Michael G. Schwern [Thu, 26 Jul 2012 23:22:25 +0000 (16:22 -0700)]
Move initialization of Git::SVN variables into Git::SVN.
Also it can compile on its own now, yay!
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Thu, 26 Jul 2012 23:22:24 +0000 (16:22 -0700)]
Extract Git::SVN from git-svn into its own .pm file.
Except for adding the 1; at the end, this is a straight copy & paste.
Tests still pass, but its doubtful Git::SVN will compile on its own
without git-svn being loaded. Next commit will fix that.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Thu, 26 Jul 2012 23:22:23 +0000 (16:22 -0700)]
Prepare Git::SVN for extraction into its own file.
This means it should be able to load without git-svn being loaded.
* Load Git.pm on its own and all the needed command functions.
* It needs to grab at a git-svn lexical $_prefix representing the --prefix
option. Provide opt_prefix() for that. This is a refactoring artifact.
The prefix should really be passed into Git::SVN->new.
* Unqualify unnecessarily fully qualified globals like
$Git::SVN::default_repo_id.
* Lexically isolate the class just to make sure nothing is leaking out.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Thu, 26 Jul 2012 23:22:22 +0000 (16:22 -0700)]
Extract some utilities from git-svn to allow extracting Git::SVN.
Put them in a new module called Git::SVN::Utils. Yeah, not terribly
original and it will be a dumping ground. But its better than having
them in the main git-svn program. At least they can be documented
and tested.
* fatal() is used by many classes.
* Change the $can_compress lexical into a function.
This should be enough to extract Git::SVN.
Signed-off-by: Michael G. Schwern <schwern@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Junio C Hamano [Fri, 27 Jul 2012 20:04:20 +0000 (20:04 +0000)]
perl: detect new files in MakeMaker builds
While Makefile.PL now finds .pm files on its own, it does not
detect new files after it generates perl/perl.mak.
[ew: commit message, minor tweaks]
ref: http://mid.gmane.org/7vlii51xz4.fsf@alter.siamese.dyndns.org
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Wed, 25 Jul 2012 03:21:11 +0000 (20:21 -0700)]
The Makefile.PL will now find .pm files itself.
It is no longer necessary to manually add new .pm files to the
Makefile.PL. This makes it easier to add modules.
It is still necessary to add them to the Makefile, but that extra work
should be removed at a future date.
Signed-off-by: Michael G Schwern <schwern@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Wed, 25 Jul 2012 03:21:10 +0000 (20:21 -0700)]
Don't lose Error.pm if $@ gets clobbered.
In older Perls, sometimes $@ can become unset between the eval and
checking $@. Its safer to check the eval directly.
Signed-off-by: Michael G Schwern <schwern@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Michael G. Schwern [Wed, 25 Jul 2012 03:21:09 +0000 (20:21 -0700)]
Quiet warning if Makefile.PL is run with -w and no --localedir
Usually it isn't, but its nice if it can be run with warnings on.
Signed-off-by: Michael G Schwern <schwern@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Junio C Hamano [Wed, 25 Jul 2012 22:47:08 +0000 (15:47 -0700)]
Merge branch 'jc/test-lib-source-build-options-early'
Reorders t/test-lib.sh so that we dot-source GIT-BUILD-OPTIONS that
records the shell and Perl the user told us to use with Git a lot
early, so that test-lib.sh script itself can use "$PERL_PATH" in
one of its early operations.
* jc/test-lib-source-build-options-early:
test-lib: reorder and include GIT-BUILD-OPTIONS a lot earlier
Junio C Hamano [Wed, 25 Jul 2012 22:47:04 +0000 (15:47 -0700)]
Merge branch 'mm/config-xdg'
Finishing touches to the XDG support (new feature for 1.7.12) and
tests.
* mm/config-xdg:
t1306: check that XDG_CONFIG_HOME works
ignore: make sure we have an xdg path before using it
attr: make sure we have an xdg path before using it
test-lib.sh: unset XDG_CONFIG_HOME
Junio C Hamano [Wed, 25 Jul 2012 22:46:59 +0000 (15:46 -0700)]
Merge branch 'cw/rebase-i-root'
Finishing touches to the "rebase -i --root" (new feature for
1.7.12).
* cw/rebase-i-root:
rebase -i: handle fixup of root commit correctly
Junio C Hamano [Wed, 25 Jul 2012 22:46:06 +0000 (15:46 -0700)]
Merge branch 'mh/maint-revisions-doc'
* mh/maint-revisions-doc:
Enumerate revision range specifiers in the documentation
Make <refname> documentation more consistent.
Junio C Hamano [Tue, 24 Jul 2012 22:03:50 +0000 (15:03 -0700)]
Enumerate revision range specifiers in the documentation
It was a bit hard to learn how <rev>^@, <rev>^! and various other
forms of range specifiers are used, because they were discussed
mostly in the prose part of the documentation, unlike various forms
of extended SHA-1 expressions that are listed in an enumerated list.
Also add a few more examples showing use of <rev>, <rev>..<rev> and
<rev>^! forms, stolen from a patch by Max Horn.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Tue, 24 Jul 2012 21:08:43 +0000 (14:08 -0700)]
Update draft release notes to 1.7.12
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Tue, 24 Jul 2012 21:05:26 +0000 (14:05 -0700)]
Merge branch 'jc/mergetool-tool-help'
"git mergetool" did not support --tool-help option to give the list
of supported backends, like "git difftool" does.
* jc/mergetool-tool-help:
mergetool: support --tool-help option like difftool does
Junio C Hamano [Tue, 24 Jul 2012 21:05:22 +0000 (14:05 -0700)]
Merge branch 'jk/maint-commit-document-editmsg'
Document $GIT_DIR/COMMIT_EDITMSG file.
* jk/maint-commit-document-editmsg:
commit: document the temporary commit message file
Junio C Hamano [Tue, 24 Jul 2012 21:05:18 +0000 (14:05 -0700)]
Merge branch 'jk/maint-commit-check-committer-early'
"git commit --amend" let the user edit the log message and then died
when the human-readable committer name was given insufficiently by
getpwent(3).
* jk/maint-commit-check-committer-early:
commit: check committer identity more strictly
Junio C Hamano [Tue, 24 Jul 2012 21:05:08 +0000 (14:05 -0700)]
Merge branch 'jk/maint-advise-vaddf'
The advise() function did not use varargs correctly to format
its message.
* jk/maint-advise-vaddf:
advice: pass varargs to strbuf_vaddf, not strbuf_addf
Jeff King [Tue, 24 Jul 2012 13:43:59 +0000 (09:43 -0400)]
t/lib-httpd: handle running under --valgrind
Running the http tests with valgrind does not work for two
reasons:
1. Apache complains about following the symbolic link from
git-http-backend to valgrind.sh.
2. Apache does not pass through the GIT_VALGRIND variable
to the backend CGI.
This patch fixes both problems. Unfortunately, there is a
slight hack we need to handle passing environment variables
through Apache. If we just tell it:
PassEnv GIT_VALGRIND
then Apache will complain when GIT_VALGRIND is not set. If
we try:
SetEnv GIT_VALGRIND ${GIT_VALGRIND}
then when GIT_VALGRIND is not set, it will pass through the
literal "${GIT_VALGRIND}". Instead, we now unconditionally
pass through GIT_VALGRIND from lib-httpd.sh into apache,
even if it is empty.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Chris Webb [Tue, 24 Jul 2012 12:17:03 +0000 (13:17 +0100)]
rebase -i: handle fixup of root commit correctly
There is a bug with git rebase -i --root when a fixup or squash line is
applied to the new root. We attempt to amend the commit onto which they
apply with git reset --soft HEAD^ followed by a normal commit. Unlike a
real commit --amend, this sequence will fail against a root commit as it
has no parent.
Fix rebase -i to use commit --amend for fixup and squash instead, and
add a test for the case of a fixup of the root commit.
Signed-off-by: Chris Webb <chris@arachsys.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Tue, 24 Jul 2012 12:27:10 +0000 (08:27 -0400)]
t1306: check that XDG_CONFIG_HOME works
This should override $HOME/.config, but we never actually tested it.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Tue, 24 Jul 2012 12:26:51 +0000 (14:26 +0200)]
ignore: make sure we have an xdg path before using it
Commit
e3ebc35 (config: fix several access(NULL) calls, 2012-07-12) was
fixing access(NULL) calls when trying to access $HOME/.config/git/config,
but missed the ones when trying to access $HOME/.config/git/ignore. Fix
and test this.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Tue, 24 Jul 2012 11:53:57 +0000 (07:53 -0400)]
attr: make sure we have an xdg path before using it
If we don't have a core.attributesfile configured, we fall
back to checking XDG config, which is usually
$HOME/.config/git/attributes.
However, if $HOME is unset, then home_config_paths will return
NULL, and we end up calling fopen(NULL).
Depending on your system, this may or may not cause the
accompanying test to fail (e.g., on Linux and glibc, the
address will go straight to open, which will return EFAULT).
However, valgrind will reliably notice the error.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Tue, 24 Jul 2012 11:53:05 +0000 (07:53 -0400)]
test-lib.sh: unset XDG_CONFIG_HOME
Now that git respects XDG_CONFIG_HOME for some lookups, we
must be sure to cleanse the test environment. Otherwise, the
user's XDG_CONFIG_HOME could influence the test results.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Tue, 24 Jul 2012 04:02:06 +0000 (21:02 -0700)]
Git 1.7.12-rc0
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Tue, 24 Jul 2012 03:56:46 +0000 (20:56 -0700)]
Merge branch 'jn/block-sha1'
The code to load a word one-byte-at-a-time was optimized into a
word-wide load instruction even when the pointer was not aligned,
which caused issues on architectures that do not like unaligned
access.
* jn/block-sha1:
Makefile: BLK_SHA1 does not require fast htonl() and unaligned loads
block-sha1: put expanded macro parameters in parentheses
block-sha1: avoid pointer conversion that violates alignment constraints
Junio C Hamano [Tue, 24 Jul 2012 03:56:39 +0000 (20:56 -0700)]
Merge branch 'jn/make-assembly-in-right-directory'
* jn/make-assembly-in-right-directory:
Makefile: fix location of listing produced by "make subdir/foo.s"
Junio C Hamano [Tue, 24 Jul 2012 03:56:17 +0000 (20:56 -0700)]
Merge branch 'th/difftool-diffall'
Finishing touches to difftool --dirdiff.
* th/difftool-diffall:
difftool: only copy back files modified during directory diff
Junio C Hamano [Tue, 24 Jul 2012 03:56:13 +0000 (20:56 -0700)]
Merge branch 'sl/autoconf'
* sl/autoconf:
build: reconfigure automatically if configure.ac changes
build: "make clean" should not remove configure-generated files
autoconf: use AC_CONFIG_COMMANDS instead of ad-hoc 'config.mak.append'
autoconf: remove few redundant semicolons
autoconf: remove some redundant shell indirections
autoconf: GIT_CONF_APPEND_LINE -> GIT_CONF_SUBST
autoconf: GIT_CONF_APPEND_LINE: change signature
Junio C Hamano [Tue, 24 Jul 2012 03:56:03 +0000 (20:56 -0700)]
Merge branch 'jv/maint-no-ext-diff'
"git diff --no-ext-diff" did not output anything for a typechange
filepair when GIT_EXTERNAL_DIFF is in effect.
* jv/maint-no-ext-diff:
diff: test precedence of external diff drivers
diff: correctly disable external_diff with --no-ext-diff
Junio C Hamano [Tue, 24 Jul 2012 03:55:53 +0000 (20:55 -0700)]
Merge branch 'kk/maint-commit-tree'
* kk/maint-commit-tree:
Revert "git-commit-tree(1): update synopsis"
commit-tree: resurrect command line parsing updates
Junio C Hamano [Tue, 24 Jul 2012 03:55:33 +0000 (20:55 -0700)]
Merge branch 'jk/mediawiki-credential'
* jk/mediawiki-credential:
mw-to-git: use git-credential's URL parser
credential: convert "url" attribute into its parsed subparts
mw-to-git: check blank credential attributes via length
docs/credential: minor clarity fixups
Junio C Hamano [Tue, 24 Jul 2012 03:55:25 +0000 (20:55 -0700)]
Merge branch 'mm/mediawiki-usability'
* mm/mediawiki-usability:
git-remote-mediawiki: allow page names with a ':'
git-remote-mediawiki: fix incorrect test usage in test
git-remote-mediawiki: properly deal with invalid remote revisions
git-remote-mediawiki: show progress information when getting last remote revision
git-remote-mediawiki: show progress information when listing pages
git-remote-mediawiki: use --force when adding notes
git-remote-mediawiki: get rid of O(N^2) loop
git-remote-mediawiki: make mediafiles export optional
git-remote-mediawiki: actually send empty comment when they're empty
git-remote-mediawiki: don't split namespaces with spaces
Junio C Hamano [Tue, 24 Jul 2012 03:55:21 +0000 (20:55 -0700)]
Merge branch 'sn/doc-typofix'
* sn/doc-typofix:
doc: A few minor copy edits.
Junio C Hamano [Tue, 24 Jul 2012 03:55:16 +0000 (20:55 -0700)]
Merge branch 'tg/ce-namelen-field'
Split lower bits of ce_flags field and creates a new ce_namelen
field in the in-core index structure.
* tg/ce-namelen-field:
Strip namelen out of ce_flags into a ce_namelen field
Junio C Hamano [Tue, 24 Jul 2012 03:55:07 +0000 (20:55 -0700)]
Merge branch 'nk/maint-gitweb-log-by-lines'
Teach gitweb to pay attention to various forms of credits that are
similar to "Signed-off-by:" lines.
* nk/maint-gitweb-log-by-lines:
gitweb: Add support to Link: tag
gitweb: Handle other types of tag in git_print_log
gitweb: Cleanup git_print_log()
Jeff King [Mon, 23 Jul 2012 21:13:12 +0000 (17:13 -0400)]
commit: document the temporary commit message file
We do not document COMMIT_EDITMSG at all, but users may want
to know about it for two reasons:
1. They may want to tell their editor to configure itself
for formatting a commit message.
2. If a commit is aborted by an error, the user may want
to recover the commit message they typed.
Let's put a note in git-commit(1).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Mon, 23 Jul 2012 21:16:12 +0000 (14:16 -0700)]
mergetool: support --tool-help option like difftool does
This way we do not have to risk the list of tools going out of sync
between the implementation and the documentation.
In the same spirit as
bf73fc2 (difftool: print list of valid tools
with '--tool-help', 2012-03-29), trim the list of merge backends in
the documentation. We do not want to have a complete list of valid
tools; we only want a list to help people guess what kind of things
the tools do to be specified there, and refer them to --tool-help
for a complete list.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Mon, 23 Jul 2012 18:50:35 +0000 (14:50 -0400)]
commit: check committer identity more strictly
The identity of the committer will ultimately be pulled from
the ident code by commit_tree(). However, we make an attempt
to check the author and committer identity early, before the
user has done any manual work like inputting a commit
message. That lets us abort without them having to worry
about salvaging the work from .git/COMMIT_EDITMSG.
The early check for committer ident does not use the
IDENT_STRICT flag, meaning that it would not find an empty
name field. The motivation was presumably because we did not
want to be too restrictive, as later calls might be more lax
(for example, when we create the reflog entry, we do not
care too much about a real name). However, because
commit_tree will always get a strict identity to put in the
commit object itself, there is no point in being lax only to
die later (and in fact it is harmful, because the user will
have wasted time typing their commit message).
Incidentally, this bug was masked prior to
060d4bb, as the
initial loose call would taint the later strict call. So the
commit would succeed (albeit with a bogus committer line in
the commit object), and nobody noticed that our early check
did not match the later one.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Mon, 23 Jul 2012 18:48:57 +0000 (14:48 -0400)]
advice: pass varargs to strbuf_vaddf, not strbuf_addf
The advise() function takes a variable number of arguments
and converts them into a va_list object to pass to strbuf
for handling. However, we accidentally called strbuf_addf
(that takes a variable number of arguments) instead of
strbuf_vaddf (that takes a va_list).
This bug dates back to
v1.7.8.1-1-g23cb5bf, but we never
noticed because none of the current callers passes a string
with a format specifier in it. And the compiler did not
notice because the format string is not available at
compile time.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jonathan Nieder [Mon, 23 Jul 2012 06:29:14 +0000 (01:29 -0500)]
Makefile: BLK_SHA1 does not require fast htonl() and unaligned loads
block-sha1/ is fast on most known platforms. Clarify the Makefile to
be less misleading about that.
Early versions of block-sha1/ explicitly relied on fast htonl() and
fast 32-bit loads with arbitrary alignment. Now it uses those on some
arches but the default behavior is byte-at-a-time access for the sake
of arches like ARM, Alpha, and their kin and it is still pretty fast
on these arches (fast enough to supersede the mozilla SHA1
implementation and the hand-written ARM assembler implementation that
were bundled before).
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jonathan Nieder [Sun, 22 Jul 2012 23:47:26 +0000 (18:47 -0500)]
Makefile: fix location of listing produced by "make subdir/foo.s"
When I invoke "make block-sha1/sha1.s", 'make' runs $(CC) -S without
specifying where it should put its output and the output ends up in
./sha1.s. Confusing.
Add an -o option to the .s rule to fix this. We were already doing
that for most compiler invocations but had forgotten it for the
assembler listings.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jonathan Nieder [Sun, 22 Jul 2012 23:40:54 +0000 (18:40 -0500)]
block-sha1: put expanded macro parameters in parentheses
't' is currently always a numeric constant, but it can't hurt to
prepare for the day that it becomes useful for a caller to pass in a
more complex expression.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jonathan Nieder [Sun, 22 Jul 2012 23:39:54 +0000 (18:39 -0500)]
block-sha1: avoid pointer conversion that violates alignment constraints
With
660231aa (block-sha1: support for architectures with memory
alignment restrictions, 2009-08-12), blk_SHA1_Update was modified to
access 32-bit chunks of memory one byte at a time on arches that
prefer that:
#define get_be32(p) ( \
(*((unsigned char *)(p) + 0) << 24) | \
(*((unsigned char *)(p) + 1) << 16) | \
(*((unsigned char *)(p) + 2) << 8) | \
(*((unsigned char *)(p) + 3) << 0) )
The code previously accessed these values by just using htonl(*p).
Unfortunately, Michael noticed on an Alpha machine that git was using
plain 32-bit reads anyway. As soon as we convert a pointer to int *,
the compiler can assume that the object pointed to is correctly
aligned as an int (C99 section 6.3.2.3 "pointer conversions"
paragraph 7), and gcc takes full advantage by using a single 32-bit
load, resulting in a whole bunch of unaligned access traps.
So we need to obey the alignment constraints even when only dealing
with pointers instead of actual values. Do so by changing the type
of 'data' to void *. This patch renames 'data' to 'block' at the same
time to make sure all references are updated to reflect the new type.
Reported-tested-and-explained-by: Michael Cree <mcree@orcon.net.nz>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Sun, 22 Jul 2012 20:20:24 +0000 (13:20 -0700)]
Update draft release notes to 1.7.12
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Sun, 22 Jul 2012 20:08:22 +0000 (13:08 -0700)]
Sync with 1.7.11.3
Junio C Hamano [Sun, 22 Jul 2012 20:07:40 +0000 (13:07 -0700)]
Git 1.7.11.3
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Sun, 22 Jul 2012 20:04:25 +0000 (13:04 -0700)]
Merge branch 'jk/push-delete-ref-error-message' into maint
The error message from "git push $there :bogo" (and its equivalent
"git push $there --delete bogo") mentioned that we tried and failed
to guess what ref is being deleted based on the LHS of the refspec,
which we don't.
* jk/push-delete-ref-error-message:
push: don't guess at qualifying remote refs on deletion
Junio C Hamano [Sun, 22 Jul 2012 20:04:05 +0000 (13:04 -0700)]
Merge branch 'ar/clone-honor-umask-at-top' into maint
A handful of files and directories we create had tighter than
necessary permission bits when the user wanted to have group
writability (e.g. by setting "umask 002").
* ar/clone-honor-umask-at-top:
add: create ADD_EDIT.patch with mode 0666
rerere: make rr-cache fanout directory honor umask
Restore umasks influence on the permissions of work tree created by clone
Junio C Hamano [Sun, 22 Jul 2012 20:03:52 +0000 (13:03 -0700)]
Merge branch 'cw/amend-commit-without-message' into maint
"commit --amend" used to refuse amending a commit with an empty log
message, with or without "--allow-empty-message".
* cw/amend-commit-without-message:
Allow edit of empty message with commit --amend
Junio C Hamano [Sun, 22 Jul 2012 20:03:29 +0000 (13:03 -0700)]
Merge branch 'jk/maint-commit-amend-only-no-paths' into maint
"git commit --amend --only --" was meant to allow "Clever" people to
rewrite the commit message without making any change even when they
have already changes for the next commit added to their index, but
it never worked as advertised since it was introduced in 1.3.0 era.
* jk/maint-commit-amend-only-no-paths:
commit: fix "--amend --only" with no pathspec
Junio C Hamano [Sun, 22 Jul 2012 20:01:56 +0000 (13:01 -0700)]
Merge branch 'tg/maint-cache-name-compare' into maint
Even though the index can record pathnames longer than 1<<12 bytes,
in some places we were not comparing them in full, potentially
replacing index entries instead of adding.
* tg/maint-cache-name-compare:
cache_name_compare(): do not truncate while comparing paths
Junio C Hamano [Sun, 22 Jul 2012 20:01:40 +0000 (13:01 -0700)]
Merge branch 'tr/maint-show-walk' into maint
"git show"'s auto-walking behaviour was an unreliable and
unpredictable hack; it now behaves just like "git log" does when it
walks.
* tr/maint-show-walk:
show: fix "range implies walking"
Demonstrate git-show is broken with ranges
Junio C Hamano [Sun, 22 Jul 2012 20:01:22 +0000 (13:01 -0700)]
Merge branch 'jc/refactor-diff-stdin' into maint
"git diff", "git status" and anything that internally uses the
comparison machinery was utterly broken when the difference
involved a file with "-" as its name. This was due to the way "git
diff --no-index" was incorrectly bolted on to the system, making
any comparison that involves a file "-" at the root level
incorrectly read from the standard input.
* jc/refactor-diff-stdin:
diff-index.c: "git diff" has no need to read blob from the standard input
diff-index.c: unify handling of command line paths
diff-index.c: do not pretend paths are pathspecs
Junio C Hamano [Sun, 22 Jul 2012 20:00:55 +0000 (13:00 -0700)]
Merge branch 'mz/empty-rebase-test' into maint
We did not have test to make sure "git rebase" without extra options
filters out an empty commit in the original history.
* mz/empty-rebase-test:
add test case for rebase of empty commit
Junio C Hamano [Sun, 22 Jul 2012 20:00:45 +0000 (13:00 -0700)]
Merge branch 'js/fast-export-paths-with-spaces' into maint
"git fast-export" produced an input stream for fast-import without
properly quoting pathnames when they contain SPs in them.
* js/fast-export-paths-with-spaces:
fast-export: quote paths with spaces
Junio C Hamano [Sun, 22 Jul 2012 20:00:31 +0000 (13:00 -0700)]
Merge branch 'cw/no-detaching-an-unborn' into maint
"git checkout --detach", when you are still on an unborn branch,
should be forbidden, but it wasn't.
* cw/no-detaching-an-unborn:
git-checkout: disallow --detach on unborn branch
Junio C Hamano [Sun, 22 Jul 2012 19:59:56 +0000 (12:59 -0700)]
Merge branch 'vr/use-our-perl-in-tests' into maint
Some implementations of Perl terminates "lines" with CRLF even when
the script is operating on just a sequence of bytes. Make sure to
use "$PERL_PATH", the version of Perl the user told Git to use, in
our tests to avoid unnecessary breakages in tests.
* vr/use-our-perl-in-tests:
t/README: add a bit more Don'ts
tests: enclose $PERL_PATH in double quotes
t/test-lib.sh: export PERL_PATH for use in scripts
t: Replace 'perl' by $PERL_PATH
Junio C Hamano [Sun, 22 Jul 2012 19:56:49 +0000 (12:56 -0700)]
Merge git://bogomips.org/git-svn
* git://bogomips.org/git-svn:
git-svn: don't create master if another head exists
Junio C Hamano [Sun, 22 Jul 2012 19:56:30 +0000 (12:56 -0700)]
Merge branch 'jk/revision-walk-stop-at-max-count'
"git log -n 1 -- rarely-touched-path" was spending unnecessary
cycles after showing the first change to find the next one, only to
discard it.
* jk/revision-walk-stop-at-max-count:
revision: avoid work after --max-count is reached
Junio C Hamano [Sun, 22 Jul 2012 19:56:27 +0000 (12:56 -0700)]
Merge branch 'mm/config-xdg'
* mm/config-xdg:
config: fix several access(NULL) calls
Junio C Hamano [Sun, 22 Jul 2012 19:56:24 +0000 (12:56 -0700)]
Merge branch 'as/t4012-style-updates'
* as/t4012-style-updates:
t4012: Use test_must_fail instead of if-else
t4012: use 'printf' instead of 'dd' to generate a binary file
t4012: Re-indent test snippets
t4012: Make --shortstat test more robust
t4012: Break up pipe into serial redirections
t4012: Actually quote the sed script
t4012: Unquote git command fragment in test title
t4012: modernize style for quoting
Junio C Hamano [Sun, 22 Jul 2012 19:56:02 +0000 (12:56 -0700)]
Merge branch 'pg/maint-1.7.9-am-where-is-patch'
When "git am" failed, old timers knew to check .git/rebase-apply/patch
to see what went wrong, but we never told the users about it.
* pg/maint-1.7.9-am-where-is-patch:
am: indicate where a failed patch is to be found
Junio C Hamano [Sun, 22 Jul 2012 19:55:47 +0000 (12:55 -0700)]
Merge branch 'jl/maint-1.7.10-recurse-submodules-with-symlink'
When "git submodule add" clones a submodule repository, it can get
confused where to store the resulting submodule repository in the
superproject's .git/ directory when there is a symbolic link in the
path to the current directory.
* jl/maint-1.7.10-recurse-submodules-with-symlink:
submodules: don't stumble over symbolic links when cloning recursively
Junio C Hamano [Sun, 22 Jul 2012 19:55:40 +0000 (12:55 -0700)]
Merge branch 'ms/daemon-doc-typo'
* ms/daemon-doc-typo:
Documentation/git-daemon: add missing word
Junio C Hamano [Sun, 22 Jul 2012 19:55:33 +0000 (12:55 -0700)]
Merge branch 'lk/more-helpful-status-hints'
* lk/more-helpful-status-hints:
status: color in-progress message like other header messages
Junio C Hamano [Sun, 22 Jul 2012 19:55:07 +0000 (12:55 -0700)]
Merge branch 'jc/sha1-name-more'
Teaches the object name parser things like a "git describe" output
is always a commit object, "A" in "git log A" must be a committish,
and "A" and "B" in "git log A...B" both must be committish, etc., to
prolong the lifetime of abbreviated object names.
* jc/sha1-name-more: (27 commits)
t1512: match the "other" object names
t1512: ignore whitespaces in wc -l output
rev-parse --disambiguate=<prefix>
rev-parse: A and B in "rev-parse A..B" refer to committish
reset: the command takes committish
commit-tree: the command wants a tree and commits
apply: --build-fake-ancestor expects blobs
sha1_name.c: add support for disambiguating other types
revision.c: the "log" family, except for "show", takes committish
revision.c: allow handle_revision_arg() to take other flags
sha1_name.c: introduce get_sha1_committish()
sha1_name.c: teach lookup context to get_sha1_with_context()
sha1_name.c: many short names can only be committish
sha1_name.c: get_sha1_1() takes lookup flags
sha1_name.c: get_describe_name() by definition groks only commits
sha1_name.c: teach get_short_sha1() a commit-only option
sha1_name.c: allow get_short_sha1() to take other flags
get_sha1(): fix error status regression
sha1_name.c: restructure disambiguation of short names
sha1_name.c: correct misnamed "canonical" and "res"
...
Junio C Hamano [Sun, 22 Jul 2012 19:54:08 +0000 (12:54 -0700)]
Merge branch 'jc/maint-filter-branch-epoch-date'
In 1.7.9 era, we taught "git rebase" about the raw timestamp format
but we did not teach the same trick to "filter-branch", which rolled
a similar logic on its own. Because of this, "filter-branch" failed
to rewrite commits with ancient timestamps.
* jc/maint-filter-branch-epoch-date:
t7003: add test to filter a branch with a commit at epoch
date.c: Fix off by one error in object-header date parsing
filter-branch: do not forget the '@' prefix to force git-timestamp
Tim Henigan [Thu, 19 Jul 2012 08:27:09 +0000 (01:27 -0700)]
difftool: only copy back files modified during directory diff
When 'difftool --dir-diff' is used to compare working tree files,
it always copies files from the tmp dir back to the working tree
when the diff tool is closed, even if the files were not modified
by the diff tool.
This causes the file timestamp to change. Files should only be
copied from the tmp dir back to the working copy if they were
actually modified.
Signed-off-by: Tim Henigan <tim.henigan@gmail.com>
Signed-off-by: David Aguilar <davvid@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:50:02 +0000 (09:50 +0200)]
build: reconfigure automatically if configure.ac changes
This provides a reduced but still useful sibling of the Automake's
"automatic Makefile rebuild" feature. It's important to note that
we take care to enable the new rules only if the tree that has already
be configured with './configure', so that users relying on manual
configuration won't be negatively impacted.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:50:01 +0000 (09:50 +0200)]
build: "make clean" should not remove configure-generated files
Those filed hold variables, settings and information set by the
configuration process run by './configure'; in Autotools-based
build system that kind of stuff should only be removed by
"make distclean". Having it removed by "make clean" is not only
inconsistent, but causes real confusion for that part of the Git
audience that is used to the Autotools semantics; for example,
an autotools old-timer that has run:
./configure --prefix /opt/git
in the past, without running "make distclean" afterwards, would
expect a "make install" issued after a "make clean" to rebuild and
install git in '/opt/git'; but with the current behaviour, the
"make clean" invocation removes (among the other things) the file
'config.mak.autogen', so that the "make install" falls back to the
default prefix of '$HOME', thus installing git in the user's home
directory -- definitely unexpected.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:50:00 +0000 (09:50 +0200)]
autoconf: use AC_CONFIG_COMMANDS instead of ad-hoc 'config.mak.append'
This will allow "./config.status --recheck; ./config.status" to work
correctly as a mean of reconfiguring the tree with the same configure
argument used in the previous "./configure" invocation.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:49:59 +0000 (09:49 +0200)]
autoconf: remove few redundant semicolons
They are merely useless now, but would get in the way of future changes.
No semantic change is intended.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:49:58 +0000 (09:49 +0200)]
autoconf: remove some redundant shell indirections
They are merely useless now, but would get in the way of future changes.
No semantic change is intended.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:49:57 +0000 (09:49 +0200)]
autoconf: GIT_CONF_APPEND_LINE -> GIT_CONF_SUBST
The new name fits better with the macro signature, and underlines the
similarities with the autoconf-provided macro AC_SUBST (which will be
made even more pronounced in planned future commits).
Once again, no semantic change is intended, and indeed no change to the
generated configure script is expected.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Stefano Lattarini [Thu, 19 Jul 2012 07:49:56 +0000 (09:49 +0200)]
autoconf: GIT_CONF_APPEND_LINE: change signature
Change one-argument GIT_CONF_APPEND_LINE([VAR=VAL]) to
GIT_CONF_APPEND_LINE([VAR], [VAL]), that is more similar to the usual
AC_SUBST macro; this is only a preparatory change in view of future
refactorings.
No semantic change is intended. In fact, the generated configure file
doesn't change at all.
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Thu, 19 Jul 2012 11:49:38 +0000 (07:49 -0400)]
diff: test precedence of external diff drivers
There are three ways to specify an external diff command:
GIT_EXTERNAL_DIFF in the environment, diff.external in the
config, or a "diff" gitattribute. The current order of
precedence is:
1. gitattribute
2. GIT_EXTERNAL_DIFF
3. diff.external
Usually our rule is that environment variables should take
precedence over on-disk config (i.e., option 2 should come
before option 1). However, this situation is trickier than
some, because option 1 is more specific to the individual
file than option 2 (which affects all files), so it might be
preferable. So the current behavior can be seen as
implementing "do the specific thing if we can, but fall back
to this general thing".
This is probably not what we would do if we were writing git
from scratch, but it has been this way for several years,
and is not worth changing. So let's at least document that
this is the way it's supposed to work with a test.
While we're there, let's also make sure that diff.external
(which was not previously tested at all) works by running it
through the same tests as GIT_EXTERNAL_DIFF.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Marcin Owsiany [Sun, 24 Jun 2012 21:40:05 +0000 (22:40 +0100)]
git-svn: don't create master if another head exists
git-svn insists on creating the "master" head (unless it exists) on every
"fetch". It is useful that it gets created initially, when no head exists
- users expect this git convention of having a "master" branch on initial
clone.
However creating it when there already is another head does not provide any
value - the ref is never updated, so it just gets stale after a while. Also,
some users find it annoying that it gets recreated, especially when they would
like the git branch names to follow SVN repository branch names. More
background in http://thread.gmane.org/gmane.comp.version-control.git/115030
Make git-svn skip the "master" creation if HEAD already points at a valid head.
This means "master" does get created on initial "clone" but does not get
recreated once a user deletes it.
Also, make post_fetch_checkout work with any head that is pointed to by HEAD,
not just "master".
Also, use fatal error handling consistent with the rest of the program for
post_fetch_checkout.
Signed-off-by: Marcin Owsiany <marcin@owsiany.pl>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Jeff King [Wed, 18 Jul 2012 13:03:08 +0000 (09:03 -0400)]
mw-to-git: use git-credential's URL parser
We can just feed our URL straight to git-credential and it
will parse it for us, saving us some code.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Wed, 18 Jul 2012 12:06:26 +0000 (08:06 -0400)]
credential: convert "url" attribute into its parsed subparts
The git-credential command requires that you feed it a
broken-down credential, which means that the client needs to
parse a URL itself. Since we have our own URL-parsing
routines, we can easily allow the caller to just give us the
URL as-is, saving them some code.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Wed, 18 Jul 2012 12:04:30 +0000 (08:04 -0400)]
mw-to-git: check blank credential attributes via length
When writing a credential to git-credential, we omit fields
that do not have a true value. This will skip empty or
undefined fields (which we want), but will also accidentally
skip usernames or passwords which happen to have a non-true
value (e.g., "0"). Be more careful by checking for non-zero
length.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Jeff King [Wed, 18 Jul 2012 12:04:02 +0000 (08:04 -0400)]
docs/credential: minor clarity fixups
The text in git-credential(1) was copied from
technical/api-credentials, so it still talks about the
input/output format as coming from git to the helper. Since
the surrounding text already indicates that this format is
used for reading and writing with git credential, we can
just remove the extraneous confusing bits.
Signed-off-by: Jeff King <peff@peff.net>
Acked-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Wed, 18 Jul 2012 05:08:59 +0000 (22:08 -0700)]
diff: correctly disable external_diff with --no-ext-diff
Upon seeing a type-change filepair, "diff --no-ext-diff" does not
show the usual "deletion followed by addition" split patch and does
not run the external diff driver either.
This is because the logic to disable external diff was placed at a
wrong level in the callchain. run_diff_cmd() decides to show the
split patch only when external diff driver is not configured or
specified via GIT_EXTERNAL_DIFF environment, but this is done before
checking if --no-ext-diff was given. To make things worse,
run_diff_cmd() checks --no-ext-diff and disables the output for such
a filepair completely, as the callchain below it (e.g. builtin_diff)
does not want to handle typechange filepairs.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Junio C Hamano [Tue, 17 Jul 2012 20:11:03 +0000 (13:11 -0700)]
Revert "git-commit-tree(1): update synopsis"
This reverts commit
d28436736a078a429213003a9472e8caeb86c286, which
was done without realizing that the updated command line argument
order was lost by mistake.
Junio C Hamano [Tue, 17 Jul 2012 20:10:49 +0000 (13:10 -0700)]
Merge branch 'kk/maint-1.7.9-commit-tree' into kk/maint-commit-tree
* kk/maint-1.7.9-commit-tree:
commit-tree: resurrect command line parsing updates
Junio C Hamano [Tue, 17 Jul 2012 20:05:13 +0000 (13:05 -0700)]
commit-tree: resurrect command line parsing updates
79a9312 (commit-tree: update the command line parsing, 2011-11-09)
updated the command line parser to understand the usual "flags first
and then non-flag arguments" order, in addition to the original and
a bit unusual "tree comes first and then zero or more -p <parent>".
Unfortunately,
ba3c69a (commit: teach --gpg-sign option, 2011-10-05)
broke it by mistake. Resurrect it, and protect the feature with a
test from future breakages.
Noticed by Keshav Kini
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Tue, 17 Jul 2012 14:06:00 +0000 (16:06 +0200)]
git-remote-mediawiki: allow page names with a ':'
Traditionnally, pages named Foo:Bar are page 'Bar' in namespace 'Foo'.
However, it is also possible to call a page Foo:Bar if 'Foo' is not a
namespace. In this case, the actual name of the page is 'Foo:Bar', in the
main namespace. Since we can't tell with only the filename, query the
wiki for a namespace 'Foo' in these cases, but deal with the case where
no such namespace is found.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Tue, 17 Jul 2012 14:05:59 +0000 (16:05 +0200)]
git-remote-mediawiki: fix incorrect test usage in test
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Alexander Strasser [Mon, 16 Jul 2012 20:47:22 +0000 (22:47 +0200)]
t4012: Use test_must_fail instead of if-else
Make the code less bulky and easier to read. Also do not overlook
failures like e.g. git failing because of unexpected signals.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Johannes Sixt [Mon, 16 Jul 2012 20:46:56 +0000 (22:46 +0200)]
t4012: use 'printf' instead of 'dd' to generate a binary file
For some reason, 'echo X | dd bs=1k seek=1' creates a file with 2050 bytes
on Windows instead of the expected 1026 bytes, so that a test fails. Since
the actual contents of the file are irrelevant as long as there is at
least one zero byte so that the diff machinery recognizes it as binary,
use printf to generate it.
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Alexander Strasser [Mon, 16 Jul 2012 20:45:40 +0000 (22:45 +0200)]
t4012: Re-indent test snippets
Most one-level indents were 1 HT (horizontal tab) followed by 1 SP.
Remove the SP.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Alexander Strasser [Mon, 16 Jul 2012 20:45:10 +0000 (22:45 +0200)]
t4012: Make --shortstat test more robust
The --shortstat test depends on the same scenario as the --stat
test. Use the part of the same expected result for the --stat test
to avoid duplicating it manually.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Alexander Strasser [Mon, 16 Jul 2012 20:44:30 +0000 (22:44 +0200)]
t4012: Break up pipe into serial redirections
Do not hide possible git errors by masking its process
exit status.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Alexander Strasser [Mon, 16 Jul 2012 20:43:58 +0000 (22:43 +0200)]
t4012: Actually quote the sed script
The nested quoting is not needed in this cases, thus the previous
version did work just fine. Nevertheless the usage is misleading,
so just achieve nested quoting by using double quotes instead. Lower
the probability of breakage in the future and make the code easier
to read.
NOTE: Just dropping the single quotes around the sed arguments would
have also been possible.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:42 +0000 (21:46 +0200)]
git-remote-mediawiki: properly deal with invalid remote revisions
Some wiki, including https://git.wiki.kernel.org/ have invalid revision
numbers (i.e. the actual revision numbers are non-contiguous). Don't die
when encountering one.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:41 +0000 (21:46 +0200)]
git-remote-mediawiki: show progress information when getting last remote revision
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:40 +0000 (21:46 +0200)]
git-remote-mediawiki: show progress information when listing pages
Initial phases of push and pull with git-remote-mediawiki can be long on
a large wiki. Let the user know what's going on.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:39 +0000 (21:46 +0200)]
git-remote-mediawiki: use --force when adding notes
When notes are created to record a push, it normally doesn't exist yet.
However, when a push is interrupted and then restarted, it may happen
that a commit already has notes attached, and we want to reflect the newly
created remote revision, hence use 'git notes add -f' to override the
existing one
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:38 +0000 (21:46 +0200)]
git-remote-mediawiki: get rid of O(N^2) loop
The algorithm to find a path from the local revision to the remote one
was calling "git rev-list" and parsing its output N times. Run rev-list
only once, and fill a hashtable with the result to optimize the body of
the loop.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Matthieu Moy [Mon, 16 Jul 2012 19:46:37 +0000 (21:46 +0200)]
git-remote-mediawiki: make mediafiles export optional
It is possible to use git-remote-mediawiki on a tree with both .mw files
and other files. Before git-remote-mediawiki learnt how to export
mediafiles, such mixed trees allowed the user to maintain both the wiki
and other files for the same project in the same repository. With the
newly added support for exporting mediafiles, pushing such mixed trees
would upload unrelated files as mediafiles, which may not be desired.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>