chore: fix Rust 1.86 type inference and update to Rust 1.86

[beanow-at-crabnebula]

We're using Rust 1.86 and one compile error showed up for ssstar.

    Checking ssstar v0.7.3 (/media/workspace/src/cn/ssstar/ssstar)
error[E0283]: type annotations needed
   --> ssstar/src/extract.rs:294:60
    |
294 |             Self::Object { key } => path.to_string_lossy() == key.as_ref(),
    |                                                            ^^ cannot infer type for reference `&_`
    |
    = note: multiple `impl`s satisfying `Cow<'_, str>: PartialEq<&_>` found in the `alloc` crate:
            - impl<'a, 'b> PartialEq<&'b str> for Cow<'a, str>;
            - impl<'a> PartialEq<&'a ByteStr> for Cow<'a, str>;

error[E0283]: type annotations needed
   --> ssstar/src/extract.rs:294:67
    |
294 |             Self::Object { key } => path.to_string_lossy() == key.as_ref(),
    |                                                                   ^^^^^^
    |
    = note: multiple `impl`s satisfying `std::string::String: AsRef<_>` found in the following crates: `alloc`, `std`:
            - impl AsRef<Path> for std::string::String;
            - impl AsRef<[u8]> for std::string::String;
            - impl AsRef<std::ffi::OsStr> for std::string::String;
            - impl AsRef<str> for std::string::String;
help: try using a fully qualified path to specify the expected types
    |
294 -             Self::Object { key } => path.to_string_lossy() == key.as_ref(),
294 +             Self::Object { key } => path.to_string_lossy() == <std::string::String as AsRef<T>>::as_ref(key),
    |

For more information about this error, try `rustc --explain E0283`.
error: could not compile `ssstar` (lib) due to 2 previous errors

I think this is a regression in the compiler, so I submitted a report rust-lang/rust#139542
But eventually when ByteStr stabilizes this problem may come back. So using as_str() should help with this.

[beanow-at-crabnebula]

As a quick summary from the upstream bug report.

Implementing unstable feature flags for new trait impls is complicated, so it's a known issue that they appear in stable Rust. In this case it did as well, and causes type inference regressions.

No patch expected from the Rust side right now, so the more explicit .as_str() should resolve this.