diff options
Diffstat (limited to 'compiler/rustc_parse/src/lexer/unescape_error_reporting.rs')
| -rw-r--r-- | compiler/rustc_parse/src/lexer/unescape_error_reporting.rs | 248 |
1 files changed, 74 insertions, 174 deletions
diff --git a/compiler/rustc_parse/src/lexer/unescape_error_reporting.rs b/compiler/rustc_parse/src/lexer/unescape_error_reporting.rs index 6373f5b4f..0d12ec608 100644 --- a/compiler/rustc_parse/src/lexer/unescape_error_reporting.rs +++ b/compiler/rustc_parse/src/lexer/unescape_error_reporting.rs @@ -3,10 +3,12 @@ use std::iter::once; use std::ops::Range; -use rustc_errors::{pluralize, Applicability, Handler}; +use rustc_errors::{Applicability, Handler}; use rustc_lexer::unescape::{EscapeError, Mode}; use rustc_span::{BytePos, Span}; +use crate::errors::{MoreThanOneCharNote, MoreThanOneCharSugg, NoBraceUnicodeSub, UnescapeError}; + pub(crate) fn emit_unescape_error( handler: &Handler, // interior part of the literal, without quotes @@ -31,53 +33,32 @@ pub(crate) fn emit_unescape_error( }; match error { EscapeError::LoneSurrogateUnicodeEscape => { - handler - .struct_span_err(span, "invalid unicode character escape") - .span_label(span, "invalid escape") - .help("unicode escape must not be a surrogate") - .emit(); + handler.emit_err(UnescapeError::InvalidUnicodeEscape { span, surrogate: true }); } EscapeError::OutOfRangeUnicodeEscape => { - handler - .struct_span_err(span, "invalid unicode character escape") - .span_label(span, "invalid escape") - .help("unicode escape must be at most 10FFFF") - .emit(); + handler.emit_err(UnescapeError::InvalidUnicodeEscape { span, surrogate: false }); } EscapeError::MoreThanOneChar => { use unicode_normalization::{char::is_combining_mark, UnicodeNormalization}; + let mut sugg = None; + let mut note = None; - let mut has_help = false; - let mut handler = handler.struct_span_err( - span_with_quotes, - "character literal may only contain one codepoint", - ); - - if lit.chars().skip(1).all(|c| is_combining_mark(c)) { - let escaped_marks = - lit.chars().skip(1).map(|c| c.escape_default().to_string()).collect::<Vec<_>>(); - handler.span_note( - span, - &format!( - "this `{}` is followed by the combining mark{} `{}`", - lit.chars().next().unwrap(), - pluralize!(escaped_marks.len()), - escaped_marks.join(""), - ), - ); + let lit_chars = lit.chars().collect::<Vec<_>>(); + let (first, rest) = lit_chars.split_first().unwrap(); + if rest.iter().copied().all(is_combining_mark) { let normalized = lit.nfc().to_string(); if normalized.chars().count() == 1 { - has_help = true; - handler.span_suggestion( - span, - &format!( - "consider using the normalized form `{}` of this character", - normalized.chars().next().unwrap().escape_default() - ), - normalized, - Applicability::MachineApplicable, - ); + let ch = normalized.chars().next().unwrap().escape_default().to_string(); + sugg = Some(MoreThanOneCharSugg::NormalizedForm { span, ch, normalized }); } + let escaped_marks = + rest.iter().map(|c| c.escape_default().to_string()).collect::<Vec<_>>(); + note = Some(MoreThanOneCharNote::AllCombining { + span, + chr: format!("{first}"), + len: escaped_marks.len(), + escaped_marks: escaped_marks.join(""), + }); } else { let printable: Vec<char> = lit .chars() @@ -87,32 +68,18 @@ pub(crate) fn emit_unescape_error( }) .collect(); - if let [ch] = printable.as_slice() { - has_help = true; - - handler.span_note( + if let &[ch] = printable.as_slice() { + sugg = + Some(MoreThanOneCharSugg::RemoveNonPrinting { span, ch: ch.to_string() }); + note = Some(MoreThanOneCharNote::NonPrinting { span, - &format!( - "there are non-printing characters, the full sequence is `{}`", - lit.escape_default(), - ), - ); - - handler.span_suggestion( - span, - "consider removing the non-printing characters", - ch, - Applicability::MaybeIncorrect, - ); + escaped: lit.escape_default().to_string(), + }); } - } - - if !has_help { - let (prefix, msg) = if mode.is_byte() { - ("b", "if you meant to write a byte string literal, use double quotes") - } else { - ("", "if you meant to write a `str` literal, use double quotes") - }; + }; + let sugg = sugg.unwrap_or_else(|| { + let is_byte = mode.is_byte(); + let prefix = if is_byte { "b" } else { "" }; let mut escaped = String::with_capacity(lit.len()); let mut chrs = lit.chars().peekable(); while let Some(first) = chrs.next() { @@ -129,54 +96,32 @@ pub(crate) fn emit_unescape_error( (c, _) => escaped.push(c), }; } - handler.span_suggestion( - span_with_quotes, - msg, - format!("{prefix}\"{escaped}\""), - Applicability::MachineApplicable, - ); - } - - handler.emit(); + let sugg = format!("{prefix}\"{escaped}\""); + MoreThanOneCharSugg::Quotes { span: span_with_quotes, is_byte, sugg } + }); + handler.emit_err(UnescapeError::MoreThanOneChar { + span: span_with_quotes, + note, + suggestion: sugg, + }); } EscapeError::EscapeOnlyChar => { let (c, char_span) = last_char(); - - let msg = if mode.is_byte() { - "byte constant must be escaped" - } else { - "character constant must be escaped" - }; - handler - .struct_span_err(span, &format!("{}: `{}`", msg, escaped_char(c))) - .span_suggestion( - char_span, - "escape the character", - c.escape_default(), - Applicability::MachineApplicable, - ) - .emit(); + handler.emit_err(UnescapeError::EscapeOnlyChar { + span, + char_span, + escaped_sugg: c.escape_default().to_string(), + escaped_msg: escaped_char(c), + byte: mode.is_byte(), + }); } EscapeError::BareCarriageReturn => { - let msg = if mode.in_double_quotes() { - "bare CR not allowed in string, use `\\r` instead" - } else { - "character constant must be escaped: `\\r`" - }; - handler - .struct_span_err(span, msg) - .span_suggestion( - span, - "escape the character", - "\\r", - Applicability::MachineApplicable, - ) - .emit(); + let double_quotes = mode.in_double_quotes(); + handler.emit_err(UnescapeError::BareCr { span, double_quotes }); } EscapeError::BareCarriageReturnInRawString => { assert!(mode.in_double_quotes()); - let msg = "bare CR not allowed in raw string"; - handler.span_err(span, msg); + handler.emit_err(UnescapeError::BareCrRawString(span)); } EscapeError::InvalidEscape => { let (c, span) = last_char(); @@ -213,22 +158,13 @@ pub(crate) fn emit_unescape_error( diag.emit(); } EscapeError::TooShortHexEscape => { - handler.span_err(span, "numeric character escape is too short"); + handler.emit_err(UnescapeError::TooShortHexEscape(span)); } EscapeError::InvalidCharInHexEscape | EscapeError::InvalidCharInUnicodeEscape => { let (c, span) = last_char(); - - let msg = if error == EscapeError::InvalidCharInHexEscape { - "invalid character in numeric character escape" - } else { - "invalid character in unicode escape" - }; - let c = escaped_char(c); - - handler - .struct_span_err(span, &format!("{}: `{}`", msg, c)) - .span_label(span, msg) - .emit(); + let is_hex = error == EscapeError::InvalidCharInHexEscape; + let ch = escaped_char(c); + handler.emit_err(UnescapeError::InvalidCharInEscape { span, is_hex, ch }); } EscapeError::NonAsciiCharInByte => { let (c, span) = last_char(); @@ -278,41 +214,22 @@ pub(crate) fn emit_unescape_error( err.emit(); } EscapeError::OutOfRangeHexEscape => { - handler - .struct_span_err(span, "out of range hex escape") - .span_label(span, "must be a character in the range [\\x00-\\x7f]") - .emit(); + handler.emit_err(UnescapeError::OutOfRangeHexEscape(span)); } EscapeError::LeadingUnderscoreUnicodeEscape => { let (c, span) = last_char(); - let msg = "invalid start of unicode escape"; - handler - .struct_span_err(span, &format!("{}: `{}`", msg, c)) - .span_label(span, msg) - .emit(); + handler.emit_err(UnescapeError::LeadingUnderscoreUnicodeEscape { + span, + ch: escaped_char(c), + }); } EscapeError::OverlongUnicodeEscape => { - handler - .struct_span_err(span, "overlong unicode escape") - .span_label(span, "must have at most 6 hex digits") - .emit(); + handler.emit_err(UnescapeError::OverlongUnicodeEscape(span)); } EscapeError::UnclosedUnicodeEscape => { - handler - .struct_span_err(span, "unterminated unicode escape") - .span_label(span, "missing a closing `}`") - .span_suggestion_verbose( - span.shrink_to_hi(), - "terminate the unicode escape", - "}", - Applicability::MaybeIncorrect, - ) - .emit(); + handler.emit_err(UnescapeError::UnclosedUnicodeEscape(span, span.shrink_to_hi())); } EscapeError::NoBraceInUnicodeEscape => { - let msg = "incorrect unicode escape sequence"; - let mut diag = handler.struct_span_err(span, msg); - let mut suggestion = "\\u{".to_owned(); let mut suggestion_len = 0; let (c, char_span) = last_char(); @@ -322,54 +239,37 @@ pub(crate) fn emit_unescape_error( suggestion_len += c.len_utf8(); } - if suggestion_len > 0 { + let (label, sub) = if suggestion_len > 0 { suggestion.push('}'); let hi = char_span.lo() + BytePos(suggestion_len as u32); - diag.span_suggestion( - span.with_hi(hi), - "format of unicode escape sequences uses braces", - suggestion, - Applicability::MaybeIncorrect, - ); + (None, NoBraceUnicodeSub::Suggestion { span: span.with_hi(hi), suggestion }) } else { - diag.span_label(span, msg); - diag.help("format of unicode escape sequences is `\\u{...}`"); - } - - diag.emit(); + (Some(span), NoBraceUnicodeSub::Help) + }; + handler.emit_err(UnescapeError::NoBraceInUnicodeEscape { span, label, sub }); } EscapeError::UnicodeEscapeInByte => { - let msg = "unicode escape in byte string"; - handler - .struct_span_err(span, msg) - .span_label(span, msg) - .help("unicode escape sequences cannot be used as a byte or in a byte string") - .emit(); + handler.emit_err(UnescapeError::UnicodeEscapeInByte(span)); } EscapeError::EmptyUnicodeEscape => { - handler - .struct_span_err(span, "empty unicode escape") - .span_label(span, "this escape must have at least 1 hex digit") - .emit(); + handler.emit_err(UnescapeError::EmptyUnicodeEscape(span)); } EscapeError::ZeroChars => { - let msg = "empty character literal"; - handler.struct_span_err(span, msg).span_label(span, msg).emit(); + handler.emit_err(UnescapeError::ZeroChars(span)); } EscapeError::LoneSlash => { - let msg = "invalid trailing slash in literal"; - handler.struct_span_err(span, msg).span_label(span, msg).emit(); + handler.emit_err(UnescapeError::LoneSlash(span)); } EscapeError::UnskippedWhitespaceWarning => { let (c, char_span) = last_char(); - let msg = - format!("non-ASCII whitespace symbol '{}' is not skipped", c.escape_unicode()); - handler.struct_span_warn(span, &msg).span_label(char_span, &msg).emit(); + handler.emit_warning(UnescapeError::UnskippedWhitespace { + span, + ch: escaped_char(c), + char_span, + }); } EscapeError::MultipleSkippedLinesWarning => { - let msg = "multiple lines skipped by escaped newline"; - let bottom_msg = "skipping everything up to and including this point"; - handler.struct_span_warn(span, msg).span_label(span, bottom_msg).emit(); + handler.emit_warning(UnescapeError::MultipleSkippedLinesWarning(span)); } } } |