diff --git a/Cargo.lock b/Cargo.lock index 045fcbac80e2a..9eba61f8d293b 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -731,9 +731,9 @@ dependencies = [ [[package]] name = "compiler_builtins" -version = "0.1.70" +version = "0.1.71" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "80873f979f0a344a4ade87c2f70d9ccf5720b83b10c97ec7cd745895d021e85a" +checksum = "163437f05ca8f29d7e9128ea728dedf5eb620e445fbca273641d3a3050305f23" dependencies = [ "cc", "rustc-std-workspace-core", diff --git a/compiler/rustc_hir/src/lang_items.rs b/compiler/rustc_hir/src/lang_items.rs index b8df163214435..9318ebb40b09b 100644 --- a/compiler/rustc_hir/src/lang_items.rs +++ b/compiler/rustc_hir/src/lang_items.rs @@ -326,6 +326,8 @@ language_item_table! { Range, sym::Range, range_struct, Target::Struct, GenericRequirement::None; RangeToInclusive, sym::RangeToInclusive, range_to_inclusive_struct, Target::Struct, GenericRequirement::None; RangeTo, sym::RangeTo, range_to_struct, Target::Struct, GenericRequirement::None; + + CStr, sym::CStr, c_str, Target::Struct, GenericRequirement::None; } pub enum GenericRequirement { diff --git a/compiler/rustc_metadata/src/creader.rs b/compiler/rustc_metadata/src/creader.rs index a9e3b55aeeedf..566d27212d08d 100644 --- a/compiler/rustc_metadata/src/creader.rs +++ b/compiler/rustc_metadata/src/creader.rs @@ -417,6 +417,7 @@ impl<'a> CrateLoader<'a> { let crate_metadata = CrateMetadata::new( self.sess, + &self.cstore, metadata, crate_root, raw_proc_macros, diff --git a/compiler/rustc_metadata/src/rmeta/decoder.rs b/compiler/rustc_metadata/src/rmeta/decoder.rs index 3933a0d19a4ad..8d0e8467404ca 100644 --- a/compiler/rustc_metadata/src/rmeta/decoder.rs +++ b/compiler/rustc_metadata/src/rmeta/decoder.rs @@ -1,6 +1,6 @@ // Decoding metadata from a single crate's metadata -use crate::creader::CrateMetadataRef; +use crate::creader::{CStore, CrateMetadataRef}; use crate::rmeta::table::{FixedSizeEncoding, Table}; use crate::rmeta::*; @@ -1737,6 +1737,7 @@ impl<'a, 'tcx> CrateMetadataRef<'a> { impl CrateMetadata { crate fn new( sess: &Session, + cstore: &CStore, blob: MetadataBlob, root: CrateRoot<'static>, raw_proc_macros: Option<&'static [ProcMacro]>, @@ -1752,11 +1753,6 @@ impl CrateMetadata { .decode((&blob, sess)) .map(|trait_impls| (trait_impls.trait_id, trait_impls.impls)) .collect(); - let incoherent_impls = root - .incoherent_impls - .decode((&blob, sess)) - .map(|incoherent_impls| (incoherent_impls.self_ty, incoherent_impls.impls)) - .collect(); let alloc_decoding_state = AllocDecodingState::new(root.interpret_alloc_index.decode(&blob).collect()); let dependencies = Lock::new(cnum_map.iter().cloned().collect()); @@ -1765,11 +1761,11 @@ impl CrateMetadata { // that does not copy any data. It just does some data verification. let def_path_hash_map = root.def_path_hash_map.decode(&blob); - CrateMetadata { + let mut cdata = CrateMetadata { blob, root, trait_impls, - incoherent_impls, + incoherent_impls: Default::default(), raw_proc_macros, source_map_import_info: OnceCell::new(), def_path_hash_map, @@ -1786,7 +1782,17 @@ impl CrateMetadata { hygiene_context: Default::default(), def_key_cache: Default::default(), def_path_hash_cache: Default::default(), - } + }; + + // Need `CrateMetadataRef` to decode `DefId`s in simplified types. + cdata.incoherent_impls = cdata + .root + .incoherent_impls + .decode(CrateMetadataRef { cdata: &cdata, cstore }) + .map(|incoherent_impls| (incoherent_impls.self_ty, incoherent_impls.impls)) + .collect(); + + cdata } crate fn dependencies(&self) -> LockGuard<'_, Vec> { diff --git a/compiler/rustc_typeck/src/check/method/probe.rs b/compiler/rustc_typeck/src/check/method/probe.rs index 83474bd7e72f7..e04cc42b6d7e5 100644 --- a/compiler/rustc_typeck/src/check/method/probe.rs +++ b/compiler/rustc_typeck/src/check/method/probe.rs @@ -644,7 +644,11 @@ impl<'a, 'tcx> ProbeContext<'a, 'tcx> { self.assemble_inherent_impl_candidates_for_type(p.def_id()); } ty::Adt(def, _) => { - self.assemble_inherent_impl_candidates_for_type(def.did()); + let def_id = def.did(); + self.assemble_inherent_impl_candidates_for_type(def_id); + if Some(def_id) == self.tcx.lang_items().c_str() { + self.assemble_inherent_candidates_for_incoherent_ty(raw_self_ty); + } } ty::Foreign(did) => { self.assemble_inherent_impl_candidates_for_type(did); diff --git a/compiler/rustc_typeck/src/coherence/inherent_impls.rs b/compiler/rustc_typeck/src/coherence/inherent_impls.rs index e11bd9355ebf9..5ad0c4ac52d49 100644 --- a/compiler/rustc_typeck/src/coherence/inherent_impls.rs +++ b/compiler/rustc_typeck/src/coherence/inherent_impls.rs @@ -55,7 +55,12 @@ impl<'tcx> ItemLikeVisitor<'_> for InherentCollect<'tcx> { let self_ty = self.tcx.type_of(item.def_id); match *self_ty.kind() { ty::Adt(def, _) => { - self.check_def_id(item, def.did()); + let def_id = def.did(); + if !def_id.is_local() && Some(def_id) == self.tcx.lang_items().c_str() { + self.check_primitive_impl(item.def_id, self_ty, items, ty.span) + } else { + self.check_def_id(item, def_id); + } } ty::Foreign(did) => { self.check_def_id(item, did); diff --git a/library/std/src/ffi/c_str.rs b/library/alloc/src/ffi/c_str.rs similarity index 63% rename from library/std/src/ffi/c_str.rs rename to library/alloc/src/ffi/c_str.rs index a68def1e83dbb..172a008e89a06 100644 --- a/library/std/src/ffi/c_str.rs +++ b/library/alloc/src/ffi/c_str.rs @@ -1,25 +1,25 @@ -#![deny(unsafe_op_in_unsafe_fn)] - #[cfg(test)] mod tests; -use crate::ascii; -use crate::borrow::{Borrow, Cow}; -use crate::cmp::Ordering; -use crate::error::Error; -use crate::fmt::{self, Write}; -use crate::io; -use crate::mem; -use crate::num::NonZeroU8; -use crate::ops; -use crate::os::raw::c_char; -use crate::ptr; +use crate::borrow::{Cow, ToOwned}; +use crate::boxed::Box; use crate::rc::Rc; -use crate::slice; -use crate::str::{self, Utf8Error}; +use crate::slice::hack::into_vec; +use crate::string::String; +use crate::vec::Vec; +use core::borrow::Borrow; +use core::ffi::{c_char, CStr}; +use core::fmt; +use core::mem; +use core::num::NonZeroU8; +use core::ops; +use core::ptr; +use core::slice; +use core::slice::memchr; +use core::str::{self, Utf8Error}; + +#[cfg(target_has_atomic = "ptr")] use crate::sync::Arc; -use crate::sys; -use crate::sys_common::memchr; /// A type representing an owned, C-compatible, nul-terminated string with no nul bytes in the /// middle. @@ -108,7 +108,7 @@ use crate::sys_common::memchr; /// and other memory errors. #[derive(PartialEq, PartialOrd, Eq, Ord, Hash, Clone)] #[cfg_attr(not(test), rustc_diagnostic_item = "cstring_type")] -#[stable(feature = "rust1", since = "1.0.0")] +#[unstable(feature = "alloc_c_string", issue = "94079")] pub struct CString { // Invariant 1: the slice ends with a zero byte and has a length of at least one. // Invariant 2: the slice contains only one zero byte. @@ -116,90 +116,6 @@ pub struct CString { inner: Box<[u8]>, } -/// Representation of a borrowed C string. -/// -/// This type represents a borrowed reference to a nul-terminated -/// array of bytes. It can be constructed safely from a &[[u8]] -/// slice, or unsafely from a raw `*const c_char`. It can then be -/// converted to a Rust &[str] by performing UTF-8 validation, or -/// into an owned [`CString`]. -/// -/// `&CStr` is to [`CString`] as &[str] is to [`String`]: the former -/// in each pair are borrowed references; the latter are owned -/// strings. -/// -/// Note that this structure is **not** `repr(C)` and is not recommended to be -/// placed in the signatures of FFI functions. Instead, safe wrappers of FFI -/// functions may leverage the unsafe [`CStr::from_ptr`] constructor to provide -/// a safe interface to other consumers. -/// -/// # Examples -/// -/// Inspecting a foreign C string: -/// -/// ```ignore (extern-declaration) -/// use std::ffi::CStr; -/// use std::os::raw::c_char; -/// -/// extern "C" { fn my_string() -> *const c_char; } -/// -/// unsafe { -/// let slice = CStr::from_ptr(my_string()); -/// println!("string buffer size without nul terminator: {}", slice.to_bytes().len()); -/// } -/// ``` -/// -/// Passing a Rust-originating C string: -/// -/// ```ignore (extern-declaration) -/// use std::ffi::{CString, CStr}; -/// use std::os::raw::c_char; -/// -/// fn work(data: &CStr) { -/// extern "C" { fn work_with(data: *const c_char); } -/// -/// unsafe { work_with(data.as_ptr()) } -/// } -/// -/// let s = CString::new("data data data data").expect("CString::new failed"); -/// work(&s); -/// ``` -/// -/// Converting a foreign C string into a Rust [`String`]: -/// -/// ```ignore (extern-declaration) -/// use std::ffi::CStr; -/// use std::os::raw::c_char; -/// -/// extern "C" { fn my_string() -> *const c_char; } -/// -/// fn my_string_safe() -> String { -/// unsafe { -/// CStr::from_ptr(my_string()).to_string_lossy().into_owned() -/// } -/// } -/// -/// println!("string: {}", my_string_safe()); -/// ``` -/// -/// [str]: prim@str "str" -#[derive(Hash)] -#[cfg_attr(not(test), rustc_diagnostic_item = "CStr")] -#[stable(feature = "rust1", since = "1.0.0")] -// FIXME: -// `fn from` in `impl From<&CStr> for Box` current implementation relies -// on `CStr` being layout-compatible with `[u8]`. -// When attribute privacy is implemented, `CStr` should be annotated as `#[repr(transparent)]`. -// Anyway, `CStr` representation and layout are considered implementation detail, are -// not documented and must not be relied upon. -pub struct CStr { - // FIXME: this should not be represented with a DST slice but rather with - // just a raw `c_char` along with some form of marker to make - // this an unsized type. Essentially `sizeof(&CStr)` should be the - // same as `sizeof(&c_char)` but `CStr` should be an unsized type. - inner: [c_char], -} - /// An error indicating that an interior nul byte was found. /// /// While Rust strings may contain nul bytes in the middle, C strings @@ -216,28 +132,13 @@ pub struct CStr { /// let _: NulError = CString::new(b"f\0oo".to_vec()).unwrap_err(); /// ``` #[derive(Clone, PartialEq, Eq, Debug)] -#[stable(feature = "rust1", since = "1.0.0")] +#[unstable(feature = "alloc_c_string", issue = "94079")] pub struct NulError(usize, Vec); -/// An error indicating that a nul byte was not in the expected position. -/// -/// The slice used to create a [`CStr`] must have one and only one nul byte, -/// positioned at the end. -/// -/// This error is created by the [`CStr::from_bytes_with_nul`] method. -/// See its documentation for more. -/// -/// # Examples -/// -/// ``` -/// use std::ffi::{CStr, FromBytesWithNulError}; -/// -/// let _: FromBytesWithNulError = CStr::from_bytes_with_nul(b"f\0oo").unwrap_err(); -/// ``` #[derive(Clone, PartialEq, Eq, Debug)] -#[stable(feature = "cstr_from_bytes", since = "1.10.0")] -pub struct FromBytesWithNulError { - kind: FromBytesWithNulErrorKind, +enum FromBytesWithNulErrorKind { + InteriorNul(usize), + NotNulTerminated, } /// An error indicating that a nul byte was not in the expected position. @@ -256,27 +157,12 @@ pub struct FromBytesWithNulError { /// let _: FromVecWithNulError = CString::from_vec_with_nul(b"f\0oo".to_vec()).unwrap_err(); /// ``` #[derive(Clone, PartialEq, Eq, Debug)] -#[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] +#[unstable(feature = "alloc_c_string", issue = "94079")] pub struct FromVecWithNulError { error_kind: FromBytesWithNulErrorKind, bytes: Vec, } -#[derive(Clone, PartialEq, Eq, Debug)] -enum FromBytesWithNulErrorKind { - InteriorNul(usize), - NotNulTerminated, -} - -impl FromBytesWithNulError { - fn interior_nul(pos: usize) -> FromBytesWithNulError { - FromBytesWithNulError { kind: FromBytesWithNulErrorKind::InteriorNul(pos) } - } - fn not_nul_terminated() -> FromBytesWithNulError { - FromBytesWithNulError { kind: FromBytesWithNulErrorKind::NotNulTerminated } - } -} - #[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] impl FromVecWithNulError { /// Returns a slice of [`u8`]s bytes that were attempted to convert to a [`CString`]. @@ -328,27 +214,6 @@ impl FromVecWithNulError { } } -/// An error indicating that no nul byte was present. -/// -/// A slice used to create a [`CStr`] must contain a nul byte somewhere -/// within the slice. -/// -/// This error is created by the [`CStr::from_bytes_until_nul`] method. -/// -#[derive(Clone, PartialEq, Eq, Debug)] -#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] -pub struct FromBytesUntilNulError(()); - -#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] -impl Error for FromBytesUntilNulError {} - -#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] -impl fmt::Display for FromBytesUntilNulError { - fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - write!(f, "data provided does not contain a nul") - } -} - /// An error indicating invalid UTF-8 when converting a [`CString`] into a [`String`]. /// /// `CString` is just a wrapper over a buffer of bytes with a nul terminator; @@ -358,7 +223,7 @@ impl fmt::Display for FromBytesUntilNulError { /// This `struct` is created by [`CString::into_string()`]. See /// its documentation for more. #[derive(Clone, PartialEq, Eq, Debug)] -#[stable(feature = "cstring_into", since = "1.7.0")] +#[unstable(feature = "alloc_c_string", issue = "94079")] pub struct IntoStringError { inner: CString, error: Utf8Error, @@ -536,7 +401,11 @@ impl CString { // information about the size of the allocation is correct on Rust's // side. unsafe { - let len = sys::strlen(ptr) + 1; // Including the NUL byte + extern "C" { + /// Provided by libc or compiler_builtins. + fn strlen(s: *const c_char) -> usize; + } + let len = strlen(ptr) + 1; // Including the NUL byte let slice = slice::from_raw_parts_mut(ptr, len as usize); CString { inner: Box::from_raw(slice as *mut [c_char] as *mut [u8]) } } @@ -626,7 +495,7 @@ impl CString { #[must_use = "`self` will be dropped if the result is not used"] #[stable(feature = "cstring_into", since = "1.7.0")] pub fn into_bytes(self) -> Vec { - let mut vec = self.into_inner().into_vec(); + let mut vec = into_vec(self.into_inner()); let _nul = vec.pop(); debug_assert_eq!(_nul, Some(0u8)); vec @@ -647,7 +516,7 @@ impl CString { #[must_use = "`self` will be dropped if the result is not used"] #[stable(feature = "cstring_into", since = "1.7.0")] pub fn into_bytes_with_nul(self) -> Vec { - self.into_inner().into_vec() + into_vec(self.into_inner()) } /// Returns the contents of this `CString` as a slice of bytes. @@ -842,7 +711,7 @@ impl ops::Deref for CString { #[inline] fn deref(&self) -> &CStr { - unsafe { CStr::_from_bytes_with_nul_unchecked(self.as_bytes_with_nul()) } + unsafe { CStr::from_bytes_with_nul_unchecked(self.as_bytes_with_nul()) } } } @@ -864,25 +733,6 @@ impl From for Vec { } } -#[stable(feature = "cstr_debug", since = "1.3.0")] -impl fmt::Debug for CStr { - fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - write!(f, "\"")?; - for byte in self.to_bytes().iter().flat_map(|&b| ascii::escape_default(b)) { - f.write_char(byte as char)?; - } - write!(f, "\"") - } -} - -#[stable(feature = "cstr_default", since = "1.10.0")] -impl Default for &CStr { - fn default() -> Self { - const SLICE: &[c_char] = &[0]; - unsafe { CStr::from_ptr(SLICE.as_ptr()) } - } -} - #[stable(feature = "cstr_default", since = "1.10.0")] impl Default for CString { /// Creates an empty `CString`. @@ -910,6 +760,7 @@ impl<'a> From> for CString { } } +#[cfg(not(test))] #[stable(feature = "box_from_c_str", since = "1.17.0")] impl From<&CStr> for Box { /// Converts a `&CStr` into a `Box`, @@ -938,7 +789,8 @@ impl From> for CString { /// Converts a [Box]<[CStr]> into a [`CString`] without copying or allocating. #[inline] fn from(s: Box) -> CString { - s.into_c_string() + let raw = Box::into_raw(s) as *mut [u8]; + CString { inner: unsafe { Box::from_raw(raw) } } } } @@ -964,6 +816,7 @@ impl From> for CString { } } +#[cfg(not(test))] #[stable(feature = "more_box_slice_clone", since = "1.29.0")] impl Clone for Box { #[inline] @@ -1008,6 +861,7 @@ impl<'a> From<&'a CString> for Cow<'a, CStr> { } } +#[cfg(target_has_atomic = "ptr")] #[stable(feature = "shared_from_slice2", since = "1.24.0")] impl From for Arc { /// Converts a [`CString`] into an [Arc]<[CStr]> by moving the [`CString`] @@ -1019,6 +873,7 @@ impl From for Arc { } } +#[cfg(target_has_atomic = "ptr")] #[stable(feature = "shared_from_slice2", since = "1.24.0")] impl From<&CStr> for Arc { /// Converts a `&CStr` into a `Arc`, @@ -1052,6 +907,7 @@ impl From<&CStr> for Rc { } } +#[cfg(not(test))] #[stable(feature = "default_box_extra", since = "1.17.0")] impl Default for Box { fn default() -> Box { @@ -1099,14 +955,6 @@ impl NulError { } } -#[stable(feature = "rust1", since = "1.0.0")] -impl Error for NulError { - #[allow(deprecated)] - fn description(&self) -> &str { - "nul byte found in data" - } -} - #[stable(feature = "rust1", since = "1.0.0")] impl fmt::Display for NulError { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { @@ -1114,42 +962,6 @@ impl fmt::Display for NulError { } } -#[stable(feature = "rust1", since = "1.0.0")] -impl From for io::Error { - /// Converts a [`NulError`] into a [`io::Error`]. - fn from(_: NulError) -> io::Error { - io::const_io_error!(io::ErrorKind::InvalidInput, "data provided contains a nul byte") - } -} - -#[stable(feature = "frombyteswithnulerror_impls", since = "1.17.0")] -impl Error for FromBytesWithNulError { - #[allow(deprecated)] - fn description(&self) -> &str { - match self.kind { - FromBytesWithNulErrorKind::InteriorNul(..) => { - "data provided contains an interior nul byte" - } - FromBytesWithNulErrorKind::NotNulTerminated => "data provided is not nul terminated", - } - } -} - -#[stable(feature = "frombyteswithnulerror_impls", since = "1.17.0")] -impl fmt::Display for FromBytesWithNulError { - #[allow(deprecated, deprecated_in_future)] - fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - f.write_str(self.description())?; - if let FromBytesWithNulErrorKind::InteriorNul(pos) = self.kind { - write!(f, " at byte pos {pos}")?; - } - Ok(()) - } -} - -#[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] -impl Error for FromVecWithNulError {} - #[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] impl fmt::Display for FromVecWithNulError { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { @@ -1179,18 +991,18 @@ impl IntoStringError { pub fn utf8_error(&self) -> Utf8Error { self.error } + + #[doc(hidden)] + #[unstable(feature = "cstr_internals", issue = "none")] + pub fn __source(&self) -> &Utf8Error { + &self.error + } } -#[stable(feature = "cstring_into", since = "1.7.0")] -impl Error for IntoStringError { - #[allow(deprecated)] +impl IntoStringError { fn description(&self) -> &str { "C string contained non-utf8 bytes" } - - fn source(&self) -> Option<&(dyn Error + 'static)> { - Some(&self.error) - } } #[stable(feature = "cstring_into", since = "1.7.0")] @@ -1201,327 +1013,126 @@ impl fmt::Display for IntoStringError { } } -impl CStr { - /// Wraps a raw C string with a safe C string wrapper. - /// - /// This function will wrap the provided `ptr` with a `CStr` wrapper, which - /// allows inspection and interoperation of non-owned C strings. The total - /// size of the raw C string must be smaller than `isize::MAX` **bytes** - /// in memory due to calling the `slice::from_raw_parts` function. - /// This method is unsafe for a number of reasons: - /// - /// * There is no guarantee to the validity of `ptr`. - /// * The returned lifetime is not guaranteed to be the actual lifetime of - /// `ptr`. - /// * There is no guarantee that the memory pointed to by `ptr` contains a - /// valid nul terminator byte at the end of the string. - /// * It is not guaranteed that the memory pointed by `ptr` won't change - /// before the `CStr` has been destroyed. - /// - /// > **Note**: This operation is intended to be a 0-cost cast but it is - /// > currently implemented with an up-front calculation of the length of - /// > the string. This is not guaranteed to always be the case. - /// - /// # Examples - /// - /// ```ignore (extern-declaration) - /// # fn main() { - /// use std::ffi::CStr; - /// use std::os::raw::c_char; - /// - /// extern "C" { - /// fn my_string() -> *const c_char; - /// } - /// - /// unsafe { - /// let slice = CStr::from_ptr(my_string()); - /// println!("string returned: {}", slice.to_str().unwrap()); - /// } - /// # } - /// ``` - #[inline] - #[must_use] - #[stable(feature = "rust1", since = "1.0.0")] - pub unsafe fn from_ptr<'a>(ptr: *const c_char) -> &'a CStr { - // SAFETY: The caller has provided a pointer that points to a valid C - // string with a NUL terminator of size less than `isize::MAX`, whose - // content remain valid and doesn't change for the lifetime of the - // returned `CStr`. - // - // Thus computing the length is fine (a NUL byte exists), the call to - // from_raw_parts is safe because we know the length is at most `isize::MAX`, meaning - // the call to `from_bytes_with_nul_unchecked` is correct. - // - // The cast from c_char to u8 is ok because a c_char is always one byte. - unsafe { - let len = sys::strlen(ptr); - let ptr = ptr as *const u8; - Self::_from_bytes_with_nul_unchecked(slice::from_raw_parts(ptr, len as usize + 1)) - } - } +#[stable(feature = "cstr_borrow", since = "1.3.0")] +impl ToOwned for CStr { + type Owned = CString; - /// Creates a C string wrapper from a byte slice. - /// - /// This method will create a `CStr` from any byte slice that contains at - /// least one nul byte. The caller does not need to know or specify where - /// the nul byte is located. - /// - /// If the first byte is a nul character, this method will return an - /// empty `CStr`. If multiple nul characters are present, the `CStr` will - /// end at the first one. - /// - /// If the slice only has a single nul byte at the end, this method is - /// equivalent to [`CStr::from_bytes_with_nul`]. - /// - /// # Examples - /// ``` - /// #![feature(cstr_from_bytes_until_nul)] - /// - /// use std::ffi::CStr; - /// - /// let mut buffer = [0u8; 16]; - /// unsafe { - /// // Here we might call an unsafe C function that writes a string - /// // into the buffer. - /// let buf_ptr = buffer.as_mut_ptr(); - /// buf_ptr.write_bytes(b'A', 8); - /// } - /// // Attempt to extract a C nul-terminated string from the buffer. - /// let c_str = CStr::from_bytes_until_nul(&buffer[..]).unwrap(); - /// assert_eq!(c_str.to_str().unwrap(), "AAAAAAAA"); - /// ``` - /// - #[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] - pub fn from_bytes_until_nul(bytes: &[u8]) -> Result<&CStr, FromBytesUntilNulError> { - let nul_pos = memchr::memchr(0, bytes); - match nul_pos { - Some(nul_pos) => { - // SAFETY: We know there is a nul byte at nul_pos, so this slice - // (ending at the nul byte) is a well-formed C string. - let subslice = &bytes[..nul_pos + 1]; - Ok(unsafe { CStr::from_bytes_with_nul_unchecked(subslice) }) - } - None => Err(FromBytesUntilNulError(())), - } + fn to_owned(&self) -> CString { + CString { inner: self.to_bytes_with_nul().into() } } - /// Creates a C string wrapper from a byte slice. - /// - /// This function will cast the provided `bytes` to a `CStr` - /// wrapper after ensuring that the byte slice is nul-terminated - /// and does not contain any interior nul bytes. - /// - /// If the nul byte may not be at the end, - /// [`CStr::from_bytes_until_nul`] can be used instead. - /// - /// # Examples - /// - /// ``` - /// use std::ffi::CStr; - /// - /// let cstr = CStr::from_bytes_with_nul(b"hello\0"); - /// assert!(cstr.is_ok()); - /// ``` - /// - /// Creating a `CStr` without a trailing nul terminator is an error: - /// - /// ``` - /// use std::ffi::CStr; - /// - /// let cstr = CStr::from_bytes_with_nul(b"hello"); - /// assert!(cstr.is_err()); - /// ``` - /// - /// Creating a `CStr` with an interior nul byte is an error: - /// - /// ``` - /// use std::ffi::CStr; - /// - /// let cstr = CStr::from_bytes_with_nul(b"he\0llo\0"); - /// assert!(cstr.is_err()); - /// ``` - #[stable(feature = "cstr_from_bytes", since = "1.10.0")] - pub fn from_bytes_with_nul(bytes: &[u8]) -> Result<&Self, FromBytesWithNulError> { - let nul_pos = memchr::memchr(0, bytes); - match nul_pos { - Some(nul_pos) if nul_pos + 1 == bytes.len() => { - // SAFETY: We know there is only one nul byte, at the end - // of the byte slice. - Ok(unsafe { Self::_from_bytes_with_nul_unchecked(bytes) }) - } - Some(nul_pos) => Err(FromBytesWithNulError::interior_nul(nul_pos)), - None => Err(FromBytesWithNulError::not_nul_terminated()), - } + fn clone_into(&self, target: &mut CString) { + let mut b = into_vec(mem::take(&mut target.inner)); + self.to_bytes_with_nul().clone_into(&mut b); + target.inner = b.into_boxed_slice(); } +} - /// Unsafely creates a C string wrapper from a byte slice. - /// - /// This function will cast the provided `bytes` to a `CStr` wrapper without - /// performing any sanity checks. The provided slice **must** be nul-terminated - /// and not contain any interior nul bytes. - /// - /// # Examples - /// - /// ``` - /// use std::ffi::{CStr, CString}; - /// - /// unsafe { - /// let cstring = CString::new("hello").expect("CString::new failed"); - /// let cstr = CStr::from_bytes_with_nul_unchecked(cstring.to_bytes_with_nul()); - /// assert_eq!(cstr, &*cstring); - /// } - /// ``` - #[inline] - #[must_use] - #[stable(feature = "cstr_from_bytes", since = "1.10.0")] - #[rustc_const_stable(feature = "const_cstr_unchecked", since = "1.59.0")] - pub const unsafe fn from_bytes_with_nul_unchecked(bytes: &[u8]) -> &CStr { - // We're in a const fn, so this is the best we can do - debug_assert!(!bytes.is_empty() && bytes[bytes.len() - 1] == 0); - unsafe { Self::_from_bytes_with_nul_unchecked(bytes) } +#[stable(feature = "cstring_asref", since = "1.7.0")] +impl From<&CStr> for CString { + fn from(s: &CStr) -> CString { + s.to_owned() } +} + +#[stable(feature = "cstring_asref", since = "1.7.0")] +impl ops::Index for CString { + type Output = CStr; #[inline] - const unsafe fn _from_bytes_with_nul_unchecked(bytes: &[u8]) -> &Self { - // SAFETY: Casting to CStr is safe because its internal representation - // is a [u8] too (safe only inside std). - // Dereferencing the obtained pointer is safe because it comes from a - // reference. Making a reference is then safe because its lifetime - // is bound by the lifetime of the given `bytes`. - unsafe { &*(bytes as *const [u8] as *const Self) } + fn index(&self, _index: ops::RangeFull) -> &CStr { + self } +} - /// Returns the inner pointer to this C string. - /// - /// The returned pointer will be valid for as long as `self` is, and points - /// to a contiguous region of memory terminated with a 0 byte to represent - /// the end of the string. - /// - /// **WARNING** - /// - /// The returned pointer is read-only; writing to it (including passing it - /// to C code that writes to it) causes undefined behavior. - /// - /// It is your responsibility to make sure that the underlying memory is not - /// freed too early. For example, the following code will cause undefined - /// behavior when `ptr` is used inside the `unsafe` block: - /// - /// ```no_run - /// # #![allow(unused_must_use)] #![allow(temporary_cstring_as_ptr)] - /// use std::ffi::CString; - /// - /// let ptr = CString::new("Hello").expect("CString::new failed").as_ptr(); - /// unsafe { - /// // `ptr` is dangling - /// *ptr; - /// } - /// ``` - /// - /// This happens because the pointer returned by `as_ptr` does not carry any - /// lifetime information and the [`CString`] is deallocated immediately after - /// the `CString::new("Hello").expect("CString::new failed").as_ptr()` - /// expression is evaluated. - /// To fix the problem, bind the `CString` to a local variable: - /// - /// ```no_run - /// # #![allow(unused_must_use)] - /// use std::ffi::CString; - /// - /// let hello = CString::new("Hello").expect("CString::new failed"); - /// let ptr = hello.as_ptr(); - /// unsafe { - /// // `ptr` is valid because `hello` is in scope - /// *ptr; - /// } - /// ``` - /// - /// This way, the lifetime of the [`CString`] in `hello` encompasses - /// the lifetime of `ptr` and the `unsafe` block. +#[stable(feature = "cstring_asref", since = "1.7.0")] +impl AsRef for CString { #[inline] - #[must_use] - #[stable(feature = "rust1", since = "1.0.0")] - #[rustc_const_stable(feature = "const_str_as_ptr", since = "1.32.0")] - pub const fn as_ptr(&self) -> *const c_char { - self.inner.as_ptr() + fn as_ref(&self) -> &CStr { + self } +} - /// Converts this C string to a byte slice. +#[cfg(bootstrap)] +#[doc(hidden)] +#[unstable(feature = "cstr_internals", issue = "none")] +pub trait CStrExt { + /// Converts a `CStr` into a [Cow]<[str]>. /// - /// The returned slice will **not** contain the trailing nul terminator that this C - /// string has. + /// If the contents of the `CStr` are valid UTF-8 data, this + /// function will return a [Cow]::[Borrowed]\(&[str]) + /// with the corresponding &[str] slice. Otherwise, it will + /// replace any invalid UTF-8 sequences with + /// [`U+FFFD REPLACEMENT CHARACTER`][U+FFFD] and return a + /// [Cow]::[Owned]\(&[str]) with the result. /// - /// > **Note**: This method is currently implemented as a constant-time - /// > cast, but it is planned to alter its definition in the future to - /// > perform the length calculation whenever this method is called. + /// [str]: prim@str "str" + /// [Borrowed]: Cow::Borrowed + /// [Owned]: Cow::Owned + /// [U+FFFD]: crate::char::REPLACEMENT_CHARACTER "std::char::REPLACEMENT_CHARACTER" /// /// # Examples /// + /// Calling `to_string_lossy` on a `CStr` containing valid UTF-8: + /// /// ``` + /// use std::borrow::Cow; /// use std::ffi::CStr; /// - /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); - /// assert_eq!(cstr.to_bytes(), b"foo"); + /// let cstr = CStr::from_bytes_with_nul(b"Hello World\0") + /// .expect("CStr::from_bytes_with_nul failed"); + /// assert_eq!(cstr.to_string_lossy(), Cow::Borrowed("Hello World")); /// ``` - #[inline] - #[must_use = "this returns the result of the operation, \ - without modifying the original"] - #[stable(feature = "rust1", since = "1.0.0")] - pub fn to_bytes(&self) -> &[u8] { - let bytes = self.to_bytes_with_nul(); - // SAFETY: to_bytes_with_nul returns slice with length at least 1 - unsafe { bytes.get_unchecked(..bytes.len() - 1) } - } - - /// Converts this C string to a byte slice containing the trailing 0 byte. /// - /// This function is the equivalent of [`CStr::to_bytes`] except that it - /// will retain the trailing nul terminator instead of chopping it off. - /// - /// > **Note**: This method is currently implemented as a 0-cost cast, but - /// > it is planned to alter its definition in the future to perform the - /// > length calculation whenever this method is called. - /// - /// # Examples + /// Calling `to_string_lossy` on a `CStr` containing invalid UTF-8: /// /// ``` + /// use std::borrow::Cow; /// use std::ffi::CStr; /// - /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); - /// assert_eq!(cstr.to_bytes_with_nul(), b"foo\0"); + /// let cstr = CStr::from_bytes_with_nul(b"Hello \xF0\x90\x80World\0") + /// .expect("CStr::from_bytes_with_nul failed"); + /// assert_eq!( + /// cstr.to_string_lossy(), + /// Cow::Owned(String::from("Hello �World")) as Cow<'_, str> + /// ); /// ``` - #[inline] #[must_use = "this returns the result of the operation, \ without modifying the original"] - #[stable(feature = "rust1", since = "1.0.0")] - pub fn to_bytes_with_nul(&self) -> &[u8] { - unsafe { &*(&self.inner as *const [c_char] as *const [u8]) } - } + #[stable(feature = "cstr_to_str", since = "1.4.0")] + fn to_string_lossy(&self) -> Cow<'_, str>; - /// Yields a &[str] slice if the `CStr` contains valid UTF-8. - /// - /// If the contents of the `CStr` are valid UTF-8 data, this - /// function will return the corresponding &[str] slice. Otherwise, - /// it will return an error with details of where UTF-8 validation failed. - /// - /// [str]: prim@str "str" + /// Converts a [Box]<[CStr]> into a [`CString`] without copying or allocating. /// /// # Examples /// /// ``` - /// use std::ffi::CStr; + /// use std::ffi::CString; /// - /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); - /// assert_eq!(cstr.to_str(), Ok("foo")); + /// let c_string = CString::new(b"foo".to_vec()).expect("CString::new failed"); + /// let boxed = c_string.into_boxed_c_str(); + /// assert_eq!(boxed.into_c_string(), CString::new("foo").expect("CString::new failed")); /// ``` - #[stable(feature = "cstr_to_str", since = "1.4.0")] - pub fn to_str(&self) -> Result<&str, str::Utf8Error> { - // N.B., when `CStr` is changed to perform the length check in `.to_bytes()` - // instead of in `from_ptr()`, it may be worth considering if this should - // be rewritten to do the UTF-8 check inline with the length calculation - // instead of doing it afterwards. - str::from_utf8(self.to_bytes()) + #[must_use = "`self` will be dropped if the result is not used"] + #[stable(feature = "into_boxed_c_str", since = "1.20.0")] + fn into_c_string(self: Box) -> CString; +} + +#[cfg(bootstrap)] +#[unstable(feature = "cstr_internals", issue = "none")] +impl CStrExt for CStr { + fn to_string_lossy(&self) -> Cow<'_, str> { + String::from_utf8_lossy(self.to_bytes()) } + fn into_c_string(self: Box) -> CString { + CString::from(self) + } +} + +#[cfg(not(bootstrap))] +#[cfg(not(test))] +impl CStr { /// Converts a `CStr` into a [Cow]<[str]>. /// /// If the contents of the `CStr` are valid UTF-8 data, this @@ -1534,7 +1145,7 @@ impl CStr { /// [str]: prim@str "str" /// [Borrowed]: Cow::Borrowed /// [Owned]: Cow::Owned - /// [U+FFFD]: crate::char::REPLACEMENT_CHARACTER "std::char::REPLACEMENT_CHARACTER" + /// [U+FFFD]: core::char::REPLACEMENT_CHARACTER "std::char::REPLACEMENT_CHARACTER" /// /// # Examples /// @@ -1562,6 +1173,7 @@ impl CStr { /// Cow::Owned(String::from("Hello �World")) as Cow<'_, str> /// ); /// ``` + #[rustc_allow_incoherent_impl] #[must_use = "this returns the result of the operation, \ without modifying the original"] #[stable(feature = "cstr_to_str", since = "1.4.0")] @@ -1580,101 +1192,10 @@ impl CStr { /// let boxed = c_string.into_boxed_c_str(); /// assert_eq!(boxed.into_c_string(), CString::new("foo").expect("CString::new failed")); /// ``` + #[rustc_allow_incoherent_impl] #[must_use = "`self` will be dropped if the result is not used"] #[stable(feature = "into_boxed_c_str", since = "1.20.0")] - pub fn into_c_string(self: Box) -> CString { - let raw = Box::into_raw(self) as *mut [u8]; - CString { inner: unsafe { Box::from_raw(raw) } } - } -} - -#[stable(feature = "rust1", since = "1.0.0")] -impl PartialEq for CStr { - fn eq(&self, other: &CStr) -> bool { - self.to_bytes().eq(other.to_bytes()) - } -} -#[stable(feature = "rust1", since = "1.0.0")] -impl Eq for CStr {} -#[stable(feature = "rust1", since = "1.0.0")] -impl PartialOrd for CStr { - fn partial_cmp(&self, other: &CStr) -> Option { - self.to_bytes().partial_cmp(&other.to_bytes()) - } -} -#[stable(feature = "rust1", since = "1.0.0")] -impl Ord for CStr { - fn cmp(&self, other: &CStr) -> Ordering { - self.to_bytes().cmp(&other.to_bytes()) - } -} - -#[stable(feature = "cstr_borrow", since = "1.3.0")] -impl ToOwned for CStr { - type Owned = CString; - - fn to_owned(&self) -> CString { - CString { inner: self.to_bytes_with_nul().into() } - } - - fn clone_into(&self, target: &mut CString) { - let mut b = Vec::from(mem::take(&mut target.inner)); - self.to_bytes_with_nul().clone_into(&mut b); - target.inner = b.into_boxed_slice(); - } -} - -#[stable(feature = "cstring_asref", since = "1.7.0")] -impl From<&CStr> for CString { - /// Copies the contents of the `&CStr` into a newly allocated `CString`. - fn from(s: &CStr) -> CString { - s.to_owned() - } -} - -#[stable(feature = "cstring_asref", since = "1.7.0")] -impl ops::Index for CString { - type Output = CStr; - - #[inline] - fn index(&self, _index: ops::RangeFull) -> &CStr { - self - } -} - -#[stable(feature = "cstr_range_from", since = "1.47.0")] -impl ops::Index> for CStr { - type Output = CStr; - - fn index(&self, index: ops::RangeFrom) -> &CStr { - let bytes = self.to_bytes_with_nul(); - // we need to manually check the starting index to account for the null - // byte, since otherwise we could get an empty string that doesn't end - // in a null. - if index.start < bytes.len() { - unsafe { CStr::_from_bytes_with_nul_unchecked(&bytes[index.start..]) } - } else { - panic!( - "index out of bounds: the len is {} but the index is {}", - bytes.len(), - index.start - ); - } - } -} - -#[stable(feature = "cstring_asref", since = "1.7.0")] -impl AsRef for CStr { - #[inline] - fn as_ref(&self) -> &CStr { - self - } -} - -#[stable(feature = "cstring_asref", since = "1.7.0")] -impl AsRef for CString { - #[inline] - fn as_ref(&self) -> &CStr { - self + pub fn into_c_string(self: Box) -> CString { + CString::from(self) } } diff --git a/library/std/src/ffi/c_str/tests.rs b/library/alloc/src/ffi/c_str/tests.rs similarity index 87% rename from library/std/src/ffi/c_str/tests.rs rename to library/alloc/src/ffi/c_str/tests.rs index c20da138a18d0..0b7476d5cc7b7 100644 --- a/library/std/src/ffi/c_str/tests.rs +++ b/library/alloc/src/ffi/c_str/tests.rs @@ -1,10 +1,12 @@ use super::*; -use crate::borrow::Cow::{Borrowed, Owned}; -use crate::collections::hash_map::DefaultHasher; -use crate::hash::{Hash, Hasher}; -use crate::os::raw::c_char; use crate::rc::Rc; use crate::sync::Arc; +use core::assert_matches::assert_matches; +use core::ffi::FromBytesUntilNulError; +use core::hash::{Hash, Hasher}; + +#[allow(deprecated)] +use core::hash::SipHasher13 as DefaultHasher; #[test] fn c_to_rust() { @@ -47,22 +49,6 @@ fn borrowed() { } } -#[test] -fn to_str() { - let data = b"123\xE2\x80\xA6\0"; - let ptr = data.as_ptr() as *const c_char; - unsafe { - assert_eq!(CStr::from_ptr(ptr).to_str(), Ok("123…")); - assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Borrowed("123…")); - } - let data = b"123\xE2\0"; - let ptr = data.as_ptr() as *const c_char; - unsafe { - assert!(CStr::from_ptr(ptr).to_str().is_err()); - assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Owned::(format!("123\u{FFFD}"))); - } -} - #[test] fn to_owned() { let data = b"123\0"; @@ -78,9 +64,11 @@ fn equal_hash() { let ptr = data.as_ptr() as *const c_char; let cstr: &'static CStr = unsafe { CStr::from_ptr(ptr) }; + #[allow(deprecated)] let mut s = DefaultHasher::new(); cstr.hash(&mut s); let cstr_hash = s.finish(); + #[allow(deprecated)] let mut s = DefaultHasher::new(); CString::new(&data[..data.len() - 1]).unwrap().hash(&mut s); let cstring_hash = s.finish(); @@ -122,11 +110,11 @@ fn cstr_from_bytes_until_nul() { // Test an empty slice. This should fail because it // does not contain a nul byte. let b = b""; - assert_eq!(CStr::from_bytes_until_nul(&b[..]), Err(FromBytesUntilNulError(()))); + assert_matches!(CStr::from_bytes_until_nul(&b[..]), Err(FromBytesUntilNulError { .. })); // Test a non-empty slice, that does not contain a nul byte. let b = b"hello"; - assert_eq!(CStr::from_bytes_until_nul(&b[..]), Err(FromBytesUntilNulError(()))); + assert_matches!(CStr::from_bytes_until_nul(&b[..]), Err(FromBytesUntilNulError { .. })); // Test an empty nul-terminated string let b = b"\0"; diff --git a/library/alloc/src/ffi/mod.rs b/library/alloc/src/ffi/mod.rs new file mode 100644 index 0000000000000..eed2851c1539b --- /dev/null +++ b/library/alloc/src/ffi/mod.rs @@ -0,0 +1,91 @@ +//! Utilities related to FFI bindings. +//! +//! This module provides utilities to handle data across non-Rust +//! interfaces, like other programming languages and the underlying +//! operating system. It is mainly of use for FFI (Foreign Function +//! Interface) bindings and code that needs to exchange C-like strings +//! with other languages. +//! +//! # Overview +//! +//! Rust represents owned strings with the [`String`] type, and +//! borrowed slices of strings with the [`str`] primitive. Both are +//! always in UTF-8 encoding, and may contain nul bytes in the middle, +//! i.e., if you look at the bytes that make up the string, there may +//! be a `\0` among them. Both `String` and `str` store their length +//! explicitly; there are no nul terminators at the end of strings +//! like in C. +//! +//! C strings are different from Rust strings: +//! +//! * **Encodings** - Rust strings are UTF-8, but C strings may use +//! other encodings. If you are using a string from C, you should +//! check its encoding explicitly, rather than just assuming that it +//! is UTF-8 like you can do in Rust. +//! +//! * **Character size** - C strings may use `char` or `wchar_t`-sized +//! characters; please **note** that C's `char` is different from Rust's. +//! The C standard leaves the actual sizes of those types open to +//! interpretation, but defines different APIs for strings made up of +//! each character type. Rust strings are always UTF-8, so different +//! Unicode characters will be encoded in a variable number of bytes +//! each. The Rust type [`char`] represents a '[Unicode scalar +//! value]', which is similar to, but not the same as, a '[Unicode +//! code point]'. +//! +//! * **Nul terminators and implicit string lengths** - Often, C +//! strings are nul-terminated, i.e., they have a `\0` character at the +//! end. The length of a string buffer is not stored, but has to be +//! calculated; to compute the length of a string, C code must +//! manually call a function like `strlen()` for `char`-based strings, +//! or `wcslen()` for `wchar_t`-based ones. Those functions return +//! the number of characters in the string excluding the nul +//! terminator, so the buffer length is really `len+1` characters. +//! Rust strings don't have a nul terminator; their length is always +//! stored and does not need to be calculated. While in Rust +//! accessing a string's length is an *O*(1) operation (because the +//! length is stored); in C it is an *O*(*n*) operation because the +//! length needs to be computed by scanning the string for the nul +//! terminator. +//! +//! * **Internal nul characters** - When C strings have a nul +//! terminator character, this usually means that they cannot have nul +//! characters in the middle — a nul character would essentially +//! truncate the string. Rust strings *can* have nul characters in +//! the middle, because nul does not have to mark the end of the +//! string in Rust. +//! +//! # Representations of non-Rust strings +//! +//! [`CString`] and [`CStr`] are useful when you need to transfer +//! UTF-8 strings to and from languages with a C ABI, like Python. +//! +//! * **From Rust to C:** [`CString`] represents an owned, C-friendly +//! string: it is nul-terminated, and has no internal nul characters. +//! Rust code can create a [`CString`] out of a normal string (provided +//! that the string doesn't have nul characters in the middle), and +//! then use a variety of methods to obtain a raw \*mut [u8] that can +//! then be passed as an argument to functions which use the C +//! conventions for strings. +//! +//! * **From C to Rust:** [`CStr`] represents a borrowed C string; it +//! is what you would use to wrap a raw \*const [u8] that you got from +//! a C function. A [`CStr`] is guaranteed to be a nul-terminated array +//! of bytes. Once you have a [`CStr`], you can convert it to a Rust +//! &[str] if it's valid UTF-8, or lossily convert it by adding +//! replacement characters. +//! +//! [`String`]: crate::string::String +//! [`CStr`]: core::ffi::CStr + +#![unstable(feature = "alloc_ffi", issue = "94079")] + +#[cfg(bootstrap)] +#[unstable(feature = "cstr_internals", issue = "none")] +pub use self::c_str::CStrExt; +#[unstable(feature = "alloc_c_string", issue = "94079")] +pub use self::c_str::FromVecWithNulError; +#[unstable(feature = "alloc_c_string", issue = "94079")] +pub use self::c_str::{CString, IntoStringError, NulError}; + +mod c_str; diff --git a/library/alloc/src/lib.rs b/library/alloc/src/lib.rs index c54001dceea49..4d2dc4ecee0b8 100644 --- a/library/alloc/src/lib.rs +++ b/library/alloc/src/lib.rs @@ -86,11 +86,13 @@ #![allow(explicit_outlives_requirements)] // // Library features: +#![cfg_attr(not(no_global_oom_handling), feature(alloc_c_string))] #![feature(alloc_layout_extra)] #![feature(allocator_api)] #![feature(array_chunks)] #![feature(array_methods)] #![feature(array_windows)] +#![feature(assert_matches)] #![feature(async_iterator)] #![feature(coerce_unsized)] #![cfg_attr(not(no_global_oom_handling), feature(const_alloc_error))] @@ -104,9 +106,12 @@ #![feature(const_maybe_uninit_write)] #![feature(const_maybe_uninit_as_mut_ptr)] #![feature(const_refs_to_cell)] +#![feature(core_c_str)] #![feature(core_intrinsics)] +#![feature(core_ffi_c)] #![feature(const_eval_select)] #![feature(const_pin)] +#![feature(cstr_from_bytes_until_nul)] #![feature(dispatch_from_dyn)] #![feature(exact_size_is_empty)] #![feature(extend_one)] @@ -152,6 +157,7 @@ #![feature(exclusive_range_pattern)] #![feature(fundamental)] #![cfg_attr(not(test), feature(generator_trait))] +#![feature(hashmap_internals)] #![feature(lang_items)] #![feature(let_else)] #![feature(min_specialization)] @@ -160,6 +166,7 @@ #![feature(nll)] // Not necessary, but here to test the `nll` feature. #![feature(rustc_allow_const_fn_unstable)] #![feature(rustc_attrs)] +#![feature(slice_internals)] #![feature(staged_api)] #![cfg_attr(test, feature(test))] #![feature(unboxed_closures)] @@ -205,6 +212,8 @@ mod boxed { } pub mod borrow; pub mod collections; +#[cfg(not(no_global_oom_handling))] +pub mod ffi; pub mod fmt; pub mod rc; pub mod slice; diff --git a/library/alloc/src/slice.rs b/library/alloc/src/slice.rs index 31edbe0c5afd5..02a47c57b8ae3 100644 --- a/library/alloc/src/slice.rs +++ b/library/alloc/src/slice.rs @@ -153,7 +153,7 @@ pub use hack::to_vec; // functions are actually methods that are in `impl [T]` but not in // `core::slice::SliceExt` - we need to supply these functions for the // `test_permutations` test -mod hack { +pub(crate) mod hack { use core::alloc::Allocator; use crate::boxed::Box; diff --git a/library/alloc/tests/c_str.rs b/library/alloc/tests/c_str.rs new file mode 100644 index 0000000000000..8fbb10e1d5cb6 --- /dev/null +++ b/library/alloc/tests/c_str.rs @@ -0,0 +1,18 @@ +use std::borrow::Cow::{Borrowed, Owned}; +use std::ffi::{c_char, CStr}; + +#[test] +fn to_str() { + let data = b"123\xE2\x80\xA6\0"; + let ptr = data.as_ptr() as *const c_char; + unsafe { + assert_eq!(CStr::from_ptr(ptr).to_str(), Ok("123…")); + assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Borrowed("123…")); + } + let data = b"123\xE2\0"; + let ptr = data.as_ptr() as *const c_char; + unsafe { + assert!(CStr::from_ptr(ptr).to_str().is_err()); + assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Owned::(format!("123\u{FFFD}"))); + } +} diff --git a/library/alloc/tests/lib.rs b/library/alloc/tests/lib.rs index 16d3b36859570..8de159246c6e0 100644 --- a/library/alloc/tests/lib.rs +++ b/library/alloc/tests/lib.rs @@ -12,6 +12,8 @@ #![feature(const_nonnull_slice_from_raw_parts)] #![feature(const_ptr_write)] #![feature(const_try)] +#![feature(core_c_str)] +#![feature(core_ffi_c)] #![feature(core_intrinsics)] #![feature(drain_filter)] #![feature(exact_size_is_empty)] @@ -49,6 +51,7 @@ mod binary_heap; mod borrow; mod boxed; mod btree_set_hash; +mod c_str; mod const_fns; mod cow_str; mod fmt; diff --git a/library/core/src/ffi/c_str.rs b/library/core/src/ffi/c_str.rs new file mode 100644 index 0000000000000..15d9d013997be --- /dev/null +++ b/library/core/src/ffi/c_str.rs @@ -0,0 +1,570 @@ +use crate::ascii; +use crate::cmp::Ordering; +use crate::ffi::c_char; +use crate::fmt::{self, Write}; +use crate::ops; +use crate::slice; +use crate::slice::memchr; +use crate::str; + +/// Representation of a borrowed C string. +/// +/// This type represents a borrowed reference to a nul-terminated +/// array of bytes. It can be constructed safely from a &[[u8]] +/// slice, or unsafely from a raw `*const c_char`. It can then be +/// converted to a Rust &[str] by performing UTF-8 validation, or +/// into an owned `CString`. +/// +/// `&CStr` is to `CString` as &[str] is to `String`: the former +/// in each pair are borrowed references; the latter are owned +/// strings. +/// +/// Note that this structure is **not** `repr(C)` and is not recommended to be +/// placed in the signatures of FFI functions. Instead, safe wrappers of FFI +/// functions may leverage the unsafe [`CStr::from_ptr`] constructor to provide +/// a safe interface to other consumers. +/// +/// # Examples +/// +/// Inspecting a foreign C string: +/// +/// ```ignore (extern-declaration) +/// use std::ffi::CStr; +/// use std::os::raw::c_char; +/// +/// extern "C" { fn my_string() -> *const c_char; } +/// +/// unsafe { +/// let slice = CStr::from_ptr(my_string()); +/// println!("string buffer size without nul terminator: {}", slice.to_bytes().len()); +/// } +/// ``` +/// +/// Passing a Rust-originating C string: +/// +/// ```ignore (extern-declaration) +/// use std::ffi::{CString, CStr}; +/// use std::os::raw::c_char; +/// +/// fn work(data: &CStr) { +/// extern "C" { fn work_with(data: *const c_char); } +/// +/// unsafe { work_with(data.as_ptr()) } +/// } +/// +/// let s = CString::new("data data data data").expect("CString::new failed"); +/// work(&s); +/// ``` +/// +/// Converting a foreign C string into a Rust `String`: +/// +/// ```ignore (extern-declaration) +/// use std::ffi::CStr; +/// use std::os::raw::c_char; +/// +/// extern "C" { fn my_string() -> *const c_char; } +/// +/// fn my_string_safe() -> String { +/// unsafe { +/// CStr::from_ptr(my_string()).to_string_lossy().into_owned() +/// } +/// } +/// +/// println!("string: {}", my_string_safe()); +/// ``` +/// +/// [str]: prim@str "str" +#[derive(Hash)] +#[cfg_attr(not(test), rustc_diagnostic_item = "CStr")] +#[unstable(feature = "core_c_str", issue = "94079")] +#[cfg_attr(not(bootstrap), lang = "CStr")] +// FIXME: +// `fn from` in `impl From<&CStr> for Box` current implementation relies +// on `CStr` being layout-compatible with `[u8]`. +// When attribute privacy is implemented, `CStr` should be annotated as `#[repr(transparent)]`. +// Anyway, `CStr` representation and layout are considered implementation detail, are +// not documented and must not be relied upon. +pub struct CStr { + // FIXME: this should not be represented with a DST slice but rather with + // just a raw `c_char` along with some form of marker to make + // this an unsized type. Essentially `sizeof(&CStr)` should be the + // same as `sizeof(&c_char)` but `CStr` should be an unsized type. + inner: [c_char], +} + +/// An error indicating that a nul byte was not in the expected position. +/// +/// The slice used to create a [`CStr`] must have one and only one nul byte, +/// positioned at the end. +/// +/// This error is created by the [`CStr::from_bytes_with_nul`] method. +/// See its documentation for more. +/// +/// # Examples +/// +/// ``` +/// use std::ffi::{CStr, FromBytesWithNulError}; +/// +/// let _: FromBytesWithNulError = CStr::from_bytes_with_nul(b"f\0oo").unwrap_err(); +/// ``` +#[derive(Clone, PartialEq, Eq, Debug)] +#[unstable(feature = "core_c_str", issue = "94079")] +pub struct FromBytesWithNulError { + kind: FromBytesWithNulErrorKind, +} + +#[derive(Clone, PartialEq, Eq, Debug)] +enum FromBytesWithNulErrorKind { + InteriorNul(usize), + NotNulTerminated, +} + +impl FromBytesWithNulError { + fn interior_nul(pos: usize) -> FromBytesWithNulError { + FromBytesWithNulError { kind: FromBytesWithNulErrorKind::InteriorNul(pos) } + } + fn not_nul_terminated() -> FromBytesWithNulError { + FromBytesWithNulError { kind: FromBytesWithNulErrorKind::NotNulTerminated } + } + + #[doc(hidden)] + #[unstable(feature = "cstr_internals", issue = "none")] + pub fn __description(&self) -> &str { + match self.kind { + FromBytesWithNulErrorKind::InteriorNul(..) => { + "data provided contains an interior nul byte" + } + FromBytesWithNulErrorKind::NotNulTerminated => "data provided is not nul terminated", + } + } +} + +/// An error indicating that no nul byte was present. +/// +/// A slice used to create a [`CStr`] must contain a nul byte somewhere +/// within the slice. +/// +/// This error is created by the [`CStr::from_bytes_until_nul`] method. +/// +#[derive(Clone, PartialEq, Eq, Debug)] +#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] +pub struct FromBytesUntilNulError(()); + +#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] +impl fmt::Display for FromBytesUntilNulError { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "data provided does not contain a nul") + } +} + +#[stable(feature = "cstr_debug", since = "1.3.0")] +impl fmt::Debug for CStr { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "\"")?; + for byte in self.to_bytes().iter().flat_map(|&b| ascii::escape_default(b)) { + f.write_char(byte as char)?; + } + write!(f, "\"") + } +} + +#[stable(feature = "cstr_default", since = "1.10.0")] +impl Default for &CStr { + fn default() -> Self { + const SLICE: &[c_char] = &[0]; + // SAFETY: `SLICE` is indeed pointing to a valid nul-terminated string. + unsafe { CStr::from_ptr(SLICE.as_ptr()) } + } +} + +#[stable(feature = "frombyteswithnulerror_impls", since = "1.17.0")] +impl fmt::Display for FromBytesWithNulError { + #[allow(deprecated, deprecated_in_future)] + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + f.write_str(self.__description())?; + if let FromBytesWithNulErrorKind::InteriorNul(pos) = self.kind { + write!(f, " at byte pos {pos}")?; + } + Ok(()) + } +} + +impl CStr { + /// Wraps a raw C string with a safe C string wrapper. + /// + /// This function will wrap the provided `ptr` with a `CStr` wrapper, which + /// allows inspection and interoperation of non-owned C strings. The total + /// size of the raw C string must be smaller than `isize::MAX` **bytes** + /// in memory due to calling the `slice::from_raw_parts` function. + /// This method is unsafe for a number of reasons: + /// + /// * There is no guarantee to the validity of `ptr`. + /// * The returned lifetime is not guaranteed to be the actual lifetime of + /// `ptr`. + /// * There is no guarantee that the memory pointed to by `ptr` contains a + /// valid nul terminator byte at the end of the string. + /// * It is not guaranteed that the memory pointed by `ptr` won't change + /// before the `CStr` has been destroyed. + /// + /// > **Note**: This operation is intended to be a 0-cost cast but it is + /// > currently implemented with an up-front calculation of the length of + /// > the string. This is not guaranteed to always be the case. + /// + /// # Examples + /// + /// ```ignore (extern-declaration) + /// # fn main() { + /// use std::ffi::CStr; + /// use std::os::raw::c_char; + /// + /// extern "C" { + /// fn my_string() -> *const c_char; + /// } + /// + /// unsafe { + /// let slice = CStr::from_ptr(my_string()); + /// println!("string returned: {}", slice.to_str().unwrap()); + /// } + /// # } + /// ``` + #[inline] + #[must_use] + #[stable(feature = "rust1", since = "1.0.0")] + pub unsafe fn from_ptr<'a>(ptr: *const c_char) -> &'a CStr { + // SAFETY: The caller has provided a pointer that points to a valid C + // string with a NUL terminator of size less than `isize::MAX`, whose + // content remain valid and doesn't change for the lifetime of the + // returned `CStr`. + // + // Thus computing the length is fine (a NUL byte exists), the call to + // from_raw_parts is safe because we know the length is at most `isize::MAX`, meaning + // the call to `from_bytes_with_nul_unchecked` is correct. + // + // The cast from c_char to u8 is ok because a c_char is always one byte. + unsafe { + extern "C" { + /// Provided by libc or compiler_builtins. + fn strlen(s: *const c_char) -> usize; + } + let len = strlen(ptr); + let ptr = ptr as *const u8; + CStr::from_bytes_with_nul_unchecked(slice::from_raw_parts(ptr, len as usize + 1)) + } + } + + /// Creates a C string wrapper from a byte slice. + /// + /// This method will create a `CStr` from any byte slice that contains at + /// least one nul byte. The caller does not need to know or specify where + /// the nul byte is located. + /// + /// If the first byte is a nul character, this method will return an + /// empty `CStr`. If multiple nul characters are present, the `CStr` will + /// end at the first one. + /// + /// If the slice only has a single nul byte at the end, this method is + /// equivalent to [`CStr::from_bytes_with_nul`]. + /// + /// # Examples + /// ``` + /// #![feature(cstr_from_bytes_until_nul)] + /// + /// use std::ffi::CStr; + /// + /// let mut buffer = [0u8; 16]; + /// unsafe { + /// // Here we might call an unsafe C function that writes a string + /// // into the buffer. + /// let buf_ptr = buffer.as_mut_ptr(); + /// buf_ptr.write_bytes(b'A', 8); + /// } + /// // Attempt to extract a C nul-terminated string from the buffer. + /// let c_str = CStr::from_bytes_until_nul(&buffer[..]).unwrap(); + /// assert_eq!(c_str.to_str().unwrap(), "AAAAAAAA"); + /// ``` + /// + #[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] + pub fn from_bytes_until_nul(bytes: &[u8]) -> Result<&CStr, FromBytesUntilNulError> { + let nul_pos = memchr::memchr(0, bytes); + match nul_pos { + Some(nul_pos) => { + let subslice = &bytes[..nul_pos + 1]; + // SAFETY: We know there is a nul byte at nul_pos, so this slice + // (ending at the nul byte) is a well-formed C string. + Ok(unsafe { CStr::from_bytes_with_nul_unchecked(subslice) }) + } + None => Err(FromBytesUntilNulError(())), + } + } + + /// Creates a C string wrapper from a byte slice. + /// + /// This function will cast the provided `bytes` to a `CStr` + /// wrapper after ensuring that the byte slice is nul-terminated + /// and does not contain any interior nul bytes. + /// + /// If the nul byte may not be at the end, + /// [`CStr::from_bytes_until_nul`] can be used instead. + /// + /// # Examples + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"hello\0"); + /// assert!(cstr.is_ok()); + /// ``` + /// + /// Creating a `CStr` without a trailing nul terminator is an error: + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"hello"); + /// assert!(cstr.is_err()); + /// ``` + /// + /// Creating a `CStr` with an interior nul byte is an error: + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"he\0llo\0"); + /// assert!(cstr.is_err()); + /// ``` + #[stable(feature = "cstr_from_bytes", since = "1.10.0")] + pub fn from_bytes_with_nul(bytes: &[u8]) -> Result<&Self, FromBytesWithNulError> { + let nul_pos = memchr::memchr(0, bytes); + match nul_pos { + Some(nul_pos) if nul_pos + 1 == bytes.len() => { + // SAFETY: We know there is only one nul byte, at the end + // of the byte slice. + Ok(unsafe { Self::from_bytes_with_nul_unchecked(bytes) }) + } + Some(nul_pos) => Err(FromBytesWithNulError::interior_nul(nul_pos)), + None => Err(FromBytesWithNulError::not_nul_terminated()), + } + } + + /// Unsafely creates a C string wrapper from a byte slice. + /// + /// This function will cast the provided `bytes` to a `CStr` wrapper without + /// performing any sanity checks. The provided slice **must** be nul-terminated + /// and not contain any interior nul bytes. + /// + /// # Examples + /// + /// ``` + /// use std::ffi::{CStr, CString}; + /// + /// unsafe { + /// let cstring = CString::new("hello").expect("CString::new failed"); + /// let cstr = CStr::from_bytes_with_nul_unchecked(cstring.to_bytes_with_nul()); + /// assert_eq!(cstr, &*cstring); + /// } + /// ``` + #[inline] + #[must_use] + #[stable(feature = "cstr_from_bytes", since = "1.10.0")] + #[rustc_const_stable(feature = "const_cstr_unchecked", since = "1.59.0")] + pub const unsafe fn from_bytes_with_nul_unchecked(bytes: &[u8]) -> &CStr { + // We're in a const fn, so this is the best we can do + debug_assert!(!bytes.is_empty() && bytes[bytes.len() - 1] == 0); + // SAFETY: Calling an inner function with the same prerequisites. + unsafe { Self::_from_bytes_with_nul_unchecked(bytes) } + } + + #[inline] + const unsafe fn _from_bytes_with_nul_unchecked(bytes: &[u8]) -> &CStr { + // SAFETY: Casting to CStr is safe because its internal representation + // is a [u8] too (safe only inside std). + // Dereferencing the obtained pointer is safe because it comes from a + // reference. Making a reference is then safe because its lifetime + // is bound by the lifetime of the given `bytes`. + unsafe { &*(bytes as *const [u8] as *const CStr) } + } + + /// Returns the inner pointer to this C string. + /// + /// The returned pointer will be valid for as long as `self` is, and points + /// to a contiguous region of memory terminated with a 0 byte to represent + /// the end of the string. + /// + /// **WARNING** + /// + /// The returned pointer is read-only; writing to it (including passing it + /// to C code that writes to it) causes undefined behavior. + /// + /// It is your responsibility to make sure that the underlying memory is not + /// freed too early. For example, the following code will cause undefined + /// behavior when `ptr` is used inside the `unsafe` block: + /// + /// ```no_run + /// # #![allow(unused_must_use)] #![allow(temporary_cstring_as_ptr)] + /// use std::ffi::CString; + /// + /// let ptr = CString::new("Hello").expect("CString::new failed").as_ptr(); + /// unsafe { + /// // `ptr` is dangling + /// *ptr; + /// } + /// ``` + /// + /// This happens because the pointer returned by `as_ptr` does not carry any + /// lifetime information and the `CString` is deallocated immediately after + /// the `CString::new("Hello").expect("CString::new failed").as_ptr()` + /// expression is evaluated. + /// To fix the problem, bind the `CString` to a local variable: + /// + /// ```no_run + /// # #![allow(unused_must_use)] + /// use std::ffi::CString; + /// + /// let hello = CString::new("Hello").expect("CString::new failed"); + /// let ptr = hello.as_ptr(); + /// unsafe { + /// // `ptr` is valid because `hello` is in scope + /// *ptr; + /// } + /// ``` + /// + /// This way, the lifetime of the `CString` in `hello` encompasses + /// the lifetime of `ptr` and the `unsafe` block. + #[inline] + #[must_use] + #[stable(feature = "rust1", since = "1.0.0")] + #[rustc_const_stable(feature = "const_str_as_ptr", since = "1.32.0")] + pub const fn as_ptr(&self) -> *const c_char { + self.inner.as_ptr() + } + + /// Converts this C string to a byte slice. + /// + /// The returned slice will **not** contain the trailing nul terminator that this C + /// string has. + /// + /// > **Note**: This method is currently implemented as a constant-time + /// > cast, but it is planned to alter its definition in the future to + /// > perform the length calculation whenever this method is called. + /// + /// # Examples + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); + /// assert_eq!(cstr.to_bytes(), b"foo"); + /// ``` + #[inline] + #[must_use = "this returns the result of the operation, \ + without modifying the original"] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn to_bytes(&self) -> &[u8] { + let bytes = self.to_bytes_with_nul(); + // SAFETY: to_bytes_with_nul returns slice with length at least 1 + unsafe { bytes.get_unchecked(..bytes.len() - 1) } + } + + /// Converts this C string to a byte slice containing the trailing 0 byte. + /// + /// This function is the equivalent of [`CStr::to_bytes`] except that it + /// will retain the trailing nul terminator instead of chopping it off. + /// + /// > **Note**: This method is currently implemented as a 0-cost cast, but + /// > it is planned to alter its definition in the future to perform the + /// > length calculation whenever this method is called. + /// + /// # Examples + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); + /// assert_eq!(cstr.to_bytes_with_nul(), b"foo\0"); + /// ``` + #[inline] + #[must_use = "this returns the result of the operation, \ + without modifying the original"] + #[stable(feature = "rust1", since = "1.0.0")] + pub fn to_bytes_with_nul(&self) -> &[u8] { + // SAFETY: Transmuting a slice of `c_char`s to a slice of `u8`s + // is safe on all supported targets. + unsafe { &*(&self.inner as *const [c_char] as *const [u8]) } + } + + /// Yields a &[str] slice if the `CStr` contains valid UTF-8. + /// + /// If the contents of the `CStr` are valid UTF-8 data, this + /// function will return the corresponding &[str] slice. Otherwise, + /// it will return an error with details of where UTF-8 validation failed. + /// + /// [str]: prim@str "str" + /// + /// # Examples + /// + /// ``` + /// use std::ffi::CStr; + /// + /// let cstr = CStr::from_bytes_with_nul(b"foo\0").expect("CStr::from_bytes_with_nul failed"); + /// assert_eq!(cstr.to_str(), Ok("foo")); + /// ``` + #[stable(feature = "cstr_to_str", since = "1.4.0")] + pub fn to_str(&self) -> Result<&str, str::Utf8Error> { + // N.B., when `CStr` is changed to perform the length check in `.to_bytes()` + // instead of in `from_ptr()`, it may be worth considering if this should + // be rewritten to do the UTF-8 check inline with the length calculation + // instead of doing it afterwards. + str::from_utf8(self.to_bytes()) + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl PartialEq for CStr { + fn eq(&self, other: &CStr) -> bool { + self.to_bytes().eq(other.to_bytes()) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl Eq for CStr {} +#[stable(feature = "rust1", since = "1.0.0")] +impl PartialOrd for CStr { + fn partial_cmp(&self, other: &CStr) -> Option { + self.to_bytes().partial_cmp(&other.to_bytes()) + } +} +#[stable(feature = "rust1", since = "1.0.0")] +impl Ord for CStr { + fn cmp(&self, other: &CStr) -> Ordering { + self.to_bytes().cmp(&other.to_bytes()) + } +} + +#[stable(feature = "cstr_range_from", since = "1.47.0")] +impl ops::Index> for CStr { + type Output = CStr; + + fn index(&self, index: ops::RangeFrom) -> &CStr { + let bytes = self.to_bytes_with_nul(); + // we need to manually check the starting index to account for the null + // byte, since otherwise we could get an empty string that doesn't end + // in a null. + if index.start < bytes.len() { + // SAFETY: Non-empty tail of a valid `CStr` is still a valid `CStr`. + unsafe { CStr::from_bytes_with_nul_unchecked(&bytes[index.start..]) } + } else { + panic!( + "index out of bounds: the len is {} but the index is {}", + bytes.len(), + index.start + ); + } + } +} + +#[stable(feature = "cstring_asref", since = "1.7.0")] +impl AsRef for CStr { + #[inline] + fn as_ref(&self) -> &CStr { + self + } +} diff --git a/library/core/src/ffi/mod.rs b/library/core/src/ffi/mod.rs index e61ea9ce87f55..6c49521c22385 100644 --- a/library/core/src/ffi/mod.rs +++ b/library/core/src/ffi/mod.rs @@ -14,6 +14,11 @@ use crate::marker::PhantomData; use crate::num::*; use crate::ops::{Deref, DerefMut}; +#[unstable(feature = "core_c_str", issue = "94079")] +pub use self::c_str::{CStr, FromBytesUntilNulError, FromBytesWithNulError}; + +mod c_str; + macro_rules! type_alias_no_nz { { $Docfile:tt, $Alias:ident = $Real:ty; diff --git a/library/core/src/lib.rs b/library/core/src/lib.rs index 6546a5244fd03..78d01c268f53e 100644 --- a/library/core/src/lib.rs +++ b/library/core/src/lib.rs @@ -20,7 +20,7 @@ // FIXME: Fill me in with more detail when the interface settles //! This library is built on the assumption of a few existing symbols: //! -//! * `memcpy`, `memcmp`, `memset` - These are core memory routines which are +//! * `memcpy`, `memcmp`, `memset`, `strlen` - These are core memory routines which are //! often generated by LLVM. Additionally, this library can make explicit //! calls to these functions. Their signatures are the same as found in C. //! These functions are often provided by the system libc, but can also be diff --git a/library/std/Cargo.toml b/library/std/Cargo.toml index dd9b035e86df1..f2d6583206035 100644 --- a/library/std/Cargo.toml +++ b/library/std/Cargo.toml @@ -16,7 +16,7 @@ panic_unwind = { path = "../panic_unwind", optional = true } panic_abort = { path = "../panic_abort" } core = { path = "../core" } libc = { version = "0.2.116", default-features = false, features = ['rustc-dep-of-std'] } -compiler_builtins = { version = "0.1.69" } +compiler_builtins = { version = "0.1.71" } profiler_builtins = { path = "../profiler_builtins", optional = true } unwind = { path = "../unwind" } hashbrown = { version = "0.12", default-features = false, features = ['rustc-dep-of-std'] } diff --git a/library/std/src/error.rs b/library/std/src/error.rs index 4fb94908c80fd..3f85c2095cbd4 100644 --- a/library/std/src/error.rs +++ b/library/std/src/error.rs @@ -26,6 +26,7 @@ use crate::borrow::Cow; use crate::cell; use crate::char; use crate::fmt::{self, Debug, Display, Write}; +use crate::io; use crate::mem::transmute; use crate::num; use crate::str; @@ -612,6 +613,48 @@ impl Error for alloc::collections::TryReserveError {} #[unstable(feature = "duration_checked_float", issue = "83400")] impl Error for time::FromFloatSecsError {} +#[stable(feature = "rust1", since = "1.0.0")] +impl Error for alloc::ffi::NulError { + #[allow(deprecated)] + fn description(&self) -> &str { + "nul byte found in data" + } +} + +#[stable(feature = "rust1", since = "1.0.0")] +impl From for io::Error { + /// Converts a [`alloc::ffi::NulError`] into a [`io::Error`]. + fn from(_: alloc::ffi::NulError) -> io::Error { + io::const_io_error!(io::ErrorKind::InvalidInput, "data provided contains a nul byte") + } +} + +#[stable(feature = "frombyteswithnulerror_impls", since = "1.17.0")] +impl Error for core::ffi::FromBytesWithNulError { + #[allow(deprecated)] + fn description(&self) -> &str { + self.__description() + } +} + +#[unstable(feature = "cstr_from_bytes_until_nul", issue = "95027")] +impl Error for core::ffi::FromBytesUntilNulError {} + +#[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] +impl Error for alloc::ffi::FromVecWithNulError {} + +#[stable(feature = "cstring_into", since = "1.7.0")] +impl Error for alloc::ffi::IntoStringError { + #[allow(deprecated)] + fn description(&self) -> &str { + "C string contained non-utf8 bytes" + } + + fn source(&self) -> Option<&(dyn Error + 'static)> { + Some(self.__source()) + } +} + // Copied from `any.rs`. impl dyn Error + 'static { /// Returns `true` if the inner type is the same as `T`. diff --git a/library/std/src/ffi/mod.rs b/library/std/src/ffi/mod.rs index 13e3dacc30d63..0141a2bccdf16 100644 --- a/library/std/src/ffi/mod.rs +++ b/library/std/src/ffi/mod.rs @@ -146,12 +146,24 @@ #![stable(feature = "rust1", since = "1.0.0")] -#[stable(feature = "cstr_from_bytes", since = "1.10.0")] -pub use self::c_str::FromBytesWithNulError; +/// See [alloc::ffi::FromVecWithNulError]. #[stable(feature = "cstring_from_vec_with_nul", since = "1.58.0")] -pub use self::c_str::FromVecWithNulError; +pub type FromVecWithNulError = alloc::ffi::FromVecWithNulError; +/// See [alloc::ffi::CString]. +#[stable(feature = "rust1", since = "1.0.0")] +pub type CString = alloc::ffi::CString; +/// See [alloc::ffi::IntoStringError]. +#[stable(feature = "rust1", since = "1.0.0")] +pub type IntoStringError = alloc::ffi::IntoStringError; +/// See [alloc::ffi::NulError]. #[stable(feature = "rust1", since = "1.0.0")] -pub use self::c_str::{CStr, CString, IntoStringError, NulError}; +pub type NulError = alloc::ffi::NulError; +/// See [core::ffi::CStr]. +#[stable(feature = "rust1", since = "1.0.0")] +pub type CStr = core::ffi::CStr; +/// See [core::ffi::FromBytesWithNulError]. +#[stable(feature = "cstr_from_bytes", since = "1.10.0")] +pub type FromBytesWithNulError = core::ffi::FromBytesWithNulError; #[stable(feature = "rust1", since = "1.0.0")] pub use self::os_str::{OsStr, OsString}; @@ -176,5 +188,4 @@ pub use core::ffi::{c_ptrdiff_t, c_size_t, c_ssize_t}; )] pub use core::ffi::{VaList, VaListImpl}; -mod c_str; mod os_str; diff --git a/library/std/src/lib.rs b/library/std/src/lib.rs index 60e7c2af8e4a5..039e3d59a4d40 100644 --- a/library/std/src/lib.rs +++ b/library/std/src/lib.rs @@ -260,7 +260,10 @@ #![feature(atomic_mut_ptr)] #![feature(char_error_internals)] #![feature(char_internals)] +#![feature(core_c_str)] #![feature(core_intrinsics)] +#![feature(cstr_from_bytes_until_nul)] +#![feature(cstr_internals)] #![feature(duration_checked_float)] #![feature(duration_constants)] #![feature(exact_size_is_empty)] @@ -286,6 +289,7 @@ // // Library features (alloc): #![feature(alloc_layout_extra)] +#![feature(alloc_c_string)] #![feature(allocator_api)] #![feature(get_mut_unchecked)] #![feature(map_try_insert)] diff --git a/library/std/src/prelude/v1.rs b/library/std/src/prelude/v1.rs index 0226c4d7a2581..f7533696df4af 100644 --- a/library/std/src/prelude/v1.rs +++ b/library/std/src/prelude/v1.rs @@ -95,3 +95,7 @@ pub use crate::string::{String, ToString}; #[stable(feature = "rust1", since = "1.0.0")] #[doc(no_inline)] pub use crate::vec::Vec; + +#[cfg(bootstrap)] +#[unstable(feature = "cstr_internals", issue = "none")] +pub use alloc::ffi::CStrExt; diff --git a/library/std/src/sys/hermit/mod.rs b/library/std/src/sys/hermit/mod.rs index 08eca42380204..60b7a973cc2c1 100644 --- a/library/std/src/sys/hermit/mod.rs +++ b/library/std/src/sys/hermit/mod.rs @@ -71,16 +71,6 @@ pub fn unsupported_err() -> crate::io::Error { ) } -pub unsafe fn strlen(start: *const c_char) -> usize { - let mut str = start; - - while *str != 0 { - str = str.offset(1); - } - - (str as usize) - (start as usize) -} - #[no_mangle] pub extern "C" fn floor(x: f64) -> f64 { unsafe { intrinsics::floorf64(x) } diff --git a/library/std/src/sys/sgx/mod.rs b/library/std/src/sys/sgx/mod.rs index 1333edb9881a1..696400670e04d 100644 --- a/library/std/src/sys/sgx/mod.rs +++ b/library/std/src/sys/sgx/mod.rs @@ -5,7 +5,6 @@ #![deny(unsafe_op_in_unsafe_fn)] use crate::io::ErrorKind; -use crate::os::raw::c_char; use crate::sync::atomic::{AtomicBool, Ordering}; pub mod abi; @@ -130,15 +129,6 @@ pub fn decode_error_kind(code: i32) -> ErrorKind { } } -pub unsafe fn strlen(mut s: *const c_char) -> usize { - let mut n = 0; - while unsafe { *s } != 0 { - n += 1; - s = unsafe { s.offset(1) }; - } - return n; -} - pub fn abort_internal() -> ! { abi::usercalls::exit(true) } diff --git a/library/std/src/sys/solid/mod.rs b/library/std/src/sys/solid/mod.rs index 492b1a55475a2..5ffa381f2e50f 100644 --- a/library/std/src/sys/solid/mod.rs +++ b/library/std/src/sys/solid/mod.rs @@ -99,5 +99,3 @@ pub fn hashmap_random_keys() -> (u64, u64) { (x1, x2) } } - -pub use libc::strlen; diff --git a/library/std/src/sys/unix/mod.rs b/library/std/src/sys/unix/mod.rs index 3ad03a8868199..aedeb02e656d6 100644 --- a/library/std/src/sys/unix/mod.rs +++ b/library/std/src/sys/unix/mod.rs @@ -3,7 +3,6 @@ use crate::io::ErrorKind; pub use self::rand::hashmap_random_keys; -pub use libc::strlen; #[cfg(not(target_os = "espidf"))] #[macro_use] diff --git a/library/std/src/sys/unsupported/common.rs b/library/std/src/sys/unsupported/common.rs index 5274f53a7dbdb..4c9ade4a8c790 100644 --- a/library/std/src/sys/unsupported/common.rs +++ b/library/std/src/sys/unsupported/common.rs @@ -4,10 +4,6 @@ pub mod memchr { pub use core::slice::memchr::{memchr, memrchr}; } -// This is not necessarily correct. May want to consider making it part of the -// spec definition? -use crate::os::raw::c_char; - // SAFETY: must be called only once during runtime initialization. // NOTE: this is not guaranteed to run, for example when Rust code is called externally. pub unsafe fn init(_argc: isize, _argv: *const *const u8) {} @@ -38,15 +34,3 @@ pub fn abort_internal() -> ! { pub fn hashmap_random_keys() -> (u64, u64) { (1, 2) } - -pub unsafe fn strlen(mut s: *const c_char) -> usize { - // SAFETY: The caller must guarantee `s` points to a valid 0-terminated string. - unsafe { - let mut n = 0; - while *s != 0 { - n += 1; - s = s.offset(1); - } - n - } -} diff --git a/library/std/src/sys/windows/mod.rs b/library/std/src/sys/windows/mod.rs index 47917e57b19ed..31c7208bbf1ee 100644 --- a/library/std/src/sys/windows/mod.rs +++ b/library/std/src/sys/windows/mod.rs @@ -7,7 +7,6 @@ use crate::path::PathBuf; use crate::time::Duration; pub use self::rand::hashmap_random_keys; -pub use libc::strlen; #[macro_use] pub mod compat; diff --git a/src/tools/clippy/clippy_utils/src/paths.rs b/src/tools/clippy/clippy_utils/src/paths.rs index 79e6e92dc0aaf..e5fa6deefc5de 100644 --- a/src/tools/clippy/clippy_utils/src/paths.rs +++ b/src/tools/clippy/clippy_utils/src/paths.rs @@ -23,7 +23,7 @@ pub const BTREEMAP_ENTRY: [&str; 6] = ["alloc", "collections", "btree", "map", " pub const BTREEMAP_INSERT: [&str; 6] = ["alloc", "collections", "btree", "map", "BTreeMap", "insert"]; pub const CLONE_TRAIT_METHOD: [&str; 4] = ["core", "clone", "Clone", "clone"]; pub const COW: [&str; 3] = ["alloc", "borrow", "Cow"]; -pub const CSTRING_AS_C_STR: [&str; 5] = ["std", "ffi", "c_str", "CString", "as_c_str"]; +pub const CSTRING_AS_C_STR: [&str; 5] = ["alloc", "ffi", "c_str", "CString", "as_c_str"]; pub const DEFAULT_TRAIT_METHOD: [&str; 4] = ["core", "default", "Default", "default"]; pub const DEREF_MUT_TRAIT_METHOD: [&str; 5] = ["core", "ops", "deref", "DerefMut", "deref_mut"]; /// Preferably use the diagnostic item `sym::deref_method` where possible