Hi all,
I've noticed a problem with using the suggested approach for excluding
files in .formatter.exs:
E.g.
[
inputs:
Enum.flat_map(
["{mix,.formatter}.exs", "{config,lib,test}/**/*.{ex,exs}"],
&Path.wildcard(&1, match_dot: true)
) -- Path.wildcard("test/example_app/**/*.*", match_dot: true)
]
Evaluating wildcards in .formatter.exs like this interferes with the way
the formatter caches inputs in its manifest file. The first time the
formatter runs, it evaluates .formatter.exs and stores the content it in a
manifest file, which is used from then-on for each formatter invocation,
until the .formatter.exs file is modified, or the manifest file is deleted.
- When a simple list of wildcards is provided like this inputs:
["{mix,.formatter}.exs", "{config,lib,test}/**/*.{ex,exs}"], the
formatter stores this list of wildcards in its manifest file which are
evaluated on each invocation of the formatter. This is good, because it
takes into account the current state of the filesystem on every formatter
run.
- When evaluating the wildcards using Path.wildcard() in the
.formatter.exs file itself, files are evaulated and unrolled *before*
the formatter gets its hands on it, so the formatter stores this unrolled
list of files in its manifest file, and the wildcards are never again
evaluated. This is bad, as the list of files to be formatted is "locked in"
the first time the formatter is run.
This results in newly created files getting completely ignored by the
formatter until the .formatter.exs file is modified, or the manifest file
is deleted!
Perhaps there's an easy way around this, but I was bitten by this issue
recently, and the exclude feature in this PR would neatly solve this
problem IMO. What do you guys think?
Cheers,
Steve
On Wednesday, 27 April 2022 at 15:24:35 UTC+1 an.le...@gmail.com wrote:
> Fair enough. If you wanted to use a simpler string/regex match for the
> original use case, you might want to write something like this:
>
> [
> inputs:
> ["{mix,.formatter}.exs", "{config,lib,test}/**/*.{ex,exs}"]
>
> |> Enum.flat_map(&Path.wildcard(&1, match_dot: true)
> |> Enum.reject(&(&1 =~ "test/example_app"))
> ]
>
>
> On Wed, Apr 27, 2022, at 3:13 PM, Ignacio Aguirrezabal wrote:
>
> Makes sense José <3
> Thanks everyone for your input
>
> On Wednesday, April 27, 2022 at 10:10:40 AM UTC-3 José Valim wrote:
>
> Thank you for the proposal, however I would prefer to not add this
> behaviour.
>
> It is definitely possible to do it today and at the same time we don't
> want to encourage excluding some files, unless absolutely necessary. And
> for the cases it is absolutely necessary, the current ways suffice. :)
>
>
> On Wed, Apr 27, 2022 at 3:03 PM Andrea Leopardi <an.le...@gmail.com>
> wrote:
>
>
> Hey Ignacio, thanks for the proposal.
>
> I don't feel too strongly about this, but I think it could be a useful
> addition.
>
> One note: I think :exclude should rather be exclude_pattern and be a
> regex. If an expanded filename matches the regex, it gets excluded. I think
> having a wildcard limits us unnecessarily since we have to expand the
> inputs wildcard anyways first.
>
> Unless other members of the team are against this feature, a PR would be
> welcome 😉
>
> Andrea
>
> On Wed, Apr 27, 2022, at 2:51 PM, Bruce Tate wrote:
>
> +1 from me. Providing convenient ways to opt out will only increase
> formatter adoption.
>
> -bt
>
> On Tue, Apr 26, 2022 at 4:35 PM Jon Rowe <ma...@jonrowe.co.uk> wrote:
>
>
> I wanted this feature recently as with the addition of the heex formatter
> I wanted to format most of an app, but not one file temporarily due to a
> whitespace wrapping issue (thats getting addressed :)) but still wanted the
> rest of the app formatted.
>
> So I'm in favour.
>
> Cheers
> Jon
>
> On Tue, 26 Apr 2022, at 6:33 PM, Ignacio Aguirrezabal wrote:
>
> Hello everyone,
>
> Currently, the only way to exclude files from the formatter is to do
> something like this:
>
> [
> inputs:
> Enum.flat_map(
> ["{mix,.formatter}.exs", "{config,lib,test}/**/*.{ex,exs}"],
> &Path.wildcard(&1, match_dot: true)
> ) -- Path.wildcard("test/example_app/**/*.*", match_dot: true) ]
>
>
>
> So I think it would be nice to have a more convenient way to do it. Maybe
> something like:
> [
> inputs: ["{mix,.formatter}.exs", "{config,lib,test}/**/*.{ex,exs}"],
> exclude: ["test/example_app/**/*.*"]
> ]
>
> I have a working implementation in this branch but I didn't want to open a
> PR before checking this is actually useful:
> https://github.com/elixir-lang/elixir/compare/main...iaguirre88:formatter-exclude-dir?expand=1
>
> Thanks,
> Ignacio
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elixir-lang-co...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elixir-lang-core/27d4a121-1f5f-4483-94e2-b43643dbe315n%40googlegroups.com
>
> <https://groups.google.com/d/msgid/elixir-lang-core/27d4a121-1f5f-4483-94e2-b43643dbe315n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elixir-lang-co...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elixir-lang-core/c1efa923-947d-4fea-abd8-2778c31cad10%40www.fastmail.com
>
> <https://groups.google.com/d/msgid/elixir-lang-core/c1efa923-947d-4fea-abd8-2778c31cad10%40www.fastmail.com?utm_medium=email&utm_source=footer>
> .
>
>
>
> --
>
> Regards,
> Bruce Tate
> CEO
>
>
> <https://bowtie.mailbutler.io/tracking/hit/f8218219-d2a8-4de4-9fef-1cdde6e723f6/c7c97460-016e-45fb-a4ab-0a70318c7b97>
>
>
> Groxio, LLC.
> 512.799.9366 <(512)%20799-9366>
> br...@grox.io
> grox.io
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elixir-lang-co...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elixir-lang-core/CAFXvW-5S4nHpN1ohrTYVG-F6hsPT3KDEa%3DyzJ-2B1911wupyXA%40mail.gmail.com
>
> <https://groups.google.com/d/msgid/elixir-lang-core/CAFXvW-5S4nHpN1ohrTYVG-F6hsPT3KDEa%3DyzJ-2B1911wupyXA%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elixir-lang-co...@googlegroups.com.
>
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elixir-lang-core/b7cade68-af9b-4ed8-86da-16e62e8450f0%40www.fastmail.com
>
> <https://groups.google.com/d/msgid/elixir-lang-core/b7cade68-af9b-4ed8-86da-16e62e8450f0%40www.fastmail.com?utm_medium=email&utm_source=footer>
> .
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "elixir-lang-core" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to elixir-lang-co...@googlegroups.com.
>
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/elixir-lang-core/cc1b078b-ea6b-46a3-8ff6-313594edfa05n%40googlegroups.com
>
> <https://groups.google.com/d/msgid/elixir-lang-core/cc1b078b-ea6b-46a3-8ff6-313594edfa05n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
>
>
--
You received this message because you are subscribed to the Google Groups
"elixir-lang-core" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to elixir-lang-core+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elixir-lang-core/7bfa3c11-033a-4daa-bef1-0204e1748a90n%40googlegroups.com.
