threads / discuss / 7814

gitattributes

Subject: gitattributes

## tl;dr

5 messages between Apr 24, 2007 and Apr 24, 2007.

replies: 4people: 3as markdown or json

Matthias Kestenholz· Apr 24, 2007, 11:36 UTC · lore

The .gitattributes system together with input/output filters has already proven exceptionally useful for me. I've got some co-workers which do not seem to care about CRLF line endings or whitespace damage.

core.autocrlf = input

and

[filter "whitespace"]
  clean = stripspace
  smudge = cat

and

*.php filter=whitespace *.tpl filter=whitespace

saves me minutes every day when I review changes made by others, because otherwise the diffs are getting illegible (Sometimes, the whole file is shown twice only because the line-ending changed)

It would be really great if these features were included in the next release.

Thanks, Matthias

Alex Riesen· Apr 24, 2007, 16:18 UTC · re: Matthias Kestenholz · lore

Re: gitattributes

On 4/24/07, Matthias Kestenholz <lists@spinlock.ch> wrote:

Show 12 quoted lines

> The .gitattributes system together with input/output filters has already
> proven exceptionally useful for me. I've got some co-workers which
> do not seem to care about CRLF line endings or whitespace damage.
>
> core.autocrlf = input
>
> and
>
> [filter "whitespace"]
>   clean = stripspace
>   smudge = cat
>

Why didn't --ignore-space-at-eol, -b and -w work for you?

Junio C Hamano· Apr 24, 2007, 21:40 UTC · re: Alex Riesen · lore

Re: gitattributes

"Alex Riesen" <raa.lkml@gmail.com> writes:

Show 15 quoted lines

> On 4/24/07, Matthias Kestenholz <lists@spinlock.ch> wrote:
>> The .gitattributes system together with input/output filters has already
>> proven exceptionally useful for me. I've got some co-workers which
>> do not seem to care about CRLF line endings or whitespace damage.
>>
>> core.autocrlf = input
>>
>> and
>>
>> [filter "whitespace"]
>>   clean = stripspace
>>   smudge = cat
>>
>
> Why didn't --ignore-space-at-eol, -b and -w work for you?

It should. What Matthias would want more is probably to have others to have the "clean" side of the filter in their working repository.

Matthias Kestenholz· Apr 24, 2007, 22:06 UTC · re: Junio C Hamano · lore

Re: gitattributes

2007/4/24, Junio C Hamano <junkio@cox.net>:

Show 21 quoted lines

> "Alex Riesen" <raa.lkml@gmail.com> writes:
>
> > On 4/24/07, Matthias Kestenholz <lists@spinlock.ch> wrote:
> >> The .gitattributes system together with input/output filters has already
> >> proven exceptionally useful for me. I've got some co-workers which
> >> do not seem to care about CRLF line endings or whitespace damage.
> >>
> >> core.autocrlf = input
> >>
> >> and
> >>
> >> [filter "whitespace"]
> >>   clean = stripspace
> >>   smudge = cat
> >>
> >
> > Why didn't --ignore-space-at-eol, -b and -w work for you?
>
> It should.  What Matthias would want more is probably to have
> others to have the "clean" side of the filter in their working
> repository.

Yes, that's the reason. Furthermore, I like the approach of the filter better than adding command line arguments (which I and others will always forget on the first try)

I've just copy-pasted the smudge filter from an example I saw on the list. I'll retest the filter setup tomorrow without "smudge = cat" and look for problems.

Junio C Hamano· Apr 24, 2007, 17:15 UTC · re: Matthias Kestenholz · lore

Re: gitattributes

"Matthias Kestenholz" <lists@spinlock.ch> writes:

Show 11 quoted lines

> The .gitattributes system together with input/output filters has already
> proven exceptionally useful for me. I've got some co-workers which
> do not seem to care about CRLF line endings or whitespace damage.
>
> core.autocrlf = input
>
> and
>
> [filter "whitespace"]
>  clean = stripspace
>  smudge = cat

You shouldn't need to say "smudge = cat"; if it does not work without that line you found a bug, so please report.