When the user selects a number of files they would typically expect
to be able to act on that selection, such as by including those files
into the next commit.
So we now have a menu option under the Commit menu that lets the user
include only the selection, rather than everything. If there is no
selection but there is a file in the diff viewer than we consider that
to be the selection (a selection of 1). Unfortunately we don't disable
this option yet when there's nothing selected to include, but this is
probably not a big deal as there are very few situations where there
are no selected files.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
It just felt wrong to me that I was using _ as part of the mode argument
to display_file to mean "don't care/use existing" and * as part of
the mode argument to mean "force to _".
So instead use ? to mean "don't care/use existing" and _ to mean
"force to _". The code is a lot clearer this way and hopefully it
won't drive another developer insane, as it did me.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I noticed that we were reshowing the current diff during a commit;
this occurs because we feed every added and modified file through
update-index just before commit. During the update-index process
we reshow the current diff if the current file in the diff pane
was one of those added or modified files we reprocessed. This
just slows down the UI more than is necessary.
So refactoring update_index so that we don't call reshow_diff
from within that code; instead we do it at a higher level.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Apparently I never really tested the logic for making or amending an
initial commit, so although most of the code was here in git-gui it
didn't quite work as it was intended to.
So this is all just bug fixes to make initial commits correctly
generate the list of files going into the initial commit, or to
show a newly added file's diff, and to amend an initial commit.
Because we really want to diff the index against a tree-ish and
there is no such tree-ish on an initial commit we create an empty
tree through git-mktree and diff against that. This unfortunately
creates a dangling tree, which may confuse a new user who uses
git-gui to make a new commit and then immediately afterwards runs
git fsck-objects to see if their object database is corrupt or not.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If we can't locate a .git directory for the given directory we need to
show a message to the user to let them know the directory wasn't found.
But since this is before we have shown our main application window we
cannot use that as the parent for the error popup; on Mac OS X this
causes an error and prevents the dialog from showing.
Instead only add -parent . to the popup call if we have mapped (shown)
the main window.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If a user works with a repository frequently they may want to just
create an icon they can use to launch git-gui against that repository.
Since we already support this concept on Windows we can do the same on
Mac OS X by creating a .app file with a tiny shell script in it that
sets up the necessary environment then invokes our script.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Don't offer to fetch from a remote unless we have at least one Pull:
line in its .git/remotes/<name> file or at least one configuration
value for remote.<name>.fetch. Ditto for push.
Users shouldn't be fetching or pushing branch groups unless they
have them configured; anything else is just crazy.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since we have some serious problems with the GIT_DIR environment variable
on Windows we cannot let the user use a non-standard GIT_DIR with their
working directory.
So require that the GIT_DIR name is actually ".git", that it exists,
and that its parent directory is our working directory.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If we are running on Windows we now offer a 'Create Desktop Icon' menu
item under the Project menu. This pops up a save dialog box letting
the user create a .bat file on their desktop (or somewhere else). The
.bat script will startup Cygwin with a login shell then launch git-gui
in the current working directory.
This is very useful for Windows users who have little to no desire to
start a command window just to run a git-gui session.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Some users may want to start us by running "git --git-dir=... gui"
rather than trying to cd into the directory first. This is especially
true if they want to just make a shortcut to our executable on Windows
and always have that associated with a certain repository.
Since Tcl on Windows throws away our environment and doesn't pass it
down to the child process correctly we cannot call git-rev-parse to
get the GIT_DIR environment variable. So instead we ask for it
specifically ourselves; if its not defined then we ask rev-parse.
This should actually reduce startup by 1 fork/exec if we were started
as "git gui" as GIT_DIR will be set for us.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
There is no reason why the user should be able to operate on the diff
buffer if there is no currently selected diff; likewise the "File:"
label text appears rather silly looking all by itself when no diff
is being shown in the diff buffer.
So now we only enable widgets (like menu items) if there is a diff
currently showing, and we disable them when a diff isn't showing.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If the user has "Allow Partially Included Files" disabled (and most
probably will as its the default setting) we should run update-index
on every included file before commit to make sure that any changes
made by the user since the last rescan will still be part of this
commit.
If we don't update-index every modified file the user will likely
become confused when part of their changes were committed and other
parts weren't; and those other parts won't show up until a later
rescan occurs. Since we don't rescan immediately after a commit
this may be a while.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Like rescan we also have cases where we need to perform a script
after we have finished updating a number of files in the index. By
changing the parameter structure of update_index we can easily pass
through any script we need to run afterwards, such as picking up
in the middle of a commit, or finishing what is left of a rescan.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
There are some situations where we need to run rescan and have it do
more than just updating the status in the UI when its complete. To
help with that this changes the rescan procedure to take a script which
it will run at the global level as soon as the rescan is done and the
UI has finished updating with the results. This is useful for example
if we performed a rescan as part of a commit operation; we can go back
to the commit where we left off when the rescan got initiated.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since we refer to the act of updating our memory structures with index
and working directory differences as a rescan in the UI its probably
a good idea to make the related procedures have the same name.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because I want to let users apply actions to more than one file at
a time we really needed a concept of "the current selection" from
the two file lists.
Since I'm abusing a Tk text widget for the file displays I can't
really use the Tk selection to track which files are picked and
which aren't. So instead we keep this in an array to tell us
which paths are currently selected and we use an inverse fg/bg
for the selected file display. This is common most operating
systems as a selection indicator.
The selection works like most users would expect; single click will
clear the selection and pick only that file, M1-click (aka Ctrl-click
or Cmd-click) will toggle the one file in/out of the selection, and
Shift-click will select the range between the last clicked file and
the currently clicked file.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
On Mac OS X the no differences informational message was linewrapped
at the wrong points due to the limited width of the system dialog,
yet the LFs embedded in the message (where I linewrapped it manually)
were also being honored. This resulted in a very difficult to read
paragraph of text.
So this narrows the text down by another 10 columns or so, making it
more readable.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I'm not a huge fan of putting the left and right mouse actions into
the same procedure. Originally this is how Paul had implemented the
logic in gitool and I had carried some of that over into git-gui, but
now that I'm getting ready to implement right mouse click features to
act on files I really should split this apart.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The concept of the Git index is confusing for many users, especially
those who are newer to Git.
Since git-gui is (at least partially) intended to be used by newer
users who don't need the complexity of the index to be put in front
of them early on, we should hide it by making any partially included
file fully included as soon as we identify it. To do this we just
run a quick update_index pass on any file which differs both in the
index and the working directory, as these files have already been
at least partially included by the user.
A new option has been added in the options dialog (gui.partialinclude)
which lets the user enable accessing the index from git-gui. This
just disables the automatic update_index pass on partially included
files.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
So although a text field with a flat relief looks like a label on
Windows it doesn't on Mac OS X. The Aqua version of Tk is still
drawing a border around the text field and that makes the diff pane
header look pretty ugly.
Earlier I had made the file name area into a text widget so the user
could highlight parts of it and copy them onto the clipboard; but with
the context menu being present this isn't quite as necessary as the user
can copy the file name to the clipboard using that instead. So although
this is a small loss in functionality for non-Mac OS X systems I think it
is still reasonable.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Moved the Close button over to the lower right corner where our
Cancel/Save buttons are in the options dialog. This should fit
better with our own look and feel as well as that of most apps
on Mac OS X and Windows.
Also set the lower status bar in a console window to indicate the
process is working and that the user should wait for it to finish.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because the Tk pack layout manager gives all space to the right/bottom
most widget during expand/contract of the frame we were adding and
removing all space from the status area of the bar and not from the
file name, which is what we actually wanted.
A simple enough fix is to just put the status of the given file on
the left side of the diff viewer header rather than on the right.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When I changed from 'check in' to 'include' I missed the human friendly
status displayed in the right side of the diff viewer heading. It was
still reporting 'Checked in' for a fully included file, which is not
what we wanted it to say.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
There's a lot of reasons why the user might need to obtain the
complete (or just part of) path of a file which they are currently
viewing in the diff viewer pane. So now we allow selection on this
widget by using a text widget instead of a label. We also offer a
context menu which has actions for copying the selection or the entire
value onto the clipboard.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we shove a large number of files at update-index and they have
very short path names we are likely going to fit a large number of
them into the pipe buffer very early; thereby seeing a huge progress
update followed by lots of waiting between progress updates due to
the latency of update-index.
Using a smaller buffer should help smooth out the progress updates
as we are better able to keep tabs on the update-index process'
progress through our list of paths.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Its a little surprising to see the UI update the icons for files
in random order, due to the fact that the files are updating in
the order they appear within the array (which is based on a hash
function and not order). So sort the list of files before we send
any to update-index so the order of operation is means something to
the user.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When displaying a diff the Git default of 3 line of context may not be
enough for a user to see what has actually changed. Consequently we
set our own program default to 5 lines of context and then allow the
user to adjust this on a per-repository and global level through our
options dialog.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I'd like to allow the user to have more control over how we format
the diff in the diff viewer; to that end we need to add additional
options to the diff-index command line as we construct the command
for execution.
So cleanup the command handling code now to use lappend so we can
come back and add in our additional options.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
We can't ask the diff viewer to recompute the diff until after our
update-index child process terminates, as the diff programs need to
be able to read the updated index in order to generate the correct
diff. This is actually why we prevent diffs from being generated
while there is an update lock on the index, which is why we ignored
our own show_diff invocation in the middle of the write_update_index
event handler.
So now we mark a flag if we identify that the file currently in the
diff viewer was also sent to update-index; then later when the
update-index process has terminated we update the diff viewer if
the flag is true.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
This is one of those stupid Tcl mistakes that an experienced Tcl
programmer just wouldn't make. We should always use eq and ne to
compare string values (and never == or !=) as when we use ==/!=
Tcl will attempt to convert either side to numeric if one of the
two sides looks like a numeric. This could cause some trouble if
a file named "1" exists and a different file named "1.0" also exists;
their paths are equal according to == but not according to eq.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since git-commit.sh invokes hooks/post-commit after running git rerere
we should do the same if its available and executable.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
We were originally trying to use $commit_active to tell us if there was
a commit currently in progress, just so we didn't attempt to start a
second (parallel) one by mistake. But really the index lock handles
this for us as it won't let us lock the index if it is already locked
for update. So this can't happen.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I started to notice on Windows that commits took a lot longer to get
going than on my Mac OS X system. The real reason is the repositories
that I'm testing with on Windows all enabled the standard pre-commit hook
while my test repository on Mac OS X doesn't have it executable (so its
not running). So the Windows repositories are spending this
lag time running that hook.
Now we run the pre-commit hook in the background, allowing the UI to
update and tell the user we are busy doing things.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because the pull diffstat summary can take as long as the pull itself
some users may just choose to disable the summary and save themselves
an extra few seconds during each pull. This is especially true if the
user really doesn't care about the other files being modified, as due
to their project organizational structure they aren't really responsible
for their content.
This adds an option to the options panel which lets the user disable
the diffstat summary (and thus we pass --no-summary to git-pull) but
there does appear to be a bug in the config saving code where we did
not set the local repo config differently from the global config.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since git-repo-config will supply us a union of both the global and
the local repository configuration data when we invoke it during startup
there is no reason to go get the global configuration with an extra call
to repo-config unless the user is trying to view & edit all options in
the options dialog.
Since skipping this extra repo-config invocation save us a little bit of
time its nice to be able to avoid it when we are invoked as git-citool
and won't be running very long.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If the user is invoking us as git-citool then they want to perform a
single commit and exit quickly. Since we are about to be a very short
lived process we should do what we can to avoid spending CPU time setting
up menus which the user will never use, like the fetch/push/pull menus.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Apparently Tcl is being helpful on Windows during exec and is throwing a
\ in front of every { it finds in the string. I'm guessing they think
the value might be read by another Tcl program? Anyway, Git faithfully
stores the \{ sequence and sends it back that way to Tcl, at which point
Tcl parses the list wrong and starts to break it in the middle of any
element which contains spaces. Therefore a list such as:
-family {Times New Roman}
gets broken up into the pairs:
{-family \{Times}
{New Roman}
which is very incorrect. So now we replace all { and } with "", at which
point Tcl doesn't throw \ in front of the " on the way out to Git yet it
reads it correctly as a list on the way back in.
I also found and fixed a bug in the way we restored the fonts when the
user presses Restore Defaults in the options dialog.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The tkwait visibility command and Windows doesn't seem to realize
the window is visible, consequently we are never finishing our
initialization by calling update_status.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since the font name can only be chosen from within the options dialog
giving the user fast access to this dialog from within a context menu
that already talks about increasing and decreasing the font size may
help users to locate the font name setting as well.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Display the name of "this" repository rather than the quite ambiguous
string "This". The idea is that seeing the name of the directory the
repository is stored in should help jog the user's memory about what
they are setting options for.
Also place the options dialog immediately over the git-gui main window
when it gets opened. This way the user isn't scrolling very far away
to gain access to the window. At least on my Mac OS X system not doing
this makes the options dialog open rather far away, thus requiring lots
of mouse activity to reach it.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The git-update-index process can take a while to process a large
number of files; for example my laptop would probably need almost
an hour to chug through 20,000 modified files. In these incredibly
large cases the user should be given at least some feedback to let
them know the application is still working on their behalf, even if
it won't them do anything else (as the index is locked).
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
This turned out to take a lot more time than I thought it would take;
but now users can edit the main UI font and the diff/fixed with font
by changing both the family name and/or the point size of the text.
We save the complete Tk font specification to the user's ~/.gitconfig
file upon saving options. This is probably more verbose than it needs
to be as there are many useless options recorded (e.g. -overstrike 0)
that a user won't really want to use in this application.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I decided that the options menu was going to turn into a mess after
a while as I start to add additional features to git-gui. The better
approach would be to create a dialog that lets the user edit the options,
including their --global options.
We also wisely let the user press Cancel (or destroy the window) to abort
any sort of option editing session, without the options being changed.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Arrow is available on all Tk platforms and is mapped to the native
system cursor on Windows and Mac OS X. Consequently its the better
cursor choice as it should match whatever the system has configured
for the standard pointing thingy.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Apparently <Button-3> doesn't work on my single button PowerBook
mouse under Mac OS X. I'm guessing this is because Tk is stealing
every event and doesn't realize that Control-Button-1 is actually
supposed to invoke the context menu on this platform.
So now we have a utility procedure is_MacOSX to guess if we are
running on a Mac OS X system, and if so setup Control-Button-1 to
also activate what Button-3 should have. This does mean that I need
to stay away from using Control-Button-1 as a binding in any other
context. Of course we should use $M1B for that, which is M1 (aka
Command) on Mac OS X so that shouldn't prove to be a problem.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The native Tk font system is actually quite powerful and has the nice
property that modifications to a named font are immediately reflected
throughout all widgets currently displayed. This really saves us
from needing to write all of the reconfigure code as part of our font
display.
I also fixed the way we detect and apply the system font on the main
UI widgets as although it worked on a Windows 2000 system it does not
work at all on my Mac OS 10.4 system.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When the user has enabled the Trust File Modification Timestamp option
then we may display a file as being modified yet that file may not have
a difference. When the user clicks on that file we wind up displaying
an empty diff viewer, which makes no sense to the user.
So instead if we get an empty diff and the user has this option enabled
and the file's current state is _M (no change in index but the working
file appears modified) then run a quick update-index on just that file
and remove it from the list of modified files. We also give the user
a quick dialog stating we are removing it, and why.
Usually I don't run into this situation when I have the Trust File
Modification Timestamp option enabled, so its not that annoying to
have a dialog pop open to remind me why there are no differences.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because the diff area is one of the most important areas to be able to
read users should be able to increase or decrease the size of the font
used within that area.
Currently we save that back to the global configuration, even if it
may have originated from the local repository configuration. This
is probably going to be considered to be a bug by at least one user
who wants some sort of different font within a given repository, but
I'm just going to ignore the problem for now.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Rather than hardcoding our fonts to something that I thought was
reasonable, guess font_ui off the font used by the system in the
menu bar. This way the application conforms by default to whatever
the user's desktop is setup to.
We also now let the user supply font configuration through their
repository configuration as gui.fontui (the overall UI font) and
gui.fontdiff (the font used for the diff viewer).
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we sign off on a commit we want to add a blank line between
whatever is in the commit buffer and the new Signed-off-by line,
unless there already is a Signed-off-by (or Acked-by) tag at the end
of the buffer already. This change makes us do the right thing more
often.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The mouse cursor (at least on Windows) seemed to be picking up the
cursor from the sash controls and then never resetting itself back
to the standard text cursor (the I-beam) when it was over a text area
that the user can edit (like the commit buffer) or over a text area
the user can copy from (like the diff viewer).
So now we always set the cursor to left_ptr (which according to the Tk
documentation should be available everywhere) and only for the two text
areas which we use to list file names, as the user clicks in these but
is not permitted to select text.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
This change adds a context menu to the commit message buffer providing
fast access to the contents of the Edit menu, and to the console text
buffer, providing easy ways to copy selections of the buffer or the
entire buffer.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Make sure the file_lists array has both elements set at all times,
otherwise we get Tcl errors during mouse clicks in the file list
areas due to the list not being defined.
Also added M1-A as a keyboard binding within the console window
text area. This lets users select all text easily and copy it
to the clipboard.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
A number of lines were line wrapping in a rather ugly way when opened
in vim with line numbers enabled, so I split most of these lines over
two lines using a sensible wrapping policy.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The gui.geometry config value was starting to contain
the odd string \\{ as part of its value due to the
way the Tcl lists were being supplied to git repo-config.
Now we write out only three values: the overall window
geomtry, the y position of the horizontal sash, and
the x position of the vertical sash. All other data is
skipped, which makes the gui.geometry value simpler.
While debugging this I noticed that the save_my_config
procedure was being invoked multiple times during exit
due to do_quit getting invoked over and over again. So
now we set a flag in do_quit and don't perform any of our
"at exit" type of logic if we've already been through the
do_quit procedure once.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Added an extra blank line between the first line of each error message
and the rest of the message, as usually the rest of the message is
coming from Tcl or is the stderr output of a git command we tried to
invoke. This makes it easier to read the output (if any).
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Rather than drawing our own toplevel for error messages we really
should just use the the native tk_messageBox command to display
any error messages.
Major benefits for doing so are:
- automatically centers over main window;
- less code required on our part in git-gui;
- includes a nifty error icon on most systems;
- better fits the look-and-feel of the operating system.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I found difffont to be a very awkward varible name, due to the use
of three f's in a row. So use easier to read variable names.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we reshow the current diff file it can be faster to just fetch
the value from the file_states array than it is to ask for all paths
whose name exactly matches the one we want to show. This is because
[array names -exact] is O(n) in the number of files.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we commit we know that whatever was in the index went as part
of the commit. Since we generally assume that the user does not
update the index except through our user interface we can be reasonably
certain that any file which was marked as A/M/D in the index will have
had that A/M/D state changed to an _ (not different) by the commit.
We can use this knowledge to update the user interface post commit
by simply updating the index part of the file state of all files whose
index state was A/M/D to _ and then removing any file memory any which
wound up with a final state of __ (not different anywhere). Finally we
redraw the file lists and update the diff view.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If we are given a file whose path name contains an LF (\n) we now
escape it by inserting the common escape string \n instead of the
LF character whenever we display the name in the UI. This way the
text fields don't start to span multiple lines just to display one
file, and it keeps the line numbers correct within the file lists.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we did a rescan to update the file lists we lost the tag which
indicated which file was currently in the diff viewer. This occurs
because we delete every line from the two file list boxes (thus
removing the tag) and then redisplay the diff in the diff viewer
but then fail to restore the tag in the file list.
Now we restore that tag by searching for the file in the file lists
and adding the tag back when the diff viewer displays something.
We also no longer obtain the file path directly from the file list
text box. Instead we now keep two Tcl lists, one for each file list,
holding the file names in sorted order. These lists can be searched
with the native [lsearch -sorted] operation (which should be faster
than our crude bsearch) or can be quickly accessed by index to return
the file path. This should help make things safer should we ever be
given a file name which contains an LF within it (as that would span
two lines in the file list, not one).
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If we load a message file (e.g. MERGE_MSG) or we have just finished
making a commit and are clearing out the commit buffer we should also
clear out the undo/redo stack associated with that buffer. The prior
undo/redo stack has no associated with the new content and therefore
is not useful to the user.
Also modified the sign-off operation to perform the entire update in
a single undo/redo operation, allowing the user to undo the signoff
in case they didn't actually want to do that.
I also noticed what may be a crash on Windows related to the up and
down arrow keys navigating within the diff viewer. Since I got back
no stack trace (just an application exit with a loss of the commit
message) I suspect that the binding to scroll the text widget crashed
with an error and the wish process just terminated. So now we catch
(and ignore) any sort of error related to the arrow keys in the diff
viewer.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Users have come to expect basic editing features within their
applications, such as cut/copy/paste/undo/redo/select-all. I
found these features to be lacking in git-gui so now we have
them.
I also added basic keyboard bindings for the diff viewing area
so that the arrow keys move around single units (lines or columns)
and the M1-X/C keys will copy the selected text and the M1-A key
will select the entire diff.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Now that we call the update-index all files action "Include All" it
makes more sense to make this M1-I (so Control-I or Command-I depending
on platform) than M1-U, which stood for update but is somewhat confusing
to users.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
I started to find it very annoying that my test application kept
opening at the wrong location on my desktop, so now we save the
basic window geometry and sash positions into the config file as
gui.geometry.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Caching the Signed-Off-By line isn't very important (as its not
performance critical). The major improvement here is that we
now report an error to the user if we can't obtain their name
from git-var.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
There's really no great reason to show the entire commit object id
within the GUI, especially if the user is unable to copy and paste
it into another interface such as gitk or a terminal window. So
we'll just show them the first 8 digits and hope that is unique
within their repository.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
At least one user was confused by the term 'check-in'; he thought that
clicking that button would commit just that one file, but he wanted to
include all modified files into his next commit.
Since git doesn't really have a check-in concept this really was poor
language to use. Git does have an update-index concept but that is a
little too low level to show to the user.
So instead we now talk about including files in a commit.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When a file jumps between the file lists due to its state changing we
crashed thanks to a stale variable reference within the procedure as we
tried to setup the new icon.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If there are modified files present in the working directory then we
should not let the user perform a pull as it may fail due to the modified
files being uncommitted but needing to be merged at the file level.
Yes there are many cases where a merge will complete successfully even
though there are modified or untracked files sitting in the working
directory. But users generally shouldn't be attempting merges like that,
and if they are they probably are advanced enough to just use the command
line and bypass this little safety check.
We also no longer run a rescan after a successful pull has completed.
Usually this is unnecessary as a successful pull won't leave modified
files laying around. Instead we just update our HEAD and PARENT values
with the new commit, if there is one.
Unfortunately this does let the user get into an insane state as there
are bugs in core Git's git-pull and git-merge programs where the exit
status is sent back as a 0 rather than non-0 when a failure is detected.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If we have the index locked then no pull command is allowed to proceed
(as it would fail to get the index lock itself). So disable the pull
menu items when we are doing any index based operations.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The user must not modify the index while a git pull operation is running,
doing so might cause problems for the merge driver and specific strategy
being used. Normally on the command line people are just really good and
don't try to run index altering operations while they are also running a
pull. But in a slick GUI like git-gui we can't trust the user quite as
much.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
On very large projects (~1000 files) on Windows systems the update-index
--refresh stage of the rescan process takes a nontrival amount of time.
If the user is generally very careful with their file modification such
that the modification timestamp of the file differs only when the content
also differs then we can skip this somewhat expensive step and go right
to the diff-index and diff-files processes.
We save the user's prefernce in the current repository if they modify the
setting during a git-gui session, but we also load it through our dump of
repo-config --list so the user could move it to their ~/.gitconfig file
if they wanted it globally disabled.
We still keep update-index --refresh enabled by default however, as most
users will probably want it.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Most users of git-gui probably shouldn't be invoking git repack directly;
instead we should be looking at how many loose objects they have and
how many active packs they have and making the decision for them. But
that's more work to code, and there's always going to be discussion about
what is the right default threshold and how do we know that the user is
willing to do the repack when we decide its time, etc.
So instead we'll just keep it simple and offer up the menu option.
Unfortunately right now we get now progress indication back from
git-pack-objects as its being invoked not through a tty, which makes
it disable progress output and the git-repack.sh wrapper won't let us
pass through --progress.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Since Tk will only supply new space gained from growing the top level to
the bottom/right most widget within a panedwindow and most users will be
growing a git-gui main window for the purposes of seeing more of the
currently shown diff, flipping the order around makes Tk do what the
user wants by default.
Of course because we also removed the paned window from the commit buffer
area it is now impossible to increase the visible space for the commit
message. But I don't see this as a huge concern right now as its actually
very awkward to try and balance three paned window dividers within the
same top level window. We could always add it back if users really want
to expand the commit buffer and see more.
I also corrected a number of bugs that I accidentally introduced in the
last commit.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Removed as much as possible from the merge_state proc, which is where
we spent most of our time before UI update. This change makes our
running time match that of git status, except that we then need about
7 additional seconds to draw 6900 files on screen.
Apparently the [array names a -exact $v] operator in Tcl is O(n) rather
than O(1), which is really quite disappointing given that each array can
only have one entry for a given value. Switching to a lookup with a
catch (whose error we ignore) runs in O(1) time and bought us most of
that improvement.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Loading 6900 newly added files required about 90 seconds on one system.
This is just far too long to perform a "status" type of operation.
git-status on the same system completes in just 8.2 seconds if it is
redirected to /dev/null.
Most of our performance improvement comes from moving all of the UI
updating out of the main fileevent handlers for the status process.
Instead we are only updating the file_states array and then only doing
the UI update when all states are known and have been finally determined.
The rescan execution is now down to almost 30 seconds for the same case,
a good (but not really all that impressive) improvement.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we were receiving a lot of output from diff-index we split records
at the wrong locations and started using the file status information
(mode and SHA1s) as path names, and then proceeded to try to use part
of the path names as status data. This caused all sorts of havoc.
So I rewrote the parsing implementation to scan for the pair of null
bytes along the buffer and stop scanning (waiting for more data) if
both can't be found during this event. This seems to work well under
high load (like when processing 6,983 added files).
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
We now create one menu entry per remote listing the first Pull: or fetch
entry associated with that remote as the branch to pull into the current
branch.
This is actually quite incorrect as we should be using the default
remote branch name listed in branch.<name>.merge for a new-style remote
described in the config file. But its a good default to get started with.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
We're likely going to need key/value pairs from the repo-config beyond
just remote.*.url, so cache them all up front into a Tcl array where we
have fast access to them without needing to refork a repo-config --list
process.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
If the user closes a console and we get more ouptut for it then we
will get a Tcl error in the readable event handle for the file channel.
Since this loses the actual output and is quite unfriendly to the end
user instead reopen any console which the user closed prior to the
additional output arriving.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
The user might be using the new style config syntax remote.name.url
rather than the older standalone remote file.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because the remote end is likely to send us progress meters by
resetting each line with a CR (and no LF) we should display those
meters by replacing the last line of text with the next line,
just like a normal xterm would do.
This makes the output of fetch look about the same as if we ran it
from within an xterm.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Apparently accelerators really only work correctly for function keys
(F1-F12) and "Cmd-q". Apparently wish on Mac OS X reports itself
as unix and the OS is Darwin, this makes it a little difficult to
be sure we are running under Aqua.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Because we cd after getting the cdup value from Git we can't try
to get the gitdir until after we perform the cd, as usually the
gitdir is relative to the current working directory.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Apparently the Cygwin tclsh/wish executables don't pass the environment
that they inherited onto any children that they invoke. This causes a
problem for some users during 'git fetch' or 'git push' as critical
environment variables like GIT_SSH and SSH_AUTH_SOCK aren't available
to the git processes.
So we work around this by forcing sh to start a login shell, thus
reloading the user's environment, then cd to the current directory,
and finally start the requested process. Of course this won't
correctly handle any transient environment variables that were
inherited but were not supplied by the user's login shell.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
* Make sure we are in the top level working directory. This
way we can access files using their repository path.
* Reload the diff viewer if the current file's status has changed;
as the diff may now be different.
* Correctly handle the 'AD' file state: added but now gone from
the working directory.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Also fixed a bug related that caused a crash if the file currently
in the diff viewer is no longer modified after the commit.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
We can now commit any type of commit (initial, normal or merge) using
the same techniques as git-commit.sh does for these types of things.
If invoked as git-citool we run exit immediately after the commit was
finished. If invoked as git-gui then we stay running.
Also fixed a bug which caused the commit message buffer to be lost
when the application shutdown and restarted.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
A user shouldn't perform a commit if any of the following are true:
* The repository state has changed since the last rescan.
* There are no files updated in the index to commit.
* There are unmerged stages still in the index.
* The commit message has not been provided.
* The pre-commit hook is executable and declined.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
When we are refreshing from the index or updating the index we shouldn't
let the user cause other index based operations to occur as these would
likely conflict with the currently running operations possibly causing
some index changes to be lost.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
* Run refresh before diff-index.
* Load saved commit message during rescan.
* Save current commit message (if any) during quit.
* Add Signed-off-by line to commit buffer.
* Batch update-index invocations through --stdin.
* Better highlight which file is in the diff viewer.
* Key bindings for signoff, check-in all and commit.
* Improved formatting of status table within source.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
This is based on Paul Mackerras' gitool prototype which he offered up
to the community earlier in 2006. Its mostly however a rewrite from
scratch of a Tcl/Tk based graphical interface for Git and the most
common commands users might need to perform.
Currently it can display the status of the current repository, and not
much else.
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
