From 1dd4e2dc87843646d5bea67fd351927699b5eb3e Mon Sep 17 00:00:00 2001
From: Andreas Kling <kling@serenityos.org>
Date: Thu, 15 Sep 2022 13:51:30 +0200
Subject: [PATCH] LibWeb: Cache lowercased names in SimpleSelector

When matching selectors in HTML documents, we know that all the elements
have lowercase local names already (the parser makes sure of this.)

Style sheets still need to remember the original name strings, in case
we want to match against non-HTML content like XML/SVG. To make the
common HTML case faster, we now cache a lowercase version of the name
with each type/class/id SimpleSelector.

This makes tag type checks O(1) instead of O(n).
---
 .../Libraries/LibWeb/CSS/Parser/Parser.cpp    |  6 +++---
 Userland/Libraries/LibWeb/CSS/Selector.h      | 20 ++++++++++++++++---
 .../Libraries/LibWeb/CSS/SelectorEngine.cpp   |  6 +++---
 Userland/Libraries/LibWeb/Dump.cpp            |  2 +-
 4 files changed, 24 insertions(+), 10 deletions(-)

diff --git a/Userland/Libraries/LibWeb/CSS/Parser/Parser.cpp b/Userland/Libraries/LibWeb/CSS/Parser/Parser.cpp
index f29328c723..754182d5bb 100644
--- a/Userland/Libraries/LibWeb/CSS/Parser/Parser.cpp
+++ b/Userland/Libraries/LibWeb/CSS/Parser/Parser.cpp
@@ -695,7 +695,7 @@ Parser::ParseErrorOr<Optional<Selector::SimpleSelector>> Parser::parse_simple_se
             }
             return Selector::SimpleSelector {
                 .type = Selector::SimpleSelector::Type::Class,
-                .value = FlyString { class_name_value.token().ident() }
+                .value = Selector::SimpleSelector::Name { class_name_value.token().ident() }
             };
         }
         case '>':
@@ -719,13 +719,13 @@ Parser::ParseErrorOr<Optional<Selector::SimpleSelector>> Parser::parse_simple_se
         }
         return Selector::SimpleSelector {
             .type = Selector::SimpleSelector::Type::Id,
-            .value = FlyString { first_value.token().hash_value() }
+            .value = Selector::SimpleSelector::Name { first_value.token().hash_value() }
         };
     }
     if (first_value.is(Token::Type::Ident)) {
         return Selector::SimpleSelector {
             .type = Selector::SimpleSelector::Type::TagName,
-            .value = FlyString { first_value.token().ident() }
+            .value = Selector::SimpleSelector::Name { first_value.token().ident() }
         };
     }
     if (first_value.is_block() && first_value.block().is_square())
diff --git a/Userland/Libraries/LibWeb/CSS/Selector.h b/Userland/Libraries/LibWeb/CSS/Selector.h
index 4d9a1c0ee2..14ebcdadca 100644
--- a/Userland/Libraries/LibWeb/CSS/Selector.h
+++ b/Userland/Libraries/LibWeb/CSS/Selector.h
@@ -142,8 +142,19 @@ public:
             CaseType case_type;
         };
 
+        struct Name {
+            Name(FlyString n)
+                : name(move(n))
+                , lowercase_name(name.to_lowercase())
+            {
+            }
+
+            FlyString name;
+            FlyString lowercase_name;
+        };
+
         Type type;
-        Variant<Empty, Attribute, PseudoClass, PseudoElement, FlyString> value {};
+        Variant<Empty, Attribute, PseudoClass, PseudoElement, Name> value {};
 
         Attribute const& attribute() const { return value.get<Attribute>(); }
         Attribute& attribute() { return value.get<Attribute>(); }
@@ -151,8 +162,11 @@ public:
         PseudoClass& pseudo_class() { return value.get<PseudoClass>(); }
         PseudoElement const& pseudo_element() const { return value.get<PseudoElement>(); }
         PseudoElement& pseudo_element() { return value.get<PseudoElement>(); }
-        FlyString const& name() const { return value.get<FlyString>(); }
-        FlyString& name() { return value.get<FlyString>(); }
+
+        FlyString const& name() const { return value.get<Name>().name; }
+        FlyString& name() { return value.get<Name>().name; }
+        FlyString const& lowercase_name() const { return value.get<Name>().lowercase_name; }
+        FlyString& lowercase_name() { return value.get<Name>().lowercase_name; }
 
         String serialize() const;
     };
diff --git a/Userland/Libraries/LibWeb/CSS/SelectorEngine.cpp b/Userland/Libraries/LibWeb/CSS/SelectorEngine.cpp
index dfdf426bf8..265b8df8aa 100644
--- a/Userland/Libraries/LibWeb/CSS/SelectorEngine.cpp
+++ b/Userland/Libraries/LibWeb/CSS/SelectorEngine.cpp
@@ -353,9 +353,9 @@ static inline bool matches(CSS::Selector::SimpleSelector const& component, DOM::
         return element.has_class(component.name());
     case CSS::Selector::SimpleSelector::Type::TagName:
         // See https://html.spec.whatwg.org/multipage/semantics-other.html#case-sensitivity-of-selectors
-        if (is<HTML::HTMLElement>(element) && element.document().document_type() != DOM::Document::Type::XML)
-            return component.name().equals_ignoring_case(element.local_name());
-        return component.name() == element.local_name();
+        if (element.document().document_type() == DOM::Document::Type::HTML)
+            return component.lowercase_name() == element.local_name();
+        return component.name().equals_ignoring_case(element.local_name());
     case CSS::Selector::SimpleSelector::Type::Attribute:
         return matches_attribute(component.attribute(), element);
     case CSS::Selector::SimpleSelector::Type::PseudoClass:
diff --git a/Userland/Libraries/LibWeb/Dump.cpp b/Userland/Libraries/LibWeb/Dump.cpp
index 7b2d642c2b..92e13a5a75 100644
--- a/Userland/Libraries/LibWeb/Dump.cpp
+++ b/Userland/Libraries/LibWeb/Dump.cpp
@@ -372,7 +372,7 @@ void dump_selector(StringBuilder& builder, CSS::Selector const& selector)
 
             builder.appendff("{}:", type_description);
             // FIXME: This is goofy
-            if (simple_selector.value.has<FlyString>())
+            if (simple_selector.value.has<CSS::Selector::SimpleSelector::Name>())
                 builder.append(simple_selector.name());
 
             if (simple_selector.type == CSS::Selector::SimpleSelector::Type::PseudoClass) {