Rollup merge of rust-lang#62393 - petrochenkov:notto-disu, r=Mark-Simulacrum

Fix pretty-printing of `$crate` (take 4)

Pretty-print `$crate` as `crate` or `crate_name` in unstructured tokens like `a $crate c` in `foo!(a $crate c)`, but only if those tokens are printed as a part of AST pretty-printing, rather than as a standalone token stream.

Fixes rust-lang#62325
Previous iterations - rust-lang#56647, rust-lang#57155, rust-lang#57915.

Author: Centril

src/librustc_resolve/build_reduced_graph.rs

@@ -758,7 +758,10 @@ impl<'a> Resolver<'a> {
     }
 
     pub fn macro_def_scope(&mut self, expansion: Mark) -> Module<'a> {
-        let def_id = self.macro_defs[&expansion];
+        let def_id = match self.macro_defs.get(&expansion) {
+            Some(def_id) => *def_id,
+            None => return self.graph_root,
+        };
         if let Some(id) = self.definitions.as_local_node_id(def_id) {
             self.local_macro_def_scopes[&id]
         } else if def_id.krate == CrateNum::BuiltinMacros {

src/librustc_resolve/macros.rs

@@ -17,12 +17,11 @@ use syntax::errors::DiagnosticBuilder;
 use syntax::ext::base::{self, Determinacy};
 use syntax::ext::base::{MacroKind, SyntaxExtension};
 use syntax::ext::expand::{AstFragment, Invocation, InvocationKind};
-use syntax::ext::hygiene::Mark;
+use syntax::ext::hygiene::{self, Mark};
 use syntax::ext::tt::macro_rules;
 use syntax::feature_gate::{feature_err, emit_feature_err, is_builtin_attr_name};
 use syntax::feature_gate::{AttributeGate, GateIssue, Stability, BUILTIN_ATTRIBUTES};
 use syntax::symbol::{Symbol, kw, sym};
-use syntax::visit::Visitor;
 use syntax::util::lev_distance::find_best_match_for_name;
 use syntax_pos::{Span, DUMMY_SP};
 use errors::Applicability;
@@ -146,24 +145,14 @@ impl<'a> base::Resolver for Resolver<'a> {
         mark
     }
 
-    fn resolve_dollar_crates(&mut self, fragment: &AstFragment) {
-        struct ResolveDollarCrates<'a, 'b> {
-            resolver: &'a mut Resolver<'b>
-        }
-        impl<'a> Visitor<'a> for ResolveDollarCrates<'a, '_> {
-            fn visit_ident(&mut self, ident: Ident) {
-                if ident.name == kw::DollarCrate {
-                    let name = match self.resolver.resolve_crate_root(ident).kind {
-                        ModuleKind::Def(.., name) if name != kw::Invalid => name,
-                        _ => kw::Crate,
-                    };
-                    ident.span.ctxt().set_dollar_crate_name(name);
-                }
+    fn resolve_dollar_crates(&mut self) {
+        hygiene::update_dollar_crate_names(|ctxt| {
+            let ident = Ident::new(kw::DollarCrate, DUMMY_SP.with_ctxt(ctxt));
+            match self.resolve_crate_root(ident).kind {
+                ModuleKind::Def(.., name) if name != kw::Invalid => name,
+                _ => kw::Crate,
             }
-            fn visit_mac(&mut self, _: &ast::Mac) {}
-        }
-
-        fragment.visit_with(&mut ResolveDollarCrates { resolver: self });
+        });
     }
 
     fn visit_ast_fragment_with_placeholders(&mut self, mark: Mark, fragment: &AstFragment,

src/libsyntax/ext/base.rs

@@ -701,7 +701,7 @@ pub trait Resolver {
 
     fn get_module_scope(&mut self, id: ast::NodeId) -> Mark;
 
-    fn resolve_dollar_crates(&mut self, fragment: &AstFragment);
+    fn resolve_dollar_crates(&mut self);
     fn visit_ast_fragment_with_placeholders(&mut self, mark: Mark, fragment: &AstFragment,
                                             derives: &[Mark]);
     fn add_builtin(&mut self, ident: ast::Ident, ext: Lrc<SyntaxExtension>);

src/libsyntax/ext/expand.rs

@@ -429,7 +429,7 @@ impl<'a, 'b> MacroExpander<'a, 'b> {
     fn collect_invocations(&mut self, mut fragment: AstFragment, derives: &[Mark])
                            -> (AstFragment, Vec<Invocation>) {
         // Resolve `$crate`s in the fragment for pretty-printing.
-        self.cx.resolver.resolve_dollar_crates(&fragment);
+        self.cx.resolver.resolve_dollar_crates();
 
         let invocations = {
             let mut collector = InvocationCollector {

src/libsyntax/print/pprust.rs

@@ -18,7 +18,7 @@ use crate::tokenstream::{self, TokenStream, TokenTree};
 
 use rustc_target::spec::abi::{self, Abi};
 use syntax_pos::{self, BytePos};
-use syntax_pos::{DUMMY_SP, FileName};
+use syntax_pos::{DUMMY_SP, FileName, Span};
 
 use std::borrow::Cow;
 use std::io::Read;
@@ -181,7 +181,46 @@ pub fn literal_to_string(lit: token::Lit) -> String {
     out
 }
 
+fn ident_to_string(ident: ast::Ident, is_raw: bool) -> String {
+    ident_to_string_ext(ident.name, is_raw, Some(ident.span))
+}
+
+// AST pretty-printer is used as a fallback for turning AST structures into token streams for
+// proc macros. Additionally, proc macros may stringify their input and expect it survive the
+// stringification (especially true for proc macro derives written between Rust 1.15 and 1.30).
+// So we need to somehow pretty-print `$crate` in a way preserving at least some of its
+// hygiene data, most importantly name of the crate it refers to.
+// As a result we print `$crate` as `crate` if it refers to the local crate
+// and as `::other_crate_name` if it refers to some other crate.
+// Note, that this is only done if the ident token is printed from inside of AST pretty-pringing,
+// but not otherwise. Pretty-printing is the only way for proc macros to discover token contents,
+// so we should not perform this lossy conversion if the top level call to the pretty-printer was
+// done for a token stream or a single token.
+fn ident_to_string_ext(
+    name: ast::Name, is_raw: bool, convert_dollar_crate: Option<Span>
+) -> String {
+    if is_raw {
+        format!("r#{}", name)
+    } else {
+        if name == kw::DollarCrate {
+            if let Some(span) = convert_dollar_crate {
+                let converted = span.ctxt().dollar_crate_name();
+                return if converted.is_path_segment_keyword() {
+                    converted.to_string()
+                } else {
+                    format!("::{}", converted)
+                }
+            }
+        }
+        name.to_string()
+    }
+}
+
 pub fn token_kind_to_string(tok: &TokenKind) -> String {
+    token_kind_to_string_ext(tok, None)
+}
+
+fn token_kind_to_string_ext(tok: &TokenKind, convert_dollar_crate: Option<Span>) -> String {
     match *tok {
         token::Eq                   => "=".to_string(),
         token::Lt                   => "<".to_string(),
@@ -227,8 +266,7 @@ pub fn token_kind_to_string(tok: &TokenKind) -> String {
         token::Literal(lit) => literal_to_string(lit),
 
         /* Name components */
-        token::Ident(s, false)      => s.to_string(),
-        token::Ident(s, true)       => format!("r#{}", s),
+        token::Ident(s, is_raw)     => ident_to_string_ext(s, is_raw, convert_dollar_crate),
         token::Lifetime(s)          => s.to_string(),
 
         /* Other */
@@ -243,7 +281,12 @@ pub fn token_kind_to_string(tok: &TokenKind) -> String {
 }
 
 pub fn token_to_string(token: &Token) -> String {
-    token_kind_to_string(&token.kind)
+    token_to_string_ext(token, false)
+}
+
+fn token_to_string_ext(token: &Token, convert_dollar_crate: bool) -> String {
+    let convert_dollar_crate = if convert_dollar_crate { Some(token.span) } else { None };
+    token_kind_to_string_ext(&token.kind, convert_dollar_crate)
 }
 
 crate fn nonterminal_to_string(nt: &Nonterminal) -> String {
@@ -256,9 +299,8 @@ crate fn nonterminal_to_string(nt: &Nonterminal) -> String {
         token::NtBlock(ref e)       => block_to_string(e),
         token::NtStmt(ref e)        => stmt_to_string(e),
         token::NtPat(ref e)         => pat_to_string(e),
-        token::NtIdent(e, false)    => ident_to_string(e),
-        token::NtIdent(e, true)     => format!("r#{}", ident_to_string(e)),
-        token::NtLifetime(e)        => ident_to_string(e),
+        token::NtIdent(e, is_raw)   => ident_to_string(e, is_raw),
+        token::NtLifetime(e)        => e.to_string(),
         token::NtLiteral(ref e)     => expr_to_string(e),
         token::NtTT(ref tree)       => tt_to_string(tree.clone()),
         token::NtImplItem(ref e)    => impl_item_to_string(e),
@@ -293,15 +335,15 @@ pub fn lifetime_to_string(lt: &ast::Lifetime) -> String {
 }
 
 pub fn tt_to_string(tt: tokenstream::TokenTree) -> String {
-    to_string(|s| s.print_tt(tt))
+    to_string(|s| s.print_tt(tt, false))
 }
 
 pub fn tts_to_string(tts: &[tokenstream::TokenTree]) -> String {
-    to_string(|s| s.print_tts(tts.iter().cloned().collect()))
+    tokens_to_string(tts.iter().cloned().collect())
 }
 
 pub fn tokens_to_string(tokens: TokenStream) -> String {
-    to_string(|s| s.print_tts(tokens))
+    to_string(|s| s.print_tts_ext(tokens, false))
 }
 
 pub fn stmt_to_string(stmt: &ast::Stmt) -> String {
@@ -344,10 +386,6 @@ pub fn path_segment_to_string(p: &ast::PathSegment) -> String {
     to_string(|s| s.print_path_segment(p, false))
 }
 
-pub fn ident_to_string(id: ast::Ident) -> String {
-    to_string(|s| s.print_ident(id))
-}
-
 pub fn vis_to_string(v: &ast::Visibility) -> String {
     to_string(|s| s.print_visibility(v))
 }
@@ -629,11 +667,7 @@ pub trait PrintState<'a> {
                 self.writer().word("::");
             }
             if segment.ident.name != kw::PathRoot {
-                if segment.ident.name == kw::DollarCrate {
-                    self.print_dollar_crate(segment.ident);
-                } else {
-                    self.writer().word(segment.ident.as_str().to_string());
-                }
+                self.writer().word(ident_to_string(segment.ident, segment.ident.is_raw_guess()));
             }
         }
     }
@@ -707,10 +741,10 @@ pub trait PrintState<'a> {
     /// appropriate macro, transcribe back into the grammar we just parsed from,
     /// and then pretty-print the resulting AST nodes (so, e.g., we print
     /// expression arguments as expressions). It can be done! I think.
-    fn print_tt(&mut self, tt: tokenstream::TokenTree) {
+    fn print_tt(&mut self, tt: tokenstream::TokenTree, convert_dollar_crate: bool) {
         match tt {
             TokenTree::Token(ref token) => {
-                self.writer().word(token_to_string(&token));
+                self.writer().word(token_to_string_ext(&token, convert_dollar_crate));
                 match token.kind {
                     token::DocComment(..) => {
                         self.writer().hardbreak()
@@ -729,12 +763,16 @@ pub trait PrintState<'a> {
     }
 
     fn print_tts(&mut self, tts: tokenstream::TokenStream) {
+        self.print_tts_ext(tts, true)
+    }
+
+    fn print_tts_ext(&mut self, tts: tokenstream::TokenStream, convert_dollar_crate: bool) {
         self.ibox(0);
         for (i, tt) in tts.into_trees().enumerate() {
             if i != 0 {
                 self.writer().space();
             }
-            self.print_tt(tt);
+            self.print_tt(tt, convert_dollar_crate);
         }
         self.end();
     }
@@ -744,21 +782,6 @@ pub trait PrintState<'a> {
     }
 
     fn nbsp(&mut self) { self.writer().word(" ") }
-
-    // AST pretty-printer is used as a fallback for turning AST structures into token streams for
-    // proc macros. Additionally, proc macros may stringify their input and expect it survive the
-    // stringification (especially true for proc macro derives written between Rust 1.15 and 1.30).
-    // So we need to somehow pretty-print `$crate` in paths in a way preserving at least some of
-    // its hygiene data, most importantly name of the crate it refers to.
-    // As a result we print `$crate` as `crate` if it refers to the local crate
-    // and as `::other_crate_name` if it refers to some other crate.
-    fn print_dollar_crate(&mut self, ident: ast::Ident) {
-        let name = ident.span.ctxt().dollar_crate_name();
-        if !ast::Ident::with_empty_ctxt(name).is_path_segment_keyword() {
-            self.writer().word("::");
-        }
-        self.writer().word(name.as_str().to_string())
-    }
 }
 
 impl<'a> PrintState<'a> for State<'a> {
@@ -2287,11 +2310,7 @@ impl<'a> State<'a> {
     }
 
     crate fn print_ident(&mut self, ident: ast::Ident) {
-        if ident.is_raw_guess() {
-            self.s.word(format!("r#{}", ident));
-        } else {
-            self.s.word(ident.as_str().to_string());
-        }
+        self.s.word(ident_to_string(ident, ident.is_raw_guess()));
         self.ann.post(self, AnnNode::Ident(&ident))
     }
 
@@ -2322,11 +2341,7 @@ impl<'a> State<'a> {
                           segment: &ast::PathSegment,
                           colons_before_params: bool) {
         if segment.ident.name != kw::PathRoot {
-            if segment.ident.name == kw::DollarCrate {
-                self.print_dollar_crate(segment.ident);
-            } else {
-                self.print_ident(segment.ident);
-            }
+            self.print_ident(segment.ident);
             if let Some(ref args) = segment.args {
                 self.print_generic_args(args, colons_before_params);
             }

src/libsyntax_pos/hygiene.rs

@@ -33,7 +33,7 @@ use crate::symbol::{kw, Symbol};
 use serialize::{Encodable, Decodable, Encoder, Decoder};
 use rustc_data_structures::fx::{FxHashMap, FxHashSet};
 use rustc_data_structures::sync::Lrc;
-use std::{fmt, mem};
+use std::fmt;
 
 /// A SyntaxContext represents a chain of macro expansions (represented by marks).
 #[derive(Clone, Copy, PartialEq, Eq, Default, PartialOrd, Ord, Hash)]
@@ -387,6 +387,23 @@ pub fn walk_chain(span: Span, to: SyntaxContext) -> Span {
     HygieneData::with(|data| data.walk_chain(span, to))
 }
 
+pub fn update_dollar_crate_names(mut get_name: impl FnMut(SyntaxContext) -> Symbol) {
+    // The new contexts that need updating are at the end of the list and have `$crate` as a name.
+    let (len, to_update) = HygieneData::with(|data| (
+        data.syntax_contexts.len(),
+        data.syntax_contexts.iter().rev()
+            .take_while(|scdata| scdata.dollar_crate_name == kw::DollarCrate).count()
+    ));
+    // The callback must be called from outside of the `HygieneData` lock,
+    // since it will try to acquire it too.
+    let range_to_update = len - to_update .. len;
+    let names: Vec<_> =
+        range_to_update.clone().map(|idx| get_name(SyntaxContext::from_u32(idx as u32))).collect();
+    HygieneData::with(|data| range_to_update.zip(names.into_iter()).for_each(|(idx, name)| {
+        data.syntax_contexts[idx].dollar_crate_name = name;
+    }))
+}
+
 impl SyntaxContext {
     #[inline]
     pub const fn empty() -> Self {
@@ -614,17 +631,6 @@ impl SyntaxContext {
     pub fn dollar_crate_name(self) -> Symbol {
         HygieneData::with(|data| data.syntax_contexts[self.0 as usize].dollar_crate_name)
     }
-
-    pub fn set_dollar_crate_name(self, dollar_crate_name: Symbol) {
-        HygieneData::with(|data| {
-            let prev_dollar_crate_name = mem::replace(
-                &mut data.syntax_contexts[self.0 as usize].dollar_crate_name, dollar_crate_name
-            );
-            assert!(dollar_crate_name == prev_dollar_crate_name ||
-                    prev_dollar_crate_name == kw::DollarCrate,
-                    "$crate name is reset for a syntax context");
-        })
-    }
 }
 
 impl fmt::Debug for SyntaxContext {

src/test/ui/proc-macro/auxiliary/dollar-crate-external.rs

@@ -14,3 +14,9 @@ macro_rules! external {
         struct D($crate::S);
     };
 }
+
+#[macro_export]
+macro_rules! issue_62325 { () => {
+    #[print_attr]
+    struct B(identity!($crate::S));
+}}

src/test/ui/proc-macro/dollar-crate-issue-57089.stdout

@@ -1,4 +1,4 @@
-PRINT-BANG INPUT (DISPLAY): struct M ( $crate :: S ) ;
+PRINT-BANG INPUT (DISPLAY): struct M ( crate :: S ) ;
 PRINT-BANG INPUT (DEBUG): TokenStream [
     Ident {
         ident: "struct",
@@ -39,7 +39,7 @@ PRINT-BANG INPUT (DEBUG): TokenStream [
     },
 ]
 PRINT-ATTR INPUT (DISPLAY): struct A(crate::S);
-PRINT-ATTR RE-COLLECTED (DISPLAY): struct A ( $crate :: S ) ;
+PRINT-ATTR RE-COLLECTED (DISPLAY): struct A ( crate :: S ) ;
 PRINT-ATTR INPUT (DEBUG): TokenStream [
     Ident {
         ident: "struct",

src/test/ui/proc-macro/dollar-crate-issue-62325.rs

@@ -0,0 +1,27 @@
+// check-pass
+// edition:2018
+// aux-build:test-macros.rs
+// aux-build:dollar-crate-external.rs
+
+// Anonymize unstable non-dummy spans while still showing dummy spans `0..0`.
+// normalize-stdout-test "bytes\([^0]\w*\.\.(\w+)\)" -> "bytes(LO..$1)"
+// normalize-stdout-test "bytes\((\w+)\.\.[^0]\w*\)" -> "bytes($1..HI)"
+
+#![feature(proc_macro_hygiene)]
+
+#[macro_use]
+extern crate test_macros;
+extern crate dollar_crate_external;
+
+type S = u8;
+
+macro_rules! m { () => {
+    #[print_attr]
+    struct A(identity!($crate::S));
+}}
+
+m!();
+
+dollar_crate_external::issue_62325!();
+
+fn main() {}