git/git.git
15 years agoMerge branch 'maint'
Junio C Hamano [Sat, 4 Mar 2006 05:36:52 +0000 (21:36 -0800)]
Merge branch 'maint'

* maint:
  tar-tree: file/dirmode fix.

15 years agotar-tree: file/dirmode fix.
Junio C Hamano [Sat, 4 Mar 2006 05:34:14 +0000 (21:34 -0800)]
tar-tree: file/dirmode fix.

This fixes two bugs introduced when we switched to generic tree
traversal code.

 (1) directory mode recorded silently became 0755, not 0777

 (2) if passed a tree object (not a commit), it emitted an
     alarming error message (but proceeded anyway).

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: fix a copied-tree bug in an overzealous assertion
Eric Wong [Fri, 3 Mar 2006 21:35:48 +0000 (13:35 -0800)]
contrib/git-svn: fix a copied-tree bug in an overzealous assertion

I thought passing --stop-on-copy to svn would save us from all
the trouble svn-arch-mirror had with directory (project) copies.
I was wrong, there was one thing I overlooked.

If a tree was moved from /foo/trunk to /bar/foo/trunk with no
other changes in r10, but the last change was done in r5, the
Last Changed Rev (from svn info) in /bar/foo/trunk will still be
r5, even though the copy in the repository didn't exist until
r10.

Now, if we ever detect that the Last Changed Rev isn't what
we're expecting, we'll run svn diff and only croak if there are
differences between them.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoshow-branch --topics: omit more uninteresting commits.
Junio C Hamano [Fri, 3 Mar 2006 22:34:40 +0000 (14:34 -0800)]
show-branch --topics: omit more uninteresting commits.

When inspecting contents of topic branches for yet-to-be-merged
commits, a commit that is in the release/master branch is
uninteresting.  Previous round still showed them, especially,
the ones before a topic branch that was forked from the
release/master later than other topic branches.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoworkaround fat/ntfs deficiencies for t3600-rm.sh (git-rm)
Alex Riesen [Fri, 3 Mar 2006 10:20:18 +0000 (11:20 +0100)]
workaround fat/ntfs deficiencies for t3600-rm.sh (git-rm)

Signed-off-by: Alex Riesen <ariesen@harmanbecker.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-mv: fix moves into a subdir from outside
Josef Weidendorfer [Fri, 3 Mar 2006 16:23:32 +0000 (17:23 +0100)]
git-mv: fix moves into a subdir from outside

git-mv needs to be run from the base directory so that
the check if a file is under revision also covers files
outside of a subdirectory. Previously, e.g. in the git repo,

  cd Documentation; git-mv ../README .

produced the error

  Error: '../README' not under version control

The test is extended for this case; it previously only tested
one direction.

Signed-off-by: Josef Weidendorfer <Josef.Weidendorfer@gmx.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agosend-email: accept --no-signed-off-by-cc as the documentation states
Eric Wong [Fri, 3 Mar 2006 09:28:48 +0000 (01:28 -0800)]
send-email: accept --no-signed-off-by-cc as the documentation states

--no-signed-off-cc is still supported, for backwards compatibility

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: better documenting of CLI switches
Eric Wong [Fri, 3 Mar 2006 09:20:09 +0000 (01:20 -0800)]
contrib/git-svn: better documenting of CLI switches

Also, fix a asciidoc formatting error

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: add --id/-i=$GIT_SVN_ID command-line switch
Eric Wong [Fri, 3 Mar 2006 09:20:09 +0000 (01:20 -0800)]
contrib/git-svn: add --id/-i=$GIT_SVN_ID command-line switch

I ended up using GIT_SVN_ID far more than I ever thought I
would.  Typing less is good.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: avoid re-reading the repository uuid, it never changes
Eric Wong [Fri, 3 Mar 2006 09:20:09 +0000 (01:20 -0800)]
contrib/git-svn: avoid re-reading the repository uuid, it never changes

If it does change, we're screwed anyways as SVN will refuse to
commit or update.  We also never access more than one SVN
repository per-invocation, so we can store it as a global, too.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: create a more recent master if one does not exist
Eric Wong [Fri, 3 Mar 2006 09:20:09 +0000 (01:20 -0800)]
contrib/git-svn: create a more recent master if one does not exist

In a new repository, the initial fetch creates a master branch
if one does not exist so HEAD has something to point to.

It now creates a master at the end of the initial fetch run,
pointing to the latest revision.  Previously it pointed to the
first revision imported, which is generally less useful.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: cleanup option parsing
Eric Wong [Fri, 3 Mar 2006 09:20:08 +0000 (01:20 -0800)]
contrib/git-svn: cleanup option parsing

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: allow --authors-file to be specified
Eric Wong [Fri, 3 Mar 2006 09:20:08 +0000 (01:20 -0800)]
contrib/git-svn: allow --authors-file to be specified

Syntax is compatible with git-svnimport and git-cvsimport:

normalperson = Eric Wong <normalperson@yhbt.net>

If this option is specified and git-svn encounters an SVN
committer name that it cannot parse, it git-svn will abort.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: strip 'git-svn-id:' when commiting to SVN
Eric Wong [Fri, 3 Mar 2006 09:20:08 +0000 (01:20 -0800)]
contrib/git-svn: strip 'git-svn-id:' when commiting to SVN

We regenerate and use git-svn-id: whenever we fetch or otherwise
commit to remotes/git-svn.  We don't actually know what revision
number we'll commit to SVN at commit time, so this is useless.
It won't throw off things like 'rebuild', though, which knows to
only use the last instance of git-svn-id: in a log message

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: several small bug fixes and changes
Eric Wong [Fri, 3 Mar 2006 09:20:07 +0000 (01:20 -0800)]
contrib/git-svn: several small bug fixes and changes

 * Fixed manually-edited commit messages not going to
   remotes/git-svn on sequential commits after the sequential
   commit optimization.
 * format help correctly after adding 'show-ignore'
 * sha1_short regexp matches down to 4 hex characters
   (from git-rev-parse --short documentation)
 * Print the first line of the commit message when we commit to
   SVN next to the sha1.
 * Document 'T' (type change) in the comments

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: add -b/--branch switch for branch detection
Eric Wong [Fri, 3 Mar 2006 09:20:07 +0000 (01:20 -0800)]
contrib/git-svn: add -b/--branch switch for branch detection

I've said I don't like branches in Subversion, and I still don't.
This is a bit more flexible, though, as the argument for -b is any
arbitrary git head/tag reference.

This makes some things easier:
 * Importing git history into a brand new SVN branch.
 * Tracking multiple SVN branches via GIT_SVN_ID, even from multiple
   repositories.
 * Adding tags from SVN (still need to use GIT_SVN_ID, though).
 * Even merge tracking is supported, if and only the heads end up with
   100% equivalent tree objects.  This is more stricter but more robust
   and foolproof than parsing commit messages, imho.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoPrevent --index-info from ignoring -z.
Shawn Pearce [Thu, 2 Mar 2006 17:21:33 +0000 (12:21 -0500)]
Prevent --index-info from ignoring -z.

If git-update-index --index-info -z is used only the first
record given to the process will actually be updated as
the -z option is ignored until after all index records
have been read and processed.  This meant that multiple
null terminated records were seen as a single record which
was lacking a trailing LF, however since the first record
ended in a null the C string handling functions ignored the
trailing garbage.  So --index-info should be required to be
the last command line option, much as --stdin is required
to be the last command line option.  Because --index-info
implies --stdin this isn't an issue as the user shouldn't
be passing --stdin when also passing --index-info.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agomanpages: insert two missing [verse] markers for multi-line SYNOPSIS
Jonas Fonseca [Thu, 2 Mar 2006 16:10:49 +0000 (17:10 +0100)]
manpages: insert two missing [verse] markers for multi-line SYNOPSIS

Found with:

for i in *.txt; do
grep -A 2 "SYNOPSIS" "$i" | grep -q "^\[verse\]$" && continue
multiline=$(grep -A 3 "SYNOPSIS" "$i" | tail -n 1)
test -n "$multiline" && echo "$i: $multiline"
done

Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: pass the missing argument _show_clicked_cb.
Aneesh Kumar K.V [Thu, 2 Mar 2006 11:47:39 +0000 (17:17 +0530)]
gitview: pass the missing argument _show_clicked_cb.

In our last update to use the encoding while showing the commit
diff we added a new argument to this function. But we missed
updating all the callers.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoFix test case for some sed
Johannes Schindelin [Thu, 2 Mar 2006 11:43:51 +0000 (12:43 +0100)]
Fix test case for some sed

Some versions of sed lack the "-i" option.

Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-branch: add -r switch to list refs/remotes/*
Eric Wong [Thu, 2 Mar 2006 20:23:17 +0000 (12:23 -0800)]
git-branch: add -r switch to list refs/remotes/*

If we decide to use refs/remotes/, having a convenient way to
list them would be nice.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocvsserver: fix checkouts with -d <somedir>
Martin Langhoff [Fri, 3 Mar 2006 03:57:03 +0000 (16:57 +1300)]
cvsserver: fix checkouts with -d <somedir>

A recent Eclipse compat fix broke checkouts with -d. Fix it so that the server
sends the correct module name instead of the destination directory name.

15 years agocvsserver: checkout faster by sending files in a sensible order
Martin Langhoff [Fri, 3 Mar 2006 03:38:03 +0000 (16:38 +1300)]
cvsserver: checkout faster by sending files in a sensible order

Just by sending the files in an ordered fashion, clients can process them
much faster. And we can optimize our check of whether we created this
directory already -- faster.

Timings for a checkout on a commandline cvs client for a project with
~13K files totalling ~100MB:

Unsorted:
  603.12 real        16.89 user        42.88 sys

Sorted:
  298.19 real        26.37 user        42.42 sys

15 years agogit-commit: make sure we protect against races.
Junio C Hamano [Fri, 3 Mar 2006 05:13:24 +0000 (21:13 -0800)]
git-commit: make sure we protect against races.

An earlier commit 8098a178b26dc7a158d129a092a5b78da6d12b72
accidentally lost race protection from git-commit command.
This commit reinstates it.  When something else updates HEAD
pointer while you were editing your commit message, the command
would notice and abort the commit.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-commit --amend
Junio C Hamano [Fri, 3 Mar 2006 05:04:05 +0000 (21:04 -0800)]
git-commit --amend

The new flag is used to amend the tip of the current branch.  Prepare
the tree object you would want to replace the latest commit as usual
(this includes the usual -i/-o and explicit paths), and the commit log
editor is seeded with the commit message from the tip of the current
branch.  The commit you create replaces the current tip -- if it was a
merge, it will have the parents of the current tip as parents -- so the
current top commit is discarded.

It is a rough equivalent for:

$ git reset --soft HEAD^
$ ... do something else to come up with the right tree ...
$ git commit -c ORIG_HEAD

but can be used to amend a merge commit.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoshow-branch --topics
Junio C Hamano [Fri, 3 Mar 2006 01:14:00 +0000 (17:14 -0800)]
show-branch --topics

This adds a new flag, --topics, to help managing topic
branches.  When you have topic branches forked some time ago
from your primary line of development, show-branch would show
many "uninteresting" things that happend on the primary line of
development when trying to see what are still not merged from
the topic branches.

With this flag, the first ref given to show-branch is taken as
the primary branch, and the rest are taken as the topic
branches.  Output from the command is modified so that commits
only on the primary branch are not shown.  In other words,

$ git show-branch --topics master topic1 topic2 ...

shows an (almost) equivalent of

$ git rev-list ^master topic1 topic2 ...

The major differences are that (1) you can tell which commits
are on which branch, and (2) the commit at the fork point is
shown.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoGIT-VERSION-GEN: squelch unneeded error from "cat version"
Junio C Hamano [Thu, 2 Mar 2006 22:38:44 +0000 (14:38 -0800)]
GIT-VERSION-GEN: squelch unneeded error from "cat version"

Now this is really a corner case, but if you have the git source
tree from somewhere other than the official tarball, you do not
have version file.  And if git-describe does not work for you
(maybe you do not have git yet), we spilled an error message
from "cat version".

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'ml/cvsserver'
Junio C Hamano [Thu, 2 Mar 2006 22:19:33 +0000 (14:19 -0800)]
Merge branch 'ml/cvsserver'

* ml/cvsserver:
  cvsserver: Eclipse compat -- now "compare with latest from HEAD" works
  cvsserver: Checkout correctly on Eclipse

15 years agoannotate: resurrect raw timestamps.
Junio C Hamano [Thu, 2 Mar 2006 09:50:09 +0000 (01:50 -0800)]
annotate: resurrect raw timestamps.

For scripted use this is quite useful.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: rev-list --objects-edge
Junio C Hamano [Thu, 2 Mar 2006 09:11:31 +0000 (01:11 -0800)]
Documentation: rev-list --objects-edge

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDocumentation: read-tree --aggressive
Junio C Hamano [Thu, 2 Mar 2006 09:11:05 +0000 (01:11 -0800)]
Documentation: read-tree --aggressive

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agowar on whitespaces: documentation.
Junio C Hamano [Thu, 2 Mar 2006 08:45:48 +0000 (00:45 -0800)]
war on whitespaces: documentation.

We were missing the --whitespace option in the usage string for
git-apply and git-am, so this commit adds them.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: use refs/remotes/git-svn instead of git-svn-HEAD
Eric Wong [Thu, 2 Mar 2006 05:58:31 +0000 (21:58 -0800)]
contrib/git-svn: use refs/remotes/git-svn instead of git-svn-HEAD

After reading a lengthy discussion on the list, I've come to the
conclusion that creating a 'remotes' directory in refs isn't
such a bad idea.

You can still branch from it by specifying remotes/git-svn (not
needing the leading 'refs/'), and the documentation has been
updated to reflect that.

The 'git-svn' part of the ref can of course be set to whatever
you want by using the GIT_SVN_ID environment variable, as
before.

I'm using refs/remotes/git-svn, and not going with something
like refs/remotes/git-svn/HEAD as it's redundant for Subversion
where there's zero distinction between branches and directories.

Run git-svn rebuild --upgrade to upgrade your repository to use
the new head.  git-svn-HEAD must be manually deleted for safety
reasons.

Side note: if you ever (and I hope you never) want to run
git-update-refs on a 'remotes/' ref, make sure you have the
'refs/' prefix as you don't want to be clobbering your
'remotes/' in $GIT_DIR (where remote URLs are stored).

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'maint'
Junio C Hamano [Thu, 2 Mar 2006 07:20:31 +0000 (23:20 -0800)]
Merge branch 'maint'

* maint:
  read-tree --aggressive: remove deleted entry from the working tree.

15 years agoread-tree --aggressive: remove deleted entry from the working tree.
Junio C Hamano [Thu, 2 Mar 2006 07:10:00 +0000 (23:10 -0800)]
read-tree --aggressive: remove deleted entry from the working tree.

When both heads deleted, or our side deleted while the other
side did not touch, we did not have to update the working tree.

However, we forgot to remove existing working tree file when we
did not touch and the other side did.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'jc/tag'
Junio C Hamano [Thu, 2 Mar 2006 06:17:44 +0000 (22:17 -0800)]
Merge branch 'jc/tag'

* jc/tag:
  Pretty-print tagger dates.

15 years agoMerge part of 'jc/diff'
Junio C Hamano [Thu, 2 Mar 2006 06:15:38 +0000 (22:15 -0800)]
Merge part of 'jc/diff'

15 years agoMerge branch 'kh/svnimport'
Junio C Hamano [Thu, 2 Mar 2006 05:46:01 +0000 (21:46 -0800)]
Merge branch 'kh/svnimport'

* kh/svnimport:
  Save username -> Full Name <email@addr.es> map file

15 years agoMerge branch 'js/refs'
Junio C Hamano [Thu, 2 Mar 2006 05:45:56 +0000 (21:45 -0800)]
Merge branch 'js/refs'

* js/refs:
  Warn about invalid refs

15 years agocvsserver: Eclipse compat -- now "compare with latest from HEAD" works
Martin Langhoff [Thu, 2 Mar 2006 05:42:01 +0000 (18:42 +1300)]
cvsserver: Eclipse compat -- now "compare with latest from HEAD" works

The Eclipse client uses cvs update when that menu option is triggered.
And doesn't like the standard cvs update response. Give it *exactly* what
it wants.

And hope the other clients don't lose the plot too badly.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoannotate: fix -S parameter to take a string
Martin Langhoff [Thu, 2 Mar 2006 04:24:30 +0000 (17:24 +1300)]
annotate: fix -S parameter to take a string

In the conversion to Getopt::Long, the -S / --rev-list parameter stopped
working. We need to tell Getopt::Long that it is a string.

As a bonus, the open() now does some useful error handling.

Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoannotate: Add a basic set of test cases.
Ryan Anderson [Thu, 2 Mar 2006 05:23:37 +0000 (00:23 -0500)]
annotate: Add a basic set of test cases.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoannotate: handle \No newline at end of file.
Ryan Anderson [Thu, 2 Mar 2006 05:23:37 +0000 (00:23 -0500)]
annotate: handle \No newline at end of file.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Use horizontal scroll bar in the tree view
Aneesh Kumar K.V [Thu, 2 Mar 2006 04:29:53 +0000 (09:59 +0530)]
gitview: Use horizontal scroll bar in the tree view

Earlier we set up the window to never scroll
horizontally, which made it harder to use on a narrow screen.
This patch allows scrollbar to be used as needed by Gtk

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocvsserver: Checkout correctly on Eclipse
Martin Langhoff [Thu, 2 Mar 2006 00:58:57 +0000 (13:58 +1300)]
cvsserver: Checkout correctly on Eclipse

Initial checkouts were failing to create Entries files under Eclipse.
Eclipse was waiting for two non-standard directory-resets to prepare for a new
directory from the server.

This patch is tricky, because the same directory resets tend to confuse other
clients. It's taken a bit of fiddling to get the commandline cvs client and
Eclipse to get a good, clean checkout.

Signed-off-by: Martin Langhoff <martin@catalyst.net.nz>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoRe-fix compilation warnings.
Luck, Tony [Wed, 1 Mar 2006 23:01:53 +0000 (15:01 -0800)]
Re-fix compilation warnings.

Commit 8fcf1ad9c68e15d881194c8544e7c11d33529c2b has a
combination of double cast and Andreas' switch to using
unsigned long ... just the latter is sufficient (and a lot less
ugly than using the double cast).

Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoUp to date with GIT 1.2.4 fixes
Junio C Hamano [Thu, 2 Mar 2006 01:07:42 +0000 (17:07 -0800)]
Up to date with GIT 1.2.4 fixes

15 years agoPretty-print tagger dates.
Junio C Hamano [Thu, 2 Mar 2006 00:43:19 +0000 (16:43 -0800)]
Pretty-print tagger dates.

We can show commit objects with human readable dates using
various --pretty options, but there was no way to do so with
tags.  This introduces two such ways:

$ git-cat-file -p v1.2.3

shows the tag object with tagger dates in human readable format.

$ git-verify-tag --verbose v1.2.3

uses it to show the contents of the tag object as well as doing
GPG verification.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'lt/fix-apply' into maint v1.2.4
Junio C Hamano [Thu, 2 Mar 2006 01:06:12 +0000 (17:06 -0800)]
Merge branch 'lt/fix-apply' into maint

* lt/fix-apply:
  git-am: --whitespace=x option.
  git-apply: war on whitespace -- finishing touches.
  git-apply --whitespace=nowarn
  apply --whitespace: configuration option.
  apply: squelch excessive errors and --whitespace=error-all
  apply --whitespace fixes and enhancements.
  The war on trailing whitespace

15 years agoMerge branch 'lt/apply'
Junio C Hamano [Thu, 2 Mar 2006 01:05:57 +0000 (17:05 -0800)]
Merge branch 'lt/apply'

* lt/apply:
  git-am: --whitespace=x option.
  git-apply: war on whitespace -- finishing touches.
  git-apply --whitespace=nowarn
  apply --whitespace: configuration option.
  apply: squelch excessive errors and --whitespace=error-all
  apply --whitespace fixes and enhancements.
  The war on trailing whitespace

15 years agoMerge early parts of 'np/delta' branch
Junio C Hamano [Thu, 2 Mar 2006 00:55:51 +0000 (16:55 -0800)]
Merge early parts of 'np/delta' branch

15 years agoMerge git-mv fixes from 'maint'
Junio C Hamano [Wed, 1 Mar 2006 20:16:25 +0000 (12:16 -0800)]
Merge git-mv fixes from 'maint'

15 years agogit-mv: fixes for path handling
Josef Weidendorfer [Wed, 1 Mar 2006 18:09:23 +0000 (19:09 +0100)]
git-mv: fixes for path handling

Moving a directory ending in a slash was not working as the
destination was not calculated correctly.
E.g. in the git repo,

 git-mv t/ Documentation

gave the error

 Error: destination 'Documentation' already exists

To get rid of this problem, strip trailing slashes from all arguments.
The comment in cg-mv made me curious about this issue; Pasky, thanks!
As result, the workaround in cg-mv is not needed any more.

Also, another bug was shown by cg-mv. When moving files outside of
a subdirectory, it typically calls git-mv with something like

 git-mv Documentation/git.txt Documentation/../git-mv.txt

which triggers the following error from git-update-index:

 Ignoring path Documentation/../git-mv.txt

The result is a moved file, removed from git revisioning, but not
added again. To fix this, the paths have to be normalized not have ".."
in the middle. This was already done in git-mv, but only for
a better visual appearance :(

Signed-off-by: Josef Weidendorfer <Josef.Weidendorfer@gmx.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-mv: Allow -h without repo & fix error message
Josef Weidendorfer [Wed, 1 Mar 2006 17:16:36 +0000 (18:16 +0100)]
git-mv: Allow -h without repo & fix error message

This fixes "git-mv -h" to output the usage without the need
to be in a git repository.
Additionally:
- fix confusing error message when only one arg was given
- fix typo in error message

Signed-off-by: Josef Weidendorfer <Josef.Weidendorfer@gmx.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoAllow git-mv to accept ./ in paths.
Junio C Hamano [Sun, 19 Feb 2006 07:42:03 +0000 (23:42 -0800)]
Allow git-mv to accept ./ in paths.

Signed-off-by: Junio C Hamano <junkio@cox.net>
(cherry picked from 9a0e6731c632c841cd2de9dec0b9091b2f10c6fd commit)

15 years agocombine-diff: Honour -z option correctly.
Mark Wooding [Mon, 27 Feb 2006 12:52:52 +0000 (12:52 +0000)]
combine-diff: Honour -z option correctly.

Combined diffs don't null terminate things in the same way as standard
diffs.  This is presumably wrong.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
(cherry picked from 6baf0484efcd29bb5e58ccd5ea0379481d4a83f4 commit)

15 years agocombine-diff: Honour --full-index.
Mark Wooding [Mon, 27 Feb 2006 12:52:50 +0000 (12:52 +0000)]
combine-diff: Honour --full-index.

For some reason, combined diffs don't honour the --full-index flag when
emitting patches.  Fix this.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
(cherry picked from e70c6b35749c316f6e97099bd6bdac895c9d6f68 commit)

15 years agodiffcore-break: micro-optimize by avoiding delta between identical files.
Junio C Hamano [Wed, 1 Mar 2006 04:19:47 +0000 (20:19 -0800)]
diffcore-break: micro-optimize by avoiding delta between identical files.

We did not check if we have the same file on both sides when
computing break score.  This is usually not a problem, but if
the user said --find-copies-harde with -B, we ended up trying a
delta between the same data even when we know the SHA1 hash of
both sides match.

Signed-off-by: Junio C Hamano <junkio@cox.net>
(cherry picked from aeecd23ae2785a0462d42191974e9d9a8e439fbe commit)

15 years agoWarn about invalid refs
Johannes Schindelin [Tue, 28 Feb 2006 21:16:01 +0000 (22:16 +0100)]
Warn about invalid refs

Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocvsserver: Eclipse compat - browsing 'modules' (heads in our case) works
Martin Langhoff [Wed, 1 Mar 2006 07:03:58 +0000 (20:03 +1300)]
cvsserver: Eclipse compat - browsing 'modules' (heads in our case) works

Eclipse CVS clients have an odd way of perusing the top level of
the repository, by calling update on module "". So reproduce cvs'
odd behaviour in the interest of compatibility.

It makes it much easier to get a checkout when using Eclipse.

15 years agocvsserver: Eclipse compat fixes - implement Questionable, alias rlog, add a space...
Martin Langhoff [Wed, 1 Mar 2006 06:30:35 +0000 (19:30 +1300)]
cvsserver: Eclipse compat fixes - implement Questionable, alias rlog, add a space after the U

A few things to satisfy Eclipse's strange habits as a cvs client:

- Implement Questionable
- Aliased rlog to log, but more work may be needed
- Add a space after the U that indicates updated

15 years agocvsserver: add notes on how to get a checkout under Eclipse
Martin Langhoff [Thu, 23 Feb 2006 23:20:51 +0000 (12:20 +1300)]
cvsserver: add notes on how to get a checkout under Eclipse

15 years agogit-am: --whitespace=x option.
Junio C Hamano [Wed, 1 Mar 2006 04:26:25 +0000 (20:26 -0800)]
git-am: --whitespace=x option.

This is passed down to git-apply to override the built-in
default and per-repository configuration at runtime.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-am: --whitespace=x option.
Junio C Hamano [Wed, 1 Mar 2006 04:26:25 +0000 (20:26 -0800)]
git-am: --whitespace=x option.

This is passed down to git-apply to override the built-in
default and per-repository configuration at runtime.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agodiffcore-rename: split out the delta counting code.
Junio C Hamano [Wed, 1 Mar 2006 00:01:36 +0000 (16:01 -0800)]
diffcore-rename: split out the delta counting code.

This is to rework diffcore break/rename/copy detection code
so that it does not affected when deltifier code gets improved.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agodiffcore-break: micro-optimize by avoiding delta between identical files.
Junio C Hamano [Wed, 1 Mar 2006 04:19:47 +0000 (20:19 -0800)]
diffcore-break: micro-optimize by avoiding delta between identical files.

We did not check if we have the same file on both sides when
computing break score.  This is usually not a problem, but if
the user said --find-copies-harde with -B, we ended up trying a
delta between the same data even when we know the SHA1 hash of
both sides match.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoDarwin: Ignore missing /sw/lib
Shawn Pearce [Tue, 28 Feb 2006 14:03:48 +0000 (09:03 -0500)]
Darwin: Ignore missing /sw/lib

When on Darwin platforms don't include Fink or DarwinPorts
into the link path unless the related library directory
is actually present.  The linker on MacOS 10.4 complains
if it is given a directory which does not exist.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Set the default width of graph cell
Aneesh Kumar K.V [Tue, 28 Feb 2006 14:40:28 +0000 (20:10 +0530)]
gitview: Set the default width of graph cell

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Some window layout changes.
Aneesh Kumar K.V [Tue, 28 Feb 2006 13:42:18 +0000 (19:12 +0530)]
gitview: Some window layout changes.

This makes menubar look nice

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Select the text color based on whether the entry in highlighted. Use standar...
Pavel Roskin [Tue, 28 Feb 2006 13:41:08 +0000 (19:11 +0530)]
gitview: Select the text color based on whether the entry in highlighted. Use standard font.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-apply: war on whitespace -- finishing touches.
Junio C Hamano [Tue, 28 Feb 2006 09:12:52 +0000 (01:12 -0800)]
git-apply: war on whitespace -- finishing touches.

This changes the default --whitespace policy to nowarn when we
are only getting --stat, --summary etc. IOW when not applying
the patch.  When applying the patch, the default is warn (spit
out warning message but apply the patch).

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-apply: war on whitespace -- finishing touches.
Junio C Hamano [Tue, 28 Feb 2006 09:12:52 +0000 (01:12 -0800)]
git-apply: war on whitespace -- finishing touches.

This changes the default --whitespace policy to nowarn when we
are only getting --stat, --summary etc. IOW when not applying
the patch.  When applying the patch, the default is warn (spit
out warning message but apply the patch).

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-apply --whitespace=nowarn
Junio C Hamano [Tue, 28 Feb 2006 01:07:16 +0000 (17:07 -0800)]
git-apply --whitespace=nowarn

Andrew insists --whitespace=warn should be the default, and I
tend to agree.  This introduces --whitespace=warn, so if your
project policy is more lenient, you can squelch them by having
apply.whitespace=nowarn in your configuration file.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply --whitespace: configuration option.
Junio C Hamano [Mon, 27 Feb 2006 22:47:45 +0000 (14:47 -0800)]
apply --whitespace: configuration option.

The new configuration option apply.whitespace can take one of
"warn", "error", "error-all", or "strip".  When git-apply is run
to apply the patch to the index, they are used as the default
value if there is no command line --whitespace option.

Andrew can now tell people who feed him git trees to update to
this version and say:

git repo-config apply.whitespace error

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply: squelch excessive errors and --whitespace=error-all
Junio C Hamano [Mon, 27 Feb 2006 22:16:30 +0000 (14:16 -0800)]
apply: squelch excessive errors and --whitespace=error-all

This by default makes --whitespace=warn, error, and strip to
warn only the first 5 additions of trailing whitespaces.  A new
option --whitespace=error-all can be used to view all of them
before applying.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply --whitespace fixes and enhancements.
Junio C Hamano [Mon, 27 Feb 2006 02:13:25 +0000 (18:13 -0800)]
apply --whitespace fixes and enhancements.

In addition to fixing obvious command line parsing bugs in the
previous round, this changes the following:

 * Adds "--whitespace=strip".  This applies after stripping the
   new trailing whitespaces introduced to the patch.

 * The output error message format is changed to say
   "patch-filename:linenumber:contents of the line".  This makes
   it similar to typical compiler error message format, and
   helps C-x ` (next-error) in Emacs compilation buffer.

 * --whitespace=error and --whitespace=warn do not stop at the
   first error.  We might want to limit the output to say first
   20 such lines to prevent cluttering, but on the other hand if
   you are willing to hand-fix after inspecting them, getting
   everything with a single run might be easier to work with.
   After all, somebody has to do the clean-up work somewhere.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoThe war on trailing whitespace
Linus Torvalds [Sun, 26 Feb 2006 17:29:00 +0000 (09:29 -0800)]
The war on trailing whitespace

On Sat, 25 Feb 2006, Andrew Morton wrote:
>
> I'd suggest a) git will simply refuse to apply such a patch unless given a
> special `forcing' flag, b) even when thus forced, it will still warn and c)
> with a different flag, it will strip-then-apply, without generating a
> warning.

This doesn't do the "strip-then-apply" thing, but it allows you to make
git-apply generate a warning or error on extraneous whitespace.

Use --whitespace=warn to warn, and (surprise, surprise) --whitespace=error
to make it a fatal error to have whitespace at the end.

Totally untested, of course. But it compiles, so it must be fine.

HOWEVER! Note that this literally will check every single patch-line with
"+" at the beginning. Which means that if you fix a simple typo, and the
line had a space at the end before, and you didn't remove it, that's still
considered a "new line with whitespace at the end", even though obviously
the line wasn't really new.

I assume this is what you wanted, and there isn't really any sane
alternatives (you could make the warning activate only for _pure_
additions with no deletions at all in that hunk, but that sounds a bit
insane).

Linus

15 years agogit-apply --whitespace=nowarn
Junio C Hamano [Tue, 28 Feb 2006 01:07:16 +0000 (17:07 -0800)]
git-apply --whitespace=nowarn

Andrew insists --whitespace=warn should be the default, and I
tend to agree.  This introduces --whitespace=warn, so if your
project policy is more lenient, you can squelch them by having
apply.whitespace=nowarn in your configuration file.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge part of kh/svnimport branch into master
Junio C Hamano [Mon, 27 Feb 2006 23:46:39 +0000 (15:46 -0800)]
Merge part of kh/svnimport branch into master

15 years agocontrib/git-svn: correct commit example in manpage
Eric Wong [Mon, 27 Feb 2006 20:55:45 +0000 (12:55 -0800)]
contrib/git-svn: correct commit example in manpage

Thanks to Nicolas Vilz <niv@iaglans.de> for noticing this.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoSave username -> Full Name <email@addr.es> map file
Karl Hasselström [Mon, 27 Feb 2006 23:08:19 +0000 (00:08 +0100)]
Save username -> Full Name <email@addr.es> map file

When the user specifies a username -> Full Name <email@addr.es> map
file with the -A option, save a copy of that file as
$git_dir/svn-authors. When running git-svnimport with an existing GIT
directory, use $git_dir/svn-authors (if it exists) unless a file was
explicitly specified with -A.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoLet git-svnimport's author file use same syntax as git-cvsimport's
Karl Hasselström [Mon, 27 Feb 2006 23:08:15 +0000 (00:08 +0100)]
Let git-svnimport's author file use same syntax as git-cvsimport's

git-cvsimport uses a username => Full Name <email@addr.es> mapping
file with this syntax:

  kha=Karl Hasselström <kha@treskal.com>

Since there is no reason to use another format for git-svnimport, use
the same format.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply --whitespace: configuration option.
Junio C Hamano [Mon, 27 Feb 2006 22:47:45 +0000 (14:47 -0800)]
apply --whitespace: configuration option.

The new configuration option apply.whitespace can take one of
"warn", "error", "error-all", or "strip".  When git-apply is run
to apply the patch to the index, they are used as the default
value if there is no command line --whitespace option.

Andrew can now tell people who feed him git trees to update to
this version and say:

git repo-config apply.whitespace error

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply: squelch excessive errors and --whitespace=error-all
Junio C Hamano [Mon, 27 Feb 2006 22:16:30 +0000 (14:16 -0800)]
apply: squelch excessive errors and --whitespace=error-all

This by default makes --whitespace=warn, error, and strip to
warn only the first 5 additions of trailing whitespaces.  A new
option --whitespace=error-all can be used to view all of them
before applying.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocontrib/git-svn: tell the user to not modify git-svn-HEAD directly
Eric Wong [Mon, 27 Feb 2006 19:04:02 +0000 (11:04 -0800)]
contrib/git-svn: tell the user to not modify git-svn-HEAD directly

As a rule, interface branches to different SCMs should never be modified
directly by the user.  They are used exclusively for talking to the
foreign SCM.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Remove trailing white space
Aneesh Kumar K.V [Mon, 27 Feb 2006 17:12:07 +0000 (22:42 +0530)]
gitview: Remove trailing white space

Do the cleanup using Dave jones vim script

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogitview: Fix the encoding related bug
Aneesh Kumar K.V [Mon, 27 Feb 2006 15:55:13 +0000 (21:25 +0530)]
gitview: Fix the encoding related bug

Get the encoding information from repository and convert it to utf-8 before
passing to gtk.TextBuffer.set_text. gtk.TextBuffer.set_text work only with utf-8

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agogit-format-patch: Always add a blank line between headers and body.
Alexandre Julliard [Mon, 27 Feb 2006 13:09:56 +0000 (14:09 +0100)]
git-format-patch: Always add a blank line between headers and body.

If the second line of the commit message isn't empty, git-format-patch
needs to add an empty line in order to generate a properly formatted
mail. Otherwise git-rebase drops the rest of the commit message.

Signed-off-by: Alexandre Julliard <julliard@winehq.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocombine-diff: Honour -z option correctly.
Mark Wooding [Mon, 27 Feb 2006 12:52:52 +0000 (12:52 +0000)]
combine-diff: Honour -z option correctly.

Combined diffs don't null terminate things in the same way as standard
diffs.  This is presumably wrong.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agocombine-diff: Honour --full-index.
Mark Wooding [Mon, 27 Feb 2006 12:52:50 +0000 (12:52 +0000)]
combine-diff: Honour --full-index.

For some reason, combined diffs don't honour the --full-index flag when
emitting patches.  Fix this.

Signed-off-by: Mark Wooding <mdw@distorted.org.uk>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoapply --whitespace fixes and enhancements.
Junio C Hamano [Mon, 27 Feb 2006 02:13:25 +0000 (18:13 -0800)]
apply --whitespace fixes and enhancements.

In addition to fixing obvious command line parsing bugs in the
previous round, this changes the following:

 * Adds "--whitespace=strip".  This applies after stripping the
   new trailing whitespaces introduced to the patch.

 * The output error message format is changed to say
   "patch-filename:linenumber:contents of the line".  This makes
   it similar to typical compiler error message format, and
   helps C-x ` (next-error) in Emacs compilation buffer.

 * --whitespace=error and --whitespace=warn do not stop at the
   first error.  We might want to limit the output to say first
   20 such lines to prevent cluttering, but on the other hand if
   you are willing to hand-fix after inspecting them, getting
   everything with a single run might be easier to work with.
   After all, somebody has to do the clean-up work somewhere.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoThe war on trailing whitespace
Linus Torvalds [Sun, 26 Feb 2006 17:29:00 +0000 (09:29 -0800)]
The war on trailing whitespace

On Sat, 25 Feb 2006, Andrew Morton wrote:
>
> I'd suggest a) git will simply refuse to apply such a patch unless given a
> special `forcing' flag, b) even when thus forced, it will still warn and c)
> with a different flag, it will strip-then-apply, without generating a
> warning.

This doesn't do the "strip-then-apply" thing, but it allows you to make
git-apply generate a warning or error on extraneous whitespace.

Use --whitespace=warn to warn, and (surprise, surprise) --whitespace=error
to make it a fatal error to have whitespace at the end.

Totally untested, of course. But it compiles, so it must be fine.

HOWEVER! Note that this literally will check every single patch-line with
"+" at the beginning. Which means that if you fix a simple typo, and the
line had a space at the end before, and you didn't remove it, that's still
considered a "new line with whitespace at the end", even though obviously
the line wasn't really new.

I assume this is what you wanted, and there isn't really any sane
alternatives (you could make the warning activate only for _pure_
additions with no deletions at all in that hunk, but that sounds a bit
insane).

Linus

15 years agosvnimport: Read author names and emails from a file
Karl Hasselström [Sun, 26 Feb 2006 05:11:31 +0000 (06:11 +0100)]
svnimport: Read author names and emails from a file

Read a file with lines on the form

  username User's Full Name <email@addres.org>

and use "User's Full Name <email@addres.org>" as the GIT author and
committer for Subversion commits made by "username". If encountering a
commit made by a user not in the list, abort.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agosvnimport: Convert the svn:ignore property
Karl Hasselström [Sun, 26 Feb 2006 05:11:29 +0000 (06:11 +0100)]
svnimport: Convert the svn:ignore property

Put the value of the svn:ignore property in a regular file when
converting a Subversion repository to GIT. The Subversion and GIT
ignore syntaxes are similar enough that it often just works to set the
filename to .gitignore and do nothing else.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agosvnimport: Convert executable flag
Karl Hasselström [Sun, 26 Feb 2006 05:11:27 +0000 (06:11 +0100)]
svnimport: Convert executable flag

Convert the svn:executable property to file mode 755 when converting
an SVN repository to GIT.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agosvnimport: Mention -r in usage summary
Karl Hasselström [Sun, 26 Feb 2006 05:11:24 +0000 (06:11 +0100)]
svnimport: Mention -r in usage summary

I added the -r option to git-svnimport some time ago, but forgot to
update the usage summary in the documentation.

Signed-off-by: Karl Hasselström <kha@treskal.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMake git diff-generation use a simpler spawn-like interface
Linus Torvalds [Sun, 26 Feb 2006 23:51:24 +0000 (15:51 -0800)]
Make git diff-generation use a simpler spawn-like interface

Instead of depending of fork() and execve() and doing things in between
the two, make the git diff functions do everything up front, and then do
a single "spawn_prog()" invocation to run the actual external diff
program (if any is even needed).

This actually ends up simplifying the code, and should make it much
easier to make it efficient under broken operating systems (read: Windows).

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoMerge branch 'maint'
Junio C Hamano [Sun, 26 Feb 2006 23:25:52 +0000 (15:25 -0800)]
Merge branch 'maint'

* maint:
  sample hooks template.

15 years agosample hooks template.
Junio C Hamano [Sun, 26 Feb 2006 23:16:41 +0000 (15:16 -0800)]
sample hooks template.

These two sample hooks try to detect and use the corresponding
commit hook from the same repository.  However, they forgot to
set up GIT_DIR for their own use, so was not in effect.

Signed-off-by: Junio C Hamano <junkio@cox.net>
15 years agoTeach the "git" command to handle some commands internally
Linus Torvalds [Sun, 26 Feb 2006 20:34:51 +0000 (12:34 -0800)]
Teach the "git" command to handle some commands internally

This is another patch in the "prepare to do more in C" series, where the
git wrapper command is taught about the notion of handling some
functionality internally.

Right now, the only internal commands are "version" and "help", but the
point being that we can now easily extend it to handle some of the trivial
scripts internally. Things like "git log" and "git diff" wouldn't need
separate external scripts any more.

This also implies that to support the old "git-log" and "git-diff" syntax,
the "git" wrapper now automatically looks at the name it was executed as,
and if it is "git-xxxx", it will assume that it is to internally do what
"git xxxx" would do.

In other words, you can (once you implement an internal command) soft- or
hard-link that command to the "git" wrapper command, and it will do the
right thing, whether you use the "git xxxx" or the "git-xxxx" format.

There's one other change: the search order for external programs is
modified slightly, so that the first entry remains GIT_EXEC_DIR, but the
second entry is the same directory as the git wrapper itself was executed
out of - if we can figure it out from argv[0], of course.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>