Move sys::pal::os::Env into sys::env

Although `Env` (as `Vars`), `Args`, path functions, and OS constants are
publicly exposed via `std::env`, their implementations are each
self-contained. Keep them separate in `std::sys` and make a new module,
`sys::env`, for `Env`.

Author: thaliaarchi

library/std/src/env.rs

@@ -13,7 +13,7 @@
 use crate::error::Error;
 use crate::ffi::{OsStr, OsString};
 use crate::path::{Path, PathBuf};
-use crate::sys::os as os_imp;
+use crate::sys::{env as env_imp, os as os_imp};
 use crate::{fmt, io, sys};
 
 /// Returns the current working directory as a [`PathBuf`].
@@ -96,7 +96,7 @@ pub struct Vars {
 /// [`env::vars_os()`]: vars_os
 #[stable(feature = "env", since = "1.0.0")]
 pub struct VarsOs {
-    inner: os_imp::Env,
+    inner: env_imp::Env,
 }
 
 /// Returns an iterator of (variable, value) pairs of strings, for all the
@@ -150,7 +150,7 @@ pub fn vars() -> Vars {
 #[must_use]
 #[stable(feature = "env", since = "1.0.0")]
 pub fn vars_os() -> VarsOs {
-    VarsOs { inner: os_imp::env() }
+    VarsOs { inner: env_imp::env() }
 }
 
 #[stable(feature = "env", since = "1.0.0")]
@@ -259,7 +259,7 @@ pub fn var_os<K: AsRef<OsStr>>(key: K) -> Option<OsString> {
 }
 
 fn _var_os(key: &OsStr) -> Option<OsString> {
-    os_imp::getenv(key)
+    env_imp::getenv(key)
 }
 
 /// The error type for operations interacting with environment variables.
@@ -363,7 +363,7 @@ impl Error for VarError {
 #[stable(feature = "env", since = "1.0.0")]
 pub unsafe fn set_var<K: AsRef<OsStr>, V: AsRef<OsStr>>(key: K, value: V) {
     let (key, value) = (key.as_ref(), value.as_ref());
-    unsafe { os_imp::setenv(key, value) }.unwrap_or_else(|e| {
+    unsafe { env_imp::setenv(key, value) }.unwrap_or_else(|e| {
         panic!("failed to set environment variable `{key:?}` to `{value:?}`: {e}")
     })
 }
@@ -434,7 +434,7 @@ pub unsafe fn set_var<K: AsRef<OsStr>, V: AsRef<OsStr>>(key: K, value: V) {
 #[stable(feature = "env", since = "1.0.0")]
 pub unsafe fn remove_var<K: AsRef<OsStr>>(key: K) {
     let key = key.as_ref();
-    unsafe { os_imp::unsetenv(key) }
+    unsafe { env_imp::unsetenv(key) }
         .unwrap_or_else(|e| panic!("failed to remove environment variable `{key:?}`: {e}"))
 }
 

library/std/src/sys/env/hermit.rs

@@ -0,0 +1,116 @@
+use core::slice::memchr;
+
+use crate::collections::HashMap;
+use crate::ffi::{CStr, OsStr, OsString, c_char};
+use crate::os::hermit::ffi::OsStringExt;
+use crate::sync::Mutex;
+use crate::{fmt, io, vec};
+
+static ENV: Mutex<Option<HashMap<OsString, OsString>>> = Mutex::new(None);
+
+pub fn init(env: *const *const c_char) {
+    let mut guard = ENV.lock().unwrap();
+    let map = guard.insert(HashMap::new());
+
+    if env.is_null() {
+        return;
+    }
+
+    unsafe {
+        let mut environ = env;
+        while !(*environ).is_null() {
+            if let Some((key, value)) = parse(CStr::from_ptr(*environ).to_bytes()) {
+                map.insert(key, value);
+            }
+            environ = environ.add(1);
+        }
+    }
+
+    fn parse(input: &[u8]) -> Option<(OsString, OsString)> {
+        // Strategy (copied from glibc): Variable name and value are separated
+        // by an ASCII equals sign '='. Since a variable name must not be
+        // empty, allow variable names starting with an equals sign. Skip all
+        // malformed lines.
+        if input.is_empty() {
+            return None;
+        }
+        let pos = memchr::memchr(b'=', &input[1..]).map(|p| p + 1);
+        pos.map(|p| {
+            (
+                OsStringExt::from_vec(input[..p].to_vec()),
+                OsStringExt::from_vec(input[p + 1..].to_vec()),
+            )
+        })
+    }
+}
+
+pub struct Env {
+    iter: vec::IntoIter<(OsString, OsString)>,
+}
+
+// FIXME(https://github.com/rust-lang/rust/issues/114583): Remove this when <OsStr as Debug>::fmt matches <str as Debug>::fmt.
+pub struct EnvStrDebug<'a> {
+    slice: &'a [(OsString, OsString)],
+}
+
+impl fmt::Debug for EnvStrDebug<'_> {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let Self { slice } = self;
+        f.debug_list()
+            .entries(slice.iter().map(|(a, b)| (a.to_str().unwrap(), b.to_str().unwrap())))
+            .finish()
+    }
+}
+
+impl Env {
+    pub fn str_debug(&self) -> impl fmt::Debug + '_ {
+        let Self { iter } = self;
+        EnvStrDebug { slice: iter.as_slice() }
+    }
+}
+
+impl fmt::Debug for Env {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let Self { iter } = self;
+        f.debug_list().entries(iter.as_slice()).finish()
+    }
+}
+
+impl !Send for Env {}
+impl !Sync for Env {}
+
+impl Iterator for Env {
+    type Item = (OsString, OsString);
+    fn next(&mut self) -> Option<(OsString, OsString)> {
+        self.iter.next()
+    }
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        self.iter.size_hint()
+    }
+}
+
+/// Returns a vector of (variable, value) byte-vector pairs for all the
+/// environment variables of the current process.
+pub fn env() -> Env {
+    let guard = ENV.lock().unwrap();
+    let env = guard.as_ref().unwrap();
+
+    let result = env.iter().map(|(key, value)| (key.clone(), value.clone())).collect::<Vec<_>>();
+
+    Env { iter: result.into_iter() }
+}
+
+pub fn getenv(k: &OsStr) -> Option<OsString> {
+    ENV.lock().unwrap().as_ref().unwrap().get(k).cloned()
+}
+
+pub unsafe fn setenv(k: &OsStr, v: &OsStr) -> io::Result<()> {
+    let (k, v) = (k.to_owned(), v.to_owned());
+    ENV.lock().unwrap().as_mut().unwrap().insert(k, v);
+    Ok(())
+}
+
+pub unsafe fn unsetenv(k: &OsStr) -> io::Result<()> {
+    ENV.lock().unwrap().as_mut().unwrap().remove(k);
+    Ok(())
+}

library/std/src/sys/env/mod.rs

@@ -0,0 +1,37 @@
+//! Platform-dependent environment variables abstraction.
+
+#![deny(unsafe_op_in_unsafe_fn)]
+
+cfg_if::cfg_if! {
+    if #[cfg(target_family = "unix")] {
+        mod unix;
+        pub use unix::*;
+    } else if #[cfg(target_family = "windows")] {
+        mod windows;
+        pub use windows::*;
+    } else if #[cfg(target_os = "hermit")] {
+        mod hermit;
+        pub use hermit::*;
+    } else if #[cfg(all(target_vendor = "fortanix", target_env = "sgx"))] {
+        mod sgx;
+        pub use sgx::*;
+    } else if #[cfg(target_os = "solid_asp3")] {
+        mod solid;
+        pub use solid::*;
+    } else if #[cfg(target_os = "uefi")] {
+        mod uefi;
+        pub use uefi::*;
+    } else if #[cfg(target_os = "wasi")] {
+        mod wasi;
+        pub use wasi::*;
+    } else if #[cfg(target_os = "xous")] {
+        mod xous;
+        pub use xous::*;
+    } else if #[cfg(target_os = "zkvm")] {
+        mod zkvm;
+        pub use zkvm::*;
+    } else {
+        mod unsupported;
+        pub use unsupported::*;
+    }
+}

library/std/src/sys/env/sgx.rs

@@ -0,0 +1,102 @@
+#![allow(fuzzy_provenance_casts)] // FIXME: this module systematically confuses pointers and integers
+
+use crate::collections::HashMap;
+use crate::ffi::{OsStr, OsString};
+use crate::sync::atomic::{AtomicUsize, Ordering};
+use crate::sync::{Mutex, Once};
+use crate::{fmt, io, vec};
+
+// Specifying linkage/symbol name is solely to ensure a single instance between this crate and its unit tests
+#[cfg_attr(test, linkage = "available_externally")]
+#[unsafe(export_name = "_ZN16__rust_internals3std3sys3pal3sgx2os3ENVE")]
+static ENV: AtomicUsize = AtomicUsize::new(0);
+// Specifying linkage/symbol name is solely to ensure a single instance between this crate and its unit tests
+#[cfg_attr(test, linkage = "available_externally")]
+#[unsafe(export_name = "_ZN16__rust_internals3std3sys3pal3sgx2os8ENV_INITE")]
+static ENV_INIT: Once = Once::new();
+type EnvStore = Mutex<HashMap<OsString, OsString>>;
+
+fn get_env_store() -> Option<&'static EnvStore> {
+    unsafe { (ENV.load(Ordering::Relaxed) as *const EnvStore).as_ref() }
+}
+
+fn create_env_store() -> &'static EnvStore {
+    ENV_INIT.call_once(|| {
+        ENV.store(Box::into_raw(Box::new(EnvStore::default())) as _, Ordering::Relaxed)
+    });
+    unsafe { &*(ENV.load(Ordering::Relaxed) as *const EnvStore) }
+}
+
+pub struct Env {
+    iter: vec::IntoIter<(OsString, OsString)>,
+}
+
+// FIXME(https://github.com/rust-lang/rust/issues/114583): Remove this when <OsStr as Debug>::fmt matches <str as Debug>::fmt.
+pub struct EnvStrDebug<'a> {
+    slice: &'a [(OsString, OsString)],
+}
+
+impl fmt::Debug for EnvStrDebug<'_> {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let Self { slice } = self;
+        f.debug_list()
+            .entries(slice.iter().map(|(a, b)| (a.to_str().unwrap(), b.to_str().unwrap())))
+            .finish()
+    }
+}
+
+impl Env {
+    pub fn str_debug(&self) -> impl fmt::Debug + '_ {
+        let Self { iter } = self;
+        EnvStrDebug { slice: iter.as_slice() }
+    }
+}
+
+impl fmt::Debug for Env {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let Self { iter } = self;
+        f.debug_list().entries(iter.as_slice()).finish()
+    }
+}
+
+impl !Send for Env {}
+impl !Sync for Env {}
+
+impl Iterator for Env {
+    type Item = (OsString, OsString);
+    fn next(&mut self) -> Option<(OsString, OsString)> {
+        self.iter.next()
+    }
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        self.iter.size_hint()
+    }
+}
+
+pub fn env() -> Env {
+    let clone_to_vec = |map: &HashMap<OsString, OsString>| -> Vec<_> {
+        map.iter().map(|(k, v)| (k.clone(), v.clone())).collect()
+    };
+
+    let iter = get_env_store()
+        .map(|env| clone_to_vec(&env.lock().unwrap()))
+        .unwrap_or_default()
+        .into_iter();
+    Env { iter }
+}
+
+pub fn getenv(k: &OsStr) -> Option<OsString> {
+    get_env_store().and_then(|s| s.lock().unwrap().get(k).cloned())
+}
+
+pub unsafe fn setenv(k: &OsStr, v: &OsStr) -> io::Result<()> {
+    let (k, v) = (k.to_owned(), v.to_owned());
+    create_env_store().lock().unwrap().insert(k, v);
+    Ok(())
+}
+
+pub unsafe fn unsetenv(k: &OsStr) -> io::Result<()> {
+    if let Some(env) = get_env_store() {
+        env.lock().unwrap().remove(k);
+    }
+    Ok(())
+}