"GOT", but the "O" is a cute, smiling pufferfish. Index | Thread | Search

From:
Omar Polo <op@omarpolo.com>
Subject:
Re: got patch: add flag to ignore whitespace?
To:
Stefan Sperling <stsp@stsp.name>
Cc:
gameoftrees@openbsd.org
Date:
Sat, 02 Jul 2022 23:27:42 +0200

Download raw body.

Thread
  • Christian Weisgerber:

    got patch: add flag to ignore whitespace?

  • Stefan Sperling <stsp@stsp.name> wrote:
    > On Sat, Jul 02, 2022 at 10:22:30PM +0200, Omar Polo wrote:
    > > Stefan Sperling <stsp@stsp.name> wrote:
    > > >  	@@ -295,7 +305,7 @@ hunk contains mangled whitespace
    > > 
    > > I like the idea!  Diff belows keeps the behavior, drops -w and adds this
    > > visual aid.  I've added it with a lower precedence, so if a hunk is
    > > applied at a different offset _and_ with mangled whitespaces only the
    > > "applied at offset" info will be printed.
    > > 
    > > We avoid adding a flag, love it :)
    > 
    > Sure. Less options is better.
    >  
    > > Rather than a flag specific for whitespaces, a "strict mode" that
    > > considers an error a non-perfect application of the diff seems more
    > > useful.
    > 
    > Makes sense. ok stsp, with two nits below:
    > 
    > > diff refs/heads/main refs/heads/pw
    > > commit - f5b0315f0e07bfd36a4eb37d91884fcd8614745a
    > > commit + 393c10527c09d54aa19c1f121b2cc74c31ac4746
    > > blob - d2db3f11b55b0dcd7008e9f4662210887aa2742b
    > > blob + a79c8a428a21ddd30f49aa132dd4457a73135fa1
    > > --- got/got.1
    > > +++ got/got.1
    > > @@ -1347,6 +1347,9 @@ If a change does not match at its exact line number, a
    > >  apply it somewhere else in the file if a good spot can be found.
    > >  Otherwise, the patch will fail to apply.
    > >  .Pp
    > > +Whitespaces may be ignored when trying to match the context of a
    > > +change, as they may have been mangled.
    > > +.Pp
    > 
    > Are you sure this is worth documenting? Diffs with some whitespace
    > issues should now "just work", and we already print a hint about it
    > during regular operation. People will expect that 'got patch' will do
    > its best to make sense of the provided input. They won't look into the
    > man page to check details about whitespace before trying to apply a patch.
    
    agreed, i felt into the trap of over-documenting.
    
    > Once we add a 'strict' mode, that would be a good place to document
    > this. We will need to explain how behaviour changes in strict mode, and
    > should then of course mention differences in treatment of whitespace.
    > 
    > > +		else
    > > +			printf("hunk contains mangled whitespaces\n");
    > 
    > The above should say just "whitespace" (singular), not "whitespaces".
    
    Sorry, you wrote it in the sigular form in the previous email but that
    's' slipped.  Committed in the singular form.
    
    > > +	echo 'M  hello.c' > $testroot/stdout.expected
    > > +	echo '@@ -5,5 +5,5 @@ hunk contains mangled whitespaces' \
    > 
    > And of course the test would need the same change here.
    
    Thanks!
    
    
  • Christian Weisgerber:

    got patch: add flag to ignore whitespace?