Prepare for invisible delimiters.

nnethercote · nnethercote · commit 95ab2220ab6e · 2024-11-19T09:57:11.000+11:00
Current places where `Interpolated` is used are going to change to
instead use invisible delimiters. This prepares for that.
- It adds invisible delimiter cases to the `can_begin_*`/`may_be_*`
  methods and the `failed_to_match_macro` that are equivalent to the
  existing `Interpolated` cases.
- It adds panics/asserts in some places where invisible delimiters
  should never occur.
- In `Parser::parse_struct_fields` it excludes an ident + invisible
  delimiter from special consideration in an error message, because
  that's quite different to an ident + paren/brace/bracket.
diff --git a/compiler/rustc_ast/src/token.rs b/compiler/rustc_ast/src/token.rs
@@ -599,10 +599,11 @@ impl Token {
     /// **NB**: Take care when modifying this function, since it will change
     /// the stable set of tokens that are allowed to match an expr nonterminal.
     pub fn can_begin_expr(&self) -> bool {
+        use Delimiter::*;
         match self.uninterpolate().kind {
             Ident(name, is_raw)              =>
                 ident_can_begin_expr(name, self.span, is_raw), // value name or keyword
-            OpenDelim(..)                     | // tuple, array or block
+            OpenDelim(Parenthesis | Brace | Bracket) | // tuple, array or block
             Literal(..)                       | // literal
             Not                               | // operator not
             BinOp(Minus)                      | // unary minus
@@ -613,7 +614,7 @@ impl Token {
             // DotDotDot is no longer supported, but we need some way to display the error
             DotDot | DotDotDot | DotDotEq     | // range notation
             Lt | BinOp(Shl)                   | // associated path
-            PathSep                            | // global path
+            PathSep                           | // global path
             Lifetime(..)                      | // labeled loop
             Pound                             => true, // expression attributes
             Interpolated(ref nt) =>
@@ -623,6 +624,12 @@ impl Token {
                     NtLiteral(..) |
                     NtPath(..)
                 ),
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(
+                MetaVarKind::Block |
+                MetaVarKind::Expr { .. } |
+                MetaVarKind::Literal |
+                MetaVarKind::Path
+            ))) => true,
             _ => false,
         }
     }
@@ -656,6 +663,14 @@ impl Token {
                     | NtPath(..)
                     | NtTy(..)
                 ),
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(
+                MetaVarKind::Expr { .. } |
+                MetaVarKind::Literal |
+                MetaVarKind::Meta |
+                MetaVarKind::Pat(_) |
+                MetaVarKind::Path |
+                MetaVarKind::Ty
+            ))) => true,
             _ => false,
         }
     }
@@ -676,6 +691,10 @@ impl Token {
             Lt | BinOp(Shl)             | // associated path
             PathSep                      => true, // global path
             Interpolated(ref nt) => matches!(&**nt, NtTy(..) | NtPath(..)),
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(
+                MetaVarKind::Ty |
+                MetaVarKind::Path
+            ))) => true,
             // For anonymous structs or unions, which only appear in specific positions
             // (type of struct fields or union fields), we don't consider them as regular types
             _ => false,
@@ -688,6 +707,9 @@ impl Token {
             OpenDelim(Delimiter::Brace) | Literal(..) | BinOp(Minus) => true,
             Ident(name, IdentIsRaw::No) if name.is_bool_lit() => true,
             Interpolated(ref nt) => matches!(&**nt, NtExpr(..) | NtBlock(..) | NtLiteral(..)),
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(
+                MetaVarKind::Expr { .. } | MetaVarKind::Block | MetaVarKind::Literal,
+            ))) => true,
             _ => false,
         }
     }
@@ -744,6 +766,13 @@ impl Token {
                 },
                 _ => false,
             },
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(mv_kind))) => match mv_kind {
+                MetaVarKind::Literal => true,
+                MetaVarKind::Expr { can_begin_literal_maybe_minus, .. } => {
+                    can_begin_literal_maybe_minus
+                }
+                _ => false,
+            },
             _ => false,
         }
     }
@@ -759,6 +788,11 @@ impl Token {
                 },
                 _ => false,
             },
+            OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(mv_kind))) => match mv_kind {
+                MetaVarKind::Literal => true,
+                MetaVarKind::Expr { can_begin_string_literal, .. } => can_begin_string_literal,
+                _ => false,
+            },
             _ => false,
         }
     }
diff --git a/compiler/rustc_expand/src/mbe/diagnostics.rs b/compiler/rustc_expand/src/mbe/diagnostics.rs
@@ -1,6 +1,6 @@
 use std::borrow::Cow;
 
-use rustc_ast::token::{self, Token, TokenKind};
+use rustc_ast::token::{self, Delimiter, Token, TokenKind};
 use rustc_ast::tokenstream::TokenStream;
 use rustc_errors::{Applicability, Diag, DiagCtxtHandle, DiagMessage};
 use rustc_macros::Subdiagnostic;
@@ -68,7 +68,9 @@ pub(super) fn failed_to_match_macro(
 
     if let MatcherLoc::Token { token: expected_token } = &remaining_matcher
         && (matches!(expected_token.kind, TokenKind::Interpolated(_))
-            || matches!(token.kind, TokenKind::Interpolated(_)))
+            || matches!(token.kind, TokenKind::Interpolated(_))
+            || matches!(expected_token.kind, TokenKind::OpenDelim(Delimiter::Invisible(_)))
+            || matches!(token.kind, TokenKind::OpenDelim(Delimiter::Invisible(_))))
     {
         err.note("captured metavariables except for `:tt`, `:ident` and `:lifetime` cannot be compared to other tokens");
         err.note("see <https://doc.rust-lang.org/nightly/reference/macros-by-example.html#forwarding-a-matched-fragment> for more information");
diff --git a/compiler/rustc_parse/src/lexer/tokentrees.rs b/compiler/rustc_parse/src/lexer/tokentrees.rs
@@ -43,11 +43,19 @@ impl<'psess, 'src> TokenTreesReader<'psess, 'src> {
         let mut buf = Vec::new();
         loop {
             match self.token.kind {
-                token::OpenDelim(delim) => buf.push(match self.lex_token_tree_open_delim(delim) {
-                    Ok(val) => val,
-                    Err(errs) => return (open_spacing, TokenStream::new(buf), Err(errs)),
-                }),
+                token::OpenDelim(delim) => {
+                    // Invisible delimiters cannot occur here because `TokenTreesReader` parses
+                    // code directly from strings, with no macro expansion involved.
+                    debug_assert!(!matches!(delim, Delimiter::Invisible(_)));
+                    buf.push(match self.lex_token_tree_open_delim(delim) {
+                        Ok(val) => val,
+                        Err(errs) => return (open_spacing, TokenStream::new(buf), Err(errs)),
+                    })
+                }
                 token::CloseDelim(delim) => {
+                    // Invisible delimiters cannot occur here because `TokenTreesReader` parses
+                    // code directly from strings, with no macro expansion involved.
+                    debug_assert!(!matches!(delim, Delimiter::Invisible(_)));
                     return (
                         open_spacing,
                         TokenStream::new(buf),
diff --git a/compiler/rustc_parse/src/parser/expr.rs b/compiler/rustc_parse/src/parser/expr.rs
@@ -3591,11 +3591,19 @@ impl<'a> Parser<'a> {
                         && !self.token.is_reserved_ident()
                         && self.look_ahead(1, |t| {
                             AssocOp::from_token(t).is_some()
-                                || matches!(t.kind, token::OpenDelim(_))
+                                || matches!(
+                                    t.kind,
+                                    token::OpenDelim(
+                                        Delimiter::Parenthesis
+                                            | Delimiter::Bracket
+                                            | Delimiter::Brace
+                                    )
+                                )
                                 || *t == token::Dot
                         })
                     {
-                        // Looks like they tried to write a shorthand, complex expression.
+                        // Looks like they tried to write a shorthand, complex expression,
+                        // E.g.: `n + m`, `f(a)`, `a[i]`, `S { x: 3 }`, or `x.y`.
                         e.span_suggestion_verbose(
                             self.token.span.shrink_to_lo(),
                             "try naming a field",
diff --git a/compiler/rustc_parse/src/parser/nonterminal.rs b/compiler/rustc_parse/src/parser/nonterminal.rs
@@ -3,7 +3,9 @@ use rustc_ast::ptr::P;
 use rustc_ast::token::Nonterminal::*;
 use rustc_ast::token::NtExprKind::*;
 use rustc_ast::token::NtPatKind::*;
-use rustc_ast::token::{self, Delimiter, NonterminalKind, Token};
+use rustc_ast::token::{
+    self, Delimiter, InvisibleOrigin, MetaVarKind, Nonterminal, NonterminalKind, Token,
+};
 use rustc_ast_pretty::pprust;
 use rustc_data_structures::sync::Lrc;
 use rustc_errors::PResult;
@@ -22,7 +24,29 @@ impl<'a> Parser<'a> {
     #[inline]
     pub fn nonterminal_may_begin_with(kind: NonterminalKind, token: &Token) -> bool {
         /// Checks whether the non-terminal may contain a single (non-keyword) identifier.
-        fn may_be_ident(nt: &token::Nonterminal) -> bool {
+        fn may_be_ident(kind: MetaVarKind) -> bool {
+            use MetaVarKind::*;
+            match kind {
+                Stmt
+                | Pat(_)
+                | Expr { .. }
+                | Ty
+                | Literal // `true`, `false`
+                | Meta
+                | Path => true,
+
+                Item
+                | Block
+                | Vis => false,
+
+                Ident
+                | Lifetime
+                | TT => unreachable!(),
+            }
+        }
+
+        /// Old variant of `may_be_ident`. Being phased out.
+        fn nt_may_be_ident(nt: &Nonterminal) -> bool {
             match nt {
                 NtStmt(_)
                 | NtPat(_)
@@ -69,7 +93,8 @@ impl<'a> Parser<'a> {
                 | token::Ident(..)
                 | token::NtIdent(..)
                 | token::NtLifetime(..)
-                | token::Interpolated(_) => true,
+                | token::Interpolated(_)
+                | token::OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(_))) => true,
                 _ => token.can_begin_type(),
             },
             NonterminalKind::Block => match &token.kind {
@@ -79,11 +104,29 @@ impl<'a> Parser<'a> {
                     NtBlock(_) | NtStmt(_) | NtExpr(_) | NtLiteral(_) => true,
                     NtItem(_) | NtPat(_) | NtTy(_) | NtMeta(_) | NtPath(_) | NtVis(_) => false,
                 },
+                token::OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(k))) => match k {
+                    MetaVarKind::Block
+                    | MetaVarKind::Stmt
+                    | MetaVarKind::Expr { .. }
+                    | MetaVarKind::Literal => true,
+                    MetaVarKind::Item
+                    | MetaVarKind::Pat(_)
+                    | MetaVarKind::Ty
+                    | MetaVarKind::Meta
+                    | MetaVarKind::Path
+                    | MetaVarKind::Vis => false,
+                    MetaVarKind::Lifetime | MetaVarKind::Ident | MetaVarKind::TT => {
+                        unreachable!()
+                    }
+                },
                 _ => false,
             },
             NonterminalKind::Path | NonterminalKind::Meta => match &token.kind {
                 token::PathSep | token::Ident(..) | token::NtIdent(..) => true,
-                token::Interpolated(nt) => may_be_ident(nt),
+                token::Interpolated(nt) => nt_may_be_ident(nt),
+                token::OpenDelim(Delimiter::Invisible(InvisibleOrigin::MetaVar(kind))) => {
+                    may_be_ident(*kind)
+                }
                 _ => false,
             },
             NonterminalKind::Pat(pat_kind) => token.can_begin_pattern(pat_kind),
