On Thu, Oct 02, 2008 at 09:27:36PM +0300, Imre Deak wrote:
> $ git apply patch
> fatal: git diff header lacks filename information (line 4)
> $ cat patch
> diff --git a/dev/null b/a
> new file mode 100644
> index 0000000000000000000000000000000000000000..1f2a4f5ef3df7f7456d91c961da36fc58904f2f1
> GIT binary patch
Hmm. The problem is that "git apply" doesn't accept that "a/dev/null"
and "b/a" are the same, so it rejects them as a name. I guess on a text
patch, we would just pull that information from the "---" and "+++"
lines, so we don't care that it's not on the diff commandline.
However, a _non_ --no-index patch doesn't produce the same output. It
will actually produce the line:
diff --git a/a b/a
even if it is a creation patch. So I'm not sure which piece of code is
at fault. Either:
1. git apply is right to reject, and "git diff --no-index" should be
putting the actual filename on the commandline of a binary patch
instead of /dev/null, even if it is a creation patch.
or
2. git apply should accept this construct. Perhaps we should relax the
"both names must be the same" rule if one of the names is /dev/null
(and we would take the other)?
Linus, the "both names must be the same" code in git_header_name blames
to you (5041aa70). Thoughts on number 2?
-Peff
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
