Revert "Rollup merge of #136355 - GuillaumeGomez:proc-macro_add_value_retrieval_methods, r=Amanieu"

This reverts commit 08dfbf49e3, reversing
changes made to 10bcdad7df.
This commit is contained in:
Ralf Jung 2025-03-18 13:28:56 +01:00
parent 75530e9f72
commit 20d04d8a40
27 changed files with 15 additions and 253 deletions

View file

@ -2082,13 +2082,6 @@ version = "0.7.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4ee93343901ab17bd981295f2cf0026d4ad018c7c31ba84549a4ddbb47a45104"
[[package]]
name = "literal-escaper"
version = "0.0.0"
dependencies = [
"rustc-std-workspace-std 1.0.1 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
name = "lld-wrapper"
version = "0.1.0"
@ -3155,12 +3148,6 @@ version = "1.0.1"
name = "rustc-std-workspace-std"
version = "1.0.1"
[[package]]
name = "rustc-std-workspace-std"
version = "1.0.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "aba676a20abe46e5b0f1b0deae474aaaf31407e6c71147159890574599da04ef"
[[package]]
name = "rustc_abi"
version = "0.0.0"
@ -3199,7 +3186,6 @@ name = "rustc_ast"
version = "0.0.0"
dependencies = [
"bitflags",
"literal-escaper",
"memchr",
"rustc_ast_ir",
"rustc_data_structures",
@ -3909,7 +3895,6 @@ name = "rustc_lexer"
version = "0.0.0"
dependencies = [
"expect-test",
"literal-escaper",
"memchr",
"unicode-properties",
"unicode-xid",
@ -4172,7 +4157,6 @@ name = "rustc_parse"
version = "0.0.0"
dependencies = [
"bitflags",
"literal-escaper",
"rustc_ast",
"rustc_ast_pretty",
"rustc_data_structures",
@ -4195,7 +4179,6 @@ dependencies = [
name = "rustc_parse_format"
version = "0.0.0"
dependencies = [
"literal-escaper",
"rustc_index",
"rustc_lexer",
]

View file

@ -6,7 +6,6 @@ edition = "2024"
[dependencies]
# tidy-alphabetical-start
bitflags = "2.4.1"
literal-escaper = { path = "../../library/literal-escaper" }
memchr = "2.7.4"
rustc_ast_ir = { path = "../rustc_ast_ir" }
rustc_data_structures = { path = "../rustc_data_structures" }

View file

@ -2,7 +2,7 @@
use std::{ascii, fmt, str};
use literal_escaper::{
use rustc_lexer::unescape::{
MixedUnit, Mode, byte_from_char, unescape_byte, unescape_char, unescape_mixed, unescape_unicode,
};
use rustc_span::{Span, Symbol, kw, sym};

View file

@ -16,7 +16,6 @@ Rust lexer used by rustc. No stability guarantees are provided.
[dependencies]
memchr = "2.7.4"
unicode-xid = "0.2.0"
literal-escaper = { path = "../../library/literal-escaper" }
[dependencies.unicode-properties]
version = "0.1.0"

View file

@ -26,13 +26,11 @@
// tidy-alphabetical-end
mod cursor;
pub mod unescape;
#[cfg(test)]
mod tests;
// FIXME: This is needed for rust-analyzer. Remove this dependency once rust-analyzer uses
// `literal-escaper`.
pub use literal_escaper as unescape;
use unicode_properties::UnicodeEmoji;
pub use unicode_xid::UNICODE_VERSION as UNICODE_XID_VERSION;

View file

@ -6,7 +6,6 @@ edition = "2024"
[dependencies]
# tidy-alphabetical-start
bitflags = "2.4.1"
literal-escaper = { path = "../../library/literal-escaper" }
rustc_ast = { path = "../rustc_ast" }
rustc_ast_pretty = { path = "../rustc_ast_pretty" }
rustc_data_structures = { path = "../rustc_data_structures" }

View file

@ -1,12 +1,12 @@
use std::ops::Range;
use literal_escaper::{self, EscapeError, Mode};
use rustc_ast::ast::{self, AttrStyle};
use rustc_ast::token::{self, CommentKind, Delimiter, IdentIsRaw, Token, TokenKind};
use rustc_ast::tokenstream::TokenStream;
use rustc_ast::util::unicode::contains_text_flow_control_chars;
use rustc_errors::codes::*;
use rustc_errors::{Applicability, Diag, DiagCtxtHandle, StashKey};
use rustc_lexer::unescape::{self, EscapeError, Mode};
use rustc_lexer::{Base, Cursor, DocStyle, LiteralKind, RawStrError};
use rustc_session::lint::BuiltinLintDiag;
use rustc_session::lint::builtin::{
@ -970,7 +970,7 @@ impl<'psess, 'src> Lexer<'psess, 'src> {
postfix_len: u32,
) -> (token::LitKind, Symbol) {
self.cook_common(kind, mode, start, end, prefix_len, postfix_len, |src, mode, callback| {
literal_escaper::unescape_unicode(src, mode, &mut |span, result| {
unescape::unescape_unicode(src, mode, &mut |span, result| {
callback(span, result.map(drop))
})
})
@ -986,7 +986,7 @@ impl<'psess, 'src> Lexer<'psess, 'src> {
postfix_len: u32,
) -> (token::LitKind, Symbol) {
self.cook_common(kind, mode, start, end, prefix_len, postfix_len, |src, mode, callback| {
literal_escaper::unescape_mixed(src, mode, &mut |span, result| {
unescape::unescape_mixed(src, mode, &mut |span, result| {
callback(span, result.map(drop))
})
})

View file

@ -3,8 +3,8 @@
use std::iter::once;
use std::ops::Range;
use literal_escaper::{EscapeError, Mode};
use rustc_errors::{Applicability, DiagCtxtHandle, ErrorGuaranteed};
use rustc_lexer::unescape::{EscapeError, Mode};
use rustc_span::{BytePos, Span};
use tracing::debug;

View file

@ -6,7 +6,6 @@ use core::ops::{Bound, ControlFlow};
use ast::mut_visit::{self, MutVisitor};
use ast::token::{IdentIsRaw, MetaVarKind};
use ast::{CoroutineKind, ForLoopKind, GenBlockKind, MatchKind, Pat, Path, PathSegment, Recovered};
use literal_escaper::unescape_char;
use rustc_ast::ptr::P;
use rustc_ast::token::{self, Delimiter, Token, TokenKind};
use rustc_ast::tokenstream::TokenTree;
@ -22,6 +21,7 @@ use rustc_ast::{
use rustc_ast_pretty::pprust;
use rustc_data_structures::stack::ensure_sufficient_stack;
use rustc_errors::{Applicability, Diag, PResult, StashKey, Subdiagnostic};
use rustc_lexer::unescape::unescape_char;
use rustc_macros::Subdiagnostic;
use rustc_session::errors::{ExprParenthesesNeeded, report_lit_error};
use rustc_session::lint::BuiltinLintDiag;

View file

@ -5,7 +5,6 @@ edition = "2024"
[dependencies]
# tidy-alphabetical-start
literal-escaper = { path = "../../library/literal-escaper" }
rustc_index = { path = "../rustc_index", default-features = false }
rustc_lexer = { path = "../rustc_lexer" }
# tidy-alphabetical-end

View file

@ -18,6 +18,7 @@
pub use Alignment::*;
pub use Count::*;
pub use Position::*;
use rustc_lexer::unescape;
// Note: copied from rustc_span
/// Range inside of a `Span` used for diagnostics when we only have access to relative positions.
@ -1093,14 +1094,12 @@ fn find_width_map_from_snippet(
fn unescape_string(string: &str) -> Option<String> {
let mut buf = String::new();
let mut ok = true;
literal_escaper::unescape_unicode(
string,
literal_escaper::Mode::Str,
&mut |_, unescaped_char| match unescaped_char {
unescape::unescape_unicode(string, unescape::Mode::Str, &mut |_, unescaped_char| {
match unescaped_char {
Ok(c) => buf.push(c),
Err(_) => ok = false,
},
);
}
});
ok.then_some(buf)
}

View file

@ -165,13 +165,6 @@ dependencies = [
"rustc-std-workspace-core",
]
[[package]]
name = "literal-escaper"
version = "0.0.0"
dependencies = [
"rustc-std-workspace-std",
]
[[package]]
name = "memchr"
version = "2.7.4"
@ -243,7 +236,6 @@ name = "proc_macro"
version = "0.0.0"
dependencies = [
"core",
"literal-escaper",
"std",
]

View file

@ -8,7 +8,6 @@ members = [
]
exclude = [
"literal-escaper",
# stdarch has its own Cargo workspace
"stdarch",
"windows_targets"

View file

@ -1,10 +0,0 @@
[package]
name = "literal-escaper"
version = "0.0.0"
edition = "2021"
[dependencies]
std = { version = '1.0.0', optional = true, package = 'rustc-std-workspace-std' }
[features]
rustc-dep-of-std = ["dep:std"]

View file

@ -1,4 +0,0 @@
# literal-escaper
This crate provides code to unescape string literals. It is used by `rustc_lexer`
and `proc_macro`.

View file

@ -4,7 +4,6 @@ version = "0.0.0"
edition = "2024"
[dependencies]
literal-escaper = { path = "../literal-escaper", features = ["rustc-dep-of-std"] }
std = { path = "../std" }
# Workaround: when documenting this crate rustdoc will try to load crate named
# `core` when resolving doc links. Without this line a different `core` will be

View file

@ -28,7 +28,6 @@
#![feature(restricted_std)]
#![feature(rustc_attrs)]
#![feature(extend_one)]
#![feature(stmt_expr_attributes)]
#![recursion_limit = "256"]
#![allow(internal_features)]
#![deny(ffi_unwind_calls)]
@ -52,24 +51,11 @@ use std::{error, fmt};
#[unstable(feature = "proc_macro_diagnostic", issue = "54140")]
pub use diagnostic::{Diagnostic, Level, MultiSpan};
#[unstable(feature = "proc_macro_value", issue = "136652")]
pub use literal_escaper::EscapeError;
use literal_escaper::{MixedUnit, Mode, byte_from_char, unescape_mixed, unescape_unicode};
#[unstable(feature = "proc_macro_totokens", issue = "130977")]
pub use to_tokens::ToTokens;
use crate::escape::{EscapeOptions, escape_bytes};
/// Errors returned when trying to retrieve a literal unescaped value.
#[unstable(feature = "proc_macro_value", issue = "136652")]
#[derive(Debug, PartialEq, Eq)]
pub enum ConversionErrorKind {
/// The literal failed to be escaped, take a look at [`EscapeError`] for more information.
FailedToUnescape(EscapeError),
/// Trying to convert a literal with the wrong type.
InvalidLiteralKind,
}
/// Determines whether proc_macro has been made accessible to the currently
/// running program.
///
@ -1465,107 +1451,6 @@ impl Literal {
}
})
}
/// Returns the unescaped string value if the current literal is a string or a string literal.
#[unstable(feature = "proc_macro_value", issue = "136652")]
pub fn str_value(&self) -> Result<String, ConversionErrorKind> {
self.0.symbol.with(|symbol| match self.0.kind {
bridge::LitKind::Str => {
if symbol.contains('\\') {
let mut buf = String::with_capacity(symbol.len());
let mut error = None;
// Force-inlining here is aggressive but the closure is
// called on every char in the string, so it can be hot in
// programs with many long strings containing escapes.
unescape_unicode(
symbol,
Mode::Str,
&mut #[inline(always)]
|_, c| match c {
Ok(c) => buf.push(c),
Err(err) => {
if err.is_fatal() {
error = Some(ConversionErrorKind::FailedToUnescape(err));
}
}
},
);
if let Some(error) = error { Err(error) } else { Ok(buf) }
} else {
Ok(symbol.to_string())
}
}
bridge::LitKind::StrRaw(_) => Ok(symbol.to_string()),
_ => Err(ConversionErrorKind::InvalidLiteralKind),
})
}
/// Returns the unescaped string value if the current literal is a c-string or a c-string
/// literal.
#[unstable(feature = "proc_macro_value", issue = "136652")]
pub fn cstr_value(&self) -> Result<Vec<u8>, ConversionErrorKind> {
self.0.symbol.with(|symbol| match self.0.kind {
bridge::LitKind::CStr => {
let mut error = None;
let mut buf = Vec::with_capacity(symbol.len());
unescape_mixed(symbol, Mode::CStr, &mut |_span, c| match c {
Ok(MixedUnit::Char(c)) => {
buf.extend_from_slice(c.encode_utf8(&mut [0; 4]).as_bytes())
}
Ok(MixedUnit::HighByte(b)) => buf.push(b),
Err(err) => {
if err.is_fatal() {
error = Some(ConversionErrorKind::FailedToUnescape(err));
}
}
});
if let Some(error) = error {
Err(error)
} else {
buf.push(0);
Ok(buf)
}
}
bridge::LitKind::CStrRaw(_) => {
// Raw strings have no escapes so we can convert the symbol
// directly to a `Lrc<u8>` after appending the terminating NUL
// char.
let mut buf = symbol.to_owned().into_bytes();
buf.push(0);
Ok(buf)
}
_ => Err(ConversionErrorKind::InvalidLiteralKind),
})
}
/// Returns the unescaped string value if the current literal is a byte string or a byte string
/// literal.
#[unstable(feature = "proc_macro_value", issue = "136652")]
pub fn byte_str_value(&self) -> Result<Vec<u8>, ConversionErrorKind> {
self.0.symbol.with(|symbol| match self.0.kind {
bridge::LitKind::ByteStr => {
let mut buf = Vec::with_capacity(symbol.len());
let mut error = None;
unescape_unicode(symbol, Mode::ByteStr, &mut |_, c| match c {
Ok(c) => buf.push(byte_from_char(c)),
Err(err) => {
if err.is_fatal() {
error = Some(ConversionErrorKind::FailedToUnescape(err));
}
}
});
if let Some(error) = error { Err(error) } else { Ok(buf) }
}
bridge::LitKind::ByteStrRaw(_) => {
// Raw strings have no escapes so we can convert the symbol
// directly to a `Lrc<u8>`.
Ok(symbol.to_owned().into_bytes())
}
_ => Err(ConversionErrorKind::InvalidLiteralKind),
})
}
}
/// Parse a single literal from its stringified representation.

View file

@ -655,7 +655,7 @@ mod dist {
let mut builder = Builder::new(&build);
builder.run_step_descriptions(
&Builder::get_step_descriptions(Kind::Build),
&["compiler/rustc".into(), "std".into()],
&["compiler/rustc".into(), "library".into()],
);
assert_eq!(builder.config.stage, 2);

View file

@ -62,11 +62,6 @@ pub fn build(build: &mut Build) {
let relative_path = krate.local_path(build);
build.crates.insert(name.clone(), krate);
let existing_path = build.crate_paths.insert(relative_path, name);
// `literal-escaper` is both a dependency of `compiler/rustc_lexer` and of
// `library/proc-macro`, making it appear multiple times in the workspace.
if existing_path.as_deref() == Some("literal-escaper") {
continue;
}
assert!(
existing_path.is_none(),
"multiple crates with the same path: {}",

View file

@ -716,7 +716,7 @@ impl Build {
features.push("llvm");
}
// keep in sync with `bootstrap/compile.rs:rustc_cargo_env`
if self.config.rust_randomize_layout && check("rustc_randomized_layouts") {
if self.config.rust_randomize_layout {
features.push("rustc_randomized_layouts");
}

View file

@ -318,7 +318,6 @@ const PERMITTED_RUSTC_DEPENDENCIES: &[&str] = &[
"libloading",
"linux-raw-sys",
"litemap",
"literal-escaper",
"lock_api",
"log",
"matchers",
@ -364,7 +363,6 @@ const PERMITTED_RUSTC_DEPENDENCIES: &[&str] = &[
"rustc-rayon",
"rustc-rayon-core",
"rustc-stable-hash",
"rustc-std-workspace-std",
"rustc_apfloat",
"rustix",
"ruzstd", // via object in thorin-dwp

View file

@ -1,3 +0,0 @@
#![crate_type = "lib"]
extern crate literal_escaper; //~ ERROR

View file

@ -1,13 +0,0 @@
error[E0658]: use of unstable library feature `rustc_private`: this crate is being loaded from the sysroot, an unstable location; did you mean to load this crate from crates.io via `Cargo.toml` instead?
--> $DIR/literal-escaper.rs:3:1
|
LL | extern crate literal_escaper;
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= note: see issue #27812 <https://github.com/rust-lang/rust/issues/27812> for more information
= help: add `#![feature(rustc_private)]` to the crate attributes to enable
= note: this compiler was built on YYYY-MM-DD; consider upgrading it if it is out of date
error: aborting due to 1 previous error
For more information about this error, try `rustc --explain E0658`.

View file

@ -1,11 +1,10 @@
// ignore-tidy-linelength
use proc_macro::{ConversionErrorKind, Literal};
use proc_macro::Literal;
pub fn test() {
test_display_literal();
test_parse_literal();
test_str_value_methods();
}
fn test_display_literal() {
@ -82,53 +81,3 @@ fn test_parse_literal() {
assert!("- 10".parse::<Literal>().is_err());
assert!("-'x'".parse::<Literal>().is_err());
}
fn test_str_value_methods() {
// Testing `str_value`
let lit = "\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.str_value(), Ok("\n".to_string()));
let lit = "r#\"\n\"#".parse::<Literal>().unwrap();
assert_eq!(lit.str_value(), Ok("\n".to_string()));
let lit = "1".parse::<Literal>().unwrap();
assert_eq!(lit.str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "b\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "c\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
// Testing `cstr_value`
let lit = "\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.cstr_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "r#\"\n\"#".parse::<Literal>().unwrap();
assert_eq!(lit.cstr_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "1".parse::<Literal>().unwrap();
assert_eq!(lit.cstr_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "b\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.cstr_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "c\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.cstr_value(), Ok(vec![b'\n', 0]));
// Testing `byte_str_value`
let lit = "\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.byte_str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "r#\"\n\"#".parse::<Literal>().unwrap();
assert_eq!(lit.byte_str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "1".parse::<Literal>().unwrap();
assert_eq!(lit.byte_str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
let lit = "b\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.byte_str_value(), Ok(vec![b'\n']));
let lit = "c\"\n\"".parse::<Literal>().unwrap();
assert_eq!(lit.byte_str_value(), Err(ConversionErrorKind::InvalidLiteralKind));
}

View file

@ -1,7 +1,6 @@
//@ edition: 2021
#![feature(proc_macro_span)]
#![feature(proc_macro_value)]
#![deny(dead_code)] // catch if a test function is never called
extern crate proc_macro;