feat(format_push_string): give a (possibly incomplete) suggestion

[ada4a]

Resolves #8824

Supersedes #15828 -- I wasn't able to get TyCtxt::in_scope_traits to work, so here's a more basic implementation that just adds a note about the possible need to import std::fmt::Write.

changelog: [format_push_string]: add a suggestion

[rustbot]

r? @samueltardieu

rustbot has assigned @samueltardieu.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[github-actions]

Lintcheck changes for 9941c97

Lint	Added	Removed	Changed
clippy::format_push_string	0	0	29

This comment will be updated if you push new changes

[samueltardieu]

I wouldn't say that the method doesn't seem to work if you call it outside of its intended use. Given its lack of documentation, it is hard to say. If you want to investigate, you might be luckier asking in a compiler-related Zulip channel.

Suggested change

	// TODO: Ideally we'd use `TyCtxt::in_scope_traits` to detect whether the trait is imported, and
	// either not emit this note if it is, or emit an automated suggestion to import it if it isn't.
	// But the method doesn't seem to work, see https://rust-lang.zulipchat.com/#narrow/channel/257328-clippy/topic/How+to+suggest+importing+a+path/with/544260181
	// TODO: omit the note if the `Write` trait is imported at point

[ada4a]

After more research, I found rust-lang/rust#54161 (comment), which is almost correct (or maybe was correct initially, has become incorrect since) -- the actual behaviour TyCtxt::in_scope_traits seems to be the following: it returns a non-empty list only when called on the HirId of a ExprKind::MethodCall that is a call of a trait method.

Added a note about this

[samueltardieu]

Suggested change

	diag.note(format!("you may need to import `{std_or_core}::fmt::Write`"));
	diag.note(format!("you may need to import the `{std_or_core}::fmt::Write` trait"));

[samueltardieu]

This might happen though:

Example:

#![feature(no_core)]
#![no_core]
#![no_std]

extern crate std as mystd;

use mystd::convert::From as _;

fn main() {
    let mut s = mystd::string::String::from("no core, no std");
    s.push_str(&mystd::format!("{}", 1+2));
    mystd::println!("s = {s}");
}

Suggested change

	let Some(std_or_core) = std_or_core(cx) else {
	// This can't really happen, as a no-core crate wouldn't have access to `String` in the first place
	return;
	};
	let Some(std_or_core) = std_or_core(cx) else {
	return;
	};

[ada4a]

Good catch. Added a test

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.

[rustbot]

This PR was rebased onto a different master commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[ada4a]

@rustbot ready