perf(css/ast): Use Atom for raw values (#6401)

crates/swc_css_ast/src/at_rule.rs

@@ -1,6 +1,6 @@
 use is_macro::Is;
 use string_enum::StringEnum;
-use swc_atoms::JsWord;
+use swc_atoms::{Atom, JsWord};
 use swc_common::{ast_node, EqIgnoreSpan, Span};
 
 use crate::{
@@ -809,8 +809,7 @@ pub struct ExtensionName {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for ExtensionName {

crates/swc_css_ast/src/selector.rs

@@ -1,6 +1,6 @@
 use is_macro::Is;
 use string_enum::StringEnum;
-use swc_atoms::JsWord;
+use swc_atoms::{Atom, JsWord};
 use swc_common::{ast_node, util::take::Take, EqIgnoreSpan, Span};
 
 use crate::{Delimiter, Ident, ListOfComponentValues, Str, TokenAndSpan};
@@ -369,11 +369,9 @@ pub enum AnPlusB {
 pub struct AnPlusBNotation {
     pub span: Span,
     pub a: Option<i32>,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub a_raw: Option<JsWord>,
+    pub a_raw: Option<Atom>,
     pub b: Option<i32>,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub b_raw: Option<JsWord>,
+    pub b_raw: Option<Atom>,
 }
 
 #[ast_node("PseudoElementSelector")]
@@ -403,8 +401,7 @@ pub struct CustomHighlightName {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for CustomHighlightName {

crates/swc_css_ast/src/token.rs

@@ -5,7 +5,7 @@ use std::{
 
 use is_macro::Is;
 use serde::{Deserialize, Serialize};
-use swc_atoms::JsWord;
+use swc_atoms::{Atom, JsWord};
 use swc_common::{ast_node, EqIgnoreSpan, Span};
 
 #[ast_node("PreservedToken")]
@@ -52,65 +52,55 @@ pub enum Token {
     Ident {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     Function {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     /// `@`
     AtKeyword {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     /// `#`
     Hash {
         is_id: bool,
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     String {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     BadString {
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_value: JsWord,
+        raw_value: Atom,
     },
 
     /// `url(value)`
     Url {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         name: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_name: JsWord,
+        raw_name: Atom,
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         value: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_value: JsWord,
+        raw_value: Atom,
     },
 
     BadUrl {
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         name: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_name: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_value: JsWord,
+        raw_name: Atom,
+        raw_value: Atom,
     },
 
     Delim {
@@ -119,34 +109,29 @@ pub enum Token {
 
     Number {
         value: f64,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
         #[serde(rename = "type")]
         type_flag: NumberType,
     },
 
     Percentage {
         value: f64,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw: JsWord,
+        raw: Atom,
     },
 
     Dimension {
         value: f64,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_value: JsWord,
+        raw_value: Atom,
         #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
         unit: JsWord,
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        raw_unit: JsWord,
+        raw_unit: Atom,
         #[serde(rename = "type")]
         type_flag: NumberType,
     },
 
     /// One or more whitespace.
     WhiteSpace {
-        #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-        value: JsWord,
+        value: Atom,
     },
 
     /// `<!--`

crates/swc_css_ast/src/value.rs

@@ -5,7 +5,7 @@ use std::{
 
 use is_macro::Is;
 use string_enum::StringEnum;
-use swc_atoms::JsWord;
+use swc_atoms::{Atom, JsWord};
 use swc_common::{ast_node, util::take::Take, EqIgnoreSpan, Span};
 
 use crate::Function;
@@ -16,8 +16,7 @@ pub struct Ident {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for Ident {
@@ -42,8 +41,7 @@ pub struct CustomIdent {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for CustomIdent {
@@ -58,8 +56,7 @@ pub struct DashedIdent {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for DashedIdent {
@@ -74,8 +71,7 @@ pub struct CustomPropertyName {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for CustomPropertyName {
@@ -91,8 +87,7 @@ pub struct Str {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for Str {
@@ -164,8 +159,7 @@ pub struct HexColor {
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
     /// Does **not** include `#`
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 #[ast_node]
@@ -326,8 +320,7 @@ pub enum TimePercentage {
 pub struct Integer {
     pub span: Span,
     pub value: i64,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl EqIgnoreSpan for Integer {
@@ -340,8 +333,7 @@ impl EqIgnoreSpan for Integer {
 pub struct Number {
     pub span: Span,
     pub value: f64,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 impl Eq for Number {}
@@ -431,8 +423,7 @@ pub struct UrlValueRaw {
     pub span: Span,
     #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
     pub value: JsWord,
-    #[cfg_attr(feature = "rkyv", with(swc_atoms::EncodeJsWord))]
-    pub raw: Option<JsWord>,
+    pub raw: Option<Atom>,
 }
 
 #[ast_node]

crates/swc_css_codegen/src/lib.rs

@@ -376,7 +376,7 @@ where
                     AtRuleName::Ident(swc_css_ast::Ident {
                         span: n.span,
                         value: js_word!("layer"),
-                        raw: Some(js_word!("layer"))
+                        raw: None
                     })
                 )
             }

crates/swc_css_minifier/src/compressor/mod.rs

@@ -400,14 +400,22 @@ impl VisitMut for Compressor {
                 | Token::Function { value, .. }
                 | Token::AtKeyword { value, .. }
                 | Token::String { value, .. }
-                | Token::BadString {
-                    raw_value: value, ..
-                }
-                | Token::Dimension { unit: value, .. }
+                | Token::Url { value, .. }
                     if !contains_only_ascii_characters(value) =>
                 {
                     self.need_utf8_at_rule = true;
                 }
+                Token::BadString {
+                    raw_value: value, ..
+                }
+                | Token::BadUrl {
+                    raw_value: value, ..
+                } if !contains_only_ascii_characters(value) => {
+                    self.need_utf8_at_rule = true;
+                }
+                Token::Dimension { unit: value, .. } if !contains_only_ascii_characters(value) => {
+                    self.need_utf8_at_rule = true;
+                }
                 _ => {}
             }
         }

crates/swc_css_parser/src/lexer/mod.rs

@@ -1,6 +1,6 @@
 use std::{cell::RefCell, char::REPLACEMENT_CHARACTER, rc::Rc};
 
-use swc_atoms::{js_word, JsWord};
+use swc_atoms::{js_word, Atom, JsWord};
 use swc_common::{input::Input, BytePos, Span};
 use swc_css_ast::{NumberType, Token, TokenAndSpan};
 
@@ -559,9 +559,9 @@ where
             // unit set initially to the empty string.
             let mut token = Token::Dimension {
                 value: number.0,
-                raw_value: number.1.into(),
+                raw_value: number.1,
                 unit: js_word!(""),
-                raw_unit: js_word!(""),
+                raw_unit: "".into(),
                 type_flag: number.2,
             };
 
@@ -592,15 +592,15 @@ where
 
             return Ok(Token::Percentage {
                 value: number.0,
-                raw: number.1.into(),
+                raw: number.1,
             });
         }
 
         // Otherwise, create a <number-token> with the same value and type flag as
         // number, and return it.
         Ok(Token::Number {
             value: number.0,
-            raw: number.1.into(),
+            raw: number.1,
             type_flag: number.2,
         })
     }
@@ -779,7 +779,7 @@ where
 
     // This section describes how to consume a url token from a stream of code
     // points. It returns either a <url-token> or a <bad-url-token>.
-    fn read_url(&mut self, name: (JsWord, JsWord), before: String) -> LexResult<Token> {
+    fn read_url(&mut self, name: (JsWord, Atom), before: String) -> LexResult<Token> {
         // Initially create a <url-token> with its value set to the empty string.
         self.with_buf_and_raw_buf(|l, out, raw| {
             raw.push_str(&before);
@@ -1178,7 +1178,7 @@ where
     // This section describes how to consume an ident sequence from a stream of code
     // points. It returns a string containing the largest name that can be formed
     // from adjacent code points in the stream, starting from the first.
-    fn read_ident_sequence(&mut self) -> LexResult<(JsWord, JsWord)> {
+    fn read_ident_sequence(&mut self) -> LexResult<(JsWord, Atom)> {
         self.with_buf_and_raw_buf(|l, buf, raw| {
             // Let result initially be an empty string.
             // Done above
@@ -1219,10 +1219,9 @@ where
 
     // This section describes how to consume a number from a stream of code points.
     // It returns a numeric value, and a type which is either "integer" or "number".
-    fn read_number(&mut self) -> LexResult<(f64, String, NumberType)> {
-        let parsed =
+    fn read_number(&mut self) -> LexResult<(f64, Atom, NumberType)> {
+        let parsed: (Atom, NumberType) = self.with_buf(|l, out| {
             // Initially set type to "integer". Let repr be the empty string.
-            self.with_buf(|l, out| {
             let mut type_flag = NumberType::Integer;
 
             // If the next input code point is U+002B PLUS SIGN (+) or U+002D HYPHEN-MINUS
@@ -1324,7 +1323,7 @@ where
         })?;
 
         // Convert repr to a number, and set the value to the returned value.
-        let value = lexical::parse(&parsed.0).unwrap_or_else(|err| {
+        let value = lexical::parse(&*parsed.0).unwrap_or_else(|err| {
             unreachable!("failed to parse `{}` using lexical: {:?}", parsed.0, err)
         });