new lint: use_crate_prefix_for_self_imports

[lengyijun]

changelog: [use_crate_prefix_for_self_imports]: new lint

fix #13645

Only check main.rs and lib.rs

Known problem

• Duplicate warning

TODO

• rm tests/ui/use_crate_prefix_for_self_imports.rs
• squash commits

Question

Although pass all tests, but if copy lintcheck to /tmp, clippy will report two warnings (expected)

warning: this import is not clear
  --> src/main.rs:42:5
   |
42 | use input::read_crates;
   |     ^^^^^ help: prefix with `crate::`: `crate::input`
   |
   = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#use_crate_prefix_for_self_imports
   = note: `#[warn(clippy::use_crate_prefix_for_self_imports)]` on by default

warning: this import is not clear
  --> src/main.rs:43:5
   |
43 | use output::{ClippyCheckOutput, ClippyWarning, RustcIce};
   |     ^^^^^^ help: prefix with `crate::`: `crate::output`
   |
   = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#use_crate_prefix_for_self_imports

[rustbot]

r? @llogiq

rustbot has assigned @llogiq.
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

[llogiq]

So, as you know we discussed this on zulip a while ago. The upshot is that we have two acceptable options:

• Put the lint in the nursery group and improve it until we deem it ready for style
• Add more checks to the lint to reduce the number of warnings that aren't too helpful. For example, if we have the respective mod near the use (where near could be measured in line number differences or in no other items but mods and uses in between), we could avoid linting, as the reader of the code will know that the item comes from the mod specified nearby (this could be done by walking the sibling items until a non-mod-or-use item is found).

Do you intend to continue working on this?

[llogiq]

@rustbot author

[lengyijun]

Don't lint if mod and use in the same block

[lengyijun]

The problem of lintcheck still exists

[llogiq]

What does that mean?

[lengyijun]

Although pass all tests, but if copy lintcheck to /tmp, clippy will report two warnings (expected)

warning: this import is not clear
  --> src/main.rs:42:5
   |
42 | use input::read_crates;
   |     ^^^^^ help: prefix with `crate::`: `crate::input`
   |
   = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#use_crate_prefix_for_self_imports
   = note: `#[warn(clippy::use_crate_prefix_for_self_imports)]` on by default

warning: this import is not clear
  --> src/main.rs:43:5
   |
43 | use output::{ClippyCheckOutput, ClippyWarning, RustcIce};
   |     ^^^^^^ help: prefix with `crate::`: `crate::output`
   |
   = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#use_crate_prefix_for_self_imports

[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.

[lengyijun]

@samueltardieu @llogiq Hi, ready for another review
I'll rebase before merge

[rustbot]

☔ The latest upstream changes (possibly e121ab8) made this pull request unmergeable. Please resolve the merge conflicts.