
                 ON_EOF
@@ -412,12 +415,12 @@ _StartOfFunction:
                 ON(0)
                 {
                     log_parse_error();
-                    m_current_token.m_comment_or_character.data.append_code_point(0xFFFD);
+                    m_current_builder.append_code_point(0xFFFD);
                     continue;
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append_code_point(current_input_character.value());
+                    m_current_builder.append_code_point(current_input_character.value());
                     continue;
                 }
             }
@@ -1346,11 +1349,12 @@ _StartOfFunction:
             {
                 ON('-')
                 {
-                    SWITCH_TO(CommentEnd);
+                    SWITCH_TO_WITH_UNCLEAN_BUILDER(CommentEnd);
                 }
                 ON('>')
                 {
                     log_parse_error();
+                    consume_current_builder();
                     SWITCH_TO_AND_EMIT_CURRENT_TOKEN(Data);
                 }
                 ON_EOF
@@ -1361,7 +1365,7 @@ _StartOfFunction:
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append('-');
+                    m_current_builder.append('-');
                     RECONSUME_IN(Comment);
                 }
             }
@@ -1371,17 +1375,17 @@ _StartOfFunction:
             {
                 ON('<')
                 {
-                    m_current_token.m_comment_or_character.data.append_code_point(current_input_character.value());
-                    SWITCH_TO(CommentLessThanSign);
+                    m_current_builder.append_code_point(current_input_character.value());
+                    SWITCH_TO_WITH_UNCLEAN_BUILDER(CommentLessThanSign);
                 }
                 ON('-')
                 {
-                    SWITCH_TO(CommentEndDash);
+                    SWITCH_TO_WITH_UNCLEAN_BUILDER(CommentEndDash);
                 }
                 ON(0)
                 {
                     log_parse_error();
-                    m_current_token.m_comment_or_character.data.append_code_point(0xFFFD);
+                    m_current_builder.append_code_point(0xFFFD);
                     continue;
                 }
                 ON_EOF
@@ -1392,7 +1396,7 @@ _StartOfFunction:
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append_code_point(current_input_character.value());
+                    m_current_builder.append_code_point(current_input_character.value());
                     continue;
                 }
             }
@@ -1402,6 +1406,7 @@ _StartOfFunction:
             {
                 ON('>')
                 {
+                    m_current_token.m_comment_or_character.data = consume_current_builder();
                     SWITCH_TO_AND_EMIT_CURRENT_TOKEN(Data);
                 }
                 ON('!')
@@ -1410,7 +1415,7 @@ _StartOfFunction:
                 }
                 ON('-')
                 {
-                    m_current_token.m_comment_or_character.data.append('-');
+                    m_current_builder.append('-');
                     continue;
                 }
                 ON_EOF
@@ -1421,7 +1426,7 @@ _StartOfFunction:
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append('-');
+                    m_current_builder.append('-');
                     RECONSUME_IN(Comment);
                 }
             }
@@ -1431,7 +1436,7 @@ _StartOfFunction:
             {
                 ON('-')
                 {
-                    m_current_token.m_comment_or_character.data.append("--!");
+                    m_current_builder.append("--!");
                     SWITCH_TO(CommentEndDash);
                 }
                 ON('>')
@@ -1447,7 +1452,7 @@ _StartOfFunction:
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append("--!");
+                    m_current_builder.append("--!");
                     RECONSUME_IN(Comment);
                 }
             }
@@ -1457,7 +1462,7 @@ _StartOfFunction:
             {
                 ON('-')
                 {
-                    SWITCH_TO(CommentEnd);
+                    SWITCH_TO_WITH_UNCLEAN_BUILDER(CommentEnd);
                 }
                 ON_EOF
                 {
@@ -1467,7 +1472,7 @@ _StartOfFunction:
                 }
                 ANYTHING_ELSE
                 {
-                    m_current_token.m_comment_or_character.data.append('-');
+                    m_current_builder.append('-');
                     RECONSUME_IN(Comment);
                 }
             }
@@ -1477,12 +1482,12 @@ _StartOfFunction:
             {
                 ON('!')
                 {
-                    m_current_token.m_comment_or_character.data.append_code_point(current_input_character.value());
-                    SWITCH_TO(CommentLessThanSignBang);
+                    m_current_builder.append_code_point(current_input_character.value());
+                    SWITCH_TO_WITH_UNCLEAN_BUILDER(CommentLessThanSignBang);
                 }
                 ON('<')
                 {
-                    m_current_token.m_comment_or_character.data.append_code_point(current_input_character.value());
+                    m_current_builder.append_code_point(current_input_character.value());
                     continue;
                 }
                 ANYTHING_ELSE

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

@@ -7,6 +7,7 @@
 #pragma once
 
 #include <AK/Queue.h>
+#include <AK/StringBuilder.h>
 #include <AK/StringView.h>
 #include <AK/Types.h>
 #include <AK/Utf8View.h>