Code

git.git
16 years agoMerge git://git.kernel.org/pub/scm/gitk/gitk
Junio C Hamano [Tue, 8 Apr 2008 04:52:23 +0000 (21:52 -0700)]
Merge git://git.kernel.org/pub/scm/gitk/gitk

* git://git.kernel.org/pub/scm/gitk/gitk:
  gitk: Fix changing colors through Edit->Preferences

16 years agogit-svn: fix following renamed paths when tracking a single path
Eric Wong [Mon, 7 Apr 2008 07:08:03 +0000 (00:08 -0700)]
git-svn: fix following renamed paths when tracking a single path

When using git-svn to follow only a single (empty) path per
svn-remote (i.e. not using --stdlayout), following the history
of a renamed path was broken in
c586879cdfa4f8181a14e953a9152a4639eef333.

This reverts the regression for the single (emtpy) path per
svn-remote case.

To avoid breaking the tests in a committed revision, this is an
addendum to a patch originally submitted by

  Santhosh Kumar Mani <santhoshmani@gmail.com>:
  > git-svn: add test for renamed directory fetch
  >
  > This test tries to fetch a directory which had renames in the
  > history from a SVN repository.

  [ew: unneccesary dependency on the starting an HTTP server
   removed from Santhosh's original test.]

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'jc/maint-apply-match-beginning'
Junio C Hamano [Mon, 7 Apr 2008 03:04:29 +0000 (20:04 -0700)]
Merge branch 'jc/maint-apply-match-beginning'

* jc/maint-apply-match-beginning:
  Fix "git apply" to correctly enforce "match at the beginning"

16 years agoAdd prefix oriented completions for diff and format-patch commands.
Pascal Obry [Sun, 6 Apr 2008 16:56:08 +0000 (18:56 +0200)]
Add prefix oriented completions for diff and format-patch commands.

Signed-off-by: Pascal Obry <pascal@obry.net>
Acked-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agotest suite: remove useless TERM cruft in "t7005-editor.sh"
Christian Couder [Sun, 6 Apr 2008 17:08:50 +0000 (19:08 +0200)]
test suite: remove useless TERM cruft in "t7005-editor.sh"

In commit 15387e3 (Test suite: reset TERM to its previous value after
testing., 2007-10-26), I added a workaround to reset TERM to its previous
value before the "test_done" at the end of "t7005-editor.sh" because
otherwise "test_done" would have printed the test result with a bad TERM
env variable (this resulted in output with no color on konsole).

But since commit c2116a1 (test-lib: fix TERM to dumb for test
repeatability, 2008-03-06), colored output is printed in a subshell with
TERM reset to its original value so the earlier workaround is not needed
anymore.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd interactive option in rebase command completion list.
Pascal Obry [Sun, 6 Apr 2008 17:32:24 +0000 (19:32 +0200)]
Add interactive option in rebase command completion list.

Signed-off-by: Pascal Obry <pascal@obry.net>
Acked-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix "git apply" to correctly enforce "match at the beginning"
Junio C Hamano [Mon, 7 Apr 2008 02:21:45 +0000 (19:21 -0700)]
Fix "git apply" to correctly enforce "match at the beginning"

An earlier commit 4be6096 (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 65aadb9 (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: Junio C Hamano <gitster@pobox.com>
16 years agoAdd description of OFS_DELTA to the pack format description
Peter Eriksen [Sun, 6 Apr 2008 20:51:49 +0000 (22:51 +0200)]
Add description of OFS_DELTA to the pack format description

Signed-off-by: Peter Eriksen <s022018@student.dtu.dk>
Acked-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogitk: Fix changing colors through Edit->Preferences
Gerrit Pape [Wed, 26 Mar 2008 18:45:26 +0000 (18:45 +0000)]
gitk: Fix changing colors through Edit->Preferences

With tcl/tk8.5 the lset command seems to behave differently.  When
changing the background color through Edit->Preferences, the changes
are applied, but new dialogs, such as View->New view... barf with

 Error: unknown color name "{#ffffff}"

Additionally when closing gitk, and starting it up again, a bad value
has been saved to ~/.gitk, preventing gitk from running properly; it
fails with

 Error in startup script: unknown color name "{#ffffff}"
 ...

This commit fixes the problem by changing the color dialogs to pass
the empty string {} as the list index to choosecolor.  This causes
the lset and lindex commands used by choosecolor to use and set the
whole variable (bgcolor, fgcolor or selectbgcolor) rather than
treating them as a 1-element list.  Tested with tcl/tk8.4 and 8.5.

Dmitry Potapov reported this problem through
 http://bugs.debian.org/472615

Signed-off-by: Gerrit Pape <pape@smarden.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agogit-pack-objects.txt: Make wording slightly less ambiguous
Peter Eriksen [Sat, 5 Apr 2008 19:46:49 +0000 (21:46 +0200)]
git-pack-objects.txt: Make wording slightly less ambiguous

It is a bit confusing on first read, that

        "The packed archive format (.pack) is designed
        to be unpackable..."

Signed-off-by: Peter Eriksen <s022018@student.dtu.dk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-fetch: Don't trigger a bus error when given the refspec "tag"
Kevin Ballard [Sat, 5 Apr 2008 18:28:53 +0000 (14:28 -0400)]
git-fetch: Don't trigger a bus error when given the refspec "tag"

When git-fetch encounters the refspec "tag" it assumes that the next
argument will be a tag name. If there is no next argument, it should
die gracefully instead of erroring.

Signed-off-by: Kevin Ballard <kevin@sb.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoRevert "gitweb: Add 'status_str' to parse_difftree_raw_line output"
Jakub Narebski [Sat, 5 Apr 2008 20:13:24 +0000 (21:13 +0100)]
Revert "gitweb: Add 'status_str' to parse_difftree_raw_line output"

This reverts commit 6aa6f92fda47cc4ee5f599895e8a5a327fb6f9ab.

It caused is_deleted() subroutine to output warnings when dealing with
old, legacy gitweb blobdiff URLs without either 'hb' or 'hpb'
parameters.

This fixes http://bugs.debian.org/469083

Signed-off-by: Jakub Narebski <jnareb@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge git://repo.or.cz/git-gui
Junio C Hamano [Sat, 5 Apr 2008 05:38:32 +0000 (22:38 -0700)]
Merge git://repo.or.cz/git-gui

* git://repo.or.cz/git-gui:
  git-gui: use +/- instead of ]/[ to show more/less context in diff
  git-gui: Update french translation
  git-gui: Switch keybindings for [ and ] to bracketleft and bracketright

16 years agogit-gui: use +/- instead of ]/[ to show more/less context in diff gitgui-0.10.1
Michele Ballabio [Fri, 4 Apr 2008 21:04:42 +0000 (23:04 +0200)]
git-gui: use +/- instead of ]/[ to show more/less context in diff

On some systems, brackets cannot be used as event details
(they don't have a keysym), so use +/- instead (both on
keyboard and keypad) and add ctrl-= as a synonym of ctrl-+
for convenience.

[sp: Had to change accelerator to show only "$M1T-="; the
     original version included "$M1T-+ $M1T-=" but this is
 not drawn at all on Mac OS X.]

Signed-off-by: Michele Ballabio <barra_cuda@katamail.com>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agogit-gui: Update french translation
Christian Couder [Sat, 5 Apr 2008 02:20:23 +0000 (04:20 +0200)]
git-gui: Update french translation

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agogit-gui: Switch keybindings for [ and ] to bracketleft and bracketright
Shawn O. Pearce [Thu, 3 Apr 2008 22:06:29 +0000 (18:06 -0400)]
git-gui: Switch keybindings for [ and ] to bracketleft and bracketright

Thanks to Michele Ballabio for the quick fix.
This resolves the error introduced by c91ee2bd61.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agoGIT 1.5.5-rc3 v1.5.5-rc3
Junio C Hamano [Wed, 2 Apr 2008 17:42:14 +0000 (10:42 -0700)]
GIT 1.5.5-rc3

The rate of fixes that trickle in has slowed and we are definitely
getting there.  Hopefully one final round and we will have the final
1.5.5 soon.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'js/filter-branch'
Junio C Hamano [Wed, 2 Apr 2008 18:13:23 +0000 (11:13 -0700)]
Merge branch 'js/filter-branch'

* js/filter-branch:
  filter-branch: Fix renaming a directory in the tree-filter
  filter-branch: Test renaming directories in a tree-filter

16 years agoDescribe the bug in handling filenames with funny characters in 'git add -i'
Teemu Likonen [Wed, 2 Apr 2008 17:06:46 +0000 (20:06 +0300)]
Describe the bug in handling filenames with funny characters in 'git add -i'

The interactive mode does not work with files whose names contain
characters that need C-quoting.  `core.quotepath` configuration can be
used to work this limitation around to some degree, but backslash,
double-quote and control characters will still have problems.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge git://repo.or.cz/git-gui
Junio C Hamano [Wed, 2 Apr 2008 17:29:10 +0000 (10:29 -0700)]
Merge git://repo.or.cz/git-gui

* git://repo.or.cz/git-gui:
  git-gui 0.10
  git-gui: Add shortcut keys for Show More/Less Context

16 years agoMerge branch 'bc/mktag'
Junio C Hamano [Wed, 2 Apr 2008 07:23:19 +0000 (00:23 -0700)]
Merge branch 'bc/mktag'

* bc/mktag:
  mktag.c: tweak validation of tagger field and adjust test script
  mktag.c: improve verification of tagger field and tests

16 years agoMerge branch 'pb/cvsserver'
Junio C Hamano [Wed, 2 Apr 2008 07:22:20 +0000 (00:22 -0700)]
Merge branch 'pb/cvsserver'

* pb/cvsserver:
  git-cvsserver: handle change type T

16 years agoMerge branch 'dd/cvsserver'
Junio C Hamano [Wed, 2 Apr 2008 07:22:15 +0000 (00:22 -0700)]
Merge branch 'dd/cvsserver'

* dd/cvsserver:
  cvsserver: Use the user part of the email in log and annotate results
  cvsserver: Add test for update -p
  cvsserver: Implement update -p (print to stdout)
  cvsserver: Add a few tests for 'status' command
  cvsserver: Do not include status output for subdirectories if -l is passed
  cvsserver: Only print the file part of the filename in status header
  cvsserver: Respond to the 'editors' and 'watchers' commands

16 years agoMerge branch 'je/cvsserver'
Junio C Hamano [Wed, 2 Apr 2008 07:22:06 +0000 (00:22 -0700)]
Merge branch 'je/cvsserver'

* je/cvsserver:
  Allow git-cvsserver database table name prefix to be specified.

16 years agot7004-tag: Skip more tests if gpg is not available.
Johannes Sixt [Wed, 2 Apr 2008 06:52:37 +0000 (08:52 +0200)]
t7004-tag: Skip more tests if gpg is not available.

This test was already careful enough to skip signed tag tests if gpg
is not available, but it must also skip all verify tests, even those
that are about non-signed tags, because they also invoke gpg.

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoverify-tag: Clean up the temporary file if gpg cannot be started.
Johannes Sixt [Wed, 2 Apr 2008 06:49:59 +0000 (08:49 +0200)]
verify-tag: Clean up the temporary file if gpg cannot be started.

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agohelp: Add a missing OPT_END().
Christian Couder [Wed, 2 Apr 2008 03:47:41 +0000 (05:47 +0200)]
help: Add a missing OPT_END().

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAccept git aliases outside a git repository
Junio C Hamano [Tue, 1 Apr 2008 04:33:09 +0000 (21:33 -0700)]
Accept git aliases outside a git repository

af05d67 (Always set *nongit_ok in setup_git_directory_gently(),
2008-03-25) had a change from the patch originally submitted that resulted
in disabling aliases outside a git repository.

It turns out that some people used "alias.fubar = diff --color-words" in
$HOME/.gitconfig to use non-index diff (or any command that do not need
git repository) outside git repositories, and this change broke them,
so this resurrects the support for such usage.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-gui 0.10 gitgui-0.10.0
Shawn O. Pearce [Wed, 2 Apr 2008 06:17:11 +0000 (02:17 -0400)]
git-gui 0.10

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agogit-gui: Add shortcut keys for Show More/Less Context
Jonathan del Strother [Tue, 1 Apr 2008 10:54:03 +0000 (11:54 +0100)]
git-gui: Add shortcut keys for Show More/Less Context

Bound to Ctrl/Cmd + left & right square brackets, depending on
your platform.

[sp: Added missing binds for . to allow shortcuts to work when
     not focused in the commit message area.]

Signed-off-by: Jonathan del Strother <jon.delStrother@bestbefore.tv>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agomktag.c: tweak validation of tagger field and adjust test script
Brandon Casey [Mon, 31 Mar 2008 23:25:23 +0000 (18:25 -0500)]
mktag.c: tweak validation of tagger field and adjust test script

Update the verify_tag() function to remove an unnecessary test, and add
additional check for angle brackets in the name and email field, and
spaces in the email field. The timestamp and timezone sections are made
more straight forward by using strspn().

Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agofilter-branch: Fix renaming a directory in the tree-filter
veillette@yahoo.ca [Mon, 31 Mar 2008 07:14:15 +0000 (09:14 +0200)]
filter-branch: Fix renaming a directory in the tree-filter

Commit d89c1df (filter-branch: don't use xargs -0, 2008-03-12) replaced a
'ls-files | xargs rm' pipeline by 'git clean'. 'git clean' however does
not recurse and remove directories by default.

Now, consider a tree-filter that renames a directory.

  1. For the first commit everything works as expected

  2. Then filter-branch checks out the files for the next commit. This
     leaves the new directory behind because there is no real "branch
     switching" involved that would notice that the directory can be
     removed.

  3. Then filter-branch invokes 'git clean' to remove exactly those
     left-overs. But here it does not remove the directory.

  4. The next tree-filter does not work as expected because there already
     exists a directory with the new name.

Just add -d to 'git clean', so that empty directories are removed.

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agofilter-branch: Test renaming directories in a tree-filter
Johannes Sixt [Mon, 31 Mar 2008 07:14:14 +0000 (09:14 +0200)]
filter-branch: Test renaming directories in a tree-filter

This test currently fails.

If b is a directory then 'mv a b' is not a plain "rename", but really a
"move", so we must also test that the directory does not exist with the
old name in the directory with the new name.

There's also some cleanup in the corresponding "rename file" test to avoid
spurious shell syntax errors and "ambigous ref" error from 'git show' (but
these should show up only if the test would fail anyway). Plus we also
test for the non-existence of the old file.

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
16 years agomktag.c: improve verification of tagger field and tests
Brandon Casey [Thu, 27 Mar 2008 16:16:04 +0000 (11:16 -0500)]
mktag.c: improve verification of tagger field and tests

Since nearly its birth, git's tags have included a "tagger" field which
describes the name of tagger, email of tagger, and date and time of tagging.
But, this field was only loosely tested by git-mktag. Provide some thorough
testing for this field and also ensure that the tag header is separated
from the tag body by an empty line to reduce the convenience of creating
a flawed tag.

Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agodiff-files: careful when inspecting work tree items
Junio C Hamano [Mon, 31 Mar 2008 00:30:08 +0000 (17:30 -0700)]
diff-files: careful when inspecting work tree items

This fixes the same breakage in diff-files.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agodiff-index: careful when inspecting work tree items
Junio C Hamano [Mon, 31 Mar 2008 00:29:48 +0000 (17:29 -0700)]
diff-index: careful when inspecting work tree items

Earlier, if you changed a staged path into a directory in the work tree,
we happily ran lstat(2) on it and found that it exists, and declared that
the user changed it to a gitlink.

This is wrong for two reasons:

 (1) It may be a directory, but it may not be a submodule, and in the
     latter case, the change we need to report is "the blob at the path
     has disappeared".  We need to check with resolve_gitlink_ref() to be
     consistent with what "git add" and "git update-index --add" does.

 (2) lstat(2) may have succeeded only because a leading component of the
     path was turned into a symbolic link that points at something that
     exists in the work tree.  In such a case, the path itself does not
     exist anymore, as far as the index is concerned.

This fixes these breakages in diff-index that the previous patch has
exposed.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd corner case tests for diff-index and diff-files
Junio C Hamano [Mon, 31 Mar 2008 00:28:05 +0000 (17:28 -0700)]
Add corner case tests for diff-index and diff-files

diff-index and diff-files can get confused in corner cases when an indexed
blob turns into something else in the work tree.  This patch adds tests to
expose such breakages.

The test is classified under t2XXX series instead of t4XXX series, because
the ultimate objective is to fix "add -u" (and "commit -a" that shares the
same issue).

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAllow git-cvsserver database table name prefix to be specified.
Josh Elsasser [Thu, 27 Mar 2008 21:02:14 +0000 (14:02 -0700)]
Allow git-cvsserver database table name prefix to be specified.

Adds a gitcvs.dbtablenameprefix config variable, the contents of which
are prepended to any database tables names used by git-cvsserver. The
same substutions as gitcvs.dbname and gitcvs.dbuser are supported, and
any non-alphabetic characters are replaced with underscores.

A typo found in contrib/completion/git-completion.bash is also fixed.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoSilence cpio's "N blocks" output when cloning locally
Bryan Donlan [Fri, 28 Mar 2008 06:19:46 +0000 (02:19 -0400)]
Silence cpio's "N blocks" output when cloning locally

Pass --quiet to cpio in git-clone to hide the (confusing) "0 blocks" message.
For compatibility with operating systems which might not support GNUisms,
the presence of --quiet is probed for by grepping cpio's --help output.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-svn: remove redundant slashes from show-ignore
Eric Wong [Sat, 29 Mar 2008 23:37:17 +0000 (16:37 -0700)]
git-svn: remove redundant slashes from show-ignore

Jonathan Scott Duff wrote:

> Recently I tried "git svn showignore" on my parrot repository and it
> failed.  I tracked it down to the prop_walk() sub.  When it recurses,
> $path has an extra / on the beginning (i.e., when it recurses, it
> tries to get the props for "//apps" instead of "/apps").   I *think*
> this is because $path is used in the recursive call rather than $p
> (which seems to contain a properly transformed $path).  Anyway, I've
> attached a patch that works for me and I think is generally the right
> thing.

Patch-submitted-by: Jonathan Scott Duff
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-p4: Handle Windows EOLs properly after removal of p4 submit template handling.
Marius Storm-Olsen [Fri, 28 Mar 2008 14:40:40 +0000 (15:40 +0100)]
git-p4: Handle Windows EOLs properly after removal of p4 submit template handling.

git-p4s handling of Windows style EOL was broken after the removal
of the p4 submit template handling in commit f2a6059. Fix that, and
make getP4OpenedType() more robust.

Signed-off-by: Marius Storm-Olsen <marius@trolltech.com>
Signed-off-by: Simon Hausmann <simon@lst.de>
16 years agogit-cvsserver: handle change type T
Paolo Bonzini [Sun, 16 Mar 2008 19:00:21 +0000 (20:00 +0100)]
git-cvsserver: handle change type T

git-cvsserver does not support changes of type T (file type change,
e.g. symlink->real file).  This patch treats them the same as changes
of type M.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoGIT 1.5.5-rc2 v1.5.5-rc2
Junio C Hamano [Fri, 28 Mar 2008 03:43:51 +0000 (20:43 -0700)]
GIT 1.5.5-rc2

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoGIT 1.5.4.5 v1.5.4.5
Junio C Hamano [Fri, 28 Mar 2008 03:36:35 +0000 (20:36 -0700)]
GIT 1.5.4.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Use the user part of the email in log and annotate results
Damien Diederen [Thu, 27 Mar 2008 22:18:35 +0000 (23:18 +0100)]
cvsserver: Use the user part of the email in log and annotate results

Generate the CVS author names by taking the first eight characters of
the user part of the email address.  The resulting names are more
likely to make sense (or at least reduce ambiguities) in "corporate"
environments.

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Add test for update -p
Damien Diederen [Thu, 27 Mar 2008 22:18:23 +0000 (23:18 +0100)]
cvsserver: Add test for update -p

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Implement update -p (print to stdout)
Damien Diederen [Thu, 27 Mar 2008 22:18:12 +0000 (23:18 +0100)]
cvsserver: Implement update -p (print to stdout)

Cvs update -p -r <rev> <path> is the documented way to retrieve a
specific revision of a file (similar to git show <rev>:<path>).
Without this patch, the -p flag is ignored and status output is
produced, causing clients to interpret it as the contents of the file.

TkCVS uses update -p as a basis for implementing its various "View"
and "Diff" commands.

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Add a few tests for 'status' command
Damien Diederen [Thu, 27 Mar 2008 22:18:02 +0000 (23:18 +0100)]
cvsserver: Add a few tests for 'status' command

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Do not include status output for subdirectories if -l is passed
Damien Diederen [Thu, 27 Mar 2008 22:17:53 +0000 (23:17 +0100)]
cvsserver: Do not include status output for subdirectories if -l is passed

This effectively implements the -l switch by pruning the entries whose
filenames contain a path separator.  It was previously ignored.

Without this, TkCVS includes strange "ghost" entries in its directory
listings.

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Only print the file part of the filename in status header
Damien Diederen [Thu, 27 Mar 2008 22:17:42 +0000 (23:17 +0100)]
cvsserver: Only print the file part of the filename in status header

The "File:" header of CVS status output only includes the basename of
the file, even when generating a recursive listing; do the same.

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocvsserver: Respond to the 'editors' and 'watchers' commands
Damien Diederen [Thu, 27 Mar 2008 22:17:26 +0000 (23:17 +0100)]
cvsserver: Respond to the 'editors' and 'watchers' commands

These commands list users editing and watching locked files.  This trivial
implementation always returns an empty response, since git-cvsserver does not
implement file locking.

Without this, TkCVS hangs at startup, waiting forever for a response.

Signed-off-by: Damien Diederen <dash@foobox.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agobuiltin-prune: protect objects listed on the command line
Junio C Hamano [Tue, 25 Mar 2008 06:20:51 +0000 (23:20 -0700)]
builtin-prune: protect objects listed on the command line

Finally, this resurrects the documented behaviour to protect other
objects listed on the command line from getting pruned.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agobuiltin-prune.c: use parse_options()
Michele Ballabio [Sun, 23 Mar 2008 20:50:29 +0000 (21:50 +0100)]
builtin-prune.c: use parse_options()

Using the OPT_DATE() introduced earlier, this updates builtin-prune to
use parse_options().

Signed-off-by: Michele Ballabio <barra_cuda@katamail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd tests for git-prune
Michele Ballabio [Sun, 23 Mar 2008 21:34:34 +0000 (22:34 +0100)]
Add tests for git-prune

It seems that git prune changed behaviour with respect to revisions added
from command line, probably when it became a builtin. Currently, it prints
a short usage and exits: instead, it should take those revisions into
account and not prune them. So add a couple of test to point this out.

We'll be fixing this by switching to parse_options(), so add tests to
detect bogus command line parameters as well, to keep ourselves from
introducing regressions.

Signed-off-by: Michele Ballabio <barra_cuda@katamail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoparse-options.c: introduce OPT_DATE
Michele Ballabio [Mon, 24 Mar 2008 14:02:21 +0000 (15:02 +0100)]
parse-options.c: introduce OPT_DATE

There are quite a few places that will need to call approxidate(),
when they'll adopt the parse-options system, so this patch adds the
function parse_opt_approxidate_cb(), used by OPT_DATE.

Signed-off-by: Michele Ballabio <barra_cuda@katamail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoUpdate draft release notes for 1.5.5
Junio C Hamano [Thu, 27 Mar 2008 20:37:29 +0000 (13:37 -0700)]
Update draft release notes for 1.5.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'maint'
Junio C Hamano [Thu, 27 Mar 2008 20:35:18 +0000 (13:35 -0700)]
Merge branch 'maint'

* maint:
  Update draft release notes for 1.5.4.5
  Documentation: clarify use of .git{ignore,attributes} versus .git/info/*
  t/t3800-mktag.sh: use test_must_fail rather than '!'

Conflicts:

t/t3800-mktag.sh

16 years agoUpdate draft release notes for 1.5.4.5
Junio C Hamano [Thu, 27 Mar 2008 20:14:20 +0000 (13:14 -0700)]
Update draft release notes for 1.5.4.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'jc/maint-fetch-regression-1.5.4' into maint
Junio C Hamano [Thu, 27 Mar 2008 20:03:56 +0000 (13:03 -0700)]
Merge branch 'jc/maint-fetch-regression-1.5.4' into maint

* 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

16 years agoDocumentation: clarify use of .git{ignore,attributes} versus .git/info/*
Jeff King [Thu, 27 Mar 2008 05:31:00 +0000 (01:31 -0400)]
Documentation: clarify use of .git{ignore,attributes} versus .git/info/*

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: Toby Corkindale <toby.corkindale@rea-group.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agotest_must_fail: 129 is a valid error code from usage()
Junio C Hamano [Tue, 25 Mar 2008 06:07:08 +0000 (23:07 -0700)]
test_must_fail: 129 is a valid error code from usage()

When a git command is run under test_must_fail to make sure that
the argument parser catches bogus command line, it exits with 129.
We need to catch it as a valid "graceful error exit".

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoimap-send: properly error out if imap.host is not set in config
Gerrit Pape [Wed, 26 Mar 2008 18:05:17 +0000 (18:05 +0000)]
imap-send: properly error out if imap.host is not set in config

If no imap host is specified in the git config, git imap-send used
to try to lookup a null pointer through gethostbyname(), causing a
segfault.  Since setting the imap.host variable is mandatory,
imap-send now properly fails with an explanatory error message.

The problem has been reported by picca through
 http://bugs.debian.org/472632

Signed-off-by: Gerrit Pape <pape@smarden.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agot9600-cvsimport.sh: set HOME before checking for cvsps availability
Frank Lichtenheld [Wed, 26 Mar 2008 17:34:20 +0000 (17:34 +0000)]
t9600-cvsimport.sh: set HOME before checking for cvsps availability

This actually sounds like a bug in cvsps, which requires an existing
home directory when asked for the usage through -h

 $ HOME=/nonexistent cvsps -h
 Cannot create the cvsps directory '.cvsps': No such file or directory

This made t9600 think that cvsps is not available if HOME did not exist,
causing the tests to be skipped

 $ HOME=/nonexistent sh t9600-cvsimport.sh
 * skipping cvsimport tests, cvsps not found
 * passed all 0 test(s)

Now t9600 sets HOME to the current working directory before checking for
the availability of the cvsps program.

This issue has been discovered by Marco Rodrigues, and fixed by Frank
Lichtenheld through
 http://bugs.debian.org/471969

Signed-off-by: Gerrit Pape <pape@smarden.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAlways set *nongit_ok in setup_git_directory_gently()
SZEDER Gábor [Tue, 25 Mar 2008 21:06:26 +0000 (22:06 +0100)]
Always set *nongit_ok in setup_git_directory_gently()

setup_git_directory_gently() only modified the value of its *nongit_ok
argument if we were not in a git repository.  Now it will always set it
to 0 when we are inside a repository.

Also remove now unnecessary initializations in the callers of this
function.

Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agot/t3800-mktag.sh: use test_must_fail rather than '!'
Brandon Casey [Wed, 26 Mar 2008 00:20:52 +0000 (19:20 -0500)]
t/t3800-mktag.sh: use test_must_fail rather than '!'

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'jc/maint-fetch-regression-1.5.4'
Junio C Hamano [Wed, 26 Mar 2008 08:49:41 +0000 (01:49 -0700)]
Merge branch 'jc/maint-fetch-regression-1.5.4'

* 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
  Fix the wrong output of `git-show v1.3.0~155^2~4` in documentation.

16 years agogit-fetch test: test tracking fetch results, not just FETCH_HEAD
Junio C Hamano [Wed, 26 Mar 2008 08:17:07 +0000 (01:17 -0700)]
git-fetch test: test tracking fetch results, not just FETCH_HEAD

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: Junio C Hamano <gitster@pobox.com>
16 years agoFix branches file configuration
Daniel Barkalow [Tue, 25 Mar 2008 23:35:28 +0000 (19:35 -0400)]
Fix branches file configuration

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: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoTighten refspec processing
Daniel Barkalow [Tue, 18 Mar 2008 02:05:23 +0000 (22:05 -0400)]
Tighten refspec processing

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 ef00d15 (Tighten refspec processing,
2008-03-17), and two fix-up commits 46220ca (remote.c: Fix overtight
refspec validation, 2008-03-20) and 7d19da4 (refspec: allow colon-less
wildcard "refs/category/*", 2008-03-25) squashed in.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix the wrong output of `git-show v1.3.0~155^2~4` in documentation.
Guanqun Lu [Mon, 24 Mar 2008 07:27:28 +0000 (15:27 +0800)]
Fix the wrong output of `git-show v1.3.0~155^2~4` in documentation.

Texts between ~ and ~ will be subscripted during the asciidoc translation.

Signed-off-by: Guanqun Lu <Guanqun.Lu@Gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
(cherry picked from commit 0c829391cfcdc57172765322575804a7ad5f3116)

16 years agorefspec: allow colon-less wildcard "refs/category/*"
Junio C Hamano [Wed, 26 Mar 2008 04:15:52 +0000 (21:15 -0700)]
refspec: allow colon-less wildcard "refs/category/*"

"git push --tags elsewhere" is implemented in terms of wildcarded refspec
"refs/tags/*" these days, and the user wants to push the tags under the
same name to the other branch.  This resurrects the support for it.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoinit: show "Reinit" message even in an (existing) empty repository
Johannes Schindelin [Mon, 24 Mar 2008 15:14:52 +0000 (16:14 +0100)]
init: show "Reinit" message even in an (existing) empty repository

Earlier, git-init tested for a valid HEAD ref, but if the repository
was empty, there was none.  Instead, test for the existence of
the file $GIT_DIR/HEAD.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation/git-checkout: Update summary to reflect current abilities
Julian Phillips [Mon, 24 Mar 2008 03:06:20 +0000 (03:06 +0000)]
Documentation/git-checkout: Update summary to reflect current abilities

For a while now, git-checkout has been more powerful than the man-page
summary would suggest (the main text does describe the new features),
so update the summary to hopefully better reflect the current
functionality.  Also update the glossary description of the word checkout.

Signed-off-by: Julian Phillips <julian@quantumfyre.co.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation: git-tag '-m'/'-F' implies '-a'
Dirk Süsserott [Mon, 24 Mar 2008 19:43:08 +0000 (15:43 -0400)]
Documentation: git-tag '-m'/'-F' implies '-a'

16 years agobuiltin-remote: Fix missing newline at end of listing of pushed branches
Johannes Sixt [Tue, 18 Mar 2008 20:52:00 +0000 (21:52 +0100)]
builtin-remote: Fix missing newline at end of listing of pushed branches

Without this the output of 'git remote show' does not end with a new-line:

bash> git remote show repo
* remote repo
  URL: repo.or.cz:/srv/git/kdbg.git
  Tracked remote branches
    maint master mob
  Local branch pushed with 'git push'
    +master:masterbash>

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Acked-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix the wrong output of `git-show v1.3.0~155^2~4` in documentation.
Guanqun Lu [Mon, 24 Mar 2008 07:27:28 +0000 (15:27 +0800)]
Fix the wrong output of `git-show v1.3.0~155^2~4` in documentation.

Texts between ~ and ~ will be subscripted during the asciidoc translation.

Signed-off-by: Guanqun Lu <Guanqun.Lu@Gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoRelNotes: mention checkout/branch's --track option, too
Johannes Schindelin [Sun, 23 Mar 2008 11:35:37 +0000 (12:35 +0100)]
RelNotes: mention checkout/branch's --track option, too

checkout and branch recently learnt to track local branches when
branch.autosetupmerge = always, but they _also_ learnt to do that when
asked explicitely with the option "--track".

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoGIT 1.5.5-rc1 v1.5.5-rc1
Junio C Hamano [Sun, 23 Mar 2008 07:21:48 +0000 (00:21 -0700)]
GIT 1.5.5-rc1

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogc --auto: raise default auto pack limit from 20 to 50
Junio C Hamano [Sun, 23 Mar 2008 07:04:48 +0000 (00:04 -0700)]
gc --auto: raise default auto pack limit from 20 to 50

Recent discussion on the list, with the improvement f7c22cc (always start
looking up objects in the last used pack first, 2007-05-30) brought in,
reached the concensus that the current default 20 is too low.

Reference: http://thread.gmane.org/gmane.comp.version-control.git/77586
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'git-p4' of git://repo.or.cz/git/git-p4
Junio C Hamano [Sun, 23 Mar 2008 07:02:06 +0000 (00:02 -0700)]
Merge branch 'git-p4' of git://repo.or.cz/git/git-p4

* 'git-p4' of git://repo.or.cz/git/git-p4:
  git-p4: Use P4EDITOR environment variable when set
  git-p4: Unset P4DIFF environment variable when using 'p4 -du diff'
  git-p4: Optimize the fetching of data from perforce.

16 years agoremote.c: Fix overtight refspec validation
Junio C Hamano [Fri, 21 Mar 2008 06:34:37 +0000 (23:34 -0700)]
remote.c: Fix overtight refspec validation

We tightened the refspec validation code in an earlier commit ef00d15
(Tighten refspec processing, 2008-03-17) per my suggestion, but the
suggestion was misguided to begin with and it broke this usage:

    $ git push origin HEAD~12:master

The syntax of push refspecs and fetch refspecs are similar in that they
are both colon separated LHS and RHS (possibly prefixed with a + to
force), but the similarity ends there.  For example, LHS in a push refspec
can be anything that evaluates to a valid object name at runtime (except
when colon and RHS is missing, or it is a glob), while it must be a
valid-looking refname in a fetch refspec.  To validate them correctly, the
caller needs to be able to say which kind of refspecs they are.  It is
unreasonable to keep a single interface that cannot tell which kind it is
dealing with, and ask it to behave sensibly.

This commit separates the parsing of the two into different functions, and
clarifies the code to implement the parsing proper (i.e. splitting into
two parts, making sure both sides are wildcard or neither side is).

This happens to also allow pushing a commit named with the esoteric "look
for that string" syntax:

    $ git push ../test.git ':/remote.c: Fix overtight refspec:master'

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agofast-import: Document the effect of "merge" with no "from" in a commit
Eyvind Bernhardsen [Fri, 21 Mar 2008 15:25:18 +0000 (16:25 +0100)]
fast-import: Document the effect of "merge" with no "from" in a commit

The fast-import documentation currently does not document the behaviour
of "merge" when there is no "from" in a commit.  This patch adds a
description of what happens: the commit is created with a parent, but
no files.  This behaviour is equivalent to "from" followed by
"filedeleteall".

Signed-off-by: Eyvind Bernhardsen <eyvind-git@orakel.ntnu.no>
Acked-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMake git-svn tests behave better on OS X
Kevin Ballard [Fri, 21 Mar 2008 07:27:35 +0000 (03:27 -0400)]
Make git-svn tests behave better on OS X

Give lib-git-svn.sh a few alternate paths to look for apache2.
Explicitly define the LockFile so httpd will actually start under OS X

Signed-off-by: Kevin Ballard <kevin@sb.org>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoImprove description of git filter-branch.
Ralf Wildenhues [Thu, 20 Mar 2008 21:30:32 +0000 (22:30 +0100)]
Improve description of git filter-branch.

Signed-off-by: Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agot/t7003-filter-branch.sh: use test_must_fail rather than '!'
Brandon Casey [Thu, 20 Mar 2008 16:54:30 +0000 (11:54 -0500)]
t/t7003-filter-branch.sh: use test_must_fail rather than '!'

Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDon't try and percent-escape existing percent escapes in git-svn URIs
Kevin Ballard [Thu, 20 Mar 2008 20:08:49 +0000 (16:08 -0400)]
Don't try and percent-escape existing percent escapes in git-svn URIs

git-svn project names are percent-escaped ever since f5530b8
(git-svn: support for funky branch and project names over HTTP(S),
2007-11-11).

Unfortunately this breaks the scenario where the user hands git-svn an
already-escaped URI.  Fix the regexp to skip over what looks like
existing percent escapes, and test this scenario.

Signed-off-by: Kevin Ballard <kevin@sb.org>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoremote show: do not show symbolic refs
Johannes Schindelin [Wed, 19 Mar 2008 00:27:42 +0000 (00:27 +0000)]
remote show: do not show symbolic refs

For symbolic refs, a sane notion of being "stale" is that the ref
they point to no longer exists.  Since this is checked already,
"remote show" does not need to show them at all.

Incidentally, this fixes the issue that "HEAD" was shown as a
stale ref by "remote show" in a freshly cloned repository.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocument the sendemail.smtpserverport config variable
Kevin Ballard [Wed, 19 Mar 2008 06:16:29 +0000 (02:16 -0400)]
Document the sendemail.smtpserverport config variable

Add sendemail.smtpserverport to the Configuration section
of the git-send-email manpage. It should probably be
referenced in the --smtp-server-port option as well.

Signed-off-by: Kevin Ballard <kevin@sb.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd --reverse to the git-rev-list usage string
Kevin Ballard [Wed, 19 Mar 2008 06:16:28 +0000 (02:16 -0400)]
Add --reverse to the git-rev-list usage string

git-rev-list accepts --reverse, as documented in
the manpage, but the usage string does not list it.

Signed-off-by: Kevin Ballard <kevin@sb.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agomake it easier for people who just want to get rid of 'git gc --auto'
Nicolas Pitre [Wed, 19 Mar 2008 21:06:11 +0000 (17:06 -0400)]
make it easier for people who just want to get rid of 'git gc --auto'

Give a direct hint to those who feel highly annoyed by the auto gc
behavior.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agobuiltin-gc.c: allow disabling all auto-gc'ing by assigning 0 to gc.auto
Brandon Casey [Wed, 19 Mar 2008 21:53:20 +0000 (16:53 -0500)]
builtin-gc.c: allow disabling all auto-gc'ing by assigning 0 to gc.auto

The gc.auto configuration variable is somewhat ambiguous now that there
is also a gc.autopacklimit setting. Some users may assume that it controls
all auto-gc'ing. Also, now users must set two configuration variables to
zero when they want to disable autopacking. Since it is unlikely that users
will want to autopack based on some threshold of pack files when they have
disabled autopacking based on the number of loose objects, be nice and allow
a setting of zero for gc.auto to disable all autopacking.

Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation/git-merge: document subtree strategy.
Miklos Vajna [Tue, 18 Mar 2008 12:26:43 +0000 (13:26 +0100)]
Documentation/git-merge: document subtree strategy.

There was already some documentation about subtree under
Documentation/howto but it was missing from git-merge manpage.

Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix tag following
Daniel Barkalow [Tue, 18 Mar 2008 02:15:02 +0000 (22:15 -0400)]
Fix tag following

Before the second fetch-pack connection in the same process, unmark
all of the objects marked in the first connection, in order that we'll
list them as things we have instead of thinking we've already
mentioned them.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMake revision limiting more robust against occasional bad commit dates
Linus Torvalds [Tue, 18 Mar 2008 01:56:33 +0000 (18:56 -0700)]
Make revision limiting more robust against occasional bad commit dates

The revision limiter uses the commit date to decide when it has seen
enough commits to finalize the revision list, but that can get confused
if there are incorrect dates far in the past on some commits.

This makes the logic a bit more robust by

 - we always walk an extra SLOP commits from the source list even if we
   decide that the source list is probably all done (unless the source is
   entirely empty, of course, because then we really can't do anything at
   all)

 - we keep track of the date of the last commit we added to the
   destination list (this will *generally* be the oldest entry we've seen
   so far)

 - we compare that with the youngest entry (the first one) of the source
   list, and if the destination is older than the source, we know we want
   to look at the source.

which causes occasional date mishaps to be handled cleanly.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix t3200 config
Daniel Barkalow [Tue, 18 Mar 2008 02:04:40 +0000 (22:04 -0400)]
Fix t3200 config

"git-config name = value" doesn't do anything most of the time. The
test meant "git-config name value", but that leaves the configuration
such that later tests will be confused, so move it to the end.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoTighten refspec processing
Daniel Barkalow [Tue, 18 Mar 2008 02:05:23 +0000 (22:05 -0400)]
Tighten refspec processing

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.

Signed-off-by: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix possible Solaris problem in 'checkout_entry()'
Linus Torvalds [Mon, 17 Mar 2008 15:56:27 +0000 (08:56 -0700)]
Fix possible Solaris problem in 'checkout_entry()'

Currently when checking out an entry "path", we try to unlink(2) it first
(because there could be stale file), and if there is a directory there,
try to deal with it (typically we run recursive rmdir).  We ignore the
error return from this unlink because there may not even be any file
there.

However if you are root on Solaris, you can unlink(2) a directory
successfully and corrupt your filesystem.

This moves the code around and check the directory first, and then
unlink(2).  Also we check the error code from it.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix read-tree not to discard errors
Junio C Hamano [Wed, 19 Mar 2008 05:01:28 +0000 (22:01 -0700)]
Fix read-tree not to discard errors

This fixes the issue identified with recently added tests to t1004

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd tests to catch problems with un-unlinkable symlinks
Junio C Hamano [Wed, 19 Mar 2008 04:59:39 +0000 (21:59 -0700)]
Add tests to catch problems with un-unlinkable symlinks

This currently fails not because we refuse to check out, but because we
detect error but incorrectly discard it in the callchain.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoTest: catch if trash cannot be removed
Junio C Hamano [Wed, 19 Mar 2008 04:58:01 +0000 (21:58 -0700)]
Test: catch if trash cannot be removed

When your test creates an unwritable directory that test framework cannot
clean out by "rm -fr trash", later tests cannot start in a fresh state
they expect to.  Detect this and error out early.

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