Rollup merge of #108801 - fee1-dead-contrib:c-str, r=compiler-errors

Implement RFC 3348, `c"foo"` literals

RFC: rust-lang/rfcs#3348
Tracking issue: #105723

compiler/rustc_ast/src/ast.rs

@@ -1821,6 +1821,8 @@ pub enum LitKind {
     /// 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 C String (`c"foo"`). Guaranteed to only have `\0` at the end.
+    CStr(Lrc<[u8]>, StrStyle),
     /// A byte char (`b'f'`).
     Byte(u8),
     /// A character literal (`'a'`).
@@ -1875,6 +1877,7 @@ impl LitKind {
             // unsuffixed variants
             LitKind::Str(..)
             | LitKind::ByteStr(..)
+            | LitKind::CStr(..)
             | LitKind::Byte(..)
             | LitKind::Char(..)
             | LitKind::Int(_, LitIntType::Unsuffixed)

compiler/rustc_ast/src/token.rs

@@ -74,6 +74,8 @@ pub enum LitKind {
     StrRaw(u8), // raw string delimited by `n` hash symbols
     ByteStr,
     ByteStrRaw(u8), // raw byte string delimited by `n` hash symbols
+    CStr,
+    CStrRaw(u8),
     Err,
 }
 
@@ -141,6 +143,10 @@ impl fmt::Display for Lit {
                 delim = "#".repeat(n as usize),
                 string = symbol
             )?,
+            CStr => write!(f, "c\"{symbol}\"")?,
+            CStrRaw(n) => {
+                write!(f, "cr{delim}\"{symbol}\"{delim}", delim = "#".repeat(n as usize))?
+            }
             Integer | Float | Bool | Err => write!(f, "{symbol}")?,
         }
 
@@ -170,6 +176,7 @@ impl LitKind {
             Float => "float",
             Str | StrRaw(..) => "string",
             ByteStr | ByteStrRaw(..) => "byte string",
+            CStr | CStrRaw(..) => "C string",
             Err => "error",
         }
     }

compiler/rustc_ast/src/util/literal.rs

@@ -2,9 +2,13 @@
 
 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_lexer::unescape::{
+    byte_from_char, unescape_byte, unescape_c_string, unescape_char, unescape_literal, CStrUnit,
+    Mode,
+};
 use rustc_span::symbol::{kw, sym, Symbol};
 use rustc_span::Span;
+use std::ops::Range;
 use std::{ascii, fmt, str};
 
 // Escapes a string, represented as a symbol. Reuses the original symbol,
@@ -35,6 +39,7 @@ pub enum LitError {
     InvalidFloatSuffix,
     NonDecimalFloat(u32),
     IntTooLarge(u32),
+    NulInCStr(Range<usize>),
 }
 
 impl LitKind {
@@ -158,6 +163,52 @@ impl LitKind {
 
                 LitKind::ByteStr(bytes.into(), StrStyle::Raw(n))
             }
+            token::CStr => {
+                let s = symbol.as_str();
+                let mut buf = Vec::with_capacity(s.len());
+                let mut error = Ok(());
+                unescape_c_string(s, Mode::CStr, &mut |span, c| match c {
+                    Ok(CStrUnit::Byte(0) | CStrUnit::Char('\0')) => {
+                        error = Err(LitError::NulInCStr(span));
+                    }
+                    Ok(CStrUnit::Byte(b)) => buf.push(b),
+                    Ok(CStrUnit::Char(c)) if c.len_utf8() == 1 => buf.push(c as u8),
+                    Ok(CStrUnit::Char(c)) => {
+                        buf.extend_from_slice(c.encode_utf8(&mut [0; 4]).as_bytes())
+                    }
+                    Err(err) => {
+                        if err.is_fatal() {
+                            error = Err(LitError::LexerError);
+                        }
+                    }
+                });
+                error?;
+                buf.push(0);
+                LitKind::CStr(buf.into(), StrStyle::Cooked)
+            }
+            token::CStrRaw(n) => {
+                let s = symbol.as_str();
+                let mut buf = Vec::with_capacity(s.len());
+                let mut error = Ok(());
+                unescape_c_string(s, Mode::RawCStr, &mut |span, c| match c {
+                    Ok(CStrUnit::Byte(0) | CStrUnit::Char('\0')) => {
+                        error = Err(LitError::NulInCStr(span));
+                    }
+                    Ok(CStrUnit::Byte(b)) => buf.push(b),
+                    Ok(CStrUnit::Char(c)) if c.len_utf8() == 1 => buf.push(c as u8),
+                    Ok(CStrUnit::Char(c)) => {
+                        buf.extend_from_slice(c.encode_utf8(&mut [0; 4]).as_bytes())
+                    }
+                    Err(err) => {
+                        if err.is_fatal() {
+                            error = Err(LitError::LexerError);
+                        }
+                    }
+                });
+                error?;
+                buf.push(0);
+                LitKind::CStr(buf.into(), StrStyle::Raw(n))
+            }
             token::Err => LitKind::Err,
         })
     }
@@ -191,6 +242,14 @@ impl fmt::Display for LitKind {
                     string = symbol
                 )?;
             }
+            LitKind::CStr(ref bytes, StrStyle::Cooked) => {
+                write!(f, "c\"{}\"", escape_byte_str_symbol(bytes))?
+            }
+            LitKind::CStr(ref bytes, StrStyle::Raw(n)) => {
+                // This can only be valid UTF-8.
+                let symbol = str::from_utf8(bytes).unwrap();
+                write!(f, "cr{delim}\"{symbol}\"{delim}", delim = "#".repeat(n as usize),)?;
+            }
             LitKind::Int(n, ty) => {
                 write!(f, "{n}")?;
                 match ty {
@@ -237,6 +296,8 @@ impl MetaItemLit {
             LitKind::Str(_, ast::StrStyle::Raw(n)) => token::StrRaw(n),
             LitKind::ByteStr(_, ast::StrStyle::Cooked) => token::ByteStr,
             LitKind::ByteStr(_, ast::StrStyle::Raw(n)) => token::ByteStrRaw(n),
+            LitKind::CStr(_, ast::StrStyle::Cooked) => token::CStr,
+            LitKind::CStr(_, ast::StrStyle::Raw(n)) => token::CStrRaw(n),
             LitKind::Byte(_) => token::Byte,
             LitKind::Char(_) => token::Char,
             LitKind::Int(..) => token::Integer,

compiler/rustc_ast_passes/src/feature_gate.rs

@@ -572,6 +572,7 @@ pub fn check_crate(krate: &ast::Crate, sess: &Session) {
             }
         };
     }
+    gate_all!(c_str_literals, "`c\"..\"` literals are experimental");
     gate_all!(
         if_let_guard,
         "`if let` guards are experimental",

compiler/rustc_ast_pretty/src/pprust/state.rs

@@ -210,6 +210,10 @@ pub fn literal_to_string(lit: token::Lit) -> String {
         token::ByteStrRaw(n) => {
             format!("br{delim}\"{string}\"{delim}", delim = "#".repeat(n as usize), string = symbol)
         }
+        token::CStr => format!("c\"{symbol}\""),
+        token::CStrRaw(n) => {
+            format!("cr{delim}\"{symbol}\"{delim}", delim = "#".repeat(n as usize))
+        }
         token::Integer | token::Float | token::Bool | token::Err => symbol.to_string(),
     };
 

compiler/rustc_builtin_macros/src/concat.rs

@@ -32,6 +32,10 @@ pub fn expand_concat(
                 Ok(ast::LitKind::Bool(b)) => {
                     accumulator.push_str(&b.to_string());
                 }
+                Ok(ast::LitKind::CStr(..)) => {
+                    cx.span_err(e.span, "cannot concatenate a C string literal");
+                    has_errors = true;
+                }
                 Ok(ast::LitKind::Byte(..) | ast::LitKind::ByteStr(..)) => {
                     cx.emit_err(errors::ConcatBytestr { span: e.span });
                     has_errors = true;

compiler/rustc_builtin_macros/src/concat_bytes.rs

@@ -18,6 +18,11 @@ fn invalid_type_err(
     };
     let snippet = cx.sess.source_map().span_to_snippet(span).ok();
     match ast::LitKind::from_token_lit(token_lit) {
+        Ok(ast::LitKind::CStr(_, _)) => {
+            // FIXME(c_str_literals): should concatenation of C string literals
+            // include the null bytes in the end?
+            cx.span_err(span, "cannot concatenate C string literals");
+        }
         Ok(ast::LitKind::Char(_)) => {
             let sugg =
                 snippet.map(|snippet| ConcatBytesInvalidSuggestion::CharLit { span, snippet });

compiler/rustc_expand/src/proc_macro_server.rs

@@ -61,6 +61,8 @@ impl FromInternal<token::LitKind> for LitKind {
             token::StrRaw(n) => LitKind::StrRaw(n),
             token::ByteStr => LitKind::ByteStr,
             token::ByteStrRaw(n) => LitKind::ByteStrRaw(n),
+            token::CStr => LitKind::CStr,
+            token::CStrRaw(n) => LitKind::CStrRaw(n),
             token::Err => LitKind::Err,
             token::Bool => unreachable!(),
         }
@@ -78,6 +80,8 @@ impl ToInternal<token::LitKind> for LitKind {
             LitKind::StrRaw(n) => token::StrRaw(n),
             LitKind::ByteStr => token::ByteStr,
             LitKind::ByteStrRaw(n) => token::ByteStrRaw(n),
+            LitKind::CStr => token::CStr,
+            LitKind::CStrRaw(n) => token::CStrRaw(n),
             LitKind::Err => token::Err,
         }
     }
@@ -436,6 +440,8 @@ impl server::FreeFunctions for Rustc<'_, '_> {
                 | token::LitKind::StrRaw(_)
                 | token::LitKind::ByteStr
                 | token::LitKind::ByteStrRaw(_)
+                | token::LitKind::CStr
+                | token::LitKind::CStrRaw(_)
                 | token::LitKind::Err => return Err(()),
                 token::LitKind::Integer | token::LitKind::Float => {}
             }

compiler/rustc_feature/src/active.rs

@@ -313,6 +313,8 @@ declare_features! (
     (active, async_closure, "1.37.0", Some(62290), None),
     /// Allows async functions to be declared, implemented, and used in traits.
     (active, async_fn_in_trait, "1.66.0", Some(91611), None),
+    /// Allows `c"foo"` literals.
+    (active, c_str_literals, "CURRENT_RUSTC_VERSION", Some(105723), None),
     /// Treat `extern "C"` function as nounwind.
     (active, c_unwind, "1.52.0", Some(74990), None),
     /// Allows using C-variadics.

compiler/rustc_hir/src/lang_items.rs

@@ -333,6 +333,7 @@ language_item_table! {
     RangeTo,                 sym::RangeTo,             range_to_struct,            Target::Struct,         GenericRequirement::None;
 
     String,                  sym::String,              string,                     Target::Struct,         GenericRequirement::None;
+    CStr,                    sym::CStr,                c_str,                      Target::Struct,         GenericRequirement::None;
 }
 
 pub enum GenericRequirement {

compiler/rustc_hir_typeck/src/fn_ctxt/checks.rs

@@ -1300,6 +1300,11 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
                 opt_ty.unwrap_or_else(|| self.next_float_var())
             }
             ast::LitKind::Bool(_) => tcx.types.bool,
+            ast::LitKind::CStr(_, _) => tcx.mk_imm_ref(
+                tcx.lifetimes.re_static,
+                tcx.type_of(tcx.require_lang_item(hir::LangItem::CStr, Some(lit.span)))
+                    .skip_binder(),
+            ),
             ast::LitKind::Err => tcx.ty_error_misc(),
         }
     }

compiler/rustc_lexer/src/lib.rs

@@ -186,12 +186,16 @@ pub enum LiteralKind {
     Str { terminated: bool },
     /// "b"abc"", "b"abc"
     ByteStr { terminated: bool },
+    /// `c"abc"`, `c"abc`
+    CStr { terminated: bool },
     /// "r"abc"", "r#"abc"#", "r####"ab"###"c"####", "r#"a". `None` indicates
     /// an invalid literal.
     RawStr { n_hashes: Option<u8> },
     /// "br"abc"", "br#"abc"#", "br####"ab"###"c"####", "br#"a". `None`
     /// indicates an invalid literal.
     RawByteStr { n_hashes: Option<u8> },
+    /// `cr"abc"`, "cr#"abc"#", `cr#"a`. `None` indicates an invalid literal.
+    RawCStr { n_hashes: Option<u8> },
 }
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq, PartialOrd, Ord)]
@@ -357,39 +361,18 @@ impl Cursor<'_> {
             },
 
             // Byte literal, byte string literal, raw byte string literal or identifier.
-            'b' => match (self.first(), self.second()) {
-                ('\'', _) => {
-                    self.bump();
-                    let terminated = self.single_quoted_string();
-                    let suffix_start = self.pos_within_token();
-                    if terminated {
-                        self.eat_literal_suffix();
-                    }
-                    let kind = Byte { terminated };
-                    Literal { kind, suffix_start }
-                }
-                ('"', _) => {
-                    self.bump();
-                    let terminated = self.double_quoted_string();
-                    let suffix_start = self.pos_within_token();
-                    if terminated {
-                        self.eat_literal_suffix();
-                    }
-                    let kind = ByteStr { terminated };
-                    Literal { kind, suffix_start }
-                }
-                ('r', '"') | ('r', '#') => {
-                    self.bump();
-                    let res = self.raw_double_quoted_string(2);
-                    let suffix_start = self.pos_within_token();
-                    if res.is_ok() {
-                        self.eat_literal_suffix();
-                    }
-                    let kind = RawByteStr { n_hashes: res.ok() };
-                    Literal { kind, suffix_start }
-                }
-                _ => self.ident_or_unknown_prefix(),
-            },
+            'b' => self.c_or_byte_string(
+                |terminated| ByteStr { terminated },
+                |n_hashes| RawByteStr { n_hashes },
+                Some(|terminated| Byte { terminated }),
+            ),
+
+            // c-string literal, raw c-string literal or identifier.
+            'c' => self.c_or_byte_string(
+                |terminated| CStr { terminated },
+                |n_hashes| RawCStr { n_hashes },
+                None,
+            ),
 
             // Identifier (this should be checked after other variant that can
             // start as identifier).
@@ -553,6 +536,47 @@ impl Cursor<'_> {
         }
     }
 
+    fn c_or_byte_string(
+        &mut self,
+        mk_kind: impl FnOnce(bool) -> LiteralKind,
+        mk_kind_raw: impl FnOnce(Option<u8>) -> LiteralKind,
+        single_quoted: Option<fn(bool) -> LiteralKind>,
+    ) -> TokenKind {
+        match (self.first(), self.second(), single_quoted) {
+            ('\'', _, Some(mk_kind)) => {
+                self.bump();
+                let terminated = self.single_quoted_string();
+                let suffix_start = self.pos_within_token();
+                if terminated {
+                    self.eat_literal_suffix();
+                }
+                let kind = mk_kind(terminated);
+                Literal { kind, suffix_start }
+            }
+            ('"', _, _) => {
+                self.bump();
+                let terminated = self.double_quoted_string();
+                let suffix_start = self.pos_within_token();
+                if terminated {
+                    self.eat_literal_suffix();
+                }
+                let kind = mk_kind(terminated);
+                Literal { kind, suffix_start }
+            }
+            ('r', '"', _) | ('r', '#', _) => {
+                self.bump();
+                let res = self.raw_double_quoted_string(2);
+                let suffix_start = self.pos_within_token();
+                if res.is_ok() {
+                    self.eat_literal_suffix();
+                }
+                let kind = mk_kind_raw(res.ok());
+                Literal { kind, suffix_start }
+            }
+            _ => self.ident_or_unknown_prefix(),
+        }
+    }
+
     fn number(&mut self, first_digit: char) -> LiteralKind {
         debug_assert!('0' <= self.prev() && self.prev() <= '9');
         let mut base = Base::Decimal;