r? @ehuss

rustbot has assigned @ehuss.
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

Some changes occurred in src/tools/cargo

cc @ehuss

How about disabling it only for stdout and stderr files in tests?

@mati865 there's a few run-make tests that use other file extensions, so collecting a comprehensive list might be a bit tricky.

i think more reliable maybe would be disabling it for all tests/ files except those written in rust and js?

@lolbinarycat that sounds sensible to me but I'd leave the decision to the reviewer.

r? compiler

I would personally also prefer to limit this restriction only to affected tests. I think it's fine for this to be an ad-hoc list of tests where it matters. I would personally love there to be a tidy check which makes sure we keep some //@ editor-no-trim-whitespace attribute in sync with an opt out in the editor config (and some way to bless the editor config file)

But 🤷 I think only listing the effected test in the editor config would also be fine, don't have any strong opinions here

I would personally also prefer to limit this restriction only to affected tests.

There are over 850 affected tests.

and these tests rely on their trailing whitespace? as in, how many tests actually intend to have trailing whitespace instead of being cases where we maybe should actually change them to not do so.

I don't want to block a change which prevents some editors from silently messing up your PR for the sake of a larger change, so I'd be fine with merging this for now and opening an issue that we should reenable this for at least ui tests and maybe even have a tidy check for it

and these tests rely on their trailing whitespace? as in, how many tests actually intend to have trailing whitespace instead of being cases where we maybe should actually change them to not do so.

There's also 4381 tests that end with multiple newlines, something that has an actual affect on UX, and something that is also (at least in emacs) also controlled by the trim_trailing_whitespace option.

I re-enabled trim_trailing_whitespace for source files in tests/, except for the two tests that intentionally have trailing whitespace. Hopefully this should be good enough to merge now.