Feature request: Provide more `const` support for string types

[seritools]

Motivation

w! allows you to create a &'static HSTRING, but .as_ptr() and other conversions don't seem to be const, so it doesn't seem possible to get a const PCWSTR from it, in const context. Currently I'm using wchar::wchz! as alternative: PCWSTR::from_raw(wchz!("my string").as_ptr()). It would be nice to have a built-in way to do this.

Another tiny one is to create a const null-PCWSTR, but PCWSTR::null is not marked as const. PCWSTR(std::ptr::null()) is the slightly uglier workaround :)
The same should be the case for the other string types (PWSTR, and the 8bit string types). null(), as_ptr() and is_null() all can be marked as const fn as far as I can tell.

Drawbacks

No response

Rationale and alternatives

No response

Additional context

No response

[kennykerr]

I'd love to get there. I think we're still waiting for things to stabilize. For example, if I try to make HSTRING::is_empty a const function I get the following error:

D:\git\windows-rs>cargo check -p windows
    Checking windows v0.40.0 (D:\git\windows-rs\crates\libs\windows)
    Finished dev [unoptimized + debuginfo] target(s) in 0.55s

D:\git\windows-rs>cargo check -p windows
    Checking windows v0.40.0 (D:\git\windows-rs\crates\libs\windows)
error: `ptr::mut_ptr::<impl *mut T>::is_null` is not yet stable as a const fn
  --> crates\libs\windows\src\core\strings\hstring.rs:19:9
   |
19 |         self.0.is_null()
   |         ^^^^^^^^^^^^^^^^
   |
   = help: add `#![feature(const_ptr_is_null)]` to the crate attributes to enable

error: could not compile `windows` due to previous error

I'm not sure when the const_ptr_is_null feature will be stabilized. It doesn't seem to have made much progress of late:

rust-lang/rust#74939

[ChrisDenton]

Hm, if you don't mind a bit of type verbosity you maybe could use Option<NonNull<Header>> as the inner type. Then it's just None for the null case.

[seritools]

Testing a bit on HSTRING (https://github.com/seritools/windows-rs/tree/hstring-tests) it seems that it's possible.

There's some ugliness around NonNull::as_ptr() returning a *mut, which rustc currently doesn't allow to deref in const context, triggering:

error[E0658]: dereferencing raw mutable pointers in constant functions is unstable
  --> crates\libs\windows\src\core\strings\hstring.rs:40:22
   |
40 |             unsafe { (*header).data }
   |                      ^^^^^^^^^^^^^^
   |
   = note: see issue #57349 <https://github.com/rust-lang/rust/issues/57349> for more information
   = help: add `#![feature(const_mut_refs)]` to the crate attributes to enable

even if the pointer is only read, but manually casting it to *const Header seems to resolve it.

With Rust 1.64 released today even as_wide() can be marked as const, as core::slice::from_raw_parts is now a const fn :)

[kennykerr]

Thanks, that seems to work:

https://github.com/microsoft/windows-rs/compare/const-hstring?expand=1

It just breaks the debug visualizer test for some reason. I'll try to figure that out and then get a PR up.

cc @ridwanabdillahi

[ridwanabdillahi]

@kennykerr I'll take a look at the failure.