Before this patch, when "git bisect start", "git bisect good" or
"git bisect skip" were called with many revisions, they could fail
after having already marked some revisions as "good", "bad" or
"skip".

This could be especilally bad for "git bisect start" because as
the file ".git/BISECT_NAMES" would not have been written, there
would have been no attempt to clear the marked revisions on a
"git bisect reset". That's because if there is no
".git/BISECT_NAMES" file, nothing is done to clean things up, as
the bisect session is not supposed to have started.

While at it, let's also create the ".git/BISECT_START" file, only
after ".git/BISECT_NAMES" as been created.
Signed-off-by: NChristian Couder <chriscool@tuxfamily.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

When the Subject: line is empty for whatever reason, git-am was fooled by
it and left an empty line at the beginning of the resulting commit log
message.

This moves the logic around so that we do not keep $SUBJECT in a separate
variable.  Instead, $dotest/msg-clean, which used to be the log message
body extracted from the message and then trailing whitespaces cleansed
out, now contains the subject line followed by a blank line at the
beginning for normal messages, and we use the first line from the file as
the summary line throughout the program.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

"git log" and friends normally skip the initial empty lines when showing
one-line summary of a commit, but merge summary didn't.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
----
 builtin-fmt-merge-msg.c |   10 +++++++++-
 1 files changed, 9 insertions(+), 1 deletions(-)
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

To get the current HEAD when we start bisecting using for example
"git bisect start", we first try "git symbolic-ref HEAD" to get a
nice name, and if it fails, we fall back to "git rev-parse
--verify HEAD".

The problem is that when "git symbolic-ref HEAD" fails, it
displays "fatal: ref HEAD not a symref", so it looks like "git
bisect start" failed and does not accept detached HEAD, even if
in fact it worked fine.

This patch adds "-q" option to the "git symbolic-ref" call to
get rid of the misleading error message.
Signed-off-by: NChristian Couder <chriscool@tuxfamily.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

When a patch can't be opened (it doesn't exist, there are permission
problems, etc.) we get the usage text, which is not a proper indication of
failure.
Signed-off-by: NAlberto Bertogli <albertito@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

For SSH clients restricted to git-shell, CVS_SERVER does not have to be
specified, because git-shell understands the default value of 'cvs' to
mean git-cvsserver'. This makes it totally transparent to CVS users, but
the instruction to set up CVS access for people with real shell access
does not apply.

Previous wording mentioning GIT_AUTHOR, GIT_COMMITTER variables was
unclear that we really meant GIT_AUTHOR_(NAME|EMAIL), etc.

Note that the .ssh/environment file is a good place to set these, and that
the .bashrc is shell-specific. Add a bit of text to differentiate cvs -d
(setting CVSROOT) from cvs co -d (setting the name of the newly checked
out directory).  Removed an extra 'Example:' string.
Signed-off-by: NScott Collins <scc@ScottCollins.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

gitk creates and uses ~/.gitk
Signed-off-by: NClifford Caoile <piyo@users.sourceforge.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Noticed by Fredrik Noring.
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The previous one overwrote the variable used to report the bad input
when the input is actually bad, and we did not give a useful enough
information.  This corrects it.
Signed-off-by: NChristian Couder <chriscool@tuxfamily.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

It seems that "git bisect good" and "git bisect skip" have never
properly checked arguments that have been passed to them. As soon
as one of them can be parsed as a SHA1, no error or warning would
be given.

This is because 'git rev-parse --revs-only --no-flags "$@"' always
"exit 0" and outputs all the SHA1 it can found from parsing "$@".

This patch fix this by using, for each "bisect good" argument, the
same logic as for the "bisect bad" argument.

While at it, this patch teaches "bisect bad" to give a meaningfull
error message when it is passed more than one argument.

Note that if "git bisect good" or "git bisect skip" is given some
proper revs and then something that is not a proper rev, then the
first proper revs will still have been marked as "good" or "skip".
Signed-off-by: NChristian Couder <chriscool@tuxfamily.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Jan Engelhardt noticed that while --topo-order can be overridden by a
subsequent --date-order, the reverse was not possible. That's because
setup_revisions() failed to set revs->lifo properly.
Signed-off-by: NMichele Ballabio <barra_cuda@katamail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The tagger is equal to the committer, not the author, so
GIT_COMMITTER_DATE is the right environment variable to use, not
GIT_AUTHOR_DATE.
Signed-off-by: NBjörn Steinbrink <B.Steinbrink@gmx.de>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Its documentation was removed by 6c96753d (Documentation/git-commit: rewrite
to make it more end-user friendly, 2006-12-08), even though it is referenced
from a few places, including builtin-commit.c (as part of the commentary in
the commit message template).
Signed-off-by: NJohannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NDirk Suesserott <newsletter@dirk.my1.cc>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The --root option from "git diff-tree" won't do nothing
when is given to commands like git-whatchanged or git-log,
because those always print the initial commit by default.

This fixes the tutorial explaining the function of the
log.showroot configuration variable.
Signed-off-by: NCarlos Rica <jasampler@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Ulrik Sverdrup noticed that git-archive doesn't correctly apply the attribute
export-subst when the option --prefix is given, too.

When it checked if a file has the attribute turned on, git-archive would try
to look up the full path -- including the prefix -- in .gitattributes.  That's
wrong, as the prefix doesn't need to have any relation to any existing
directories, tracked or not.

This patch makes git-archive ignore the prefix when looking up if value of the
attribute export-subst for a file.
Signed-off-by: NRene Scharfe <rene.scharfe@lsrfire.ath.cx>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

When an argument for an option is optional, like in -n from git-tag,
puting a space between the option and the argument is interpreted
as a missing argument for the option plus an isolated argument.
Documentation now reflects the need to write the parameter following
the option -n, as in "git tag -nARG", for instance.
Signed-off-by: NCarlos Rica <jasampler@gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

An earlier commit 4be60962 (apply --unidiff-zero: loosen sanity checks for
--unidiff=0 patches, 2006-09-17) made match_beginning and match_end
computed incorrectly.  If a hunk inserts at the beginning, old position
recorded at the hunk is line 0, and if a hunk changes at the beginning, it
is line 1.  The new test added to t4104 exposes that the old code did not
insist on matching at the beginning for a patch to add a line to an empty
file.

An even older 65aadb92 (apply: force matching at the beginning.,
2006-05-24) was equally wrong in that it tried to take hints from the
number of leading context lines, to decide if the hunk must match at the
beginning, but we can just look at the line number in the hunk to decide.
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

* jc/maint-fetch-regression-1.5.4:
  git-fetch test: test tracking fetch results, not just FETCH_HEAD
  Fix branches file configuration
  Tighten refspec processing

gitignore patterns can be read from three different
files, while gitattributes can come from two files. Let's
provide some hints to the user about the differences and how
they are typically used.

Suggested by Toby Corkindale, but gratuitously reworded by Jeff King.
Signed-off-by: NToby Corkindale <toby.corkindale@rea-group.com>
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

We really should have done this long time ago.  Existing t5515 test
was written for the specific purpose of catching regression to the
contents of generated FETCH_HEAD file, but it also is a good place
to make sure various fetch configurations do fetch what they intend
to fetch (and nothing else).
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Fetched remote branch from .git/branches/foo should fetch into
refs/heads/foo.  Also when partial URL is given, the fetched head should
always be remote HEAD, and the result should not be stored anywhere.
Signed-off-by: NDaniel Barkalow <barkalow@iabervon.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

This changes the pattern matching code to not store the required final
/ before the *, and then to require each side to be a valid ref (or
empty). In particular, any refspec that looks like it should be a
pattern but doesn't quite meet the requirements will be found to be
invalid as a fallback non-pattern.

This was cherry picked from commit ef00d150 (Tighten refspec processing,
2008-03-17), and two fix-up commits 46220ca1 (remote.c: Fix overtight
refspec validation, 2008-03-20) and 7d19da46 (refspec: allow colon-less
wildcard "refs/category/*", 2008-03-25) squashed in.
Signed-off-by: NDaniel Barkalow <barkalow@iabervon.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Texts between ~ and ~ will be subscripted during the asciidoc translation.
Signed-off-by: NGuanqun Lu <Guanqun.Lu@Gmail.com>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>
(cherry picked from commit 0c829391)

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

When rebasing changes that contain issues that the pre-commit hook flags
as problematic, the rebase cannot be continued.  However, rebase is about
transplanting commits that are already made with as little distortion as
possible, and pre-commit check should not interfere.

Earlier, c5b09feb (Avoid update hook during git-rebase --interactive,
2007-12-19) fixed "rebase -i", but "rebase -m" shared the same issue.
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Earlier, the callchain from pretty_print_commit() down to pp_title_line()
had an unwarranted assumption that the presense of "after_subject"
parameter, means the caller has already output MIME headers for
attachments.  The parameter's primary purpose is to give extra header
lines the caller wants to place after pp_title_line() generates the
"Subject: " line.

This assumption does not hold when the user used the format.header
configuration variable to pass extra headers, and caused a message with
non-ASCII character to lack proper MIME headers (e.g.  8-bit CTE header).
The earlier logic also failed to suppress duplicated MIME headers when
"format-patch -s --attach" is asked for and the signer's name demanded
8-bit clean transport.

This patch fixes the logic by introducing a separate need_8bit_cte
parameter passed down the callchain.  This can have one of these values:

 -1 : we've already done MIME crap and we do not want to add extra header
      to say this is 8bit in pp_title_line();

  0 : we haven't done MIME and we have not seen anything that is 8bit yet;

  1 : we haven't done MIME and we have seen something that is 8bit;
      pp_title_line() must add MIME header.

It adds two tests by Jeff King who independently diagnosed this issue.
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Tell xmlto to repress printing of the lines:

	Note: meta date   : No date. Using generated date       git-xyx
	Note: Writing git-xyz.1
Signed-off-by: NJonas Fonseca <fonseca@diku.dk>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Multiple work dirs with git svn caused each work dir to have its own
stale copy of the SVN meta data in .git/svn

git svn rebase updates commits with git-svn-id: in the repository and
stores the SVN meta data information only in that work dir.  Attempting to
git svn rebase in other work dirs for the same branch would fail because
the last revision fetched according to the git-svn-id is greater than the
revision in the SVN meta data for that work directory.
Signed-off-by: NBernt Hansen <bernt@norang.ca>
Acked-by: NEric Wong <normalperson@yhbt.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

I think it would make more sense for rev~ to have the same guarantees that
rev^ has, namely to always return a commit. I would also suggest that not
giving a number would have the same effect of defaulting to 1, not 0.

Right now it's a bit illogical, but at least it's an _undocumented_
illogical behaviour.

This patch makes '^' and '~' act the same for the default count (i.e. both
default to 1), and also have the same behaviour for a count of zero.

Before (no discernible pattern):

	[torvalds@woody git]$ git rev-parse v1.5.1 v1.5.1^0 v1.5.1~0 v1.5.1^ v1.5.1~
	45354a57ee7e3e42c7137db6c94fa968c6babe8d
	89815cab
	45354a57ee7e3e42c7137db6c94fa968c6babe8d
	045f5759
	45354a57ee7e3e42c7137db6c94fa968c6babe8d

After (fairly logical):

	[torvalds@woody git]$ git rev-parse v1.5.1 v1.5.1^0 v1.5.1~0 v1.5.1^ v1.5.1~
	45354a57ee7e3e42c7137db6c94fa968c6babe8d
	89815cab
	89815cab
	045f5759
	045f5759Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Previously, git-svn would blindly append '*' even if it was specified by
the user during initialization (for certain SVN setups, it is
necessary).

Now, the following command will work correctly:

  git svn init -T trunk/docutils \
               -t 'tags/*/docutils' \
               -b 'branches/*/docutils' \
               svn://svn.berlios.de/docutils

Thanks to martin f krafft for the bug report:
> My git-svn target configuration is
>
>   [svn-remote "svn"]
>     url = svn://svn.berlios.de/docutils
>     fetch = trunk/docutils:refs/remotes/trunk
>     branches = branches/*/docutils:refs/remotes/*
>     tags = tags/*/docutils:refs/remotes/tags/*
>
> Unfortunately, when I run
>
>   git-svn init -T trunk/docutils -t 'tags/*/docutils'
>    -b 'branches/*/docutils'
>
> then I get (note the two asterisks on the left hand side):
>
>     branches = branches/*/docutils/*:refs/remotes/*
>     tags = tags/*/docutils/*:refs/remotes/tags/*
>
> I took a brief stab at the code but I can't even figure out where
> the /* is appended, so I defer to you.
>
> It should be trivial to keep git-svn from appending /* if the left
> side already contains an asterisk.
Signed-off-by: NEric Wong <normalperson@yhbt.net>
Tested-by: Nmartin f krafft <madduck@madduck.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

* ph/maint-quiltimport:
  quiltimport: fix misquoting of parsed -p<num> parameter
  git-quiltimport: better parser to grok "enhanced" series files.

Earlier, it would error out while trying to read and/or writing them.
Now, calling merge-file with empty files is neither interesting nor
useful, but it is a bug that needed fixing.

Noticed by Clemens Buchacher.
Signed-off-by: NJohannes Schindelin <johannes.schindelin@gmx.de>

File mode changes should be handled similarly to changes of content.
That is, if the file mode changed in only one branch, keep the changed
version, and if both branch changed to different mode, mark it as a
conflict.
Signed-off-by: NClemens Buchacher <drizzd@aon.at>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The wording of the interactive help text from git-add--interactive.perl is
clearer.  Just duplicate that text here.
Signed-off-by: NVineet Kumar <vineet@doorstop.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Prior to commit 83201998 (Rewrite builtin-fetch option parsing to use
parse_options().), we understood '-n' as a short option to mean "don't
fetch tags from the remote". This patch reinstates behaviour similar,
but not identical to the pre commit 83201998 times.

Back then, -n always overrode --tags, so if both --tags and -n was
given on command-line, no tags were fetched regardless of argument
ordering. Now we use a "last entry wins" strategy, so '-n --tags'
means "fetch tags".

Since it's patently absurd to say both --tags and --no-tags, this
shouldn't matter in practice.
Spotted-by: NArtem Zolochevskiy <azol@altlinux.org>
Reported-by: NDmitry V. Levin <ldv@altlinux.org>
Signed-off-by: NJohannes Schindelin <johannes.schindelin@gmx.de>
Tested-by: NAndreas Ericsson <ae@op5.se>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Signed-off-by: NJunio C Hamano <gitster@pobox.com>