llvm dev - Jan 2021 - Clang-format: Allow alignment across empty lines and comments

Paul C. Anagnostopoulos via llvm-dev writes:
> How would the formatter know whether the declarations following the
> comment are a continuation of the preceding series of declarations?
If "across comments" is set, it would always assume that declarations
following a comment are a continuation of the previous ones. Only a line
containing anything but a declaration or a comment would break alignment.

One could further control this by distinguishing between block and line
comments. So for example:

int bar = 1;
// int baz = 2;
int longfoo = 3;

would be aligned to

int bar     = 1;
// int baz = 2;
int longfoo = 3;

But

int bar = 1;
/* This comment is meant to break alignment */
int longfoo = 3;

would stay as-is.

Would you prefer this behavior? I can see its appeal - it gets rid of
all the "superfluous change" when commenting-out an assignment in the
middle of a block, without lobbing together blocks of alignment you want
separated. However, this would make configuration more complicated. For
full flexibility, the enum would then need to have the members:

  - `<Something>_None` (the old `false`)
  - `<Something>_Consecutive` (the old `true`)
  - `<Something>_AcrossComments`
  - `<Something>_AcrossBlockComments`
  - `<Something>_AcrossLineComments`
  - `<Something>_AcrossEmptyLines`
  - `<Something>_AcrossEmptyLinesAndComments`
  - `<Something>_AcrossEmptyLinesAndBlockComments`
  - `<Something>_AcrossEmptyLinesAndLineComments`

Regards,
Lukas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 861 bytes
Desc: not available
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20210101/4782a1bf/attachment.sig>

Can I suggest you submit the patch, so we can discuss it in Phabricator,
you can put me down as a reviewer

https://llvm.org/docs/Contributing.html

On Fri, Jan 1, 2021 at 4:07 PM Lukas Barth via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
>
> Paul C. Anagnostopoulos via llvm-dev writes:
> > How would the formatter know whether the declarations following the
> > comment are a continuation of the preceding series of declarations?
>
> If "across comments" is set, it would always assume that
declarations
> following a comment are a continuation of the previous ones. Only a line
> containing anything but a declaration or a comment would break alignment.
>
> One could further control this by distinguishing between block and line
> comments. So for example:
>
> int bar = 1;
> // int baz = 2;
> int longfoo = 3;
>
> would be aligned to
>
> int bar     = 1;
> // int baz = 2;
> int longfoo = 3;
>
> But
>
> int bar = 1;
> /* This comment is meant to break alignment */
> int longfoo = 3;
>
> would stay as-is.
>
> Would you prefer this behavior? I can see its appeal - it gets rid of
> all the "superfluous change" when commenting-out an assignment in
the
> middle of a block, without lobbing together blocks of alignment you want
> separated. However, this would make configuration more complicated. For
> full flexibility, the enum would then need to have the members:
>
>   - `<Something>_None` (the old `false`)
>   - `<Something>_Consecutive` (the old `true`)
>   - `<Something>_AcrossComments`
>   - `<Something>_AcrossBlockComments`
>   - `<Something>_AcrossLineComments`
>   - `<Something>_AcrossEmptyLines`
>   - `<Something>_AcrossEmptyLinesAndComments`
>   - `<Something>_AcrossEmptyLinesAndBlockComments`
>   - `<Something>_AcrossEmptyLinesAndLineComments`
>
> Regards,
> Lukas
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20210101/b7df2522/attachment.html>