git/git.git
15 years agoMerge branch 'hold/am'
Junio C Hamano [Wed, 14 Dec 2005 21:04:43 +0000 (13:04 -0800)]
Merge branch 'hold/am'

15 years agorebase: do not get confused in fast-forward situation.
Junio C Hamano [Wed, 14 Dec 2005 11:11:37 +0000 (03:11 -0800)]
rebase: do not get confused in fast-forward situation.

When switching to another branch and rebasing it in a one-go, it
failed to update the variable that holds the branch head, and
did not detect fast-forward situation correctly.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'fk/usage'
Junio C Hamano [Wed, 14 Dec 2005 21:01:01 +0000 (13:01 -0800)]
Merge branch 'fk/usage'

15 years agoDocumentation: topic branches
Junio C Hamano [Wed, 14 Dec 2005 20:57:49 +0000 (12:57 -0800)]
Documentation: topic branches

Recommend git over ssh direct to master.kernel.org, instead of
going over rsync to public machines, since this is meant to be a
procedure for kernel subsystem maintainers.

Also fix an obvious typo.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapplypatch: no need to do non-portable [[ ... ]]
Junio C Hamano [Wed, 14 Dec 2005 02:19:12 +0000 (18:19 -0800)]
applypatch: no need to do non-portable [[ ... ]]

... when old, proven, case would do.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-repack: Usage string clean-up, emit usage at incorrect invocation
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:32 +0000 (23:30 +0100)]
git-repack: Usage string clean-up, emit usage at incorrect invocation

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-rebase: Usage string clean-up, emit usage string at incorrect invocation
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:32 +0000 (23:30 +0100)]
git-rebase: Usage string clean-up, emit usage string at incorrect invocation

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-prune: Usage string clean-up, use the 'usage' function
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:32 +0000 (23:30 +0100)]
git-prune: Usage string clean-up, use the 'usage' function

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-lost-found: Usage string clean-up, emit usage string at incorrect invocation
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-lost-found: Usage string clean-up, emit usage string at incorrect invocation

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-fetch: Usage string clean-up, emit usage string at unrecognized option
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-fetch: Usage string clean-up, emit usage string at unrecognized option

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-cherry: Usage string clean-up, use the 'usage' function
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-cherry: Usage string clean-up, use the 'usage' function

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-applypatch: Usage string clean-up, emit usage string at incorrect invocation
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-applypatch: Usage string clean-up, emit usage string at incorrect invocation

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-am: Usage string clean-up
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-am: Usage string clean-up

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-revert: Usage string clean-up
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-revert: Usage string clean-up

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-verify-tag: Usage string clean-up, emit usage string at incorrect invocation
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
git-verify-tag: Usage string clean-up, emit usage string at incorrect invocation

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoTrivial usage string clean-up
freku045@student.liu.se [Tue, 13 Dec 2005 22:30:31 +0000 (23:30 +0100)]
Trivial usage string clean-up

Signed-off-by: Fredrik Kuivinen <freku045@student.liu.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agot3200: branch --help does not die anymore.
Junio C Hamano [Tue, 13 Dec 2005 07:38:47 +0000 (23:38 -0800)]
t3200: branch --help does not die anymore.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoUsage message clean-up, take #2
Fredrik Kuivinen [Sun, 11 Dec 2005 09:55:49 +0000 (10:55 +0100)]
Usage message clean-up, take #2

There were some problems with the usage message clean-up patch
series. I hadn't realised that subdirectory aware scripts can't source
git-sh-setup. I propose that we change this and let the scripts which
are subdirectory aware set a variable, SUBDIRECTORY_OK, before they
source git-sh-setup.

The scripts will also set USAGE and possibly LONG_USAGE before they
source git-sh-setup. If LONG_USAGE isn't set it defaults to USAGE.

If we go this way it's easy to catch --help in git-sh-setup, print the
(long) usage message to stdout and exit cleanly. git-sh-setup can
define a 'usage' shell function which can be called by the scripts to
print the short usage string to stderr and exit non-cleanly. It will
also be easy to change $0 to basename $0 or something else, if would
like to do that sometime in the future.

What follows is a patch to convert a couple of the commands to this
style. If it's ok with everyone to do it this way I will convert the
rest of the scripts too.

[jc: thrown in to proposed updates queue for comments.]

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-am support for naked email messages (take 2)
H. Peter Anvin [Wed, 14 Dec 2005 06:39:23 +0000 (22:39 -0800)]
git-am support for naked email messages (take 2)

This allows git-am to accept single-message files as well as mboxes.
Unlike the previous version, this one doesn't need to be explicitly told
which one it is; rather, it looks to see if the first line is a From
line and uses it to select mbox mode or not.

I moved the logic to do all this into git-mailsplit, which got a new
user interface as result, although the old interface is still available
for backwards compatibility.

[jc: applied with two obvious fixes.]

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoEveryday: a bit more examples.
Junio C Hamano [Wed, 14 Dec 2005 08:42:45 +0000 (00:42 -0800)]
Everyday: a bit more examples.

Talk about the following as well:

 * git fetch --tags
 * Use of "git push" as a one-man distributed development vehicle.
 * Show example of remotes file for pulling and pushing.
 * Annotate git-shell setup.
 * Using Carl's update hook in a CVS-style shared repository.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocheckout-index: fix checking out specific path.
Junio C Hamano [Wed, 14 Dec 2005 05:39:56 +0000 (21:39 -0800)]
checkout-index: fix checking out specific path.

3bd348aeea24709cd9be4b9d741f79b6014cd7e3 commit broke checking
out specific paths.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMake git-send-pack exit with error when some refs couldn't be pushed out
Petr Baudis [Wed, 14 Dec 2005 00:45:40 +0000 (01:45 +0100)]
Make git-send-pack exit with error when some refs couldn't be pushed out

In case some refs couldn't be pushed out due to an error (mostly the
not-a-proper-subset error), make git-send-pack exit with non-zero status
after the push is over (that is, it still tries to push out the rest
of the refs).

[jc: I adjusted a test for this change.]

Signed-off-by: Petr Baudis <pasky@suse.cz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years ago[PATCH] allow merging any committish
Junio C Hamano [Wed, 14 Dec 2005 01:01:23 +0000 (17:01 -0800)]
[PATCH] allow merging any committish

Although "git-merge" is advertised as the end-user level command
(instead of being a "git-pull" backend), it was not prepared to
take tag objects that point at commits and barfed when fed one.
Sanitize the input while we validate them, for which we already
have a loop.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-clone: tell the user a bit more about clone-pack failure.
Junio C Hamano [Tue, 13 Dec 2005 23:58:00 +0000 (15:58 -0800)]
git-clone: tell the user a bit more about clone-pack failure.

When clone-pack has trouble with the remote, it dies unfriendly
"EOF" error message.  We cannot tell the reason why it failed
from the local end; it could be that the repository did not
exist, or configured not to serve over git-daemon, or a network
failure.  At least, saying clone-pack failed makes it a bit more
meaningful.

I am not convinced yet that removing the newly created directory
is the right thing to do, so this commit leaves the new
directory behind.

Reported by Sam Ravnborg.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: not learning core git commands.
Junio C Hamano [Tue, 13 Dec 2005 10:38:24 +0000 (02:38 -0800)]
Documentation: not learning core git commands.

The initial section of tutorial was too heavy on internal
workings for the first-time readers, so rewrite the introductory
section of git(7) to start with "not learning core git commands"
section and refer them to README to grasp the basic concepts,
then Everyday to give overview with task/role oriented examples
for minimum set of commands, and finally the tutorial.

Also add to existing note in the tutorial that many too
technical descriptions can be skipped by a casual reader.

I initially started to review the tutorial, with the objective
of ripping out the detailed technical information altogether,
but I found that the level of details in the initial couple of
sections that talk about refs and the object database in a
hands-on fashion was about rigth, and left all of them there.  I
feel that reading about fsck-index and repack is too abstract
without being aware of these directories and files.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: diff examples.
Junio C Hamano [Tue, 13 Dec 2005 09:54:15 +0000 (01:54 -0800)]
Documentation: diff examples.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: fix missing links to git(7)
Junio C Hamano [Tue, 13 Dec 2005 07:55:09 +0000 (23:55 -0800)]
Documentation: fix missing links to git(7)

Also move pack protocol description to technical/.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: more examples.
Junio C Hamano [Tue, 13 Dec 2005 07:24:06 +0000 (23:24 -0800)]
Documentation: more examples.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoEveryday: a bit more example.
Junio C Hamano [Tue, 13 Dec 2005 02:29:53 +0000 (18:29 -0800)]
Everyday: a bit more example.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agodiffcore-break: do not break too small filepair.
Junio C Hamano [Tue, 13 Dec 2005 01:15:55 +0000 (17:15 -0800)]
diffcore-break: do not break too small filepair.

Somehow we checked only one side and not the other.  By checking
the filesize upfront, we can bypass generating delta
unnecessarily.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoRevert "Add deltifier test."
Junio C Hamano [Tue, 13 Dec 2005 00:43:15 +0000 (16:43 -0800)]
Revert "Add deltifier test."

This reverts e726715a52e25d8035c89d4ea09398599610737e commit,
because reverting diff-delta emptiness change would break this test.

15 years agoRevert "diff-delta.c: allow delta with empty blob."
Junio C Hamano [Tue, 13 Dec 2005 00:42:38 +0000 (16:42 -0800)]
Revert "diff-delta.c: allow delta with empty blob."

This reverts 962537a3eb03a118cf27d9d0da365a3216ed1caa commit to
play safe.

15 years agoEveryday: some examples.
Junio C Hamano [Tue, 13 Dec 2005 00:20:21 +0000 (16:20 -0800)]
Everyday: some examples.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agodiff-delta.c: allow delta with empty blob.
Junio C Hamano [Mon, 12 Dec 2005 20:01:52 +0000 (12:01 -0800)]
diff-delta.c: allow delta with empty blob.

Delta computation with an empty blob used to punt and returned NULL.
This commit allows creation with empty blob; all combination of
empty->empty, empty->something, and something->empty are allowed.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAdd deltifier test.
Junio C Hamano [Mon, 12 Dec 2005 20:28:34 +0000 (12:28 -0800)]
Add deltifier test.

This test kicks in only if you built test-delta executable, and
makes sure that the basic delta routine is working properly even
on empty files.

This commit is to make sure we have a test to catch the
breakage.  The delitifier code is still broken, which will be
fixed with the next commit.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agodiffcore-break.c: check diff_delta() return value.
Junio C Hamano [Mon, 12 Dec 2005 20:50:40 +0000 (12:50 -0800)]
diffcore-break.c: check diff_delta() return value.

This bug caused Darrin Thompson to notice that our deltifier was
half broken and punting on an empty blob.

Signed-off-by: Junio C Hamano <junio@twinsun.com>
15 years agot/t6022: a new test for renaming merge.
Junio C Hamano [Sun, 11 Dec 2005 20:44:46 +0000 (12:44 -0800)]
t/t6022: a new test for renaming merge.

This adds a couple of tests to cover the following renaming
merge cases:

 - one side renames and the other side does not, with and without
   content conflicts.

 - both side rename to the same path, with and without content
   conflicts.

The test setup also prepares a case in which both side rename to
different destination, but currently the code collapses these
destination paths and removes the original path, which may be
wrong.  The outcome of this case is not checked by the tests in
this round.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agomerge-recursive: cleanup setIndexStages
Junio C Hamano [Sun, 11 Dec 2005 09:55:52 +0000 (01:55 -0800)]
merge-recursive: cleanup setIndexStages

Fredrik points out there is a useful wrapper runProgram() used
everywhere that we can use to feed input into subprocess.  Use
it to catch errors from the subprocess; it is a good cleanup as
well.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agomerge-recursive: leave unmerged entries in the index.
Junio C Hamano [Sun, 11 Dec 2005 06:05:01 +0000 (22:05 -0800)]
merge-recursive: leave unmerged entries in the index.

This does two things.

 - When one branch renamed and the other branch did not, the
   resulting half-merged file in the working tree used to swap
   branches around and showed as if renaming side was "ours".
   This was confusing and inconsistent (even though the conflict
   markers were marked with branch names, it was not a good
   enough excuse).  This changes the order of arguments to
   mergeFile in such a case to make sure we always see "our"
   change between <<< and ===, and "their" change between ===
   and >>>.

 - When both branches renamed to the same path, and when one
   branch renamed and the other branch did not, we attempt
   mergeFile.  When this automerge conflicted, we used to
   collapse the index.  Now we use update-index --index-info
   to inject higher stage entries to leave the index in unmerged
   state for these two cases.

What this still does _not_ do is to inject unmerged state into
the index when the structural changes conflict.  I have not
thought things through what to do in each case yet, but the
cases this commit cover are the most common ones, so this would
be a good start.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoformat-patch: use same number of digits in numbers
Junio C Hamano [Sun, 11 Dec 2005 06:22:35 +0000 (22:22 -0800)]
format-patch: use same number of digits in numbers

This would help sorting by subject in MUA work saner even though
MUA is too dumb to attempt sorting numbered subjects sanely.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agot0000: catch trivial pilot errors.
Junio C Hamano [Sun, 11 Dec 2005 04:55:32 +0000 (20:55 -0800)]
t0000: catch trivial pilot errors.

People seem to be getting test failure from t6021 not becuase
git is faulty but because they forgot to install "merge".  Check
this and other trivial pilot errors in the first test.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agot3300: funny name test
Junio C Hamano [Sun, 11 Dec 2005 02:36:15 +0000 (18:36 -0800)]
t3300: funny name test

Add double quote character to the test pattern.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAllow saving an object from a pipe
Daniel Barkalow [Sat, 10 Dec 2005 22:25:24 +0000 (17:25 -0500)]
Allow saving an object from a pipe

In order to support getting data into git with scripts, this adds a
--stdin option to git-hash-object, which will make it read from stdin.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoarchimport: fix a in new changeset applyer addition
Eric Wong [Thu, 24 Nov 2005 07:58:16 +0000 (23:58 -0800)]
archimport: fix a in new changeset applyer addition

Fix a stupid bug I introduced when splitting the
accurate and fast changeset appliers.

Also, remove an old debugging statement I added

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: Fix a bug I introduced in the new log parser
Eric Wong [Thu, 24 Nov 2005 07:56:31 +0000 (23:56 -0800)]
archimport: Fix a bug I introduced in the new log parser

This fixes the case (that worked originally in Martin's version)
where the only new/modified files are Arch control files.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: Add the accurate changeset applyer
Eric Wong [Thu, 24 Nov 2005 07:55:04 +0000 (23:55 -0800)]
archimport: Add the accurate changeset applyer

And make it the default.
This includes stats tracking to verbose mode

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: safer log file parsing
Eric Wong [Thu, 24 Nov 2005 07:53:55 +0000 (23:53 -0800)]
archimport: safer log file parsing

Better logfile parsing, no longer confused by 'headers' after the first
blank line.

Re-enabled tag-reading with abrowse (baz and tla compatible)

Remove need to quote args to external processes

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: add -D <depth> and -a switch
Eric Wong [Thu, 24 Nov 2005 07:52:43 +0000 (23:52 -0800)]
archimport: add -D <depth> and -a switch

add -D <depth> option to abrowse add -a switch to attempt to
auto-register archives at mirrors.sourcecontrol.net

(ML: Also removes some std libraries no longer in use)

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: remove git wrapper dependency
Eric Wong [Thu, 24 Nov 2005 07:51:33 +0000 (23:51 -0800)]
archimport: remove git wrapper dependency

use git-diff-files instead of git diff-files so we don't rely on the
wrapper being installed (some people may have git as GNU interactive
tools :)

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: fix -t tmpdir switch
Eric Wong [Thu, 24 Nov 2005 07:50:27 +0000 (23:50 -0800)]
archimport: fix -t tmpdir switch

set TMPDIR env correctly if -t <tmpdir> is passed from the command-line.
setting TMPDIR => 1 as an argument to tempdir() has no effect otherwise

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: remove String::ShellQuote dependency.
Eric Wong [Thu, 24 Nov 2005 07:48:57 +0000 (23:48 -0800)]
archimport: remove String::ShellQuote dependency.

use safe_pipe_capture() or system() over backticks where
shellquoting may have been necessary.
More changes planned, so I'm not touching the parts I'm
planning on replacing entirely.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoarchimport: first, make sure it still compiles
Eric Wong [Thu, 24 Nov 2005 07:47:39 +0000 (23:47 -0800)]
archimport: first, make sure it still compiles

(ML: And introduce safe_pipe_capture())

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
15 years agoLink Everyday GIT to main documentation tree.
Junio C Hamano [Sat, 10 Dec 2005 07:41:03 +0000 (23:41 -0800)]
Link Everyday GIT to main documentation tree.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoEveryday GIT with 20 commands
Junio C Hamano [Sat, 10 Dec 2005 07:07:29 +0000 (23:07 -0800)]
Everyday GIT with 20 commands

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoUse GIT_EXEC_PATH explicitly for initial git-init-db in tests.
Junio C Hamano [Sat, 10 Dec 2005 01:32:18 +0000 (17:32 -0800)]
Use GIT_EXEC_PATH explicitly for initial git-init-db in tests.

This is just a belts-and-suspenders check, but makes sure we
have both "git" and "git-init-db" built, executable, and
checking.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agouse "git init-db" in tests
Alex Riesen [Thu, 8 Dec 2005 20:25:55 +0000 (21:25 +0100)]
use "git init-db" in tests

This is to catch an error where tests are run without first
building what are being tested.  Relying on prefixing $PATH with
the build directory and expect that the PATH mechanism would
find what we just built would silently run an already installed
binaries from the PATH.

Signed-off-by: Alex Riesen <raa.lkml@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-prune: never lose objects reachable from our refs.
Junio C Hamano [Fri, 9 Dec 2005 07:18:41 +0000 (23:18 -0800)]
git-prune: never lose objects reachable from our refs.

Explicit <head> arguments to git-prune replaces, instead of
extends, the list of heads used for reachability analysis by
fsck-objects.  By giving a subset of heads by mistake, objects
reachable only from other heads can be removed, resulting in a
corrupted repository.

This commit stops replacing the list of heads, and makes the
command line arguments to add to them instead for safety.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoqsort(): ptrdiff_t may be larger than int
Junio C Hamano [Fri, 9 Dec 2005 01:29:11 +0000 (17:29 -0800)]
qsort(): ptrdiff_t may be larger than int

This is a companion patch to e23eff8be92a2a2cb66b53deef020063cff285ed
commit.  The same logic, the same rationale that a comparison
function that returns an int should not just compute a ptrdiff_t
and return it.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: git-prune
Junio C Hamano [Fri, 9 Dec 2005 00:30:24 +0000 (16:30 -0800)]
Documentation: git-prune

Not replacing but always including our own refs may be more
desirable (and unarguably much safer), but at the same time I
have a suspicion that that might be forbidding a useful usage I
haven't thought of, so...

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-read-tree.txt: Add --reset to SYNOPSIS.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:06 +0000 (00:28 +0100)]
Documentation/git-read-tree.txt: Add --reset to SYNOPSIS.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-tag.txt: Fix the order of sections (DESCRIPTION should come before...
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-tag.txt: Fix the order of sections (DESCRIPTION should come before OPTIONS).

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-update-server-info.txt: Add -f alias for --force to documentation.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-update-server-info.txt: Add -f alias for --force to documentation.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-http-fetch.txt: Document the commit-id argument.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-http-fetch.txt: Document the commit-id argument.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-repack.txt: Add -l and -n.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-repack.txt: Add -l and -n.

This adds documentation for the -l and -n options to git-repack.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-ls-remote.txt: Add -h and -t.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-ls-remote.txt: Add -h and -t.

-h and -t are aliases for --heads and --tags to git-ls-remote.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-format-patch.txt: Add --signoff, --check, and long option-names.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-format-patch.txt: Add --signoff, --check, and long option-names.

The documentation was lacking descriptions for the --signoff and --check
options to git-format-patch.  It was also missing the following long
option-names: --output-directory (-o), --numbered (-n), --keep-subject
(-k), --author (-a), --date (-d), and --mbox (-m).

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAdd documentation for the --topo-order option to git-show-branch.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Add documentation for the --topo-order option to git-show-branch.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAdd documentation for git-revert and git-cherry-pick.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Add documentation for git-revert and git-cherry-pick.

* Added the -e option to the documentation of git-cherry-pick.
* Added the -e and --no-commit option to git-revert.
* Removed redundant case expression for -n as --no-edit (already taken by
  --no-commit).

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-cherry-pick: Add --replay and --no-commit.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-cherry-pick: Add --replay and --no-commit.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-cvsimport.txt: Fix a slight glitch in description heading.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-cvsimport.txt: Fix a slight glitch in description heading.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-verify-pack.txt: added documentation for --.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-verify-pack.txt: added documentation for --.

The -- option has been added to the documentation of git-verify-pack.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation/git-commit.txt: Add long options and -- to documentation.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Documentation/git-commit.txt: Add long options and -- to documentation.

Added the following long options to documentation:

* --all
* --signoff
* --verify
* --no-verify
* --edit

Also added documentation for the -- option for terminating option parsing.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoUse uniform description for the '--' option.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Use uniform description for the '--' option.

All descriptions of the '--' option were the same except for that in
Documentation/git-merge-index.txt.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoFix the description of --utf8 and --keep to git-am.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Fix the description of --utf8 and --keep to git-am.

The git-am script actually transform --utf8 and --keep to -u and -k when
sent to git-mailinfo.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocument the --non-empty command-line option to git-pack-objects.
Nikolai Weibull [Thu, 8 Dec 2005 23:28:05 +0000 (00:28 +0100)]
Document the --non-empty command-line option to git-pack-objects.

This provides (minimal) documentation for the --non-empty command-line
option to the pack-objects command.

Signed-off-by: Nikolai Weibull <nikolai@bitwi.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoshow-branch: comment typo
Junio C Hamano [Thu, 8 Dec 2005 22:10:02 +0000 (14:10 -0800)]
show-branch: comment typo

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: recursive is the default strategy these days.
Junio C Hamano [Thu, 8 Dec 2005 22:04:33 +0000 (14:04 -0800)]
Documentation: recursive is the default strategy these days.

We still said resolve was the default in handful places.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMake sure we use compat/subprocess.py with Python 2.3 while running tests.
Junio C Hamano [Thu, 8 Dec 2005 05:52:28 +0000 (21:52 -0800)]
Make sure we use compat/subprocess.py with Python 2.3 while running tests.

Otherwise the test will not succeed without installing.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoClean up file descriptors when calling hooks.
Daniel Barkalow [Thu, 8 Dec 2005 02:04:38 +0000 (21:04 -0500)]
Clean up file descriptors when calling hooks.

When calling post-update hook, don't leave stdin and stdout connected to
the pushing connection.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agorefs.c: make sure leading directories exist before writing a ref.
Junio C Hamano [Thu, 8 Dec 2005 01:48:53 +0000 (17:48 -0800)]
refs.c: make sure leading directories exist before writing a ref.

Otherwise cloning a repository with hierarchical branch/tag
over http would fail.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation(tutorial): adjust merge example to the new merge world order.
Junio C Hamano [Thu, 8 Dec 2005 00:44:12 +0000 (16:44 -0800)]
Documentation(tutorial): adjust merge example to the new merge world order.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation(cvs-migration): minor cleanups.
Junio C Hamano [Thu, 8 Dec 2005 00:26:49 +0000 (16:26 -0800)]
Documentation(cvs-migration): minor cleanups.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation(glossary): minor formatting clean-ups.
Junio C Hamano [Thu, 8 Dec 2005 00:16:04 +0000 (16:16 -0800)]
Documentation(glossary): minor formatting clean-ups.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: git.html/git.7
Junio C Hamano [Thu, 8 Dec 2005 00:05:21 +0000 (16:05 -0800)]
Documentation: git.html/git.7

Finish each sentence with a full stop.

Instead of saying 'directory index' 'directory cache' etc,
consistently say 'index'.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge http://www.kernel.org/pub/scm/gitk/gitk
Junio C Hamano [Wed, 7 Dec 2005 19:25:42 +0000 (11:25 -0800)]
Merge http://www.kernel.org/pub/scm/gitk/gitk

15 years agogitk: Work around Tcl's non-standard names for encodings
Paul Mackerras [Wed, 7 Dec 2005 12:28:22 +0000 (23:28 +1100)]
gitk: Work around Tcl's non-standard names for encodings

This uses a table of encoding names and aliases distilled from
http://www.iana.org/assignments/character-sets plus some heuristics
to convert standard encoding names to ones that Tcl recognizes.

Signed-off-by: Paul Mackerras <paulus@samba.org>
15 years agoupdate-index: allow --index-info to add higher stages.
Junio C Hamano [Wed, 7 Dec 2005 09:45:38 +0000 (01:45 -0800)]
update-index: allow --index-info to add higher stages.

The new merge world order tells the merge strategies to leave
the cache unmerged and store the automerge result in the working
tree if automerge is not clean.  This was done for the resolve
strategy and recursive strategy when no rename is involved, but
recording a conflicting merge in the rename case could not
easily be done by the recursive strategy.

This commit adds a new input format, in addition to the exsting
two, to "update-index --index-info".

    (1) mode         SP sha1          TAB path
    The first format is what "git-apply --index-info"
    reports, and used to reconstruct a partial tree
    that is used for phony merge base tree when falling
    back on 3-way merge.

    (2) mode SP type SP sha1          TAB path
    The second format is to stuff git-ls-tree output
    into the index file.

    (3) mode         SP sha1 SP stage TAB path
    This format is to put higher order stages into the
    index file and matches git-ls-files --stage output.

To place a higher stage entry to the index, the path should
first be removed by feeding a mode=0 entry for the path, and
then feeding necessary input lines in the (3) format.

For example, starting with this index:

$ git ls-files -s
100644 8a1218a1024a212bb3db30becd860315f9f3ac52 0       frotz

$ git update-index --index-info ;# interactive session -- input follows...

0000000000000000000000000000000000000000 frotz
100644 8a1218a1024a212bb3db30becd860315f9f3ac52 1 frotz
100755 8a1218a1024a212bb3db30becd860315f9f3ac52 2 frotz

The first line of the input feeds 0 as the mode to remove the
path; the SHA1 does not matter as long as it is well formatted.
Then the second and third line feeds stage 1 and stage 2 entries
for that path.  After the above, we would end up with this:

$ git ls-files -s
100644 8a1218a1024a212bb3db30becd860315f9f3ac52 1 frotz
100755 8a1218a1024a212bb3db30becd860315f9f3ac52 2 frotz

This completes the groundwork for the new merge world order.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-merge-one: new merge world order.
Junio C Hamano [Wed, 7 Dec 2005 08:50:33 +0000 (00:50 -0800)]
git-merge-one: new merge world order.

This does two things:

 - Use new --stage=2 option to create the working tree file with
   leading paths and correct permission bits using
   checkout-index, as before.

 - Make sure we do not confuse "merge" program when the file
   being merged has an unfortunate name, '-L'.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocheckout-index: allow checking out from higher stages.
Junio C Hamano [Wed, 7 Dec 2005 08:29:51 +0000 (00:29 -0800)]
checkout-index: allow checking out from higher stages.

The new option, --stage=<n>, lets you copy out from an unmerged,
higher stage.  This is to help the new merge world order during
a nontrivial merge.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoUse printf rather than echo -n.
Jason Riedy [Tue, 6 Dec 2005 22:21:52 +0000 (14:21 -0800)]
Use printf rather than echo -n.

On AIX, there is no -n option to the system's echo.  Instead,
it needs the '\c' control character.  We could replace
  echo -n "foo"
with
  echo -e "foo\c"
but printf is recommended by most man pages.  Tested on AIX
5.3, Solaris 8, and Debian.

[jc: futureproofed two instances that uses variable with '%s'
 so later feeding different messages would not break things too
 easily; others are emitting literal so whoever changes the
 literal ought to notice more easily so they are safe.]

Signed-off-by: E. Jason Riedy <ejr@cs.berkeley.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoqsort() ptrdiff_t may be larger than int
Junio C Hamano [Tue, 6 Dec 2005 21:41:48 +0000 (13:41 -0800)]
qsort() ptrdiff_t may be larger than int

Morten Welinder <mwelinder@gmail.com> writes:

> The code looks wrong.  It assumes that pointers are no larger than ints.
> If pointers are larger than ints, the code does not necessarily compute
> a consistent ordering and qsort is allowed to do whatever it wants.
>
> Morten
>
> static int compare_object_pointers(const void *a, const void *b)
> {
>  const struct object * const *pa = a;
>  const struct object * const *pb = b;
>  return *pa - *pb;
> }

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years ago[PATCH] Initial AIX portability fixes.
Jason Riedy [Tue, 6 Dec 2005 22:20:16 +0000 (14:20 -0800)]
[PATCH] Initial AIX portability fixes.

Added an AIX clause in the Makefile; that clause likely
will be wrong for any AIX pre-5.2, but I can only test
on 5.3.  mailinfo.c was missing the compat header file,
and convert-objects.c needs to define a specific
_XOPEN_SOURCE as well as _XOPEN_SOURCE_EXTENDED.

Signed-off-by: E. Jason Riedy <ejr@cs.berkeley.edu>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-merge-one-file: resurrect leading path creation.
Junio C Hamano [Tue, 6 Dec 2005 08:07:58 +0000 (00:07 -0800)]
git-merge-one-file: resurrect leading path creation.

Since we do not use git-update-index followed by
git-checkout-index -u to create the half-merged file on
conflicting case anymore, we need to make sure the leading
directories are created here.

Maybe a better solution would be to allow update-index to add to
higher stage, and checkout-index to extract from such, but that
is a change slightly bigger than I would like to have so close
to 1.0, so this should do for now.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentaiton (read-tree): update description of 3-way
Junio C Hamano [Tue, 6 Dec 2005 07:26:10 +0000 (23:26 -0800)]
Documentaiton (read-tree): update description of 3-way

The merge-one-file used to leave the working tree intact, but
it has long been changed to leave the merge result there since
2a68a8659f7dc55fd285d235ae2d19e7a8116c30 commit.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: hash-object.
Junio C Hamano [Tue, 6 Dec 2005 06:36:22 +0000 (22:36 -0800)]
Documentation: hash-object.

The file parameter is better spelled just "file", not "any file
on the filesystem".  We stress that in the description text
later anyway.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agowrite-tree: check extra arguments and die but be a bit more helpful.
Junio C Hamano [Tue, 6 Dec 2005 06:30:07 +0000 (22:30 -0800)]
write-tree: check extra arguments and die but be a bit more helpful.

"git-write-tree junk" complains and dies, but it does not say
what option it supports.  Die with the usage string in such a
case.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoinit-db: check extra arguments and complain.
Junio C Hamano [Tue, 6 Dec 2005 06:29:36 +0000 (22:29 -0800)]
init-db: check extra arguments and complain.

"git-init-db junk" does not complain but just ignores "junk".
Die with the usage string in such a case.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agohash-object: -- and --help
Junio C Hamano [Tue, 6 Dec 2005 06:29:05 +0000 (22:29 -0800)]
hash-object: -- and --help

It was cumbersome to feed hash-object the file '-t' (you could
have said "./-t", though).  Teach it '--' that terminates the
option list, like everybody else.  There is no way to extract
usage string from the command either, so teach it "--help" as
well.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAdded documentation for few missing options.
Jon Loeliger [Tue, 6 Dec 2005 05:13:03 +0000 (23:13 -0600)]
Added documentation for few missing options.

More $ shell prompts in examples.
Minor English grammar improvements.
Added a few "See Also"s.
Use back-ticks on more command examples.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-mv to work with Perl 5.6
Junio C Hamano [Tue, 6 Dec 2005 02:14:34 +0000 (18:14 -0800)]
git-mv to work with Perl 5.6

List form of pipe open is 5.8 invention.

Signed-off-by: Junio C Hamano <junkio@cox.net>