Re: cut -DF

[Pádraig Brady]

Thanks for taking the time to consolidate options/functionality
across different implementations.  This is important for users.
Some notes below...

On 05/01/2022 16:23, Rob Landley wrote:
> Around 5 years ago toybox added the -D, -F, and -O options to cut:
>
>      -D  Don't sort/collate selections or match -fF lines without delimiter
>      -F  Select fields separated by DELIM regex
>      -O  Output delimiter (default one space for -F, input delim for -f)
>
> This lets you do:
>
>    $ echo one two three four five six seven eight nine | cut -DF 7,1-3,2
>    seven one two three two
>
> -F is a regex version of -f (defaulting to "match a run of whitespace")
>
> -D says to show the raw matches in the order requested (and ONLY those matches,
> it doesn't pass through lines with no matches)
>
> -O is -d for output.

Cool. I agree that the functionality is useful,
especially in places where awk may not be available.

As I see it, the main functionalities added here:
  - reordering of selected fields
  - adjusted suppression of lines without matching fields
  - regex delimiter support

I see regex support as less important, but still useful.

> You need all three because -F is useful by itself, and -F needs -O because when
> you're matching a regex it's not clear what to output. (Does
> "echo -e one\ttwo three' cut | -DF 3,1" glue them together with what's before
> match 1 (nothing), what's after match 3 (nothing), or an arbitrarily chosen one
> of the two different splits in between?)
>
> Elliott Hughes (the Android base OS maintainer) asked if I could get the feature
> more widely adopted:
>
>    http://lists.landley.net/pipermail/toybox-landley.net/2021-June/012453.html
>
> > your non-POSIX cut(1) extension covers 80% of the in-the-wild use of awk
> > anyway :-) if you still talk to any of the busybox folks, we should suggest
> > they copy that --- it would be nice for it to be a de facto standard so we
> > can get it into POSIX sometime around the 2040s... (and have made lives
> > better for the folks who don't care about standards and just want to "get
> > things done" in the intervening decades!)
>
> So I offered to implement it in busybox:
>
>    http://lists.busybox.net/pipermail/busybox/2021-June/088886.html
>
> And the busybox maintainer merged it here:
>
>    https://git.busybox.net/busybox/commit/?id=0068ce2fa0e3
>
> This is working and in use in Android, and now in busybox, and it would simplify
> my regression test suite if coreutils was in sync, so I thought I'd ask if you
> were interested.
>
> Thanks,
>
> Rob
>
> P.S. Somebody submitted a proposal to do this to posix way back when (see end of
> rationale at https://pubs.opengroup.org/onlinepubs/9699919799/utilities/cut.html
> -- they replied that they only standardize existing features, not take
> suggestions for new ones nobody's implemented yet). If I'd noticed I'd have used
> -o instead of -D at the time, but whoever suggested it apparently didn't try to
> make it actually work because -F is useful without -D, and -F without -O isn't
> well-defined.
>
> P.P.S. -D implying -F doesn't help because -F is the one that takes arguments,
> analogous to -f.

As for the interface, it's a bit surprising that -F wasn't used to
switch the field handling mode, rather than -D. I.e. I see the mode
change more pertaining to field handling, rather than delimiter handling.
I don't have a strong opinion on this, but it may be a bit confusing to users.

BTW it's useful to note existing edge cases in delimiter handling
when considering this new interface:
https://www.pixelbeat.org/docs/coreutils-gotchas.html#cut

So to summarize the new interface, and how it might map to / be described in 
coreutils,
I see it as:

  -D,--matched-fields
    Use the field order specified, and suppress lines without matching fields
  -F, --regex-fields=LIST
    Like -f, but interpret -d as a regular expression (defaulting
    to a run of whitespace)
  -O,--output-delimiter=STRING
    use STRING as the output delimiter
    default is one space with -F, or input delimiter for -f

cheers,
Pádraig