Rollup merge of #127233 - nnethercote:parser-cleanups, r=petrochenkov

Some parser cleanups

Cleanups I made while looking closely at this code.

r? ``@petrochenkov``

Author: matthiaskrgr

compiler/rustc_ast/src/attr/mod.rs

@@ -204,12 +204,14 @@ impl Attribute {
 
     pub fn tokens(&self) -> TokenStream {
         match &self.kind {
-            AttrKind::Normal(normal) => normal
-                .tokens
-                .as_ref()
-                .unwrap_or_else(|| panic!("attribute is missing tokens: {self:?}"))
-                .to_attr_token_stream()
-                .to_tokenstream(),
+            AttrKind::Normal(normal) => TokenStream::new(
+                normal
+                    .tokens
+                    .as_ref()
+                    .unwrap_or_else(|| panic!("attribute is missing tokens: {self:?}"))
+                    .to_attr_token_stream()
+                    .to_token_trees(),
+            ),
             &AttrKind::DocComment(comment_kind, data) => TokenStream::token_alone(
                 token::DocComment(comment_kind, self.style, data),
                 self.span,

compiler/rustc_ast/src/tokenstream.rs

@@ -23,7 +23,6 @@ use rustc_data_structures::sync::{self, Lrc};
 use rustc_macros::{Decodable, Encodable, HashStable_Generic};
 use rustc_serialize::{Decodable, Encodable};
 use rustc_span::{sym, Span, SpanDecoder, SpanEncoder, Symbol, DUMMY_SP};
-use smallvec::{smallvec, SmallVec};
 
 use std::borrow::Cow;
 use std::{cmp, fmt, iter};
@@ -180,42 +179,33 @@ impl AttrTokenStream {
         AttrTokenStream(Lrc::new(tokens))
     }
 
-    /// Converts this `AttrTokenStream` to a plain `TokenStream`.
+    /// Converts this `AttrTokenStream` to a plain `Vec<TokenTree>`.
     /// During conversion, `AttrTokenTree::Attributes` get 'flattened'
     /// back to a `TokenStream` of the form `outer_attr attr_target`.
     /// If there are inner attributes, they are inserted into the proper
     /// place in the attribute target tokens.
-    pub fn to_tokenstream(&self) -> TokenStream {
-        let trees: Vec<_> = self
-            .0
-            .iter()
-            .flat_map(|tree| match &tree {
+    pub fn to_token_trees(&self) -> Vec<TokenTree> {
+        let mut res = Vec::with_capacity(self.0.len());
+        for tree in self.0.iter() {
+            match tree {
                 AttrTokenTree::Token(inner, spacing) => {
-                    smallvec![TokenTree::Token(inner.clone(), *spacing)].into_iter()
+                    res.push(TokenTree::Token(inner.clone(), *spacing));
                 }
                 AttrTokenTree::Delimited(span, spacing, delim, stream) => {
-                    smallvec![TokenTree::Delimited(
+                    res.push(TokenTree::Delimited(
                         *span,
                         *spacing,
                         *delim,
-                        stream.to_tokenstream()
-                    ),]
-                    .into_iter()
+                        TokenStream::new(stream.to_token_trees()),
+                    ))
                 }
                 AttrTokenTree::Attributes(data) => {
                     let idx = data
                         .attrs
                         .partition_point(|attr| matches!(attr.style, crate::AttrStyle::Outer));
                     let (outer_attrs, inner_attrs) = data.attrs.split_at(idx);
 
-                    let mut target_tokens: Vec<_> = data
-                        .tokens
-                        .to_attr_token_stream()
-                        .to_tokenstream()
-                        .0
-                        .iter()
-                        .cloned()
-                        .collect();
+                    let mut target_tokens = data.tokens.to_attr_token_stream().to_token_trees();
                     if !inner_attrs.is_empty() {
                         let mut found = false;
                         // Check the last two trees (to account for a trailing semi)
@@ -251,17 +241,14 @@ impl AttrTokenStream {
                             "Failed to find trailing delimited group in: {target_tokens:?}"
                         );
                     }
-                    let mut flat: SmallVec<[_; 1]> =
-                        SmallVec::with_capacity(target_tokens.len() + outer_attrs.len());
                     for attr in outer_attrs {
-                        flat.extend(attr.tokens().0.iter().cloned());
+                        res.extend(attr.tokens().0.iter().cloned());
                     }
-                    flat.extend(target_tokens);
-                    flat.into_iter()
+                    res.extend(target_tokens);
                 }
-            })
-            .collect();
-        TokenStream::new(trees)
+            }
+        }
+        res
     }
 }
 
@@ -409,8 +396,8 @@ impl PartialEq<TokenStream> for TokenStream {
 }
 
 impl TokenStream {
-    pub fn new(streams: Vec<TokenTree>) -> TokenStream {
-        TokenStream(Lrc::new(streams))
+    pub fn new(tts: Vec<TokenTree>) -> TokenStream {
+        TokenStream(Lrc::new(tts))
     }
 
     pub fn is_empty(&self) -> bool {
@@ -461,7 +448,7 @@ impl TokenStream {
                 AttributesData { attrs: attrs.iter().cloned().collect(), tokens: tokens.clone() };
             AttrTokenStream::new(vec![AttrTokenTree::Attributes(attr_data)])
         };
-        attr_stream.to_tokenstream()
+        TokenStream::new(attr_stream.to_token_trees())
     }
 
     pub fn from_nonterminal_ast(nt: &Nonterminal) -> TokenStream {

compiler/rustc_builtin_macros/src/cfg_eval.rs

@@ -38,16 +38,14 @@ pub(crate) fn cfg_eval(
     lint_node_id: NodeId,
 ) -> Annotatable {
     let features = Some(features);
-    CfgEval { cfg: &mut StripUnconfigured { sess, features, config_tokens: true, lint_node_id } }
+    CfgEval(StripUnconfigured { sess, features, config_tokens: true, lint_node_id })
         .configure_annotatable(annotatable)
         // Since the item itself has already been configured by the `InvocationCollector`,
         // we know that fold result vector will contain exactly one element.
         .unwrap()
 }
 
-struct CfgEval<'a, 'b> {
-    cfg: &'a mut StripUnconfigured<'b>,
-}
+struct CfgEval<'a>(StripUnconfigured<'a>);
 
 fn flat_map_annotatable(
     vis: &mut impl MutVisitor,
@@ -125,9 +123,9 @@ fn has_cfg_or_cfg_attr(annotatable: &Annotatable) -> bool {
     res.is_break()
 }
 
-impl CfgEval<'_, '_> {
+impl CfgEval<'_> {
     fn configure<T: HasAttrs + HasTokens>(&mut self, node: T) -> Option<T> {
-        self.cfg.configure(node)
+        self.0.configure(node)
     }
 
     fn configure_annotatable(&mut self, mut annotatable: Annotatable) -> Option<Annotatable> {
@@ -196,7 +194,7 @@ impl CfgEval<'_, '_> {
         // Re-parse the tokens, setting the `capture_cfg` flag to save extra information
         // to the captured `AttrTokenStream` (specifically, we capture
         // `AttrTokenTree::AttributesData` for all occurrences of `#[cfg]` and `#[cfg_attr]`)
-        let mut parser = Parser::new(&self.cfg.sess.psess, orig_tokens, None);
+        let mut parser = Parser::new(&self.0.sess.psess, orig_tokens, None);
         parser.capture_cfg = true;
         match parse_annotatable_with(&mut parser) {
             Ok(a) => annotatable = a,
@@ -212,16 +210,16 @@ impl CfgEval<'_, '_> {
     }
 }
 
-impl MutVisitor for CfgEval<'_, '_> {
+impl MutVisitor for CfgEval<'_> {
     #[instrument(level = "trace", skip(self))]
     fn visit_expr(&mut self, expr: &mut P<ast::Expr>) {
-        self.cfg.configure_expr(expr, false);
+        self.0.configure_expr(expr, false);
         mut_visit::noop_visit_expr(expr, self);
     }
 
     #[instrument(level = "trace", skip(self))]
     fn visit_method_receiver_expr(&mut self, expr: &mut P<ast::Expr>) {
-        self.cfg.configure_expr(expr, true);
+        self.0.configure_expr(expr, true);
         mut_visit::noop_visit_expr(expr, self);
     }
 

compiler/rustc_expand/src/mbe/diagnostics.rs

@@ -120,21 +120,21 @@ struct CollectTrackerAndEmitter<'a, 'cx, 'matcher> {
 
 struct BestFailure {
     token: Token,
-    position_in_tokenstream: usize,
+    position_in_tokenstream: u32,
     msg: &'static str,
     remaining_matcher: MatcherLoc,
 }
 
 impl BestFailure {
-    fn is_better_position(&self, position: usize) -> bool {
+    fn is_better_position(&self, position: u32) -> bool {
         position > self.position_in_tokenstream
     }
 }
 
 impl<'a, 'cx, 'matcher> Tracker<'matcher> for CollectTrackerAndEmitter<'a, 'cx, 'matcher> {
-    type Failure = (Token, usize, &'static str);
+    type Failure = (Token, u32, &'static str);
 
-    fn build_failure(tok: Token, position: usize, msg: &'static str) -> Self::Failure {
+    fn build_failure(tok: Token, position: u32, msg: &'static str) -> Self::Failure {
         (tok, position, msg)
     }
 
@@ -211,9 +211,9 @@ impl<'matcher> FailureForwarder<'matcher> {
 }
 
 impl<'matcher> Tracker<'matcher> for FailureForwarder<'matcher> {
-    type Failure = (Token, usize, &'static str);
+    type Failure = (Token, u32, &'static str);
 
-    fn build_failure(tok: Token, position: usize, msg: &'static str) -> Self::Failure {
+    fn build_failure(tok: Token, position: u32, msg: &'static str) -> Self::Failure {
         (tok, position, msg)
     }
 

compiler/rustc_expand/src/mbe/macro_parser.rs

@@ -452,7 +452,7 @@ impl TtParser {
         &mut self,
         matcher: &'matcher [MatcherLoc],
         token: &Token,
-        approx_position: usize,
+        approx_position: u32,
         track: &mut T,
     ) -> Option<NamedParseResult<T::Failure>> {
         // Matcher positions that would be valid if the macro invocation was over now. Only

compiler/rustc_expand/src/mbe/macro_rules.rs

@@ -153,7 +153,7 @@ pub(super) trait Tracker<'matcher> {
     /// Arm failed to match. If the token is `token::Eof`, it indicates an unexpected
     /// end of macro invocation. Otherwise, it indicates that no rules expected the given token.
     /// The usize is the approximate position of the token in the input token stream.
-    fn build_failure(tok: Token, position: usize, msg: &'static str) -> Self::Failure;
+    fn build_failure(tok: Token, position: u32, msg: &'static str) -> Self::Failure;
 
     /// This is called before trying to match next MatcherLoc on the current token.
     fn before_match_loc(&mut self, _parser: &TtParser, _matcher: &'matcher MatcherLoc) {}
@@ -182,7 +182,7 @@ pub(super) struct NoopTracker;
 impl<'matcher> Tracker<'matcher> for NoopTracker {
     type Failure = ();
 
-    fn build_failure(_tok: Token, _position: usize, _msg: &'static str) -> Self::Failure {}
+    fn build_failure(_tok: Token, _position: u32, _msg: &'static str) -> Self::Failure {}
 
     fn description() -> &'static str {
         "none"