git.git
18 years agoMerge part of np/delta
Junio C Hamano [Tue, 28 Feb 2006 09:15:29 +0000 (01:15 -0800)]
Merge part of np/delta

18 years agoMerge branch 'lt/apply' into next
Junio C Hamano [Tue, 28 Feb 2006 09:14:56 +0000 (01:14 -0800)]
Merge branch 'lt/apply' into next

* lt/apply:
  git-apply: war on whitespace -- finishing touches.
  git-apply --whitespace=nowarn

18 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>
18 years agodiff-delta: bound hash list length to avoid O(m*n) behavior
Nicolas Pitre [Tue, 28 Feb 2006 04:09:55 +0000 (23:09 -0500)]
diff-delta: bound hash list length to avoid O(m*n) behavior

The diff-delta code can exhibit O(m*n) behavior with some patological
data set where most hash entries end up in the same hash bucket.

The latest code rework reduced the block size making it particularly
vulnerable to this issue, but the issue was always there and can be
triggered regardless of the block size.

This patch does two things:

1) the hashing has been reworked to offer a better distribution to
   atenuate the problem a bit, and

2) a limit is imposed to the number of entries that can exist in the
   same hash bucket.

Because of the above the code is a bit more expensive on average, but
the problematic samples used to diagnoze the issue are now orders of
magnitude less expensive to process with only a slight loss in
compression.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoRevert "Revert "diff-delta: produce optimal pack data""
Junio C Hamano [Tue, 28 Feb 2006 05:37:56 +0000 (21:37 -0800)]
Revert "Revert "diff-delta: produce optimal pack data""

18 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>
18 years agoMerge branch 'master' into next
Junio C Hamano [Mon, 27 Feb 2006 23:54:36 +0000 (15:54 -0800)]
Merge branch 'master' into next

* master:
  Merge part of kh/svnimport branch into master
  contrib/git-svn: correct commit example in manpage
  contrib/git-svn: tell the user to not modify git-svn-HEAD directly
  gitview: Remove trailing white space
  gitview: Fix the encoding related bug
  git-format-patch: Always add a blank line between headers and body.
  combine-diff: Honour -z option correctly.
  combine-diff: Honour --full-index.

18 years agoMerge branch 'lt/rev-list' into next
Junio C Hamano [Mon, 27 Feb 2006 23:48:17 +0000 (15:48 -0800)]
Merge branch 'lt/rev-list' into next

* lt/rev-list:
  Splitting rev-list into revisions lib, end of beginning.

18 years agoMerge branch 'lt/apply' into next
Junio C Hamano [Mon, 27 Feb 2006 23:48:13 +0000 (15:48 -0800)]
Merge branch 'lt/apply' into next

* lt/apply:
  apply --whitespace: configuration option.
  apply: squelch excessive errors and --whitespace=error-all

18 years agoMerge branch 'kh/svnimport' into next
Junio C Hamano [Mon, 27 Feb 2006 23:48:06 +0000 (15:48 -0800)]
Merge branch 'kh/svnimport' into next

* kh/svnimport:
  Save username -> Full Name <email@addr.es> map file
  Let git-svnimport's author file use same syntax as git-cvsimport's

18 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

18 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>
18 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>
18 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>
18 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>
18 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>
18 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>
18 years agoSplitting rev-list into revisions lib, end of beginning.
Linus Torvalds [Mon, 27 Feb 2006 16:54:36 +0000 (08:54 -0800)]
Splitting rev-list into revisions lib, end of beginning.

This makes the rewrite easier to validate in that revision flag
parsing and warlking part are now all in rev_info structure.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 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>
18 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>
18 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>
18 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>
18 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>
18 years agoMerge branch 'kh/svnimport' into next
Junio C Hamano [Mon, 27 Feb 2006 05:55:16 +0000 (21:55 -0800)]
Merge branch 'kh/svnimport' into next

* kh/svnimport:
  svnimport: Read author names and emails from a file

18 years agoMerge branch 'lt/apply' into next
Junio C Hamano [Mon, 27 Feb 2006 05:55:08 +0000 (21:55 -0800)]
Merge branch 'lt/apply' into next

* lt/apply:
  apply --whitespace fixes and enhancements.
  The war on trailing whitespace
  svnimport: Convert the svn:ignore property
  svnimport: Convert executable flag
  svnimport: Mention -r in usage summary
  Make git diff-generation use a simpler spawn-like interface

18 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>
18 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

18 years agoMerge branch 'lt/rev-list' into next
Junio C Hamano [Mon, 27 Feb 2006 05:53:56 +0000 (21:53 -0800)]
Merge branch 'lt/rev-list' into next

* lt/rev-list:
  rev-list split: minimum fixup.

18 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>
18 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>
18 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>
18 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>
18 years agorev-list split: minimum fixup.
Junio C Hamano [Mon, 27 Feb 2006 05:19:14 +0000 (21:19 -0800)]
rev-list split: minimum fixup.

This fixes "the other end has commit X but since then we tagged
that commit with tag T, and he says he wants T -- what is the
list of objects we need to send him?" question:

git-rev-list --objects ^X T

We ended up sending everything since the beginning of time X-<.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 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>
18 years agoMerge branch 'lt/rev-list' into next
Junio C Hamano [Sun, 26 Feb 2006 23:33:49 +0000 (15:33 -0800)]
Merge branch 'lt/rev-list' into next

* lt/rev-list:
  First cut at libifying revlist generation
  Merge branch 'maint'
  sample hooks template.
  Teach the "git" command to handle some commands internally
  Use setenv(), fix warnings
  contrib/git-svn: version 0.10.0
  contrib/git-svn: optimize sequential commits to svn
  contrib/git-svn: add show-ignore command
  annotate: Use qx{} for pipes on activestate.
  annotate: Convert all -| calls to use a helper open_pipe().
  annotate: Handle dirty state and arbitrary revisions.
  git-fetch: print the new and old ref when fast-forwarding

18 years agoFirst cut at libifying revlist generation
Linus Torvalds [Sun, 26 Feb 2006 00:19:46 +0000 (16:19 -0800)]
First cut at libifying revlist generation

This really just splits things up partially, and creates the
interface to set things up by parsing the command line.

No real code changes so far, although the parsing of filenames is a bit
stricter. In particular, if there is a "--", then we do not accept any
filenames before it, and if there isn't any "--", then we check that _all_
paths listed are valid, not just the first one.

The new argument parsing automatically also gives us "--default" and
"--not" handling as in git-rev-parse.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 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.

18 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>
18 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>
18 years agoUse setenv(), fix warnings
Timo Hirvonen [Sun, 26 Feb 2006 15:13:46 +0000 (17:13 +0200)]
Use setenv(), fix warnings

  - Fix -Wundef -Wold-style-definition warnings
  - Make pll_free() static

[jc: original patch by Timo had another unrelated bits:

  - Use setenv() instead of putenv()

 I'm postponing that part for now.]

Signed-off-by: Timo Hirvonen <tihirvon@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocontrib/git-svn: version 0.10.0
Eric Wong [Sun, 26 Feb 2006 10:22:27 +0000 (02:22 -0800)]
contrib/git-svn: version 0.10.0

New features deserve an increment of the minor version.  This will very
likely become 1.0.0 unless release-critical bugs are found.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocontrib/git-svn: optimize sequential commits to svn
Eric Wong [Sun, 26 Feb 2006 10:22:27 +0000 (02:22 -0800)]
contrib/git-svn: optimize sequential commits to svn

Avoid running 'svn up' to a previous revision if we know the
revision we just committed is the first descendant of the
revision we came from.

This reduces the time to do a series of commits by about 25%.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocontrib/git-svn: add show-ignore command
Eric Wong [Sun, 26 Feb 2006 10:22:27 +0000 (02:22 -0800)]
contrib/git-svn: add show-ignore command

Recursively finds and lists the svn:ignore property on
directories.  The output is suitable for appending to the
$GIT_DIR/info/exclude file.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoannotate: Use qx{} for pipes on activestate.
Ryan Anderson [Sun, 26 Feb 2006 21:09:12 +0000 (16:09 -0500)]
annotate: Use qx{} for pipes on activestate.

Note: This needs someone to tell me what the value of $^O is on ActiveState.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoannotate: Convert all -| calls to use a helper open_pipe().
Ryan Anderson [Sun, 26 Feb 2006 03:02:05 +0000 (22:02 -0500)]
annotate: Convert all -| calls to use a helper open_pipe().

When we settle on a solution for ActiveState's forking issues, all
compatibility checks can be handled inside this one function.

Also, fixed an abuse of global variables in the process of cleaning this up.

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoannotate: Handle dirty state and arbitrary revisions.
Ryan Anderson [Sun, 26 Feb 2006 01:48:33 +0000 (20:48 -0500)]
annotate: Handle dirty state and arbitrary revisions.

Also, use Getopt::Long and only process each rev once.

(Thanks to Morten Welinder for spotting the performance problems.)

Signed-off-by: Ryan Anderson <ryan@michonline.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-fetch: print the new and old ref when fast-forwarding
Lukas Sandström [Sat, 25 Feb 2006 11:20:13 +0000 (12:20 +0100)]
git-fetch: print the new and old ref when fast-forwarding

Signed-off-by: Lukas Sandström <lukass@etek.chalmers.se>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'master' into next
Junio C Hamano [Sat, 25 Feb 2006 07:47:48 +0000 (23:47 -0800)]
Merge branch 'master' into next

* master:
  fix warning from pack-objects.c
  Merge branches 'jc/rev-list' and 'jc/pack-thin'
  gitview: Fix the graph display .

18 years agofix warning from pack-objects.c
Luck, Tony [Thu, 23 Feb 2006 22:42:39 +0000 (14:42 -0800)]
fix warning from pack-objects.c

When compiling on ia64 I get this warning (from gcc 3.4.3):

gcc -o pack-objects.o -c -g -O2 -Wall -DSHA1_HEADER='<openssl/sha.h>'  pack-objects.c
pack-objects.c: In function `pack_revindex_ix':
pack-objects.c:94: warning: cast from pointer to integer of different size

A double cast (first to long, then to int) shuts gcc up, but is there
a better way?

[jc: Andreas Ericsson suggests to use ulong instead. ]

Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branches 'jc/rev-list' and 'jc/pack-thin'
Junio C Hamano [Sat, 25 Feb 2006 05:55:23 +0000 (21:55 -0800)]
Merge branches 'jc/rev-list' and 'jc/pack-thin'

* jc/rev-list:
  rev-list --objects: use full pathname to help hashing.
  rev-list --objects-edge: remove duplicated edge commit output.
  rev-list --objects-edge

* jc/pack-thin:
  pack-objects: hash basename and direname a bit differently.
  pack-objects: allow "thin" packs to exceed depth limits
  pack-objects: use full pathname to help hashing with "thin" pack.
  pack-objects: thin pack micro-optimization.
  Use thin pack transfer in "git fetch".
  Add git-push --thin.
  send-pack --thin: use "thin pack" delta transfer.
  Thin pack - create packfile with missing delta base.

Conflicts:

pack-objects.c (taking "next")
send-pack.c (taking "next")

18 years agoMerge branch 'master' into next
Junio C Hamano [Sat, 25 Feb 2006 03:01:02 +0000 (19:01 -0800)]
Merge branch 'master' into next

* master:
  Merge branches 'jc/rev-list' and 'jc/pack-thin'
  gitview: Code cleanup
  Add missing programs to ignore list
  git ls files recursively show ignored files
  Build and install git-mailinfo.
  gitview: Bump the rev
  gitview: Fix DeprecationWarning

18 years agoMerge branches 'jc/rev-list' and 'jc/pack-thin'
Junio C Hamano [Sat, 25 Feb 2006 02:55:25 +0000 (18:55 -0800)]
Merge branches 'jc/rev-list' and 'jc/pack-thin'

* jc/rev-list:
  rev-list --objects: use full pathname to help hashing.
  rev-list --objects-edge: remove duplicated edge commit output.
  rev-list --objects-edge

* jc/pack-thin:
  pack-objects: hash basename and direname a bit differently.
  pack-objects: allow "thin" packs to exceed depth limits
  pack-objects: use full pathname to help hashing with "thin" pack.
  pack-objects: thin pack micro-optimization.
  Use thin pack transfer in "git fetch".
  Add git-push --thin.
  send-pack --thin: use "thin pack" delta transfer.
  Thin pack - create packfile with missing delta base.

Conflicts:

pack-objects.c (manual adjustment for thin pack needed)
send-pack.c

18 years agogitview: Fix the graph display .
Aneesh Kumar K.V [Fri, 24 Feb 2006 16:27:51 +0000 (21:57 +0530)]
gitview: Fix the graph display .

This fix all the known issue with the graph display
The bug need to be explained graphically

                                 |
                                 a
This line need not be there ---->| \
                                 b  |
                                 | /
                                 c

c is parent of a and all a,b and c are placed on the same line and b is child of c
With my last checkin I added  a seperate line to indicate that a is
connected to c. But then we had the line connecting a and b which should
not be ther. This changes fixes the same bug

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogitview: Code cleanup
Aneesh Kumar K.V [Fri, 24 Feb 2006 16:19:54 +0000 (21:49 +0530)]
gitview: Code cleanup

Rearrange the code little bit so that it is easier to read

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoAdd missing programs to ignore list
Shawn Pearce [Fri, 24 Feb 2006 22:51:15 +0000 (17:51 -0500)]
Add missing programs to ignore list

Added recently added programs to the default exclude list.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit ls files recursively show ignored files
Shawn Pearce [Fri, 24 Feb 2006 22:02:34 +0000 (17:02 -0500)]
git ls files recursively show ignored files

Make git-ls-files --others --ignored recurse into non-excluded
subdirectories.

Typically when asking git-ls-files to display all files which are
ignored by one or more exclude patterns one would want it to recurse
into subdirectories which are not themselves excluded to see if
there are any excluded files contained within those subdirectories.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoBuild and install git-mailinfo.
Junio C Hamano [Sat, 25 Feb 2006 00:16:10 +0000 (16:16 -0800)]
Build and install git-mailinfo.

The merge 712b1dd389ad5bcdbaab0279641f0970702fc1f1 was done
incorrectly, and lost this program from Makefile.

Big thanks go to Tony Luck for noticing it, and Linus for
diagnosing it.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogitview: Bump the rev
Aneesh Kumar [Fri, 24 Feb 2006 08:38:35 +0000 (14:08 +0530)]
gitview: Bump the rev

Make the 0.7 release

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogitview: Fix DeprecationWarning
Aneesh Kumar [Fri, 24 Feb 2006 08:32:32 +0000 (14:02 +0530)]
gitview: Fix DeprecationWarning

DeprecationWarning: integer argument expected, got float

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'master' into next
Junio C Hamano [Fri, 24 Feb 2006 10:22:01 +0000 (02:22 -0800)]
Merge branch 'master' into next

* master:
  Merge fixes early for next maint series.
  Merge branch 'fix' into maint
  git-am: do not allow empty commits by mistake.

18 years agoMerge fixes early for next maint series.
Junio C Hamano [Fri, 24 Feb 2006 10:21:28 +0000 (02:21 -0800)]
Merge fixes early for next maint series.

18 years agoMerge branch 'fix' into maint
Junio C Hamano [Fri, 24 Feb 2006 10:21:00 +0000 (02:21 -0800)]
Merge branch 'fix' into maint

* fix:
  git-am: do not allow empty commits by mistake.

18 years agoMerge branches 'jc/rev-list' and 'jc/pack-thin' into next
Junio C Hamano [Fri, 24 Feb 2006 09:56:38 +0000 (01:56 -0800)]
Merge branches 'jc/rev-list' and 'jc/pack-thin' into next

* jc/rev-list:
  rev-list --objects: use full pathname to help hashing.

* jc/pack-thin:
  pack-objects: hash basename and direname a bit differently.
  pack-objects: allow "thin" packs to exceed depth limits
  pack-objects: use full pathname to help hashing with "thin" pack.

18 years agoMerge fix bits from jc/rev-list
Junio C Hamano [Fri, 24 Feb 2006 09:33:57 +0000 (01:33 -0800)]
Merge fix bits from jc/rev-list

18 years agoMerge branch 'np/delta' into next
Junio C Hamano [Fri, 24 Feb 2006 09:30:04 +0000 (01:30 -0800)]
Merge branch 'np/delta' into next

* np/delta:
  Revert "diff-delta: produce optimal pack data"
  Tweak break/merge score to adjust to the new delta generation code.
  count-delta: fix counting of copied source.

18 years agoRevert "diff-delta: produce optimal pack data"
Junio C Hamano [Fri, 24 Feb 2006 08:54:59 +0000 (00:54 -0800)]
Revert "diff-delta: produce optimal pack data"

This reverts 6b7d25d97bdb8a26719f90d17ff5c9720be68762 commit.

It turns out that the new algorithm has a really bad corner
case, that literally spends minutes for inputs that takes less
than a quater seconds to delta with the old algorithm.  The
resulting delta is 50% smaller which is admirable, but the
performance degradation is simply unacceptable for unconditional
use.

Some example cases are these blobs in Linux 2.6 repository:

    4917ec509720a42846d513addc11cbd25e0e3c4f
    9af06ba723df75fed49f7ccae5b6c9c34bc5115f
    dfc9cd58dc065d17030d875d3fea6e7862ede143

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge fixes from master
Junio C Hamano [Fri, 24 Feb 2006 09:14:41 +0000 (01:14 -0800)]
Merge fixes from master

18 years agopack-objects: hash basename and direname a bit differently.
Junio C Hamano [Fri, 24 Feb 2006 07:27:49 +0000 (23:27 -0800)]
pack-objects: hash basename and direname a bit differently.

...so that "Makefile"s from different revs are sorted together,
separate from "t/Makefile"s, but close enough.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list --objects: use full pathname to help hashing.
Junio C Hamano [Thu, 23 Feb 2006 06:10:24 +0000 (22:10 -0800)]
rev-list --objects: use full pathname to help hashing.

This helps to group the same files from different revs together,
while spreading files with the same basename in different
directories, to help pack-object.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agorev-list --objects-edge: remove duplicated edge commit output.
Junio C Hamano [Fri, 24 Feb 2006 07:44:15 +0000 (23:44 -0800)]
rev-list --objects-edge: remove duplicated edge commit output.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopack-objects: allow "thin" packs to exceed depth limits
Junio C Hamano [Fri, 24 Feb 2006 07:04:52 +0000 (23:04 -0800)]
pack-objects: allow "thin" packs to exceed depth limits

When creating a new pack to be used in .git/objects/pack/
directory, we carefully count the depth of deltified objects to
be reused, so that the generated pack does not to exceed the
specified depth limit for runtime efficiency.  However, when we
are generating a thin pack that does not contain base objects,
such a pack can only be used during network transfer that is
expanded on the other end upon reception, so being careful and
artificially cutting the delta chain does not buy us anything
except increased bandwidth requirement.  This patch disables the
delta chain depth limit check when reusing an existing delta.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'ar/win'
Junio C Hamano [Fri, 24 Feb 2006 06:35:55 +0000 (22:35 -0800)]
Merge branch 'ar/win'

* ar/win:
  PATCH: simplify calls to git programs in git-fmt-merge-msg

18 years agoMerge branch 'jc/send-insane-refs'
Junio C Hamano [Fri, 24 Feb 2006 06:34:39 +0000 (22:34 -0800)]
Merge branch 'jc/send-insane-refs'

* jc/send-insane-refs:
  send-pack: do not give up when remote has insanely large number of refs.

18 years agoMerge fixes early for next maint series.
Junio C Hamano [Fri, 24 Feb 2006 06:27:03 +0000 (22:27 -0800)]
Merge fixes early for next maint series.

18 years agoMerge branches 'jc/fix-co-candy', 'jc/fix-rename-leak' and 'ar/fix-win' into maint
Junio C Hamano [Fri, 24 Feb 2006 06:25:32 +0000 (22:25 -0800)]
Merge branches 'jc/fix-co-candy', 'jc/fix-rename-leak' and 'ar/fix-win' into maint

* jc/fix-co-candy:
  checkout - eye candy.

* jc/fix-rename-leak:
  diffcore-rename: plug memory leak.

* ar/fix-win:
  fix t5600-clone-fail-cleanup.sh on windows

18 years agoMerge branch 'ak/gitview'
Junio C Hamano [Fri, 24 Feb 2006 06:20:30 +0000 (22:20 -0800)]
Merge branch 'ak/gitview'

* ak/gitview:
  gitview: Display the lines joining commit nodes clearly.

18 years agogitview: Display the lines joining commit nodes clearly.
Aneesh Kumar K.V [Thu, 23 Feb 2006 19:29:42 +0000 (00:59 +0530)]
gitview: Display the lines joining commit nodes clearly.

Since i wanted to limit the graph box size i was resetting
the window after an index of 5. This result in line joining
commit nodes to pass over nodes which are not related. The
changes fixes the same

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@gmail.com>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agogit-am: do not allow empty commits by mistake.
Junio C Hamano [Fri, 24 Feb 2006 06:14:47 +0000 (22:14 -0800)]
git-am: do not allow empty commits by mistake.

Running "git-am --resolved" without doing anything can create an empty
commit. Prevent it.

Thanks for Eric W. Biederman for spotting this.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoTweak break/merge score to adjust to the new delta generation code.
Junio C Hamano [Fri, 24 Feb 2006 01:02:56 +0000 (17:02 -0800)]
Tweak break/merge score to adjust to the new delta generation code.

This lowers the default merge threshold score to 75% from
earlier 80%.  The break threshold stays the same at 50% for now,
but we might want to revisit it (and the rename detection limit
as well).

 * break score: this much edit (both insertion of new material
   and deletion of old material) needs to be there in the file
   before we consider this _might_ be a rewrite and break the
   filepair.

 * merge score: after a filepair is broken by the above criteria
   and goes through rename detection, if their pieces did not
   match with other files as rename/copy, we merge them back
   into one as if nothing happened.  If the filepair had at
   least this much deletion of old material, however, we say
   this is completely rewritten with dissimilarity index X% when
   we do so.

The updated delta code by Nico is so good that what we earlier
thought to be complete rewrite now reuses a lot more from the
source material (reducing the counted "delete"), so this
adjustment is needed to keep the perceived behaviour similar to
what we had earlier.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocount-delta: fix counting of copied source.
Junio C Hamano [Thu, 23 Feb 2006 10:58:37 +0000 (02:58 -0800)]
count-delta: fix counting of copied source.

The previous one wrongly coalesced a span with the next one
even though the span being added does not reach it.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoPATCH: simplify calls to git programs in git-fmt-merge-msg
Alex Riesen [Thu, 23 Feb 2006 10:26:46 +0000 (11:26 +0100)]
PATCH: simplify calls to git programs in git-fmt-merge-msg

It also makes it work on ActiveState Perl.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agofix t5600-clone-fail-cleanup.sh on windows
Alex Riesen [Thu, 23 Feb 2006 11:25:20 +0000 (12:25 +0100)]
fix t5600-clone-fail-cleanup.sh on windows

In windows you cannot remove current or opened directory,
an opened file, a running program, a loaded library, etc...

[jc: signoffs?  With a minor quoting fix.]

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge part of pack-thin branch
Junio C Hamano [Thu, 23 Feb 2006 11:00:59 +0000 (03:00 -0800)]
Merge part of pack-thin branch

18 years agoMerge branch 'np/delta' into next
Junio C Hamano [Thu, 23 Feb 2006 10:59:24 +0000 (02:59 -0800)]
Merge branch 'np/delta' into next

* np/delta:
  count-delta: tweak counting of copied source material.
  diff-delta: produce optimal pack data

18 years agocount-delta: tweak counting of copied source material.
Junio C Hamano [Thu, 23 Feb 2006 10:58:37 +0000 (02:58 -0800)]
count-delta: tweak counting of copied source material.

With the finer grained delta algorithm, count-delta algorithm
started overcounting copied source material, since the new delta
output tends to reuse the same source range more than once and
more aggressively.  This broke an earlier assumption that the
number of bytes copied out from the source buffer is a good
approximation how much source material is actually remaining in
the result.

This uses fairly inefficient algorithm to keep track of ranges
of source material that are actually copied out to the
destination buffer.  With this tweak, the obvious rename/break
detection tests in the testsuite start to work again.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopack-objects: use full pathname to help hashing with "thin" pack.
Junio C Hamano [Thu, 23 Feb 2006 06:10:24 +0000 (22:10 -0800)]
pack-objects: use full pathname to help hashing with "thin" pack.

This uses the same hashing algorithm to the "preferred base
tree" objects and the incoming pathnames, to group the same
files from different revs together, while spreading files with
the same basename in different directories.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopack-objects: thin pack micro-optimization.
Junio C Hamano [Thu, 23 Feb 2006 05:45:45 +0000 (21:45 -0800)]
pack-objects: thin pack micro-optimization.

Since we sort objects by type, hash, preferredness and then
size, after we have a delta against preferred base, there is no
point trying a delta with non-preferred base.  This seems to
save expensive calls to diff-delta and it also seems to save the
output space as well.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branches 'maint', 'jc/fix-co-candy' and 'jc/fix-rename-leak' into next
Junio C Hamano [Thu, 23 Feb 2006 03:46:59 +0000 (19:46 -0800)]
Merge branches 'maint', 'jc/fix-co-candy' and 'jc/fix-rename-leak' into next

* maint:
  Give no terminating LF to error() function.

* jc/fix-co-candy:
  checkout - eye candy.

* jc/fix-rename-leak:
  diffcore-rename: plug memory leak.

18 years agodiffcore-rename: plug memory leak.
Junio C Hamano [Thu, 23 Feb 2006 03:45:48 +0000 (19:45 -0800)]
diffcore-rename: plug memory leak.

Spotted by Nicolas Pitre.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'ml/cvs'
Junio C Hamano [Thu, 23 Feb 2006 03:20:55 +0000 (19:20 -0800)]
Merge branch 'ml/cvs'

* ml/cvs:
  Introducing git-cvsserver -- a CVS emulator for git.

18 years agoMerge branch 'ra/anno'
Junio C Hamano [Thu, 23 Feb 2006 03:20:08 +0000 (19:20 -0800)]
Merge branch 'ra/anno'

* ra/anno:
  Use Ryan's git-annotate instead of jsannotate
  Add git-annotate, a tool for assigning blame.

18 years agoGive no terminating LF to error() function.
Junio C Hamano [Thu, 23 Feb 2006 01:47:10 +0000 (17:47 -0800)]
Give no terminating LF to error() function.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agocheckout - eye candy.
Junio C Hamano [Thu, 23 Feb 2006 03:02:39 +0000 (19:02 -0800)]
checkout - eye candy.

This implements "eye candy" similar to the pack-object/unpack-object
to entertain users while a large tree is being checked out after
a clone or a pull.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge branch 'cw/remove' into next
Junio C Hamano [Thu, 23 Feb 2006 01:15:01 +0000 (17:15 -0800)]
Merge branch 'cw/remove' into next

* cw/remove:
  git-rm: Fix to properly handle files with spaces, tabs, newlines, etc.
  Add new git-rm command with documentation

18 years agogit-rm: Fix to properly handle files with spaces, tabs, newlines, etc.
Carl Worth [Thu, 23 Feb 2006 00:37:27 +0000 (16:37 -0800)]
git-rm: Fix to properly handle files with spaces, tabs, newlines, etc.

New tests are added to the git-rm test case to cover this as well.

Signed-off-by: Carl Worth <cworth@cworth.org>
Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoAdd new git-rm command with documentation
Carl Worth [Tue, 21 Feb 2006 23:04:51 +0000 (15:04 -0800)]
Add new git-rm command with documentation

This adds a git-rm command which provides convenience similar to
git-add, (and a bit more since it takes care of the rm as well if
given -f).

Like git-add, git-rm expands the given path names through
git-ls-files. This means it only acts on files listed in the
index. And it does act recursively on directories by default, (no -r
needed as in the case of rm itself). When it recurses, it does not
remove empty directories that are left behind.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agoMerge master to get fixes up to GIT 1.2.3
Junio C Hamano [Thu, 23 Feb 2006 00:38:21 +0000 (16:38 -0800)]
Merge master to get fixes up to GIT 1.2.3

18 years agoMerge fixes up to GIT 1.2.3
Junio C Hamano [Thu, 23 Feb 2006 00:15:42 +0000 (16:15 -0800)]
Merge fixes up to GIT 1.2.3

18 years agogit-fetch: follow tag only when tracking remote branch. v1.2.3
Junio C Hamano [Wed, 22 Feb 2006 21:10:37 +0000 (13:10 -0800)]
git-fetch: follow tag only when tracking remote branch.

Unless --no-tags flag was given, git-fetch tried to always
follow remote tags that point at the commits we picked up.

It is not very useful to pick up tags from remote unless storing
the fetched branch head in a local tracking branch.  This is
especially true if the fetch is done to merge the remote branch
into our current branch as one-shot basis (i.e. "please pull"),
and is even harmful if the remote repository has many irrelevant
tags.

This proposed update disables the automated tag following unless
we are storing the a fetched branch head in a local tracking
branch.

Signed-off-by: Junio C Hamano <junkio@cox.net>
18 years agopack-objects eye-candy: finishing touches.
Junio C Hamano [Thu, 23 Feb 2006 00:02:59 +0000 (16:02 -0800)]
pack-objects eye-candy: finishing touches.

This updates the progress output to match "every one second or
every percent whichever comes early" used by unpack-objects, as
discussed on the list.

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