new lint: path_ends_with_ext

[y21]

Closes #11479

Not sure if it needs more test cases. I couldn't come up with any other ones, but it is a pretty simple lint logic wise with not too many checks

changelog: new lint: [path_ends_with_ext]

[rustbot]

r? @Alexendoo

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

[Alexendoo]

This would have a fair few false positives with stuff like .git or any other dotfile type things

If we can source one shipping a big list of file extensions would probably be the ideal solution since it wouldn't take up all that much space, the problem being how easily it is to source such a list

[y21]

Maybe a configuration option for additional dotfiles to allow would also be useful.

[kornelski]

I suggest limiting it to 2-char and 3-char extensions. The longer the name, the more likely it is a dotfile.

I don't have any 2-char dotfiles. The 3-letter dotfiles I could find:

.git .svn .gem .npm .vim .env .rnd .ssh .vnc

[y21]

I suggest limiting it to 2-char and 3-char extensions. The longer the name, the more likely it is a dotfile.

I don't have any 2-char dotfiles. The 3-letter dotfiles I could find:

.git .svn .gem .npm .vim .env .rnd .ssh .vnc

That sounds great, I implemented this in the last commit (that and the ability to add custom dotfiles that clippy shouldn't warn).

[jdh8]

How is this different from #6425 🧐

[y21]

How is this different from #6425 🧐

The description and code looks like it has a little bit of overlap and sounds similar, but that one seems to look for a different kind of suspicious pattern and I would say this one is different enough to be its own lint.

This lint looks for Path::ends_with calls and checks if the argument looks like a file extension and warns the user that this isn't what Path::ends_with does, which has nothing to do with case sensitivity like the other lint IMO.
The other lint looks for str::ends_with, which at least does the right thing in that it actually checks the extension (just not case insensitive).

We may want to change the suggestion to use str::eq_ignore_ascii_case instead of == though.
(And maybe that other lint could be improved to also flag the suggestion that we're currently making here: path.extension() == "rs")

[Alexendoo]

Thanks!

@bors r+

[bors]

📌 Commit 981e960 has been approved by Alexendoo

It is now in the queue for this repository.

[bors]

⌛ Testing commit 981e960 with merge daadab5...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: Alexendoo
Pushing daadab5 to master...