Document that OsString and OsStr are bytes; provide conversions t…

…o bytes

This commit consists almost entirely of documentation, updating the
module documentation in `std::ffi` and the type documentation on
`OsString`. This only adds two methods: `OsStr::as_bytes` and
`OsString::into_vec`, and plumbs those methods down to the existing
implementations for both UNIX and Windows.

compiler/rustc_codegen_cranelift/src/debuginfo/line_info.rs

@@ -39,7 +39,7 @@ fn osstr_as_utf8_bytes(path: &OsStr) -> &[u8] {
     #[cfg(unix)]
     {
         use std::os::unix::ffi::OsStrExt;
-        path.as_bytes()
+        OsStrExt::as_bytes(path)
     }
     #[cfg(not(unix))]
     {

compiler/rustc_fs_util/src/lib.rs

@@ -82,7 +82,7 @@ pub fn path_to_c_string(p: &Path) -> CString {
     use std::ffi::OsStr;
     use std::os::unix::ffi::OsStrExt;
     let p: &OsStr = p.as_ref();
-    CString::new(p.as_bytes()).unwrap()
+    CString::new(OsStrExt::as_bytes(p)).unwrap()
 }
 #[cfg(windows)]
 pub fn path_to_c_string(p: &Path) -> CString {

library/std/src/ffi/mod.rs

@@ -90,7 +90,8 @@
 //! exists you will get a <code>[Some]\(os_string)</code>, which you can
 //! *then* try to convert to a Rust string. This yields a [`Result`], so that
 //! your code can detect errors in case the environment variable did
-//! not in fact contain valid Unicode data.
+//! not in fact contain valid Unicode data. You can also process the `OsString` directly, such as
+//! by using it as a filename.
 //!
 //! * [`OsStr`] losslessly represents a borrowed reference to a platform string.
 //! However, this representation is not necessarily in a form native to the platform.
@@ -99,17 +100,28 @@
 //!
 //! # Conversions
 //!
+//! ## On all platforms
+//!
+//! On all platforms, `OsStr` and `OsString` consist of a sequence of bytes; see [`OsString`] for
+//! more details on its encoding on different platforms.
+//!
+//! `OsStr` provides the method [`OsStr::as_bytes`], which provides a zero-cost conversion to a
+//! byte slice. (`OsString` provides this method as well, along with all other `OsStr` methods, via
+//! `Deref`.)
+//!
+//! `OsString` provides the method [`OsString::into_vec`], which provides a zero-cost conversion to
+//! `Vec<u8>`.
+//!
 //! ## On Unix
 //!
 //! On Unix, [`OsStr`] implements the
 //! <code>std::os::unix::ffi::[OsStrExt][unix.OsStrExt]</code> trait, which
-//! augments it with two methods, [`from_bytes`] and [`as_bytes`].
-//! These do inexpensive conversions from and to UTF-8 byte slices.
+//! augments it with an additional method [`from_bytes`], providing a zero-cost conversion from a
+//! byte slice.
 //!
 //! Additionally, on Unix [`OsString`] implements the
-//! <code>std::os::unix::ffi::[OsStringExt][unix.OsStringExt]</code> trait,
-//! which provides [`from_vec`] and [`into_vec`] methods that consume
-//! their arguments, and take or produce vectors of [`u8`].
+//! <code>std::os::unix::ffi::[OsStringExt][unix.OsStringExt]</code> trait, which provides the
+//! [`from_vec`] method that consumes a `Vec<u8>` and produces an `OsString`.
 //!
 //! ## On Windows
 //!
@@ -119,8 +131,8 @@
 //! On Windows, [`OsStr`] implements the
 //! <code>std::os::windows::ffi::[OsStrExt][windows.OsStrExt]</code> trait,
 //! which provides an [`encode_wide`] method. This provides an
-//! iterator that can be [`collect`]ed into a vector of [`u16`]. After a nul
-//! characters is appended, this is the same as a native Windows string.
+//! iterator that can be [`collect`]ed into a vector of [`u16`]. After a 16-bit nul
+//! character is appended, this is the same as a native Windows string.
 //!
 //! Additionally, on Windows [`OsString`] implements the
 //! <code>std::os::windows:ffi::[OsStringExt][windows.OsStringExt]</code>
@@ -133,10 +145,8 @@
 //! [`env::var_os()`]: crate::env::var_os "env::var_os"
 //! [unix.OsStringExt]: crate::os::unix::ffi::OsStringExt "os::unix::ffi::OsStringExt"
 //! [`from_vec`]: crate::os::unix::ffi::OsStringExt::from_vec "os::unix::ffi::OsStringExt::from_vec"
-//! [`into_vec`]: crate::os::unix::ffi::OsStringExt::into_vec "os::unix::ffi::OsStringExt::into_vec"
 //! [unix.OsStrExt]: crate::os::unix::ffi::OsStrExt "os::unix::ffi::OsStrExt"
 //! [`from_bytes`]: crate::os::unix::ffi::OsStrExt::from_bytes "os::unix::ffi::OsStrExt::from_bytes"
-//! [`as_bytes`]: crate::os::unix::ffi::OsStrExt::as_bytes "os::unix::ffi::OsStrExt::as_bytes"
 //! [`OsStrExt`]: crate::os::unix::ffi::OsStrExt "os::unix::ffi::OsStrExt"
 //! [windows.OsStrExt]: crate::os::windows::ffi::OsStrExt "os::windows::ffi::OsStrExt"
 //! [`encode_wide`]: crate::os::windows::ffi::OsStrExt::encode_wide "os::windows::ffi::OsStrExt::encode_wide"

library/std/src/ffi/os_str.rs

@@ -30,20 +30,27 @@ use crate::sys_common::{AsInner, FromInner, IntoInner};
 ///
 /// `OsString` and [`OsStr`] bridge this gap by simultaneously representing Rust
 /// and platform-native string values, and in particular allowing a Rust string
-/// to be converted into an "OS" string with no cost if possible. A consequence
-/// of this is that `OsString` instances are *not* `NUL` terminated; in order
-/// to pass to e.g., Unix system call, you should create a [`CStr`].
+/// to be converted into an "OS" string with no cost. A consequence of this is
+/// that `OsString` instances are *not* `NUL` terminated; in order to pass to
+/// e.g., a Unix system call, you should create a [`CStr`].
 ///
-/// `OsString` is to <code>&[OsStr]</code> as [`String`] is to <code>&[str]</code>: the former
-/// in each pair are owned strings; the latter are borrowed
-/// references.
+/// `OsString` is to <code>&[OsStr]</code> as [`String`] is to <code>&[str]</code>: `OsString` is
+/// an owned string like `String, while `&OsStr` is a borrowed reference like `&str`.
 ///
-/// Note, `OsString` and [`OsStr`] internally do not necessarily hold strings in
-/// the form native to the platform; While on Unix, strings are stored as a
-/// sequence of 8-bit values, on Windows, where strings are 16-bit value based
-/// as just discussed, strings are also actually stored as a sequence of 8-bit
-/// values, encoded in a less-strict variant of UTF-8. This is useful to
-/// understand when handling capacity and length values.
+/// Note that `OsString` and [`OsStr`] internally do not necessarily hold strings in the form
+/// native to the platform. On all platforms, `OsString` and `OsStr` consist of a sequence of
+/// bytes, in a superset of UTF-8; any valid UTF-8 sequence is a valid `OsString` or `OsStr`.
+/// * On Unix, these bytes can contain any values, in an arbitrary encoding (not necessarily
+///   UTF-8, and not necessarily the same encoding for different OS strings).
+/// * On Windows, where the native OS uses a sequence of 16-bit values, `OsString` and `OsStr`
+///   still consist of a sequence of 8-bit values, encoded in a superset of UTF-8 called
+///   ["WTF-8"](https://simonsapin.github.io/wtf-8/) ("Wobbly Translation Format 8-bit"). The
+///   WTF-8 format allows encoding arbitrary 16-bit values, including unpaired UTF-16 surrogates
+///   that do not constitute valid Unicode, since Windows accepts sequences of arbitrary 16-bit
+///   values. (In practice, Windows filenames and similar are almost always valid UTF-16.)
+///
+/// Capacity and length values are always in terms of the sequence of bytes, not characters or
+/// 16-bit values.
 ///
 /// # Creating an `OsString`
 ///
@@ -65,8 +72,16 @@ use crate::sys_common::{AsInner, FromInner, IntoInner};
 ///
 /// # Conversions
 ///
+/// `OsStr` provides the method [`OsStr::as_bytes`], which provides a zero-cost conversion to a
+/// byte slice. (`OsString` provides this method as well, along with all other `OsStr` methods, via
+/// `Deref`.)
+///
+/// `OsString` provides the method [`OsString::into_vec`], which provides a zero-cost conversion to
+/// `Vec<u8>`.
+///
 /// See the [module's toplevel documentation about conversions][conversions] for a discussion on
-/// the traits which `OsString` implements for [conversions] from/to native representations.
+/// OS-specific traits which `OsString` and `OsStr` implement for [conversions] from/to native
+/// representations.
 ///
 /// [`CStr`]: crate::ffi::CStr
 /// [conversions]: super#conversions
@@ -163,6 +178,24 @@ impl OsString {
         self.inner.into_string().map_err(|buf| OsString { inner: buf })
     }
 
+    /// Converts the `OsString` into a `Vec<u8>`.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// #![feature(osstr_bytes)]
+    /// use std::ffi::OsString;
+    ///
+    /// let os_string = OsString::from("foo");
+    /// let v = os_string.into_vec();
+    /// assert_eq!(v, b"foo");
+    /// ```
+    #[unstable(feature = "osstr_bytes", issue = "none")]
+    #[inline]
+    pub fn into_vec(self) -> Vec<u8> {
+        self.inner.into_vec()
+    }
+
     /// Extends the string with the given <code>&[OsStr]</code> slice.
     ///
     /// # Examples
@@ -667,6 +700,23 @@ impl OsStr {
         self.inner.to_str()
     }
 
+    /// Converts the `OsStr` into a `&[u8]`.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// #![feature(osstr_bytes)]
+    /// use std::ffi::OsStr;
+    ///
+    /// let os_str = OsStr::new("foo");
+    /// assert_eq!(os_str.as_bytes(), b"foo");
+    /// ```
+    #[unstable(feature = "osstr_bytes", issue = "none")]
+    #[inline]
+    pub fn as_bytes(&self) -> &[u8] {
+        self.inner.as_u8_slice()
+    }
+
     /// Converts an `OsStr` to a <code>[Cow]<[str]></code>.
     ///
     /// Any non-Unicode sequences are replaced with

library/std/src/os/unix/ffi/mod.rs

@@ -11,10 +11,6 @@
 //! // OsStringExt::from_vec
 //! let os_string = OsString::from_vec(bytes);
 //! assert_eq!(os_string.to_str(), Some("foo"));
-//!
-//! // OsStringExt::into_vec
-//! let bytes = os_string.into_vec();
-//! assert_eq!(bytes, b"foo");
 //! ```
 //!
 //! ```
@@ -26,10 +22,6 @@
 //! // OsStrExt::from_bytes
 //! let os_str = OsStr::from_bytes(bytes);
 //! assert_eq!(os_str.to_str(), Some("foo"));
-//!
-//! // OsStrExt::as_bytes
-//! let bytes = os_str.as_bytes();
-//! assert_eq!(bytes, b"foo");
 //! ```
 //!
 //! [`std::ffi`]: crate::ffi

library/std/src/sys/unix/os_str.rs

@@ -182,6 +182,11 @@ impl Buf {
     pub fn into_rc(&self) -> Rc<Slice> {
         self.as_slice().into_rc()
     }
+
+    #[inline]
+    pub fn into_vec(self) -> Vec<u8> {
+        self.inner
+    }
 }
 
 impl Slice {
@@ -190,6 +195,11 @@ impl Slice {
         unsafe { mem::transmute(s) }
     }
 
+    #[inline]
+    pub fn as_u8_slice(&self) -> &[u8] {
+        unsafe { mem::transmute(self) }
+    }
+
     #[inline]
     pub fn from_str(s: &str) -> &Slice {
         Slice::from_u8_slice(s.as_bytes())

library/std/src/sys/windows/os_str.rs

@@ -146,6 +146,11 @@ impl Buf {
     pub fn into_rc(&self) -> Rc<Slice> {
         self.as_slice().into_rc()
     }
+
+    #[inline]
+    pub fn into_vec(self) -> Vec<u8> {
+        self.inner.into_vec()
+    }
 }
 
 impl Slice {
@@ -193,6 +198,11 @@ impl Slice {
         unsafe { Rc::from_raw(Rc::into_raw(rc) as *const Slice) }
     }
 
+    #[inline]
+    pub fn as_u8_slice(&self) -> &[u8] {
+        self.inner.as_inner()
+    }
+
     #[inline]
     pub fn make_ascii_lowercase(&mut self) {
         self.inner.make_ascii_lowercase()

library/std/src/sys_common/wtf8.rs

@@ -3,13 +3,6 @@
 //! This library uses Rust’s type system to maintain
 //! [well-formedness](https://simonsapin.github.io/wtf-8/#well-formed),
 //! like the `String` and `&str` types do for UTF-8.
-//!
-//! Since [WTF-8 must not be used
-//! for interchange](https://simonsapin.github.io/wtf-8/#intended-audience),
-//! this library deliberately does not provide access to the underlying bytes
-//! of WTF-8 strings,
-//! nor can it decode WTF-8 from arbitrary bytes.
-//! WTF-8 strings can be obtained from UTF-8, UTF-16, or code points.
 
 // this module is imported from @SimonSapin's repo and has tons of dead code on
 // unix (it's mostly used on windows), so don't worry about dead code here.
@@ -399,6 +392,12 @@ impl Wtf8Buf {
         let bytes: Box<[u8]> = unsafe { mem::transmute(boxed) };
         Wtf8Buf { bytes: bytes.into_vec() }
     }
+
+    /// Converts this `Wtf8Buf` into a `Vec<u8>`.
+    #[inline]
+    pub fn into_vec(self) -> Vec<u8> {
+        self.bytes
+    }
 }
 
 /// Creates a new WTF-8 string from an iterator of code points.

src/test/ui/env-funky-keys.rs

@@ -9,14 +9,14 @@
 // no-prefer-dynamic
 
 #![feature(rustc_private)]
+#![feature(osstr_bytes)]
 
 extern crate libc;
 
 use libc::c_char;
 use libc::execve;
 use std::env;
 use std::ffi::CString;
-use std::os::unix::prelude::*;
 use std::ptr;
 
 fn main() {