Rollup merge of rust-lang#73345 - petrochenkov:nointerp, r=Aaron1011

expand: Stop using nonterminals for passing tokens to attribute and derive macros

Make one more step towards fully token-based expansion and fix issues described in rust-lang#72545 (comment).

Now `struct S;` is passed to `foo!(struct S;)` and `#[foo] struct S;` in the same way - as a token stream `struct S ;`, rather than a single non-terminal token `NtItem` which is then broken into parts later.

The cost is making pretty-printing of token streams less pretty.
Some of the pretty-printing regressions will be recovered by keeping jointness with each token, which we will need to do anyway.

Unfortunately, this is not exactly the same thing as rust-lang#73102.
One more observable effect is how `$crate` is printed in the attribute input.
Inside `NtItem` was printed as `crate` or `that_crate`, now as a part of a token stream it's printed as `$crate` (there are good reasons for these differences, see rust-lang#62393 and related PRs).
This may break old proc macros (custom derives) written before the main portion of the proc macro API (macros 1.2) was stabilized, those macros did `input.to_string()` and reparsed the result, now that result can contain `$crate` which cannot be reparsed.

So, I think we should do this regardless, but we need to run crater first.
r? @Aaron1011

Author: Manishearth

src/librustc_ast/attr/mod.rs

@@ -475,7 +475,7 @@ impl MetaItem {
                 let span = span.with_hi(segments.last().unwrap().ident.span.hi());
                 Path { span, segments }
             }
-            Some(TokenTree::Token(Token { kind: token::Interpolated(nt, _), .. })) => match *nt {
+            Some(TokenTree::Token(Token { kind: token::Interpolated(nt), .. })) => match *nt {
                 token::Nonterminal::NtMeta(ref item) => return item.meta(item.path.span),
                 token::Nonterminal::NtPath(ref path) => path.clone(),
                 _ => return None,

src/librustc_ast/mut_visit.rs

@@ -656,7 +656,7 @@ pub fn noop_visit_token<T: MutVisitor>(t: &mut Token, vis: &mut T) {
             *span = ident.span;
             return; // Avoid visiting the span for the second time.
         }
-        token::Interpolated(nt, _) => {
+        token::Interpolated(nt) => {
             let mut nt = Lrc::make_mut(nt);
             vis.visit_interpolated(&mut nt);
         }

src/librustc_ast/token.rs

@@ -11,7 +11,7 @@ use crate::tokenstream::TokenTree;
 use rustc_data_structures::stable_hasher::{HashStable, StableHasher};
 use rustc_data_structures::sync::Lrc;
 use rustc_macros::HashStable_Generic;
-use rustc_span::symbol::kw;
+use rustc_span::symbol::{kw, sym};
 use rustc_span::symbol::{Ident, Symbol};
 use rustc_span::{self, Span, DUMMY_SP};
 use std::borrow::Cow;
@@ -182,15 +182,6 @@ fn ident_can_begin_type(name: Symbol, span: Span, is_raw: bool) -> bool {
             .contains(&name)
 }
 
-/// A hack used to pass AST fragments to attribute and derive macros
-/// as a single nonterminal token instead of a token stream.
-/// FIXME: It needs to be removed, but there are some compatibility issues (see #73345).
-#[derive(Clone, PartialEq, RustcEncodable, RustcDecodable, Debug, HashStable_Generic)]
-pub enum FlattenGroup {
-    Yes,
-    No,
-}
-
 #[derive(Clone, PartialEq, RustcEncodable, RustcDecodable, Debug, HashStable_Generic)]
 pub enum TokenKind {
     /* Expression-operator symbols. */
@@ -245,7 +236,7 @@ pub enum TokenKind {
     /// treat regular and interpolated lifetime identifiers in the same way.
     Lifetime(Symbol),
 
-    Interpolated(Lrc<Nonterminal>, FlattenGroup),
+    Interpolated(Lrc<Nonterminal>),
 
     // Can be expanded into several tokens.
     /// A doc comment.
@@ -352,7 +343,7 @@ 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.span(),
             _ => self.span,
         }
     }
@@ -391,7 +382,7 @@ impl Token {
             ModSep                            | // global path
             Lifetime(..)                      | // labeled loop
             Pound                             => true, // expression attributes
-            Interpolated(ref nt, _) => match **nt {
+            Interpolated(ref nt) => match **nt {
                 NtLiteral(..) |
                 NtExpr(..)    |
                 NtBlock(..)   |
@@ -417,7 +408,7 @@ impl Token {
             Lifetime(..)                | // lifetime bound in trait object
             Lt | BinOp(Shl)             | // associated path
             ModSep                      => true, // global path
-            Interpolated(ref nt, _) => match **nt {
+            Interpolated(ref nt) => match **nt {
                 NtTy(..) | NtPath(..) => true,
                 _ => false,
             },
@@ -429,7 +420,7 @@ impl Token {
     pub fn can_begin_const_arg(&self) -> bool {
         match self.kind {
             OpenDelim(Brace) => true,
-            Interpolated(ref nt, _) => match **nt {
+            Interpolated(ref nt) => match **nt {
                 NtExpr(..) | NtBlock(..) | NtLiteral(..) => true,
                 _ => false,
             },
@@ -464,7 +455,7 @@ 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 {
                 NtLiteral(_) => true,
                 NtExpr(e) => match &e.kind {
                     ast::ExprKind::Lit(_) => true,
@@ -485,7 +476,7 @@ impl Token {
     // otherwise returns the original token.
     pub fn uninterpolate(&self) -> Cow<'_, Token> {
         match &self.kind {
-            Interpolated(nt, _) => match **nt {
+            Interpolated(nt) => match **nt {
                 NtIdent(ident, is_raw) => {
                     Cow::Owned(Token::new(Ident(ident.name, is_raw), ident.span))
                 }
@@ -532,7 +523,7 @@ impl Token {
 
     /// Returns `true` if the token is an interpolated path.
     fn is_path(&self) -> bool {
-        if let Interpolated(ref nt, _) = self.kind {
+        if let Interpolated(ref nt) = self.kind {
             if let NtPath(..) = **nt {
                 return true;
             }
@@ -544,7 +535,7 @@ impl Token {
     /// That is, is this a pre-parsed expression dropped into the token stream
     /// (which happens while parsing the result of macro expansion)?
     pub fn is_whole_expr(&self) -> bool {
-        if let Interpolated(ref nt, _) = self.kind {
+        if let Interpolated(ref nt) = self.kind {
             if let NtExpr(_) | NtLiteral(_) | NtPath(_) | NtIdent(..) | NtBlock(_) = **nt {
                 return true;
             }
@@ -555,7 +546,7 @@ impl Token {
 
     // Is the token an interpolated block (`$b:block`)?
     pub fn is_whole_block(&self) -> bool {
-        if let Interpolated(ref nt, _) = self.kind {
+        if let Interpolated(ref nt) = self.kind {
             if let NtBlock(..) = **nt {
                 return true;
             }
@@ -785,6 +776,26 @@ impl Nonterminal {
             NtTT(tt) => tt.span(),
         }
     }
+
+    /// This nonterminal looks like some specific enums from
+    /// `proc-macro-hack` and `procedural-masquerade` crates.
+    /// We need to maintain some special pretty-printing behavior for them due to incorrect
+    /// asserts in old versions of those crates and their wide use in the ecosystem.
+    /// See issue #73345 for more details.
+    /// FIXME(#73933): Remove this eventually.
+    pub fn pretty_printing_compatibility_hack(&self) -> bool {
+        if let NtItem(item) = self {
+            let name = item.ident.name;
+            if name == sym::ProceduralMasqueradeDummyType || name == sym::ProcMacroHack {
+                if let ast::ItemKind::Enum(enum_def, _) = &item.kind {
+                    if let [variant] = &*enum_def.variants {
+                        return variant.ident.name == sym::Input;
+                    }
+                }
+            }
+        }
+        false
+    }
 }
 
 impl PartialEq for Nonterminal {

src/librustc_ast/util/literal.rs

@@ -205,7 +205,7 @@ impl Lit {
                 token::Lit::new(token::Bool, name, None)
             }
             token::Literal(lit) => lit,
-            token::Interpolated(ref nt, _) => {
+            token::Interpolated(ref nt) => {
                 if let token::NtExpr(expr) | token::NtLiteral(expr) = &**nt {
                     if let ast::ExprKind::Lit(lit) = &expr.kind {
                         return Ok(lit.clone());

src/librustc_ast_lowering/lib.rs

@@ -1027,7 +1027,7 @@ impl<'a, 'hir> LoweringContext<'a, 'hir> {
 
     fn lower_token(&mut self, token: Token) -> TokenStream {
         match token.kind {
-            token::Interpolated(nt, _) => {
+            token::Interpolated(nt) => {
                 let tts = (self.nt_to_tokenstream)(&nt, &self.sess.parse_sess, token.span);
                 self.lower_token_stream(tts)
             }

src/librustc_ast_pretty/pprust.rs

@@ -148,9 +148,14 @@ pub fn to_string(f: impl FnOnce(&mut State<'_>)) -> String {
     printer.s.eof()
 }
 
-// This makes comma-separated lists look slightly nicer,
-// and also addresses a specific regression described in issue #63896.
+// This makes printed token streams look slightly nicer,
+// and also addresses some specific regressions described in #63896 and #73345.
 fn tt_prepend_space(tt: &TokenTree, prev: &TokenTree) -> bool {
+    if let TokenTree::Token(token) = prev {
+        if let token::DocComment(s) = token.kind {
+            return !s.as_str().starts_with("//");
+        }
+    }
     match tt {
         TokenTree::Token(token) => match token.kind {
             token::Comma => false,
@@ -163,7 +168,14 @@ fn tt_prepend_space(tt: &TokenTree, prev: &TokenTree) -> bool {
             },
             _ => true,
         },
-        _ => true,
+        TokenTree::Delimited(_, DelimToken::Bracket, _) => match prev {
+            TokenTree::Token(token) => match token.kind {
+                token::Pound => false,
+                _ => true,
+            },
+            _ => true,
+        },
+        TokenTree::Delimited(..) => true,
     }
 }
 
@@ -266,7 +278,7 @@ fn token_kind_to_string_ext(tok: &TokenKind, convert_dollar_crate: Option<Span>)
         token::Shebang(s) => format!("/* shebang: {}*/", s),
         token::Unknown(s) => s.to_string(),
 
-        token::Interpolated(ref nt, _) => nonterminal_to_string(nt),
+        token::Interpolated(ref nt) => nonterminal_to_string(nt),
     }
 }
 

src/librustc_expand/base.rs

@@ -4,14 +4,14 @@ use crate::module::DirectoryOwnership;
 use rustc_ast::ast::{self, Attribute, NodeId, PatKind};
 use rustc_ast::mut_visit::{self, MutVisitor};
 use rustc_ast::ptr::P;
-use rustc_ast::token::{self, FlattenGroup};
-use rustc_ast::tokenstream::{self, TokenStream, TokenTree};
+use rustc_ast::token;
+use rustc_ast::tokenstream::{self, TokenStream};
 use rustc_ast::visit::{AssocCtxt, Visitor};
 use rustc_attr::{self as attr, Deprecation, HasAttrs, Stability};
 use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::{self, Lrc};
 use rustc_errors::{DiagnosticBuilder, ErrorReported};
-use rustc_parse::{self, parser, MACRO_ARGUMENTS};
+use rustc_parse::{self, nt_to_tokenstream, parser, MACRO_ARGUMENTS};
 use rustc_session::{parse::ParseSess, Limit};
 use rustc_span::def_id::DefId;
 use rustc_span::edition::Edition;
@@ -120,10 +120,7 @@ impl Annotatable {
         }
     }
 
-    crate fn into_tokens(self) -> TokenStream {
-        // `Annotatable` can be converted into tokens directly, but we
-        // are packing it into a nonterminal as a piece of AST to make
-        // the produced token stream look nicer in pretty-printed form.
+    crate fn into_tokens(self, sess: &ParseSess) -> TokenStream {
         let nt = match self {
             Annotatable::Item(item) => token::NtItem(item),
             Annotatable::TraitItem(item) | Annotatable::ImplItem(item) => {
@@ -142,7 +139,7 @@ impl Annotatable {
             | Annotatable::StructField(..)
             | Annotatable::Variant(..) => panic!("unexpected annotatable"),
         };
-        TokenTree::token(token::Interpolated(Lrc::new(nt), FlattenGroup::Yes), DUMMY_SP).into()
+        nt_to_tokenstream(&nt, sess, DUMMY_SP)
     }
 
     pub fn expect_item(self) -> P<ast::Item> {
@@ -374,7 +371,7 @@ where
         impl MutVisitor for AvoidInterpolatedIdents {
             fn visit_tt(&mut self, tt: &mut tokenstream::TokenTree) {
                 if let tokenstream::TokenTree::Token(token) = tt {
-                    if let token::Interpolated(nt, _) = &token.kind {
+                    if let token::Interpolated(nt) = &token.kind {
                         if let token::NtIdent(ident, is_raw) = **nt {
                             *tt = tokenstream::TokenTree::token(
                                 token::Ident(ident.name, is_raw),

src/librustc_expand/expand.rs

@@ -705,7 +705,7 @@ impl<'a, 'b> MacroExpander<'a, 'b> {
                 SyntaxExtensionKind::Attr(expander) => {
                     self.gate_proc_macro_input(&item);
                     self.gate_proc_macro_attr_item(span, &item);
-                    let tokens = item.into_tokens();
+                    let tokens = item.into_tokens(self.cx.parse_sess);
                     let attr_item = attr.unwrap_normal_item();
                     if let MacArgs::Eq(..) = attr_item.args {
                         self.cx.span_err(span, "key-value macro attributes are not supported");

src/librustc_expand/mbe/macro_parser.rs

@@ -790,7 +790,7 @@ fn may_begin_with(token: &Token, name: Symbol) -> bool {
         },
         sym::block => match token.kind {
             token::OpenDelim(token::Brace) => true,
-            token::Interpolated(ref nt, _) => match **nt {
+            token::Interpolated(ref nt) => match **nt {
                 token::NtItem(_)
                 | token::NtPat(_)
                 | token::NtTy(_)
@@ -804,7 +804,7 @@ fn may_begin_with(token: &Token, name: Symbol) -> bool {
         },
         sym::path | sym::meta => match token.kind {
             token::ModSep | token::Ident(..) => true,
-            token::Interpolated(ref nt, _) => match **nt {
+            token::Interpolated(ref nt) => match **nt {
                 token::NtPath(_) | token::NtMeta(_) => true,
                 _ => may_be_ident(&nt),
             },
@@ -823,12 +823,12 @@ fn may_begin_with(token: &Token, name: Symbol) -> bool {
             token::ModSep |                     // path
             token::Lt |                         // path (UFCS constant)
             token::BinOp(token::Shl) => true,   // path (double UFCS)
-            token::Interpolated(ref nt, _) => may_be_ident(nt),
+            token::Interpolated(ref nt) => may_be_ident(nt),
             _ => false,
         },
         sym::lifetime => match token.kind {
             token::Lifetime(_) => true,
-            token::Interpolated(ref nt, _) => match **nt {
+            token::Interpolated(ref nt) => match **nt {
                 token::NtLifetime(_) | token::NtTT(_) => true,
                 _ => false,
             },

src/librustc_expand/mbe/transcribe.rs

@@ -4,7 +4,7 @@ use crate::mbe::macro_parser::{MatchedNonterminal, MatchedSeq, NamedMatch};
 
 use rustc_ast::ast::MacCall;
 use rustc_ast::mut_visit::{self, MutVisitor};
-use rustc_ast::token::{self, FlattenGroup, NtTT, Token};
+use rustc_ast::token::{self, NtTT, Token};
 use rustc_ast::tokenstream::{DelimSpan, TokenStream, TokenTree, TreeAndJoint};
 use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::Lrc;
@@ -240,10 +240,7 @@ pub(super) fn transcribe<'a>(
                             result.push(tt.clone().into());
                         } else {
                             marker.visit_span(&mut sp);
-                            let token = TokenTree::token(
-                                token::Interpolated(nt.clone(), FlattenGroup::No),
-                                sp,
-                            );
+                            let token = TokenTree::token(token::Interpolated(nt.clone()), sp);
                             result.push(token.into());
                         }
                     } else {

src/librustc_expand/proc_macro.rs

@@ -2,10 +2,11 @@ use crate::base::{self, *};
 use crate::proc_macro_server;
 
 use rustc_ast::ast::{self, ItemKind, MetaItemKind, NestedMetaItem};
-use rustc_ast::token::{self, FlattenGroup};
-use rustc_ast::tokenstream::{self, TokenStream};
+use rustc_ast::token;
+use rustc_ast::tokenstream::{TokenStream, TokenTree};
 use rustc_data_structures::sync::Lrc;
 use rustc_errors::{Applicability, ErrorReported};
+use rustc_parse::nt_to_tokenstream;
 use rustc_span::symbol::sym;
 use rustc_span::{Span, DUMMY_SP};
 
@@ -102,8 +103,12 @@ impl MultiItemModifier for ProcMacroDerive {
             }
         }
 
-        let token = token::Interpolated(Lrc::new(token::NtItem(item)), FlattenGroup::Yes);
-        let input = tokenstream::TokenTree::token(token, DUMMY_SP).into();
+        let item = token::NtItem(item);
+        let input = if item.pretty_printing_compatibility_hack() {
+            TokenTree::token(token::Interpolated(Lrc::new(item)), DUMMY_SP).into()
+        } else {
+            nt_to_tokenstream(&item, ecx.parse_sess, DUMMY_SP)
+        };
 
         let server = proc_macro_server::Rustc::new(ecx);
         let stream = match self.client.run(&EXEC_STRATEGY, server, input) {