Add StrStyle to ast::LitKind::ByteStr.

This is required to distinguish between cooked and raw byte string
literals in an `ast::LitKind`, without referring to an adjacent
`token::Lit`. It's a prerequisite for the next commit.

Author: nnethercote

compiler/rustc_ast/src/ast.rs

@@ -1796,8 +1796,9 @@ pub enum LitKind {
     /// A string literal (`"foo"`). The symbol is unescaped, and so may differ
     /// from the original token's symbol.
     Str(Symbol, StrStyle),
-    /// A byte string (`b"foo"`).
-    ByteStr(Lrc<[u8]>),
+    /// A byte string (`b"foo"`). Not stored as a symbol because it might be
+    /// non-utf8, and symbols only allow utf8 strings.
+    ByteStr(Lrc<[u8]>, StrStyle),
     /// A byte char (`b'f'`).
     Byte(u8),
     /// A character literal (`'a'`).
@@ -1822,7 +1823,7 @@ impl LitKind {
 
     /// Returns `true` if this literal is byte literal string.
     pub fn is_bytestr(&self) -> bool {
-        matches!(self, LitKind::ByteStr(_))
+        matches!(self, LitKind::ByteStr(..))
     }
 
     /// Returns `true` if this is a numeric literal.

compiler/rustc_ast/src/util/literal.rs

@@ -1,11 +1,12 @@
 //! Code related to parsing literals.
 
-use crate::ast::{self, LitKind, MetaItemLit};
+use crate::ast::{self, LitKind, MetaItemLit, StrStyle};
 use crate::token::{self, Token};
 use rustc_lexer::unescape::{byte_from_char, unescape_byte, unescape_char, unescape_literal, Mode};
 use rustc_span::symbol::{kw, sym, Symbol};
 use rustc_span::Span;
 use std::ascii;
+use std::str;
 
 #[derive(Debug)]
 pub enum LitError {
@@ -115,9 +116,9 @@ impl LitKind {
                     }
                 });
                 error?;
-                LitKind::ByteStr(buf.into())
+                LitKind::ByteStr(buf.into(), StrStyle::Cooked)
             }
-            token::ByteStrRaw(_) => {
+            token::ByteStrRaw(n) => {
                 let s = symbol.as_str();
                 let bytes = if s.contains('\r') {
                     let mut buf = Vec::with_capacity(s.len());
@@ -136,7 +137,7 @@ impl LitKind {
                     symbol.to_string().into_bytes()
                 };
 
-                LitKind::ByteStr(bytes.into())
+                LitKind::ByteStr(bytes.into(), StrStyle::Raw(n))
             }
             token::Err => LitKind::Err,
         })
@@ -155,10 +156,15 @@ impl LitKind {
                 (token::Str, symbol, None)
             }
             LitKind::Str(symbol, ast::StrStyle::Raw(n)) => (token::StrRaw(n), symbol, None),
-            LitKind::ByteStr(ref bytes) => {
+            LitKind::ByteStr(ref bytes, ast::StrStyle::Cooked) => {
                 let string = bytes.escape_ascii().to_string();
                 (token::ByteStr, Symbol::intern(&string), None)
             }
+            LitKind::ByteStr(ref bytes, ast::StrStyle::Raw(n)) => {
+                // Unwrap because raw byte string literals can only contain ASCII.
+                let string = str::from_utf8(bytes).unwrap();
+                (token::ByteStrRaw(n), Symbol::intern(&string), None)
+            }
             LitKind::Byte(byte) => {
                 let string: String = ascii::escape_default(byte).map(Into::<char>::into).collect();
                 (token::Byte, Symbol::intern(&string), None)

compiler/rustc_ast_lowering/src/expr.rs

@@ -97,7 +97,7 @@ impl<'hir> LoweringContext<'_, 'hir> {
                 }
                 ExprKind::IncludedBytes(bytes) => hir::ExprKind::Lit(respan(
                     self.lower_span(e.span),
-                    LitKind::ByteStr(bytes.clone()),
+                    LitKind::ByteStr(bytes.clone(), StrStyle::Cooked),
                 )),
                 ExprKind::Cast(expr, ty) => {
                     let expr = self.lower_expr(expr);

compiler/rustc_ast_pretty/src/pprust/state/expr.rs

@@ -323,7 +323,8 @@ impl<'a> State<'a> {
                 self.print_token_literal(*token_lit, expr.span);
             }
             ast::ExprKind::IncludedBytes(bytes) => {
-                let lit = ast::LitKind::ByteStr(bytes.clone()).synthesize_token_lit();
+                let lit = ast::LitKind::ByteStr(bytes.clone(), ast::StrStyle::Cooked)
+                    .synthesize_token_lit();
                 self.print_token_literal(lit, expr.span)
             }
             ast::ExprKind::Cast(expr, ty) => {

compiler/rustc_builtin_macros/src/concat_bytes.rs

@@ -69,7 +69,7 @@ fn invalid_type_err(
         Ok(ast::LitKind::Int(_, _)) => {
             cx.span_err(span, "numeric literal is not a `u8`");
         }
-        Ok(ast::LitKind::ByteStr(_) | ast::LitKind::Byte(_)) => unreachable!(),
+        Ok(ast::LitKind::ByteStr(..) | ast::LitKind::Byte(_)) => unreachable!(),
         Err(err) => {
             report_lit_error(&cx.sess.parse_sess, err, token_lit, span);
         }
@@ -97,7 +97,7 @@ fn handle_array_element(
             )) if val <= u8::MAX.into() => Some(val as u8),
 
             Ok(ast::LitKind::Byte(val)) => Some(val),
-            Ok(ast::LitKind::ByteStr(_)) => {
+            Ok(ast::LitKind::ByteStr(..)) => {
                 if !*has_errors {
                     cx.struct_span_err(expr.span, "cannot concatenate doubly nested array")
                         .note("byte strings are treated as arrays of bytes")
@@ -174,7 +174,7 @@ pub fn expand_concat_bytes(
                 Ok(ast::LitKind::Byte(val)) => {
                     accumulator.push(val);
                 }
-                Ok(ast::LitKind::ByteStr(ref bytes)) => {
+                Ok(ast::LitKind::ByteStr(ref bytes, _)) => {
                     accumulator.extend_from_slice(&bytes);
                 }
                 _ => {

compiler/rustc_expand/src/base.rs

@@ -1234,7 +1234,7 @@ pub fn expr_to_spanned_string<'a>(
     Err(match expr.kind {
         ast::ExprKind::Lit(token_lit) => match ast::LitKind::from_token_lit(token_lit) {
             Ok(ast::LitKind::Str(s, style)) => return Ok((s, style, expr.span)),
-            Ok(ast::LitKind::ByteStr(_)) => {
+            Ok(ast::LitKind::ByteStr(..)) => {
                 let mut err = cx.struct_span_err(expr.span, err_msg);
                 let span = expr.span.shrink_to_lo();
                 err.span_suggestion(

compiler/rustc_expand/src/build.rs

@@ -361,7 +361,7 @@ impl<'a> ExtCtxt<'a> {
     }
 
     pub fn expr_byte_str(&self, sp: Span, bytes: Vec<u8>) -> P<ast::Expr> {
-        self.expr_lit(sp, ast::LitKind::ByteStr(Lrc::from(bytes)))
+        self.expr_lit(sp, ast::LitKind::ByteStr(Lrc::from(bytes), ast::StrStyle::Cooked))
     }
 
     /// `[expr1, expr2, ...]`

compiler/rustc_expand/src/proc_macro_server.rs

@@ -526,7 +526,8 @@ impl server::TokenStream for Rustc<'_, '_> {
                 Ok(tokenstream::TokenStream::token_alone(token::Literal(*token_lit), expr.span))
             }
             ast::ExprKind::IncludedBytes(bytes) => {
-                let lit = ast::LitKind::ByteStr(bytes.clone()).synthesize_token_lit();
+                let lit = ast::LitKind::ByteStr(bytes.clone(), ast::StrStyle::Cooked)
+                    .synthesize_token_lit();
                 Ok(tokenstream::TokenStream::token_alone(token::TokenKind::Literal(lit), expr.span))
             }
             ast::ExprKind::Unary(ast::UnOp::Neg, e) => match &e.kind {

compiler/rustc_hir_typeck/src/fn_ctxt/checks.rs

@@ -1169,7 +1169,7 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
 
         match lit.node {
             ast::LitKind::Str(..) => tcx.mk_static_str(),
-            ast::LitKind::ByteStr(ref v) => {
+            ast::LitKind::ByteStr(ref v, _) => {
                 tcx.mk_imm_ref(tcx.lifetimes.re_static, tcx.mk_array(tcx.types.u8, v.len() as u64))
             }
             ast::LitKind::Byte(_) => tcx.types.u8,

compiler/rustc_hir_typeck/src/pat.rs

@@ -386,7 +386,7 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
         // Byte string patterns behave the same way as array patterns
         // They can denote both statically and dynamically-sized byte arrays.
         let mut pat_ty = ty;
-        if let hir::ExprKind::Lit(Spanned { node: ast::LitKind::ByteStr(_), .. }) = lt.kind {
+        if let hir::ExprKind::Lit(Spanned { node: ast::LitKind::ByteStr(..), .. }) = lt.kind {
             let expected = self.structurally_resolved_type(span, expected);
             if let ty::Ref(_, inner_ty, _) = expected.kind()
                 && matches!(inner_ty.kind(), ty::Slice(_))