git.git
16 years agoAdd support for GIT_CEILING_DIRECTORIES
David Reiss [Tue, 20 May 2008 06:49:26 +0000 (23:49 -0700)]
Add support for GIT_CEILING_DIRECTORIES

Make git recognize a new environment variable that prevents it from
chdir'ing up into specified directories when looking for a GIT_DIR.
Useful for avoiding slow network directories.

For example, I use git in an environment where homedirs are automounted
and "ls /home/nonexistent" takes about 9 seconds.  Setting
GIT_CEILING_DIRS="/home" allows "git help -a" (for bash completion) and
"git symbolic-ref" (for my shell prompt) to run in a reasonable time.

Signed-off-by: David Reiss <dreiss@facebook.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFold test-absolute-path into test-path-utils
David Reiss [Tue, 20 May 2008 06:49:00 +0000 (23:49 -0700)]
Fold test-absolute-path into test-path-utils

Signed-off-by: David Reiss <dreiss@facebook.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoImplement normalize_absolute_path
David Reiss [Tue, 20 May 2008 06:48:54 +0000 (23:48 -0700)]
Implement normalize_absolute_path

normalize_absolute_path removes several oddities form absolute paths,
giving nice clean paths like "/dir/sub1/sub2".  Also add a test case
for this utility, based on a new test program (in the style of test-sha1).

Signed-off-by: David Reiss <dreiss@facebook.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMakefile: update the default build options for AIX
Mike Ralphson [Wed, 7 May 2008 08:35:55 +0000 (09:35 +0100)]
Makefile: update the default build options for AIX

NO_MKDTEMP is required to build, FREAD_READS_DIRECTORIES and the definition
of _LARGE_FILES fix test suite failures and INTERNAL_QSORT is required for
adequate performance.

Tested on AIX v5.3 Maintenance Level 06

Signed-off-by: Mike Ralphson <mike@abacus.co.uk>
Tested-by: Johannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'maint'
Junio C Hamano [Sat, 17 May 2008 05:10:13 +0000 (22:10 -0700)]
Merge branch 'maint'

* maint:
  git-filter-branch: Clarify file removal example.

16 years agoReplace in-place sed in t7502-commit
Marcel Koeppen [Fri, 16 May 2008 00:21:43 +0000 (02:21 +0200)]
Replace in-place sed in t7502-commit

The in-place mode of sed used in t7502-commit is a non-POSIX extension.
That call of sed is replaced by a more portable version using a temporary file.

Signed-off-by: Marcel Koeppen <git-dev@marzelpan.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-filter-branch: Clarify file removal example.
Jon Loeliger [Fri, 16 May 2008 19:43:50 +0000 (14:43 -0500)]
git-filter-branch: Clarify file removal example.

Signed-off-by: Jon Loeliger <jdl@jdl.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agosubmodule update: add convenience option --init
Johannes Schindelin [Fri, 16 May 2008 10:23:03 +0000 (11:23 +0100)]
submodule update: add convenience option --init

When a submodule is not initialized and you do not want to change the
defaults from .gitmodules anyway, you can now say

$ git submodule update --init <name>

When "update" is called without --init on an uninitialized submodule,
a hint to use --init is printed.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-fast-import: rename cmd_*() functions to parse_*()
Miklos Vajna [Thu, 15 May 2008 22:35:56 +0000 (00:35 +0200)]
git-fast-import: rename cmd_*() functions to parse_*()

There is a cmd_merge() function in fast-import that will conflict with
builtin-merge's cmd_merge() function. To keep it consistent, rename all
cmd_*() function to parse_*()

Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
Acked-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation/git-web--browse.txt: fix small typo
Teemu Likonen [Thu, 15 May 2008 17:52:22 +0000 (20:52 +0300)]
Documentation/git-web--browse.txt: fix small typo

Change "brower.konqueror.path" to "browser.konqueror.path" in
documentation.

Signed-off-by: Teemu Likonen <tlikonen@iki.fi>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoUse '-f' option to point to the .gitmodules file
Imran M Yousuf [Thu, 15 May 2008 07:42:58 +0000 (13:42 +0600)]
Use '-f' option to point to the .gitmodules file

'git config' has a '-f' option that takes the file to parse.
Using it rather than the environment variable seems more logical
and simplified.

Signed-off-by: Imran M Yousuf <imyousuf@smartitengineering.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge git://repo.or.cz/git-gui
Junio C Hamano [Thu, 15 May 2008 08:31:15 +0000 (01:31 -0700)]
Merge git://repo.or.cz/git-gui

* git://repo.or.cz/git-gui:
  git-gui: Delete branches with 'git branch -D' to clear config
  git-gui: Setup branch.remote,merge for shorthand git-pull
  git-gui: Update German translation
  git-gui: Don't use '$$cr master' with aspell earlier than 0.60
  git-gui: Report less precise object estimates for database compression

16 years agoDocumentation/git-prune.txt: document unpacked logic
Chris Frey [Wed, 14 May 2008 23:22:18 +0000 (19:22 -0400)]
Documentation/git-prune.txt: document unpacked logic

Clarifies the git-prune man page, documenting that it only
prunes unpacked objects.

Signed-off-by: Chris Frey <cdfrey@foursquare.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge git://git.kernel.org/pub/scm/gitk/gitk
Junio C Hamano [Thu, 15 May 2008 04:27:55 +0000 (21:27 -0700)]
Merge git://git.kernel.org/pub/scm/gitk/gitk

* git://git.kernel.org/pub/scm/gitk/gitk: (44 commits)
  gitk: Add a progress bar for checking out a head
  gitk: Show current row number and total number of rows
  gitk: Allow users to view diffs in external diff viewer
  gitk: Synchronize highlighting in file view for 'f' and 'b' commands
  gitk: Make updates go faster
  gitk: Disable "Reset %s branch to here" when on a detached head
  gitk: German translation again updated
  gitk: Update German translation
  gitk: Makefile/install: force permissions when installing files and dirs
  gitk: Initial Swedish translation.
  gitk: Spanish translation of gitk
  gitk: Fix handling of tree file list with special chars in names
  gitk: Reorganize processing of arguments for git log
  gitk: Fix problem with target row not being in scroll region
  gitk: Avoid a crash in selectline if commitinfo($id) isn't set
  gitk: Fix some corner cases in computing vrowmod and displayorder
  gitk: Correct a few strings and comments to say "git log"
  gitk: Don't filter view arguments through git rev-parse
  gitk: Fix problems with target row stuff
  gitk: Handle updating with path limiting better
  ...

16 years agoMerge branch 'maint'
Junio C Hamano [Wed, 14 May 2008 20:55:17 +0000 (13:55 -0700)]
Merge branch 'maint'

* maint:
  Documentation/git-describe.txt: make description more readable

16 years agoMerge branch 'maint-1.5.4' into maint
Junio C Hamano [Wed, 14 May 2008 20:46:42 +0000 (13:46 -0700)]
Merge branch 'maint-1.5.4' into maint

* maint-1.5.4:
  Documentation/git-describe.txt: make description more readable

16 years agoMerge branch 'sb/committer'
Junio C Hamano [Wed, 14 May 2008 20:45:20 +0000 (13:45 -0700)]
Merge branch 'sb/committer'

* sb/committer:
  commit: Show committer if automatic
  commit: Show author if different from committer
  Preparation to call determine_author_info from prepare_to_commit

16 years agoMerge branch 'bd/tests'
Junio C Hamano [Wed, 14 May 2008 20:45:16 +0000 (13:45 -0700)]
Merge branch 'bd/tests'

* bd/tests:
  Rename the test trash directory to contain spaces.
  Fix tests breaking when checkout path contains shell metacharacters
  Don't use the 'export NAME=value' in the test scripts.
  lib-git-svn.sh: Fix quoting issues with paths containing shell metacharacters
  test-lib.sh: Fix some missing path quoting
  Use test_set_editor in t9001-send-email.sh
  test-lib.sh: Add a test_set_editor function to safely set $VISUAL
  git-send-email.perl: Handle shell metacharacters in $EDITOR properly
  config.c: Escape backslashes in section names properly
  git-rebase.sh: Fix --merge --abort failures when path contains whitespace

Conflicts:

t/t9115-git-svn-dcommit-funky-renames.sh

16 years agoMerge branch 'mv/format-cc'
Junio C Hamano [Wed, 14 May 2008 20:34:34 +0000 (13:34 -0700)]
Merge branch 'mv/format-cc'

* mv/format-cc:
  Add tests for sendemail.cc configuration variable
  git-send-email: add a new sendemail.cc configuration variable
  git-format-patch: add a new format.cc configuration variable

16 years agoMerge branch 'cc/hooks-doc'
Junio C Hamano [Wed, 14 May 2008 20:34:23 +0000 (13:34 -0700)]
Merge branch 'cc/hooks-doc'

* cc/hooks-doc:
  Documentation: rename "hooks.txt" to "githooks.txt" and make it a man page

16 years agoMerge branch 'jk/renamelimit' (early part)
Junio C Hamano [Wed, 14 May 2008 19:37:28 +0000 (12:37 -0700)]
Merge branch 'jk/renamelimit' (early part)

* 'jk/renamelimit' (early part):
  diff: make "too many files" rename warning optional
  bump rename limit defaults
  add merge.renamelimit config option

16 years agoDocumentation/git-describe.txt: make description more readable
Ian Hilt [Wed, 14 May 2008 18:30:55 +0000 (14:30 -0400)]
Documentation/git-describe.txt: make description more readable

Signed-off-by: Ian Hilt <ian.hilt@gmail.com>
Credit-to: Kevin Ballard <kevin@sb.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agofilter-branch: fix variable export logic
Jeff King [Tue, 13 May 2008 08:46:38 +0000 (04:46 -0400)]
filter-branch: fix variable export logic

filter-branch tries to restore "old" copies of some
environment variables by using the construct:

  unset var
  test -z "$old_var" || var="$old_var" && export var

This is just wrong.  AND-list and OR-list operators && and || have equal
precedence and they bind left to right.  The second term, var="$old"
assignment always succeeds, so we always end up exporting var.

On bash and dash, exporting an unset variable has no effect. However, on
some shells (such as FreeBSD's /bin/sh), the shell exports the empty
value.

This manifested itself in this case as git-filter-branch setting
GIT_INDEX_FILE to the empty string, which in turn caused its call to
git-read-tree to fail, leaving the working tree pointing at the original
HEAD instead of the rewritten one.

To fix this, we change the short-circuit logic to better match the intent:

  test -z "$old_var" || {
    var="$old_var" && export var
  }

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoclone: bsd shell portability fix
Jeff King [Tue, 13 May 2008 08:45:56 +0000 (04:45 -0400)]
clone: bsd shell portability fix

When using /bin/sh from FreeBSD 6.1, the value of $? is lost
when calling a function inside the 'trap' action. This
resulted in clone erroneously indicating success when it
should have reported failure.

As a workaround, we save the value of $? before calling any
functions.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agot5000: tar portability fix
Jeff King [Tue, 13 May 2008 08:45:32 +0000 (04:45 -0400)]
t5000: tar portability fix

The output of 'tar tv' varies from system to system. In
particular, the t5000 was expecting to parse the date from
something like:

  -rw-rw-r-- root/root         0 2008-05-13 04:27 file

but FreeBSD's tar produces this:

  -rw-rw-r--  0 root   root        0 May 13 04:27 file

Instead of relying on tar's output, let's just extract the
file using tar and stat the result using perl.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agofix bsd shell negation
Jeff King [Wed, 14 May 2008 04:01:22 +0000 (00:01 -0400)]
fix bsd shell negation

On some shells (notably /bin/sh on FreeBSD 6.1), the
construct

  foo && ! bar | baz

is true if

  foo && baz

whereas for most other shells (such as bash) is true if

  foo && ! baz

We can work around this by specifying

  foo && ! (bar | baz)

which works everywhere.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoImprove reporting of errors in config file routines
Alex Riesen [Mon, 12 May 2008 21:41:04 +0000 (23:41 +0200)]
Improve reporting of errors in config file routines

Signed-off-by: Alex Riesen <raa.lkml@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'gp/bisect-fix'
Junio C Hamano [Mon, 12 May 2008 22:44:43 +0000 (15:44 -0700)]
Merge branch 'gp/bisect-fix'

* gp/bisect-fix:
  bisect: print an error message when "git rev-list --bisect-vars" fails
  git-bisect.sh: don't accidentally override existing branch "bisect"

16 years agogitk: Add a progress bar for checking out a head
Paul Mackerras [Mon, 12 May 2008 00:18:38 +0000 (10:18 +1000)]
gitk: Add a progress bar for checking out a head

Now that git checkout reports progress when checking out files, we
can use that to provide a progress bar in gitk.  We re-use the green
progress bar (formerly used when reading stuff in) for that.

Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agoMerge branch 'maint'
Junio C Hamano [Sun, 11 May 2008 19:09:18 +0000 (12:09 -0700)]
Merge branch 'maint'

* maint:
  wt-status.h: declare global variables as extern
  builtin-commit.c: add -u as short name for --untracked-files
  git-repack: re-enable parsing of -n command line option

16 years agoMerge branch 'maint-1.5.4' into maint
Junio C Hamano [Sun, 11 May 2008 19:09:12 +0000 (12:09 -0700)]
Merge branch 'maint-1.5.4' into maint

* maint-1.5.4:
  wt-status.h: declare global variables as extern
  builtin-commit.c: add -u as short name for --untracked-files
  git-repack: re-enable parsing of -n command line option

16 years agoMerge branch 'lt/core-optim'
Junio C Hamano [Sun, 11 May 2008 19:08:20 +0000 (12:08 -0700)]
Merge branch 'lt/core-optim'

* lt/core-optim:
  Optimize symlink/directory detection
  Avoid some unnecessary lstat() calls
  is_racy_timestamp(): do not check timestamp for gitlinks
  diff-lib.c: rename check_work_tree_entity()
  diff: a submodule not checked out is not modified
  Add t7506 to test submodule related functions for git-status
  t4027: test diff for submodule with empty directory
  Make git-add behave more sensibly in a case-insensitive environment
  When adding files to the index, add support for case-independent matches
  Make unpack-tree update removed files before any updated files
  Make branch merging aware of underlying case-insensitive filsystems
  Add 'core.ignorecase' option
  Make hash_name_lookup able to do case-independent lookups
  Make "index_name_exists()" return the cache_entry it found
  Move name hashing functions into a file of its own
  Make unpack_trees_options bit flags actual bitfields

16 years agorev-parse --verify: do not output anything on error
Christian Couder [Sun, 11 May 2008 16:28:25 +0000 (18:28 +0200)]
rev-parse --verify: do not output anything on error

Before this patch, when "git rev-parse --verify" was passed at least one
good rev and then anything, it would output something for the good rev
even if it would latter exit on error.

With this patch, we only output something if everything is ok.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agorev-parse: fix using "--default" with "--verify"
Christian Couder [Sun, 11 May 2008 16:27:36 +0000 (18:27 +0200)]
rev-parse: fix using "--default" with "--verify"

Before this patch, something like:

$ git rev-parse --verify HEAD --default master

did not work, while:

$ git rev-parse --default master --verify HEAD

worked.

This patch fixes that, so that they both work (assuming
HEAD and master can be parsed).

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agorev-parse: add test script for "--verify"
Christian Couder [Sun, 11 May 2008 16:27:10 +0000 (18:27 +0200)]
rev-parse: add test script for "--verify"

This patch documents the current behavior of "git rev-parse --verify".

This command is tested both with and without the "--quiet" and
"--default" options.

This shows some problems with the current behavior that will be fixed
in latter patches:

- in case of errors, there should be no good rev output on
stdout,
- with "--default" one test case is broken

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAdd svn-compatible "blame" output format to git-svn
Steven Grimm [Sun, 11 May 2008 05:11:18 +0000 (22:11 -0700)]
Add svn-compatible "blame" output format to git-svn

git-svn blame produced output in the format of git blame; in environments
where there are scripts that read the output of svn blame, it's useful
to be able to use them with the output of git-svn. The git-compatible
format is still available using the new "--git-format" option.

This also fixes a bug in the initial git-svn blame implementation; it was
bombing out on uncommitted local changes.

Signed-off-by: Steven Grimm <koreth@midwinter.com>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agomergetool: Make ECMerge use the settings as specified by the user in the GUI
Sebastian Schuberth [Tue, 6 May 2008 10:53:56 +0000 (12:53 +0200)]
mergetool: Make ECMerge use the settings as specified by the user in the GUI

When run from the command line, ECMerge does not automatically use the same
settings for a merge / diff that it would use when starting the GUI and loading
files manually. In the first case the built-in factory defaults would be used,
while in the second case the settings the user has specified in the GUI would
be used, which can be misleading. Specifying the "--default" command line
option changes this behavior so that always the user specfified GUI settings
are used.

Signed-off-by: Sebastian Schuberth <sschuberth@visageimaging.com>
Tested-by: Steffen Prohaska <prohaska@zib.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-format-patch: add --no-binary to omit binary changes in the patch.
Caio Marcelo de Oliveira Filho [Fri, 9 May 2008 22:55:43 +0000 (19:55 -0300)]
git-format-patch: add --no-binary to omit binary changes in the patch.

Add a new option --no-binary to git-format-patch so that no binary
changes are included in the generated patches, only notices that those
files changed.  This generate patches that cannot be applied, but still
is useful for generating mails for code review purposes.

See also: commit e47f306d4bf964def1a0b29e8f7cea419471dffd, where --binary
option was turned on by default.

Signed-off-by: Caio Marcelo de Oliveira Filho <cmarcelo@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agowt-status.h: declare global variables as extern
Johannes Sixt [Fri, 9 May 2008 08:05:27 +0000 (10:05 +0200)]
wt-status.h: declare global variables as extern

There are linkers out there that complain if a global non-static variable
is defined multiple times.

Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agobuiltin-commit.c: add -u as short name for --untracked-files
Sitaram Chamarty [Fri, 9 May 2008 17:12:55 +0000 (22:42 +0530)]
builtin-commit.c: add -u as short name for --untracked-files

This makes the C code consistent with the documentation and the old shell
code.

Signed-off-by: Sitaram Chamarty <sitaramc@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-repack: re-enable parsing of -n command line option
A Large Angry SCM [Sat, 10 May 2008 20:52:51 +0000 (16:52 -0400)]
git-repack: re-enable parsing of -n command line option

In commit 5715d0b (Migrate git-repack.sh to use git-rev-parse --parseopt,
2007-11-04), parsing of the '-n' command line option was accidentally lost
when git-repack.sh was migrated to use git-rev-parse --parseopt. This adds
it back.

Signed-off-by: A Large Angry SCM <gitzilla@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAllow tracking branches to set up rebase by default.
Dustin Sallings [Sat, 10 May 2008 22:36:29 +0000 (15:36 -0700)]
Allow tracking branches to set up rebase by default.

Change cd67e4d4 introduced a new configuration parameter that told
pull to automatically perform a rebase instead of a merge.  This
change provides a configuration option to enable this feature
automatically when creating a new branch.

If the variable branch.autosetuprebase applies for a branch that's
being created, that branch will have branch.<name>.rebase set to true.

Signed-off-by: Dustin Sallings <dustin@spy.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-svn: fix cloning of HTTP URLs with '+' in their path
Eric Wong [Sun, 11 May 2008 00:14:49 +0000 (17:14 -0700)]
git-svn: fix cloning of HTTP URLs with '+' in their path

With this, git svn clone -s http://svn.gnome.org/svn/gtk+
is successful.

Also modified the funky rename test for this, which _does_
include escaped '+' signs for HTTP URLs.  SVN seems to accept
either "+" or "%2B" in filenames and directories (just not the
main URL), so I'll leave it alone for now.

Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoalloc_ref_from_str(): factor out a common pattern of alloc_ref from string
Krzysztof Kowalczyk [Sat, 10 May 2008 23:26:58 +0000 (16:26 -0700)]
alloc_ref_from_str(): factor out a common pattern of alloc_ref from string

Also fix an underallocation in walker.c::interpret_target().

Signed-off-by: Krzysztof Kowalczyk <kkowalczyk@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogitk: Show current row number and total number of rows
Paul Mackerras [Sun, 11 May 2008 12:13:02 +0000 (22:13 +1000)]
gitk: Show current row number and total number of rows

This adds a couple of fields in the bar just below the upper panes
that show the row number of the currently selected commit, and how
many rows are displayed in total.  The latter increments as commits
are read in, and thus functions to show that progress is being made.
This therefore also removes the code that showed progress using a
green oscillating bar in the progress bar window (which some people
disliked).

Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agoOptimize symlink/directory detection
Linus Torvalds [Fri, 9 May 2008 16:21:07 +0000 (09:21 -0700)]
Optimize symlink/directory detection

This is the base for making symlink detection in the middle fo a pathname
saner and (much) more efficient.

Under various loads, we want to verify that the full path leading up to a
filename is a real directory tree, and that when we successfully do an
'lstat()' on a filename, we don't get a false positive due to a symlink in
the middle of the path that git should have seen as a symlink, not as a
normal path component.

The 'has_symlink_leading_path()' function already did this, and cached
a single level of symlink information, but didn't cache the _lack_ of a
symlink, so the normal behaviour was actually the wrong way around, and we
ended up doing an 'lstat()' on each path component to check that it was a
real directory.

This caches the last detected full directory and symlink entries, and
speeds up especially deep directory structures a lot by avoiding to
lstat() all the directories leading up to each entry in the index.

[ This can - and should - probably be extended upon so that we eventually
  never do a bare 'lstat()' on any path entries at *all* when checking the
  index, but always check the full path carefully. Right now we do not
  generally check the whole path for all our normal quick index
  revalidation.

  We should also make sure that we're careful about all the invalidation,
  ie when we remove a link and replace it by a directory we should
  invalidate the symlink cache if it matches (and vice versa for the
  directory cache).

  But regardless, the basic function needs to be sane to do that. The old
  'has_symlink_leading_path()' was not capable enough - or indeed the code
  readable enough - to really do that sanely. So I'm pushing this as not
  just an optimization, but as a base for further work. ]

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoAvoid some unnecessary lstat() calls
Linus Torvalds [Fri, 9 May 2008 16:11:43 +0000 (09:11 -0700)]
Avoid some unnecessary lstat() calls

The commit sequence used to do

if (file_exists(p->path))
add_file_to_cache(p->path, 0);

where both "file_exists()" and "add_file_to_cache()" needed to do a
lstat() on the path to do their work.

This cuts down 'lstat()' calls for the partial commit case by two
for each path we know about (because we do this twice per path).

Just move the lstat() to the caller instead (that's all that
"file_exists()" really does), and pass the stat information down to the
add_to_cache() function.

This essentially makes 'add_to_index()' the core function that adds a path
to the index, getting the index pointer, the pathname and the stat
information as arguments. There are then shorthand helper functions that
use this core function:

 - 'add_to_cache()' is just 'add_to_index()' with the default index

 - 'add_file_to_cache/index()' is the same, but does the lstat() call
   itself, so you can pass just the pathname if you don't already have the
   stat information available.

So old users of the 'add_file_to_xyzzy()' are essentially left unchanged,
and this just exposes the more generic helper function that can take
existing stat information into account.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'py/diff-submodule'
Junio C Hamano [Sun, 11 May 2008 01:16:25 +0000 (18:16 -0700)]
Merge branch 'py/diff-submodule'

* py/diff-submodule:
  is_racy_timestamp(): do not check timestamp for gitlinks
  diff-lib.c: rename check_work_tree_entity()
  diff: a submodule not checked out is not modified
  Add t7506 to test submodule related functions for git-status
  t4027: test diff for submodule with empty directory

16 years agoMerge branch 'lt/case-insensitive'
Junio C Hamano [Sun, 11 May 2008 01:14:28 +0000 (18:14 -0700)]
Merge branch 'lt/case-insensitive'

* lt/case-insensitive:
  Make git-add behave more sensibly in a case-insensitive environment
  When adding files to the index, add support for case-independent matches
  Make unpack-tree update removed files before any updated files
  Make branch merging aware of underlying case-insensitive filsystems
  Add 'core.ignorecase' option
  Make hash_name_lookup able to do case-independent lookups
  Make "index_name_exists()" return the cache_entry it found
  Move name hashing functions into a file of its own
  Make unpack_trees_options bit flags actual bitfields

16 years agoMerge branch 'master' into dev
Paul Mackerras [Sun, 11 May 2008 00:50:14 +0000 (10:50 +1000)]
Merge branch 'master' into dev

16 years agogitk: Allow users to view diffs in external diff viewer
Thomas Arcila [Mon, 24 Mar 2008 11:55:36 +0000 (12:55 +0100)]
gitk: Allow users to view diffs in external diff viewer

This allows gitk to run an external diff viewer such as meld.

Right-click on a file in the file list view gives "External diff"
popup menu entry, which launches the selected external diff tool.
The menu entry is only active in "Patch" mode, not in "Tree" mode.

The program to run to display the diff is configurable through
Edit/Preference/External diff tool.  The program is run with two
arguments, being the names of files containing the two versions to
diff.  Gitk will create temporary directories called
.gitk-tmp.<pid>/<n> to place these files in, and remove them when
it's finished.

If the file doesn't exist in one or other revision, gitk will supply
/dev/null as the name of the file on that side of the diff.  This may
need to be adjusted for Windows or MacOS.

[paulus@samba.org - cleaned up and rewrote some parts of the patch.]

Signed-off-by: Thomas Arcila <thomas.arcila@gmail.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agogitk: Synchronize highlighting in file view for 'f' and 'b' commands
Paul Mackerras [Sat, 10 May 2008 03:15:36 +0000 (13:15 +1000)]
gitk: Synchronize highlighting in file view for 'f' and 'b' commands

This is based on a patch by Eric Raible <raible@gmail.com>, but does
things a bit more simply.

Previously, 'b', backspace, and delete all did the same thing.
This changes 'b' to perform the inverse of 'f'.  And both of
them now highlight the filename of the currently diff.

This makes it easier to review and navigate the diffs associated
with a particular commit using only f, b, and space because the
filename of the currently display diff will be dynamically
highlighted.

Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agoMerge branch 'maint'
Junio C Hamano [Fri, 9 May 2008 03:50:03 +0000 (20:50 -0700)]
Merge branch 'maint'

* maint:
  Documentation/config.txt: Mention branch.<name>.rebase applies to "git pull"
  doc: clarify definition of "update" for git-add -u

16 years agoMerge branch 'maint-1.5.4' into maint
Junio C Hamano [Fri, 9 May 2008 03:12:44 +0000 (20:12 -0700)]
Merge branch 'maint-1.5.4' into maint

* maint-1.5.4:
  Documentation/config.txt: Mention branch.<name>.rebase applies to "git pull"
  doc: clarify definition of "update" for git-add -u

16 years agoMerge branch 'sg/merge-options' (early part)
Junio C Hamano [Fri, 9 May 2008 03:06:36 +0000 (20:06 -0700)]
Merge branch 'sg/merge-options' (early part)

* 'sg/merge-options' (early part):
  merge, pull: add '--(no-)log' command line option
  fmt-merge-msg: add '--(no-)log' options and 'merge.log' config variable
  add 'merge.stat' config variable
  merge, pull: introduce '--(no-)stat' option
  doc: moved merge.* config variables into separate merge-config.txt

16 years agoMerge branch 'db/learn-HEAD'
Junio C Hamano [Fri, 9 May 2008 03:06:23 +0000 (20:06 -0700)]
Merge branch 'db/learn-HEAD'

* db/learn-HEAD:
  Make ls-remote http://... list HEAD, like for git://...
  Make walker.fetch_ref() take a struct ref.

16 years agoMerge branch 'jn/webfeed'
Junio C Hamano [Fri, 9 May 2008 03:06:15 +0000 (20:06 -0700)]
Merge branch 'jn/webfeed'

* jn/webfeed:
  gitweb: Use feed link according to current view

16 years agoMerge branch 'cc/help'
Junio C Hamano [Fri, 9 May 2008 03:06:11 +0000 (20:06 -0700)]
Merge branch 'cc/help'

* cc/help:
  documentation: web--browse: add a note about konqueror
  documentation: help: add info about "man.<tool>.cmd" config var
  help: use "man.<tool>.cmd" as custom man viewer command
  documentation: help: add "man.<tool>.path" config variable
  help: use man viewer path from "man.<tool>.path" config var

16 years agoMerge branch 'dm/cherry-pick-s'
Junio C Hamano [Fri, 9 May 2008 03:06:06 +0000 (20:06 -0700)]
Merge branch 'dm/cherry-pick-s'

* dm/cherry-pick-s:
  Allow cherry-pick (and revert) to add signoff line

16 years agoMerge branch 'lt/dirmatch-optim'
Junio C Hamano [Fri, 9 May 2008 03:05:43 +0000 (20:05 -0700)]
Merge branch 'lt/dirmatch-optim'

* lt/dirmatch-optim:
  Optimize match_pathspec() to avoid fnmatch()

16 years agobisect: print an error message when "git rev-list --bisect-vars" fails
Christian Couder [Wed, 7 May 2008 21:54:28 +0000 (23:54 +0200)]
bisect: print an error message when "git rev-list --bisect-vars" fails

Before this patch no error was printed when "git rev-list --bisect-vars"
failed. This can happen when bad and good revs are mistaken.

This patch prints an error message on stderr that describe the likely
failure cause.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocompat-util: avoid macro redefinition warning
Johannes Sixt [Thu, 8 May 2008 07:34:49 +0000 (09:34 +0200)]
compat-util: avoid macro redefinition warning

Some systems define fopen as a macro based on compiler settings, and
unconditionally redefining it triggers a compilation warning.

16 years agocompat/fopen.c: avoid clobbering the system defined fopen macro
Brandon Casey [Wed, 7 May 2008 17:34:18 +0000 (12:34 -0500)]
compat/fopen.c: avoid clobbering the system defined fopen macro

Some systems define fopen as a macro based on compiler settings.
The previous technique for reverting to the system fopen function
by merely undefining fopen is inadequate in this case. Instead,
avoid defining fopen entirely when compiling this source file.

Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Tested-by: Mike Ralphson <mike@abacus.co.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-gui: Delete branches with 'git branch -D' to clear config
Shawn O. Pearce [Fri, 9 May 2008 00:29:42 +0000 (20:29 -0400)]
git-gui: Delete branches with 'git branch -D' to clear config

If we are deleting a local branch from refs/heads/ we need to
make sure any associated configuration stored in .git/config is
also removed (such as branch.$name.remote and branch.$name.merge).
The easiest way to do this is to use git-branch as that automatically
will look for and delete configuration keys as necessary.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agogit-gui: Setup branch.remote,merge for shorthand git-pull
Shawn O. Pearce [Fri, 9 May 2008 00:16:43 +0000 (20:16 -0400)]
git-gui: Setup branch.remote,merge for shorthand git-pull

When creating new branches if branch.autosetupmerge is not set, or
is set to true or always and we have been given a remote tracking
branch as the starting point for a new branch we want to create the
necessary configuration options in .git/config for the new branch
so that a no argument git-pull on the command line pulls from the
remote repository's branch.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
16 years agogitk: Make updates go faster
Paul Mackerras [Fri, 9 May 2008 00:14:07 +0000 (10:14 +1000)]
gitk: Make updates go faster

This goes back to the method of doing updates where we translate the
revisions we're given to SHA1 ids and then remove the ones we've asked
for before or that we've already come across.  This speeds up updates
enormously in most cases since it means git log doesn't have to traverse
large parts of the tree.  We used to do this, but it had bugs, and commit
468bcaedbb1589f16955e63b6bfba01c2f53e433 (gitk: Don't filter view
arguments through git rev-parse) went to the slower method to avoid the
bugs.

In order to do this properly, we have to parse the command line and
understand all the flag arguments.  So this adds a parser that checks
all the flag arguments.  If there are any we don't know about, we
disable the optimization and just pass the whole lot to git log
(except for -d/--date-order, which we remove from the list).

With this we can then use git rev-parse on the non-flag arguments to
work out exactly what SHA1 ids are included and excluded in the list,
which then enables us to ask for just the new ones when updating.
One wrinkle is that we have to turn symmetric diff arguments (of the
form a...b) back into symmetric diff form so that --left-right still
works, as git rev parse turns a...b into a b ^merge_base(a,b).

This also updates a couple of copyright notices.

Signed-off-by: Paul Mackerras <paulus@samba.org>
16 years agoDocumentation/config.txt: Mention branch.<name>.rebase applies to "git pull"
Dustin Sallings [Thu, 8 May 2008 18:28:07 +0000 (11:28 -0700)]
Documentation/config.txt: Mention branch.<name>.rebase applies to "git pull"

Signed-off-by: Dustin Sallings <dustin@spy.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agodoc: clarify definition of "update" for git-add -u
Jeff King [Thu, 8 May 2008 17:25:06 +0000 (13:25 -0400)]
doc: clarify definition of "update" for git-add -u

The "-u" option is described only in terms of "updating"
files, which in turn is described only as "similar to what
git commit -a does". Let's be a little more specific about
what updating entails.

Suggested by Geoffrey Irving.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation: bisect: add a few "git bisect run" examples
Christian Couder [Wed, 7 May 2008 23:00:54 +0000 (01:00 +0200)]
Documentation: bisect: add a few "git bisect run" examples

Before this patch, there were no "git bisect run" example.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation/config.txt: Add git-gui options
Gustaf Hendeby [Thu, 8 May 2008 08:55:02 +0000 (10:55 +0200)]
Documentation/config.txt: Add git-gui options

The 'git gui' has a number of options that can be specified using the
options dialog. Sometimes it is convenient to be able to specify these
from the command line, therefor document these options.

Signed-off-by: Gustaf Hendeby <hendeby@isy.liu.se>
Acked-by: Shawn O. Pearce <speace@spearce.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation: improve "add", "pull" and "format-patch" examples
Christian Couder [Wed, 7 May 2008 04:29:28 +0000 (06:29 +0200)]
Documentation: improve "add", "pull" and "format-patch" examples

Before this patch in "git-add.txt" and "git-format-patch.txt", the
commands used in the examples were "git-CMD" instead of "git CMD".
This patch fixes that.

In "git-pull.txt" only the last example had the code sample in an
asciidoc "Listing Block", and in the other two files, none.
This patch fixes that by putting all code samples in listing
blocks.

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoBe more careful with objects directory permissions on clone
Mark Hills [Mon, 5 May 2008 16:46:42 +0000 (17:46 +0100)]
Be more careful with objects directory permissions on clone

Honour the setgid and umask when re-creating the objects directory
at the destination.

cpio in copy-pass mode aims to copy file permissions which causes this
problem and cannot be disabled. Be explicit by copying the directory
structure first, honouring the permissions at the destination, then copy
the files with 0444 permissions. This also avoids bugs in some versions
of cpio.

Signed-off-by: Mark Hills <mark@pogo.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocommit: Show committer if automatic
Santi Béjar [Sun, 4 May 2008 16:04:51 +0000 (18:04 +0200)]
commit: Show committer if automatic

To warn the user in case he/she might be using an unintended
committer identity.

Signed-off-by: Santi Béjar <sbejar@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agocommit: Show author if different from committer
Santi Béjar [Sun, 4 May 2008 16:04:50 +0000 (18:04 +0200)]
commit: Show author if different from committer

That would help reassure anybody while committing other's changes.

Signed-off-by: Santi Béjar <sbejar@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoPreparation to call determine_author_info from prepare_to_commit
Santi Béjar [Sun, 4 May 2008 16:04:49 +0000 (18:04 +0200)]
Preparation to call determine_author_info from prepare_to_commit

Reorder functions definitions such that determine_author_info is
defined before prepare_to_commit. No code changes.

Signed-off-by: Santi Béjar <sbejar@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoMerge branch 'jc/lstat'
Junio C Hamano [Tue, 6 May 2008 02:16:26 +0000 (19:16 -0700)]
Merge branch 'jc/lstat'

* jc/lstat:
  diff-files: mark an index entry we know is up-to-date as such
  write_index(): optimize ce_smudge_racily_clean_entry() calls with CE_UPTODATE

16 years agoMerge branch 'bc/filter-branch'
Junio C Hamano [Tue, 6 May 2008 02:16:20 +0000 (19:16 -0700)]
Merge branch 'bc/filter-branch'

* bc/filter-branch:
  filter-branch.sh: support nearly proper tag name filtering

16 years agoMerge branch 'lh/git-file'
Junio C Hamano [Tue, 6 May 2008 02:16:16 +0000 (19:16 -0700)]
Merge branch 'lh/git-file'

* lh/git-file:
  Teach GIT-VERSION-GEN about the .git file
  Teach git-submodule.sh about the .git file
  Teach resolve_gitlink_ref() about the .git file
  Add platform-independent .git "symlink"

16 years agoMerge branch 'jk/fetch-status'
Junio C Hamano [Tue, 6 May 2008 02:16:12 +0000 (19:16 -0700)]
Merge branch 'jk/fetch-status'

* jk/fetch-status:
  git-fetch: always show status of non-tracking-ref fetches

16 years agoMerge branch 'lh/branch-merged'
Junio C Hamano [Tue, 6 May 2008 02:16:06 +0000 (19:16 -0700)]
Merge branch 'lh/branch-merged'

* lh/branch-merged:
  Add tests for `branch --[no-]merged`
  git-branch.txt: compare --contains, --merged and --no-merged
  git-branch: add support for --merged and --no-merged

16 years agoMerge branch 'pb/remote-mirror-config'
Junio C Hamano [Tue, 6 May 2008 02:15:39 +0000 (19:15 -0700)]
Merge branch 'pb/remote-mirror-config'

* pb/remote-mirror-config:
  Add a remote.*.mirror configuration option

16 years agopost-merge: Add it's not executed if merge failed.
Jörg Sommer [Mon, 5 May 2008 09:06:49 +0000 (11:06 +0200)]
post-merge: Add it's not executed if merge failed.

Signed-off-by: J��rg Sommer <joerg@alea.gnuu.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-bisect.sh: don't accidentally override existing branch "bisect"
Gerrit Pape [Mon, 5 May 2008 07:43:00 +0000 (07:43 +0000)]
git-bisect.sh: don't accidentally override existing branch "bisect"

If a branch named "bisect" or "new-bisect" already was created in the
repo by other means than git bisect, doing a git bisect used to override
the branch without a warning.  Now if the branch "bisect" or
"new-bisect" already exists, and it was not created by git bisect itself,
git bisect start fails with an appropriate error message.  Additionally,
if checking out a new bisect state fails due to a merge problem, git
bisect cleans up the temporary branch "new-bisect".

The accidental override has been noticed by Andres Salomon, reported
through
 http://bugs.debian.org/478647

Signed-off-by: Gerrit Pape <pape@smarden.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation: Add create-ignore to git svn manual
Gustaf Hendeby [Sun, 4 May 2008 22:33:10 +0000 (00:33 +0200)]
Documentation: Add create-ignore to git svn manual

Signed-off-by: Gustaf Hendeby <hendeby@isy.liu.se>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-svn: Make create-ignore use git add -f
Gustaf Hendeby [Sun, 4 May 2008 22:33:09 +0000 (00:33 +0200)]
git-svn: Make create-ignore use git add -f

When having a svn:ignore that ignores the .gitignore file the -f
option to git add must be used to avoid git complaining about adding
an ignored file and hence stop the process of creating .gitignores.

Signed-off-by: Gustaf Hendeby <hendeby@isy.liu.se>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoINSTALL: add a note about GNU interactive tools has been renamed
Miklos Vajna [Sun, 4 May 2008 14:55:11 +0000 (16:55 +0200)]
INSTALL: add a note about GNU interactive tools has been renamed

In recent versions GNU's git has been renamed to gnuit, document this
while talking about how to resolve the conflict.

Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoRename the test trash directory to contain spaces.
Bryan Donlan [Sun, 4 May 2008 05:38:00 +0000 (01:38 -0400)]
Rename the test trash directory to contain spaces.

In order to help prevent regressions in the future, rename the trash directory
for all tests to contain spaces. This patch also corrects two failures that
were caused or exposed by this change.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoFix tests breaking when checkout path contains shell metacharacters
Bryan Donlan [Sun, 4 May 2008 05:37:59 +0000 (01:37 -0400)]
Fix tests breaking when checkout path contains shell metacharacters

This fixes the remainder of the issues where the test script itself is at
fault for failing when the git checkout path contains whitespace or other
shell metacharacters.

The majority of git svn tests used the idiom

  test_expect_success "title" "test script using $svnrepo"

These were changed to have the test script in single-quotes:

  test_expect_success "title" 'test script using "$svnrepo"'

which unfortunately makes the patch appear larger than it really is.

One consequence of this change is that in the verbose test output the
value of $svnrepo (and in some cases other variables, too) is no
longer expanded, i.e. previously we saw

  * expecting success:
test script using /path/to/git/t/trash/svnrepo

but now it is:

  * expecting success:
test script using "$svnrepo"

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDon't use the 'export NAME=value' in the test scripts.
Bryan Donlan [Sun, 4 May 2008 05:37:58 +0000 (01:37 -0400)]
Don't use the 'export NAME=value' in the test scripts.

This form is not portable across all shells, so replace instances of:

  export FOO=bar

with:

  FOO=bar
  export FOO

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agolib-git-svn.sh: Fix quoting issues with paths containing shell metacharacters
Bryan Donlan [Sun, 4 May 2008 05:37:57 +0000 (01:37 -0400)]
lib-git-svn.sh: Fix quoting issues with paths containing shell metacharacters

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agotest-lib.sh: Fix some missing path quoting
Bryan Donlan [Sun, 4 May 2008 05:37:56 +0000 (01:37 -0400)]
test-lib.sh: Fix some missing path quoting

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoUse test_set_editor in t9001-send-email.sh
Bryan Donlan [Sun, 4 May 2008 05:37:55 +0000 (01:37 -0400)]
Use test_set_editor in t9001-send-email.sh

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agotest-lib.sh: Add a test_set_editor function to safely set $VISUAL
Bryan Donlan [Sun, 4 May 2008 05:37:54 +0000 (01:37 -0400)]
test-lib.sh: Add a test_set_editor function to safely set $VISUAL

In particular, this function correctly handles cases where the pwd contains
spaces, quotes, and other troublesome metacharacters.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-send-email.perl: Handle shell metacharacters in $EDITOR properly
Bryan Donlan [Sun, 4 May 2008 05:37:53 +0000 (01:37 -0400)]
git-send-email.perl: Handle shell metacharacters in $EDITOR properly

This fixes the git-send-perl semantics for launching an editor when
$GIT_EDITOR (or friends) contains shell metacharacters to match
launch_editor() in builtin-tag.c. If we use the current approach
(sh -c '$0 $@' "$EDITOR" files ...), we see it fails when $EDITOR has
shell metacharacters:

  $ sh -x -c '$0 $@' "$VISUAL" "foo"
  + "$FAKE_EDITOR" foo
  "$FAKE_EDITOR": 1: "$FAKE_EDITOR": not found

Whereas builtin-tag.c will invoke sh -c "$EDITOR \"$@\"".

Thus, this patch changes git-send-email.perl to use the same method as the
C utilities, and additionally updates t/t9001-send-email.sh to test for
this bug.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoconfig.c: Escape backslashes in section names properly
Bryan Donlan [Sun, 4 May 2008 05:37:52 +0000 (01:37 -0400)]
config.c: Escape backslashes in section names properly

If an element of the configuration key name other than the first or last
contains a backslash, it is not escaped on output, but is treated as an
escape sequence on input. Thus, the backslash is lost when re-loading
the configuration.

This patch corrects this by having backslashes escaped properly, and
introduces a new test for this bug.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agogit-rebase.sh: Fix --merge --abort failures when path contains whitespace
Bryan Donlan [Sun, 4 May 2008 05:37:51 +0000 (01:37 -0400)]
git-rebase.sh: Fix --merge --abort failures when path contains whitespace

Also update t/t3407-rebase-abort.sh to expose the bug.

Signed-off-by: Bryan Donlan <bdonlan@fushizen.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agoDocumentation: rename "hooks.txt" to "githooks.txt" and make it a man page
Christian Couder [Fri, 2 May 2008 03:30:47 +0000 (05:30 +0200)]
Documentation: rename "hooks.txt" to "githooks.txt" and make it a man page

Also now "gitcli(5)" becomes "gitcli(7)".

Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agois_racy_timestamp(): do not check timestamp for gitlinks
Junio C Hamano [Sun, 4 May 2008 00:24:28 +0000 (17:24 -0700)]
is_racy_timestamp(): do not check timestamp for gitlinks

Because we do not even check the timestamp to determie if a gitlink
is up to date or not, triggering the racy-timestamp check for gitlinks
does not make sense.

This fixes the recently added test in t7506.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agodiff-lib.c: rename check_work_tree_entity()
Junio C Hamano [Sun, 4 May 2008 00:23:46 +0000 (17:23 -0700)]
diff-lib.c: rename check_work_tree_entity()

The function is about checking for removed work tree item, so name it
accordingly to avoid future confusion.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
16 years agodiff: a submodule not checked out is not modified
Junio C Hamano [Sun, 4 May 2008 00:04:42 +0000 (17:04 -0700)]
diff: a submodule not checked out is not modified

948dd34 (diff-index: careful when inspecting work tree items, 2008-03-30)
made the work tree check careful not to be fooled by a new directory that
exists at a place the index expects a blob.  For such a change to be a
typechange from blob to submodule, the new directory has to be a
repository.

However, if the index expects a submodule there, we should not insist the
work tree entity to be a repository --- a simple directory that is not a
full fledged repository (even an empty directory would do) should be
considered an unmodified subproject, because that is how a superproject
with a submodule is checked out sparsely by default.

This makes the function check_work_tree_entity() even more careful not to
report a submodule that is not checked out as removed.  It fixes the
recently added test in t4027.

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