Matthieu Moy [off-list ref] writes:
Junio C Hamano [off-list ref] writes:
quoted
David Aguilar [off-list ref] writes:
quoted
+IFS='
+'
+git ls-files *.h ewah/*.h vcs-svn/*.h xdiff/*.h |
Hmm. This is only for true developers (not one who merely compiles
after expanding a tarball), so "git ls-files" may probably be OK.
But "/bin/ls" would be equally fine for that, no?
Actually, since this is "| while read header", I have to wonder why this
is not written as
for header in .h ewah/*.h vcs-svn/*.h xdiff/*.h
do
...
done
Yes, that would be even better. Then you wouldn't even have to
worry about $IFS dance.
On Mon, Sep 08, 2014 at 12:57:46PM -0700, Junio C Hamano wrote:
Matthieu Moy [off-list ref] writes:
quoted
Junio C Hamano [off-list ref] writes:
quoted
David Aguilar [off-list ref] writes:
quoted
+IFS='
+'
+git ls-files *.h ewah/*.h vcs-svn/*.h xdiff/*.h |
Hmm. This is only for true developers (not one who merely compiles
after expanding a tarball), so "git ls-files" may probably be OK.
But "/bin/ls" would be equally fine for that, no?
Actually, since this is "| while read header", I have to wonder why this
is not written as
for header in .h ewah/*.h vcs-svn/*.h xdiff/*.h
do
...
done
Yes, that would be even better. Then you wouldn't even have to
worry about $IFS dance.
The original motivation was to avoid picking up the generated
common-cmds.h header file.
It was the N_() function that was messing it up.
Would it make sense to split out a separate patch that makes common-cmds.h
check-headers clean?
--
David