Jeff King <[email protected]> writes:
> I do strip it off, so it is OK for it to be different in both the
> pre-image and post-image. But what I can't tolerate is the
> intermingling with actual data:
>
> +\t\t\x1b[32m;foo
> +\t\x1b[32m;bar
I think that depends on the definition of "strip it off" ;-) What I
meant was that whitespace coloring can appear anywhere on the line,
e.g.
echo COPYING whitespace=tab-in-indent,-space-before-tab >.gitattributes
printf " \tHeh\n" >>COPYING
git diff
will give you
<new> + <reset> SP <wserror> HT <reset> <new> Heh <reset> LF
Obviously, stripping "+" painted in "new" is not sufficient.
Side note: hmm, shouldn't that SP painted in <new>, though?
> I think this "array of spans" is the only way to go. Otherwise whichever
> markup scheme processes the hunk first ruins the data for the next
> processor.
Yes, I agree with that 100%.
--
To unsubscribe from this list: send the line "unsubscribe git" in
