Auto merge of #114292 - estebank:issue-71039, r=<try>

More detail when expecting expression but encountering bad macro argument

Partially address #71039.

```
error: expected expression, found pattern `1 + 1`
  --> $DIR/trace_faulty_macros.rs:49:37
   |
LL |     (let $p:pat = $e:expr) => {test!(($p,$e))};
   |                   -------                -- this is interpreted as expression, but it is expected to be pattern
   |                   |
   |                   this macro fragment matcher is expression
...
LL |     (($p:pat, $e:pat)) => {let $p = $e;};
   |               ------                ^^ expected expression
   |               |
   |               this macro fragment matcher is pattern
...
LL |     test!(let x = 1+1);
   |     ------------------
   |     |             |
   |     |             this is expected to be expression
   |     in this macro invocation
   |
   = note: when forwarding a matched fragment to another macro-by-example, matchers in the second macro will see an opaque AST of the fragment type, not the underlying tokens
   = note: this error originates in the macro `test` (in Nightly builds, run with -Z macro-backtrace for more info)
```

compiler/rustc_ast/src/attr/mod.rs

@@ -342,7 +342,7 @@ impl MetaItem {
                 let span = span.with_hi(segments.last().unwrap().ident.span.hi());
                 Path { span, segments, tokens: None }
             }
-            Some(TokenTree::Token(Token { kind: token::Interpolated(nt), .. }, _)) => match &**nt {
+            Some(TokenTree::Token(Token { kind: token::Interpolated(nt), .. }, _)) => match &nt.0 {
                 token::Nonterminal::NtMeta(item) => return item.meta(item.path.span),
                 token::Nonterminal::NtPath(path) => (**path).clone(),
                 _ => return None,

compiler/rustc_ast/src/mut_visit.rs

@@ -764,7 +764,10 @@ pub fn visit_token<T: MutVisitor>(t: &mut Token, vis: &mut T) {
             return; // Avoid visiting the span for the second time.
         }
         token::Interpolated(nt) => {
-            visit_nonterminal(Lrc::make_mut(nt), vis);
+            let nt = Lrc::make_mut(nt);
+            let (nt, sp) = (&mut nt.0, &mut nt.1);
+            vis.visit_span(sp);
+            visit_nonterminal(nt, vis);
         }
         _ => {}
     }

compiler/rustc_ast/src/token.rs

@@ -110,7 +110,7 @@ impl Lit {
             Ident(name, false) if name.is_bool_lit() => Some(Lit::new(Bool, name, None)),
             Literal(token_lit) => Some(token_lit),
             Interpolated(ref nt)
-                if let NtExpr(expr) | NtLiteral(expr) = &**nt
+                if let NtExpr(expr) | NtLiteral(expr) = &nt.0
                     && let ast::ExprKind::Lit(token_lit) = expr.kind =>
             {
                 Some(token_lit)
@@ -314,7 +314,7 @@ pub enum TokenKind {
     /// - It prevents `Token` from implementing `Copy`.
     /// It adds complexity and likely slows things down. Please don't add new
     /// occurrences of this token kind!
-    Interpolated(Lrc<Nonterminal>),
+    Interpolated(Lrc<(Nonterminal, Span)>),
 
     /// A doc comment token.
     /// `Symbol` is the doc comment's data excluding its "quotes" (`///`, `/**`, etc)
@@ -421,7 +421,7 @@ impl Token {
     /// if they keep spans or perform edition checks.
     pub fn uninterpolated_span(&self) -> Span {
         match &self.kind {
-            Interpolated(nt) => nt.span(),
+            Interpolated(nt) => nt.0.use_span(),
             _ => self.span,
         }
     }
@@ -464,7 +464,7 @@ impl Token {
             ModSep                            | // global path
             Lifetime(..)                      | // labeled loop
             Pound                             => true, // expression attributes
-            Interpolated(ref nt) => matches!(**nt, NtLiteral(..) |
+            Interpolated(ref nt) => matches!(&nt.0, NtLiteral(..) |
                 NtExpr(..)    |
                 NtBlock(..)   |
                 NtPath(..)),
@@ -488,7 +488,7 @@ impl Token {
             | DotDot | DotDotDot | DotDotEq      // ranges
             | Lt | BinOp(Shl)                    // associated path
             | ModSep                    => true, // global path
-            Interpolated(ref nt) => matches!(**nt, NtLiteral(..) |
+            Interpolated(ref nt) => matches!(&nt.0, NtLiteral(..) |
                 NtPat(..)     |
                 NtBlock(..)   |
                 NtPath(..)),
@@ -511,7 +511,7 @@ impl Token {
             Lifetime(..)                | // lifetime bound in trait object
             Lt | BinOp(Shl)             | // associated path
             ModSep                      => true, // global path
-            Interpolated(ref nt) => matches!(**nt, NtTy(..) | NtPath(..)),
+            Interpolated(ref nt) => matches!(&nt.0, NtTy(..) | NtPath(..)),
             // 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,
@@ -522,7 +522,7 @@ impl Token {
     pub fn can_begin_const_arg(&self) -> bool {
         match self.kind {
             OpenDelim(Delimiter::Brace) => true,
-            Interpolated(ref nt) => matches!(**nt, NtExpr(..) | NtBlock(..) | NtLiteral(..)),
+            Interpolated(ref nt) => matches!(&nt.0, NtExpr(..) | NtBlock(..) | NtLiteral(..)),
             _ => self.can_begin_literal_maybe_minus(),
         }
     }
@@ -576,7 +576,7 @@ impl Token {
         match self.uninterpolate().kind {
             Literal(..) | BinOp(Minus) => true,
             Ident(name, false) if name.is_bool_lit() => true,
-            Interpolated(ref nt) => match &**nt {
+            Interpolated(ref nt) => match &nt.0 {
                 NtLiteral(_) => true,
                 NtExpr(e) => match &e.kind {
                     ast::ExprKind::Lit(_) => true,
@@ -597,9 +597,9 @@ impl Token {
     /// otherwise returns the original token.
     pub fn uninterpolate(&self) -> Cow<'_, Token> {
         match &self.kind {
-            Interpolated(nt) => match **nt {
+            Interpolated(nt) => match &nt.0 {
                 NtIdent(ident, is_raw) => {
-                    Cow::Owned(Token::new(Ident(ident.name, is_raw), ident.span))
+                    Cow::Owned(Token::new(Ident(ident.name, *is_raw), ident.span))
                 }
                 NtLifetime(ident) => Cow::Owned(Token::new(Lifetime(ident.name), ident.span)),
                 _ => Cow::Borrowed(self),
@@ -614,8 +614,8 @@ impl Token {
         // We avoid using `Token::uninterpolate` here because it's slow.
         match &self.kind {
             &Ident(name, is_raw) => Some((Ident::new(name, self.span), is_raw)),
-            Interpolated(nt) => match **nt {
-                NtIdent(ident, is_raw) => Some((ident, is_raw)),
+            Interpolated(nt) => match &nt.0 {
+                NtIdent(ident, is_raw) => Some((*ident, *is_raw)),
                 _ => None,
             },
             _ => None,
@@ -628,8 +628,8 @@ impl Token {
         // We avoid using `Token::uninterpolate` here because it's slow.
         match &self.kind {
             &Lifetime(name) => Some(Ident::new(name, self.span)),
-            Interpolated(nt) => match **nt {
-                NtLifetime(ident) => Some(ident),
+            Interpolated(nt) => match &nt.0 {
+                NtLifetime(ident) => Some(*ident),
                 _ => None,
             },
             _ => None,
@@ -655,7 +655,7 @@ impl Token {
     /// Returns `true` if the token is an interpolated path.
     fn is_path(&self) -> bool {
         if let Interpolated(nt) = &self.kind
-            && let NtPath(..) = **nt
+            && let NtPath(..) = &nt.0
         {
             return true;
         }
@@ -668,7 +668,7 @@ impl Token {
     /// (which happens while parsing the result of macro expansion)?
     pub fn is_whole_expr(&self) -> bool {
         if let Interpolated(nt) = &self.kind
-            && let NtExpr(_) | NtLiteral(_) | NtPath(_) | NtBlock(_) = **nt
+            && let NtExpr(_) | NtLiteral(_) | NtPath(_) | NtBlock(_) = &nt.0
         {
             return true;
         }
@@ -679,7 +679,7 @@ impl Token {
     /// Is the token an interpolated block (`$b:block`)?
     pub fn is_whole_block(&self) -> bool {
         if let Interpolated(nt) = &self.kind
-            && let NtBlock(..) = **nt
+            && let NtBlock(..) = &nt.0
         {
             return true;
         }
@@ -927,7 +927,7 @@ impl fmt::Display for NonterminalKind {
 }
 
 impl Nonterminal {
-    pub fn span(&self) -> Span {
+    pub fn use_span(&self) -> Span {
         match self {
             NtItem(item) => item.span,
             NtBlock(block) => block.span,
@@ -941,6 +941,23 @@ impl Nonterminal {
             NtVis(vis) => vis.span,
         }
     }
+
+    pub fn descr(&self) -> &'static str {
+        match self {
+            NtItem(..) => "item",
+            NtBlock(..) => "block",
+            NtStmt(..) => "statement",
+            NtPat(..) => "pattern",
+            NtExpr(..) => "expression",
+            NtLiteral(..) => "literal",
+            NtTy(..) => "type",
+            NtIdent(..) => "identifier",
+            NtLifetime(..) => "lifetime",
+            NtMeta(..) => "attribute",
+            NtPath(..) => "path",
+            NtVis(..) => "visibility",
+        }
+    }
 }
 
 impl PartialEq for Nonterminal {

compiler/rustc_ast/src/tokenstream.rs

@@ -477,13 +477,13 @@ impl TokenStream {
 
     fn flatten_token(token: &Token, spacing: Spacing) -> TokenTree {
         match &token.kind {
-            token::Interpolated(nt) if let token::NtIdent(ident, is_raw) = **nt => {
+            token::Interpolated(nt) if let token::NtIdent(ident, is_raw) = nt.0 => {
                 TokenTree::Token(Token::new(token::Ident(ident.name, is_raw), ident.span), spacing)
             }
             token::Interpolated(nt) => TokenTree::Delimited(
                 DelimSpan::from_single(token.span),
                 Delimiter::Invisible,
-                TokenStream::from_nonterminal_ast(nt).flattened(),
+                TokenStream::from_nonterminal_ast(&nt.0).flattened(),
             ),
             _ => TokenTree::Token(token.clone(), spacing),
         }

compiler/rustc_ast_pretty/src/pprust/state.rs

@@ -825,7 +825,7 @@ pub trait PrintState<'a>: std::ops::Deref<Target = pp::Printer> + std::ops::Dere
             }
             token::Eof => "<eof>".into(),
 
-            token::Interpolated(ref nt) => self.nonterminal_to_string(nt).into(),
+            token::Interpolated(ref nt) => self.nonterminal_to_string(&nt.0).into(),
         }
     }
 

compiler/rustc_expand/src/mbe/diagnostics.rs

@@ -67,6 +67,12 @@ pub(super) fn failed_to_match_macro<'cx>(
         && (matches!(expected_token.kind, TokenKind::Interpolated(_))
             || matches!(token.kind, TokenKind::Interpolated(_)))
     {
+        if let TokenKind::Interpolated(node) = &expected_token.kind {
+            err.span_label(node.1, "");
+        }
+        if let TokenKind::Interpolated(node) = &token.kind {
+            err.span_label(node.1, "");
+        }
         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");
 

compiler/rustc_expand/src/mbe/macro_parser.rs

@@ -397,7 +397,7 @@ pub(crate) enum NamedMatch {
     MatchedTokenTree(rustc_ast::tokenstream::TokenTree),
 
     // A metavar match of any type other than `tt`.
-    MatchedNonterminal(Lrc<Nonterminal>),
+    MatchedNonterminal(Lrc<(Nonterminal, rustc_span::Span)>),
 }
 
 /// Performs a token equality check, ignoring syntax context (that is, an unhygienic comparison)
@@ -692,7 +692,7 @@ impl TtParser {
                             Ok(nt) => nt,
                         };
                         let m = match nt {
-                            ParseNtResult::Nt(nt) => MatchedNonterminal(Lrc::new(nt)),
+                            ParseNtResult::Nt(nt) => MatchedNonterminal(Lrc::new((nt, span))),
                             ParseNtResult::Tt(tt) => MatchedTokenTree(tt),
                         };
                         mp.push_match(next_metavar, seq_depth, m);

compiler/rustc_expand/src/proc_macro.rs

@@ -126,7 +126,7 @@ impl MultiItemModifier for DeriveProcMacro {
                 Annotatable::Stmt(stmt) => token::NtStmt(stmt),
                 _ => unreachable!(),
             };
-            TokenStream::token_alone(token::Interpolated(Lrc::new(nt)), DUMMY_SP)
+            TokenStream::token_alone(token::Interpolated(Lrc::new((nt, span))), DUMMY_SP)
         } else {
             item.to_tokens()
         };

compiler/rustc_expand/src/proc_macro_server.rs

@@ -226,18 +226,23 @@ impl FromInternal<(TokenStream, &mut Rustc<'_, '_>)> for Vec<TokenTree<TokenStre
                     }));
                 }
 
-                Interpolated(nt) if let NtIdent(ident, is_raw) = *nt => trees
-                    .push(TokenTree::Ident(Ident { sym: ident.name, is_raw, span: ident.span })),
+                Interpolated(ref nt) if let NtIdent(ident, is_raw) = &nt.0 => {
+                    trees.push(TokenTree::Ident(Ident {
+                        sym: ident.name,
+                        is_raw: *is_raw,
+                        span: ident.span,
+                    }))
+                }
 
                 Interpolated(nt) => {
-                    let stream = TokenStream::from_nonterminal_ast(&nt);
+                    let stream = TokenStream::from_nonterminal_ast(&nt.0);
                     // A hack used to pass AST fragments to attribute and derive
                     // macros as a single nonterminal token instead of a token
                     // stream. Such token needs to be "unwrapped" and not
                     // represented as a delimited group.
                     // FIXME: It needs to be removed, but there are some
                     // compatibility issues (see #73345).
-                    if crate::base::nt_pretty_printing_compatibility_hack(&nt, rustc.sess()) {
+                    if crate::base::nt_pretty_printing_compatibility_hack(&nt.0, rustc.sess()) {
                         trees.extend(Self::from_internal((stream, rustc)));
                     } else {
                         trees.push(TokenTree::Group(Group {

compiler/rustc_parse/src/parser/attr.rs

@@ -249,7 +249,7 @@ impl<'a> Parser<'a> {
     /// The delimiters or `=` are still put into the resulting token stream.
     pub fn parse_attr_item(&mut self, capture_tokens: bool) -> PResult<'a, ast::AttrItem> {
         let item = match &self.token.kind {
-            token::Interpolated(nt) => match &**nt {
+            token::Interpolated(nt) => match &nt.0 {
                 Nonterminal::NtMeta(item) => Some(item.clone().into_inner()),
                 _ => None,
             },
@@ -369,7 +369,7 @@ impl<'a> Parser<'a> {
     /// ```
     pub fn parse_meta_item(&mut self) -> PResult<'a, ast::MetaItem> {
         let nt_meta = match &self.token.kind {
-            token::Interpolated(nt) => match &**nt {
+            token::Interpolated(nt) => match &nt.0 {
                 token::NtMeta(e) => Some(e.clone()),
                 _ => None,
             },

compiler/rustc_parse/src/parser/diagnostics.rs

@@ -24,11 +24,12 @@ use crate::parser;
 use rustc_ast as ast;
 use rustc_ast::ptr::P;
 use rustc_ast::token::{self, Delimiter, Lit, LitKind, TokenKind};
+use rustc_ast::tokenstream::AttrTokenTree;
 use rustc_ast::util::parser::AssocOp;
 use rustc_ast::{
     AngleBracketedArg, AngleBracketedArgs, AnonConst, AttrVec, BinOpKind, BindingAnnotation, Block,
-    BlockCheckMode, Expr, ExprKind, GenericArg, Generics, Item, ItemKind, Param, Pat, PatKind,
-    Path, PathSegment, QSelf, Ty, TyKind,
+    BlockCheckMode, Expr, ExprKind, GenericArg, Generics, HasTokens, Item, ItemKind, Param, Pat,
+    PatKind, Path, PathSegment, QSelf, Ty, TyKind,
 };
 use rustc_ast_pretty::pprust;
 use rustc_data_structures::fx::FxHashSet;
@@ -2252,6 +2253,59 @@ impl<'a> Parser<'a> {
             err.subdiagnostic(ExprParenthesesNeeded::surrounding(*sp));
         }
         err.span_label(span, "expected expression");
+
+        // Walk the chain of macro expansions for the current token to point at how the original
+        // code was interpreted. This helps the user realize when a macro argument of one type is
+        // later reinterpreted as a different type, like `$x:expr` being reinterpreted as `$x:pat`
+        // in a subsequent macro invocation (#71039).
+        let mut tok = self.token.clone();
+        let mut labels = vec![];
+        while let TokenKind::Interpolated(node) = &tok.kind {
+            let tokens = node.0.tokens();
+            labels.push(node.clone());
+            if let Some(tokens) = tokens
+                && let tokens = tokens.to_attr_token_stream()
+                && let tokens = tokens.0.deref()
+                && let [AttrTokenTree::Token(token, _)] = &tokens[..]
+            {
+                tok = token.clone();
+            } else {
+                break;
+            }
+        }
+        let mut iter = labels.into_iter().peekable();
+        let mut show_link = false;
+        while let Some(node) = iter.next() {
+            let descr = node.0.descr();
+            if let Some(next) = iter.peek() {
+                let next_descr = next.0.descr();
+                if next_descr != descr {
+                    err.span_label(next.1, format!("this macro fragment matcher is {next_descr}"));
+                    err.span_label(node.1, format!("this macro fragment matcher is {descr}"));
+                    err.span_label(
+                        next.0.use_span(),
+                        format!("this is expected to be {next_descr}"),
+                    );
+                    err.span_label(
+                        node.0.use_span(),
+                        format!(
+                            "this is interpreted as {}, but it is expected to be {}",
+                            next_descr, descr,
+                        ),
+                    );
+                    show_link = true;
+                } else {
+                    err.span_label(node.1, "");
+                }
+            }
+        }
+        if show_link {
+            err.note(
+                "when forwarding a matched fragment to another macro-by-example, matchers in the \
+                 second macro will see an opaque AST of the fragment type, not the underlying \
+                 tokens",
+            );
+        }
         err
     }
 

compiler/rustc_parse/src/parser/expr.rs

@@ -46,7 +46,7 @@ use thin_vec::{thin_vec, ThinVec};
 macro_rules! maybe_whole_expr {
     ($p:expr) => {
         if let token::Interpolated(nt) = &$p.token.kind {
-            match &**nt {
+            match &nt.0 {
                 token::NtExpr(e) | token::NtLiteral(e) => {
                     let e = e.clone();
                     $p.bump();
@@ -1952,7 +1952,7 @@ impl<'a> Parser<'a> {
         mk_lit_char: impl FnOnce(Symbol, Span) -> L,
     ) -> PResult<'a, L> {
         if let token::Interpolated(nt) = &self.token.kind
-            && let token::NtExpr(e) | token::NtLiteral(e) = &**nt
+            && let token::NtExpr(e) | token::NtLiteral(e) = &nt.0
             && matches!(e.kind, ExprKind::Err)
         {
             let mut err = errors::InvalidInterpolatedExpression { span: self.token.span }