From 982918f4935fd4677af06d42e0d0b298bfb1c243 Mon Sep 17 00:00:00 2001 From: Esteban Küber Date: Sat, 9 Mar 2024 00:35:57 +0000 Subject: Handle str literals written with `'` lexed as lifetime Given `'hello world'` and `'1 str', provide a structured suggestion for a valid string literal: ``` error[E0762]: unterminated character literal --> $DIR/lex-bad-str-literal-as-char-3.rs:2:26 | LL | println!('hello world'); | ^^^^ | help: if you meant to write a `str` literal, use double quotes | LL | println!("hello world"); | ~ ~ ``` ``` error[E0762]: unterminated character literal --> $DIR/lex-bad-str-literal-as-char-1.rs:2:20 | LL | println!('1 + 1'); | ^^^^ | help: if you meant to write a `str` literal, use double quotes | LL | println!("1 + 1"); | ~ ~ ``` Fix #119685. --- compiler/rustc_parse/src/errors.rs | 11 +++++++++ compiler/rustc_parse/src/lexer/mod.rs | 46 ++++++++++++++++++++++++++++++++--- 2 files changed, 53 insertions(+), 4 deletions(-) (limited to 'compiler/rustc_parse/src') diff --git a/compiler/rustc_parse/src/errors.rs b/compiler/rustc_parse/src/errors.rs index 32b56bb7e87..140eb6cd187 100644 --- a/compiler/rustc_parse/src/errors.rs +++ b/compiler/rustc_parse/src/errors.rs @@ -1994,6 +1994,17 @@ pub enum UnknownPrefixSugg { style = "verbose" )] Whitespace(#[primary_span] Span), + #[multipart_suggestion( + parse_suggestion_str, + applicability = "maybe-incorrect", + style = "verbose" + )] + MeantStr { + #[suggestion_part(code = "\"")] + start: Span, + #[suggestion_part(code = "\"")] + end: Span, + }, } #[derive(Diagnostic)] diff --git a/compiler/rustc_parse/src/lexer/mod.rs b/compiler/rustc_parse/src/lexer/mod.rs index f57945a52df..659e0b63d2e 100644 --- a/compiler/rustc_parse/src/lexer/mod.rs +++ b/compiler/rustc_parse/src/lexer/mod.rs @@ -63,6 +63,7 @@ pub(crate) fn parse_token_trees<'psess, 'src>( cursor, override_span, nbsp_is_whitespace: false, + last_lifetime: None, }; let (stream, res, unmatched_delims) = tokentrees::TokenTreesReader::parse_all_token_trees(string_reader); @@ -105,6 +106,10 @@ struct StringReader<'psess, 'src> { /// in this file, it's safe to treat further occurrences of the non-breaking /// space character as whitespace. nbsp_is_whitespace: bool, + + /// Track the `Span` for the leading `'` of the last lifetime. Used for + /// diagnostics to detect possible typo where `"` was meant. + last_lifetime: Option, } impl<'psess, 'src> StringReader<'psess, 'src> { @@ -130,6 +135,18 @@ impl<'psess, 'src> StringReader<'psess, 'src> { debug!("next_token: {:?}({:?})", token.kind, self.str_from(start)); + if let rustc_lexer::TokenKind::Semi + | rustc_lexer::TokenKind::LineComment { .. } + | rustc_lexer::TokenKind::BlockComment { .. } + | rustc_lexer::TokenKind::CloseParen + | rustc_lexer::TokenKind::CloseBrace + | rustc_lexer::TokenKind::CloseBracket = token.kind + { + // Heuristic: we assume that it is unlikely we're dealing with an unterminated + // string surrounded by single quotes. + self.last_lifetime = None; + } + // Now "cook" the token, converting the simple `rustc_lexer::TokenKind` enum into a // rich `rustc_ast::TokenKind`. This turns strings into interned symbols and runs // additional validation. @@ -247,6 +264,7 @@ impl<'psess, 'src> StringReader<'psess, 'src> { // expansion purposes. See #12512 for the gory details of why // this is necessary. let lifetime_name = self.str_from(start); + self.last_lifetime = Some(self.mk_sp(start, start + BytePos(1))); if starts_with_number { let span = self.mk_sp(start, self.pos); self.dcx().struct_err("lifetimes cannot start with a number") @@ -395,10 +413,21 @@ impl<'psess, 'src> StringReader<'psess, 'src> { match kind { rustc_lexer::LiteralKind::Char { terminated } => { if !terminated { - self.dcx() + let mut err = self + .dcx() .struct_span_fatal(self.mk_sp(start, end), "unterminated character literal") - .with_code(E0762) - .emit() + .with_code(E0762); + if let Some(lt_sp) = self.last_lifetime { + err.multipart_suggestion( + "if you meant to write a `str` literal, use double quotes", + vec![ + (lt_sp, "\"".to_string()), + (self.mk_sp(start, start + BytePos(1)), "\"".to_string()), + ], + Applicability::MaybeIncorrect, + ); + } + err.emit() } self.cook_unicode(token::Char, Mode::Char, start, end, 1, 1) // ' ' } @@ -673,7 +702,16 @@ impl<'psess, 'src> StringReader<'psess, 'src> { let sugg = if prefix == "rb" { Some(errors::UnknownPrefixSugg::UseBr(prefix_span)) } else if expn_data.is_root() { - Some(errors::UnknownPrefixSugg::Whitespace(prefix_span.shrink_to_hi())) + if self.cursor.first() == '\'' + && let Some(start) = self.last_lifetime + { + Some(errors::UnknownPrefixSugg::MeantStr { + start, + end: self.mk_sp(self.pos, self.pos + BytePos(1)), + }) + } else { + Some(errors::UnknownPrefixSugg::Whitespace(prefix_span.shrink_to_hi())) + } } else { None }; -- cgit 1.4.1-3-g733a5