diff options
| author | Esteban Küber <esteban@kuber.com.ar> | 2025-01-15 21:24:31 +0000 |
|---|---|---|
| committer | Esteban Küber <esteban@kuber.com.ar> | 2025-01-22 00:52:27 +0000 |
| commit | 57dd42d6134539f5a98f59039bcba6d93daf9d6a (patch) | |
| tree | 09e6374ea06094fc073a535434b1109a9eb34915 /compiler/rustc_parse/src/lib.rs | |
| parent | 6a64e3b89724395d87a394ed9ebd954c474b080a (diff) | |
| download | rust-57dd42d6134539f5a98f59039bcba6d93daf9d6a.tar.gz rust-57dd42d6134539f5a98f59039bcba6d93daf9d6a.zip | |
Point at invalid utf-8 span on user's source code
```
error: couldn't read `$DIR/not-utf8-bin-file.rs`: stream did not contain valid UTF-8
--> $DIR/not-utf8-2.rs:6:5
|
LL | include!("not-utf8-bin-file.rs");
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
note: `[193]` is not valid utf-8
--> $DIR/not-utf8-bin-file.rs:2:14
|
LL | let _ = "�|�␂!5�cc␕␂��";
| ^
= note: this error originates in the macro `include` (in Nightly builds, run with -Z macro-backtrace for more info)
```
When we attempt to load a Rust source code file, if there is a OS file failure we try reading the file as bytes. If that succeeds we try to turn it into UTF-8. If *that* fails, we provide additional context about *where* the file has the first invalid UTF-8 character.
Fix #76869.
Diffstat (limited to 'compiler/rustc_parse/src/lib.rs')
| -rw-r--r-- | compiler/rustc_parse/src/lib.rs | 67 |
1 files changed, 63 insertions, 4 deletions
diff --git a/compiler/rustc_parse/src/lib.rs b/compiler/rustc_parse/src/lib.rs index f963a424a7f..e681987ff07 100644 --- a/compiler/rustc_parse/src/lib.rs +++ b/compiler/rustc_parse/src/lib.rs @@ -11,18 +11,21 @@ #![feature(if_let_guard)] #![feature(iter_intersperse)] #![feature(let_chains)] +#![feature(string_from_utf8_lossy_owned)] #![warn(unreachable_pub)] // tidy-alphabetical-end -use std::path::Path; +use std::path::{Path, PathBuf}; +use std::str::Utf8Error; use rustc_ast as ast; use rustc_ast::tokenstream::TokenStream; use rustc_ast::{AttrItem, Attribute, MetaItemInner, token}; use rustc_ast_pretty::pprust; use rustc_data_structures::sync::Lrc; -use rustc_errors::{Diag, FatalError, PResult}; +use rustc_errors::{Diag, EmissionGuarantee, FatalError, PResult, pluralize}; use rustc_session::parse::ParseSess; +use rustc_span::source_map::SourceMap; use rustc_span::{FileName, SourceFile, Span}; pub use unicode_normalization::UNICODE_VERSION as UNICODE_NORMALIZATION_VERSION; @@ -73,9 +76,22 @@ pub fn new_parser_from_file<'a>( path: &Path, sp: Option<Span>, ) -> Result<Parser<'a>, Vec<Diag<'a>>> { - let source_file = psess.source_map().load_file(path).unwrap_or_else(|e| { - let msg = format!("couldn't read {}: {}", path.display(), e); + let sm = psess.source_map(); + let source_file = sm.load_file(path).unwrap_or_else(|e| { + let msg = format!("couldn't read `{}`: {}", path.display(), e); let mut err = psess.dcx().struct_fatal(msg); + if let Ok(contents) = std::fs::read(path) + && let Err(utf8err) = String::from_utf8(contents.clone()) + { + utf8_error( + sm, + &path.display().to_string(), + sp, + &mut err, + utf8err.utf8_error(), + &contents, + ); + } if let Some(sp) = sp { err.span(sp); } @@ -84,6 +100,49 @@ pub fn new_parser_from_file<'a>( new_parser_from_source_file(psess, source_file) } +pub fn utf8_error<E: EmissionGuarantee>( + sm: &SourceMap, + path: &str, + sp: Option<Span>, + err: &mut Diag<'_, E>, + utf8err: Utf8Error, + contents: &[u8], +) { + // The file exists, but it wasn't valid UTF-8. + let start = utf8err.valid_up_to(); + let note = format!("invalid utf-8 at byte `{start}`"); + let msg = if let Some(len) = utf8err.error_len() { + format!( + "byte{s} `{bytes}` {are} not valid utf-8", + bytes = if len == 1 { + format!("{:?}", contents[start]) + } else { + format!("{:?}", &contents[start..start + len]) + }, + s = pluralize!(len), + are = if len == 1 { "is" } else { "are" }, + ) + } else { + note.clone() + }; + let contents = String::from_utf8_lossy(contents).to_string(); + let source = sm.new_source_file(PathBuf::from(path).into(), contents); + let span = Span::with_root_ctxt( + source.normalized_byte_pos(start as u32), + source.normalized_byte_pos(start as u32), + ); + if span.is_dummy() { + err.note(note); + } else { + if sp.is_some() { + err.span_note(span, msg); + } else { + err.span(span); + err.span_label(span, msg); + } + } +} + /// Given a session and a `source_file`, return a parser. Returns any buffered errors from lexing /// the initial token stream. fn new_parser_from_source_file( |
