Converted all platform-specific stdin/stdout/stderr implementations to use io:: traits

[Lucretiel]

Currently, some of the platform-specific standard streams (src/libstd/sys/*/stdio.rs) manually implement parts of the io::Write interface directly as methods on the struct, rather than by actually implementing the trait. There doesn't seem to be any reason for this, other than an unused advantage of fn write(&self, ...) instead of fn write(&mut self, ...).

Unfortunately, this means that those implementations don't have the default-implemented io methods, like read_exact and write_all. This caused #72705, which adds forwarding methods to the user-facing standard stream implementations, to fail to compile on those platforms.

This change converts all such standard stream structs to use the standard library traits. This change should not cause any breakages, because the changed types are not publicly exported, and in fact are only ever used in src/libstd/io/stdio.rs.

[rust-highfive]

r? @sfackler

(rust_highfive has picked a reviewer for you, use r? to override)

[hellow554]

There are some modifications from &self to &mut self. Hopefully this won't break anything. You sure, that they aren't exported? If yes, then this LGTM.

[Lucretiel]

You sure, that they aren't exported?

I'm at 99.9% sure. Here are the checks I did:

• I know for sure that the std::sys isn't exported, and aside from the module declaration, that file makes no further reference to it. So they're definitely not publicly exported. The only issue is potential uses inside of libstd
• I know for sure that the platform specific code doesn't make any further interior references to these types (that is, for example, no other file matching /src/libstd/sys/hermit/**.rs refers to the stream types in /src/libstd/sys/hermit/stdio.rs
• I know for sure that the only meaningful references to the Stdin, Stdout, Stderr symbols are in either libstd/sys/*/stdio.rs or libstd/io/stdio.rs, and I've been combing through those files thoroughly recently as part of this PR, Added io forwarding methods to the stdio structs #72705, and Substantial refactor to the design of LineWriter #72808. This means that there shouldn't be any internal breakages.
• I know that many of the sys::stdio types do use an io trait implementation, and all of the cases I know of these types being used are seamless conditional import (that is, either you use sys::unix::stdio or sys::hermit::stdio, with the idea being that the types in those modules should export identical interfaces. This means that the only way there could be an internal breakage is if someone uses specifically one of these types (like, they specifically use hermit::stdio instead of *::stdio) and specifically depend on being able to call write on &self instead of &mut self.

[sfackler]

@bors r+

[bors]

📌 Commit 0094f44 has been approved by sfackler