Since git 2.0.0 starting git gui in a submodule using a gitfile fails with
the following error:
No working directory ../../../<path>
couldn't change working directory
to "../../../<path>": no such file or
directory
This is because "git rev-parse --show-toplevel" is only run when git gui
sees a git version of at least 1.7.0 (which is the version in which the
--show-toplevel option was introduced). But "package vsatisfies" returns
false when the major version changes, which is not what we want here.
Fix that for both places where the git version is checked using vsatisfies
by appending a '-' to the version number. This tells vsatisfies that a
change of the major version is not considered to be a problem, as long as
the new major version is larger. This is done for both the place that
caused the reported bug and another spot where the git version is tested
for another feature.
Reported-by: Chris Packham <judge.packham@gmail.com>
Reported-by: Yann Dirson <ydirson@free.fr>
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de>
Currently setting submodule.<name>.ignore and/or diff.ignoreSubmodules to
"all" suppresses all output of submodule changes for git-gui. This is
really confusing, as even when the user chooses to record a new commit for
an ignored submodule by adding it manually this change won't show up under
"Staged Changes (Will Commit)".
Fix that by using the '--ignore-submodules=dirty' option for both callers
of "git diff-index --cached" when the underlying git version supports that
option.
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
The Makefile only runs po/po2msg.sh using tclsh, but because the
script has the usual tcl preamble starting with #!/bin/sh it can also
be run directly.
The Windows git-gui wrapper is usable in-place for the same reason.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Since 918dbf58, git-gui crashes if started with Tk 8.4. The reason is that
tk < 8.5 does not support -stretch option for panedwindow.
Without the option it's not possible to properly expand the right half -
the commit area is expanded, while desired behavior is to expand the diff
area. So the whole feature should be disabled with Tk
version less than 8.5.
Signed-off-by: Max Kirillov <max@max630.net>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When using git-gui as the primary git application on Windows it can be
awkward obtaining a suitable shell. This commit adds a menu item to the
Repository menu that launches the bash shell provided with the git
installation on Windows.
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Under cygwin the _gitworktree variable needs to contain the Windows
style path string so the output provided by git rev-parse must
be converted from cygwin path style to native.
Reviewed-by: Jesse Welch <jesse.welch@baml.com>
Signed-off-by: John Patrick Murphy <john.murphy@baml.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
For long descriptions it would be nice to be able to resize
the comment text field.
Signed-off-by: Max Kirillov <max@max630.net>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When git is used to track only a subset of a directory, or
there is no sure way to divide files to ignore from files to track,
git user have to live with large number of untracked files. These files
present in file list, and should always be scrolled through
to handle real changes. Situation can become even worse, then number
of the untracked files grows above the maxfilesdisplayed limit. In the
case, even staged can be hidden by git-gui.
This change introduces new configuration variable gui.displayuntracked,
which, when set to false, instructs git-gui not to show untracked files
in files list. They can be staged from commandline or other tools (like
IDE of file manager), then they become visible. Default value of the
option is true, which is compatible with current behavior.
Signed-off-by: Max Kirillov <max@max630.net>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
The list of recently opened repositories shown when launching git-gui from
outside a repository was hard coded to only show a maximum of 10 items.
This config variable allows the user to override this default.
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When the last remote is removed on a system that has tearoff menu items
the code that adjusts the fetch and prune menus may raise an error when
probing the menu entry for a non-existing -label option.
Check the entry type to avoid this fault.
Reported-by: Vedran Miletić <rivanvx@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Commit e3d06ca (git-gui: Detect full path when parsing arguments -
2012-10-02) fixed the handling of absolute paths passed to the browser
and blame subcommands by checking whether the file exists without the
prefix before prepending the prefix and checking again. Since we have
chdir'd to the top level of the working tree before doing this, this
does not work if a file with the same name exists in a subdirectory and
at the top level (for example Makefile in git.git's t/ directory).
Instead of doing this, revert that patch and fix absolute path issue by
using "file join" to prepend the prefix to the supplied path. This will
correctly handle absolute paths by skipping the prefix in that case.
Acked-by: Andrew Wong <andrew.kw.w@gmail.com>
Signed-off-by: John Keeping <john@keeping.me.uk>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
On Mac OS X, any application that is started from the Terminal will open
behind all running applications; as a work-around, manually bring ourselves
to the front. (Stolen from gitk, commit 76bf6ff93e.)
We do this as the very first thing, so that any message boxes that might pop
up during the rest of the startup sequence are actually seen by the user.
[PT: added catch and moved down to ensure Tk has been loaded]
Signed-off-by: Stefan Haller <stefan@haller-berlin.de>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Counting of lines did not skip this line when generating the hunk
header.
Acked-by: Tobias Preuss <tobias.preuss@googlemail.com>
Signed-off-by: Heiko Voigt <hvoigt@hvoigt.net>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When using Beyond Compare as the mergetool it fails to save the merged
result correctly due to a quoting problem when executing the tool.
This patch solves the quoting problem.
Signed-off-by: Warren Falk <warren@warrenfalk.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
This is clearer to many people this way.
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Commit e3d06ca9 (git-gui: Detect full path when parsing arguments) broke
git gui blame rev path-not-present-in-worktree
in particular this does not work anymore:
# in linux.git
$ git gui blame 2bb8c26242c2393b097a993ffe9b003ec9b85395 drivers/net/sky2.c
This patch restores the original functionality in this case.
Acked-by: Andrew Wong <andrew.kw.w@gmail.com>
Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When argument parsing fails to detect a file name, "git-gui" will try to
use the previously detected "head" as the file name. We should avoid
prepending the prefix if "head" looks like a full path.
Signed-off-by: Andrew Wong <andrew.kw.w@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When running "git-gui blame" from a subfolder (which means prefix is
non-empty), if we pass a full path as argument, the argument parsing
will fail to recognize the argument as a file name, because prefix is
prepended to the argument.
This patch handles that scenario by adding an additional branch that
checks the file name without using the prefix.
Signed-off-by: Andrew Wong <andrew.kw.w@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Adding _git_ps1() to one's bash prompt displays various repo status
info after each command. After committing a git cherry-pick -n using
git-gui, the prompt still contains the "|CHERRY-PICKING" flag.
Delete the file causing this flag when cleaning up.
Signed-off-by: Beat Bolli <bbolli@ewanet.ch>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Make Ctrl+U for unstaging and Ctrl+J for reverting selection behave
more like Ctrl+T for adding.
They were working only when one area was focused (diff or commit message),
now they should work everywhere.
Acked-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Vitaly _Vi Shukela <vi0oss@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
According to the translation in git-core, we
translate "remote" as "extern".
Signed-off-by: Ralf Thielow <ralf.thielow@googlemail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
According to the translation in git-core, we
translate "bare" as "bloß".
Signed-off-by: Ralf Thielow <ralf.thielow@googlemail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
The user might not really know what hook is
actually meant if it's translated. To avoid such
a confusion we should consistently write it untranslated
within braces after.
Signed-off-by: Ralf Thielow <ralf.thielow@googlemail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
The commit message buffer is automatically preserved to a local file
but this uses the system encoding which may fail to properly encode
unicode text. Forcing this file to use utf-8 preserves the message
correctly.
Reported-by: Ángel José Riesgo <ajriesgo@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When starting a gui program on windows stdout, stderr and stdin are not
connected to the cmd console. As a workaround tk has a console window.
Lets open this when the --trace commandline option has been given.
This is helpful for debugging.
Signed-off-by: Heiko Voigt <hvoigt@hvoigt.net>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
"succeeded" was misspelled in the code, which propagated throughout the
translations.
Fixed all of them.
Signed-off-by: Benjamin Kerensa <bkerensa <at> ubuntu.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
The current working directory is set to / when git-gui is invoked
using the Git Gui.app bundle on Mac OS X. This means that if it is
launched from a directory which contains a repository then git-gui
won't automatically find it unless the repository happens to be
located in /.
The PWD environment variable is however preserved if the bundle is
invoked using open(1). If git-gui would check for PWD then a user
could for example type open -a 'Git Gui' on a command line in order to
launch the program and it would automatically find the repository.
Teach git-gui to use the PWD environment variable on Mac OS X.
Signed-off-by: Marcus Karlsson <mk@acc.umu.se>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Recently, a clone initiated via git gui on Windows crashed on me due to
an "unknown variable cdone". It turns out that there is a code path
where this variable is used uninitialized.
Signed-off-by: Clemens Buchacher <drizzd@aon.at>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
When loading a file into the blame window git-gui does all the work and
must handle the text conversion filters if defined. On Windows it is
necessary to detect the need for a shell script explicitly.
Such filter commands are run using non-blocking I/O but this has the
unfortunate side effect of losing any error that might be reported when
the pipe is closed. Switching to blocking mode just before closing
enables reporting of errors in the filter scripts to the user.
Tested-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
It would be better if the 32x32 icon was equivalent to the one used on
Windows (in git-gui.ico), but I'm not sure how that would best be done,
so I copied this code from gitk instead.
Signed-off-by: Samuel Bronson <naesten@gmail.com>
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
On startup in multicommit mode git-gui checks to see if the repository
has a lot of objects. If so it shows a dialog suggesting gc be run.
This adds 'gui.gcwarning' as a control config variable to allow this
to be disabled. The default is true (the warning is shown). Setting this
false will prevent the check being done.
Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
Jens Lehmann