We have always had a nice way to reset a working tree to another state
while carrying our changes around: "git read-tree -u -m". Yes, it fails if
the target tree is different in the paths that are dirty in the working
tree, but this is how we used to switch branches in "git checkout", and it
worked fine.
However, perhaps exactly _because_ we've supported this from very early
on, another low-level command, namely "git reset", never did.
But as time went on, 'git reset' remains as a very common command, while
'git read-tree' is now a very odd and low-level plumbing thing that nobody
sane should ever use, because it only makes sense together with other
operations like either switching branches or just rewriting HEAD.
Which means that we have effectively lost the ability to do something very
common: jump to another point in time without always dropping all our
dirty state.
So add this kind of mode to "git reset", and since it merges your changes
to what you are resetting to, just call it that: "git reset --merge".
I've wanted this for a long time, since I very commonly carry a dirty
tree while working on things. My main 'Makefile' file quite often has the
next version already modified, and sometimes I have local modifications
that I don't want to commit, but I still do pulls and patch applications,
and occasionally want to do "git reset" to undo them - while still keeping
my local modifications.
(Maybe we could eventually change it to something like "if we have a
working tree, default to --merge, otherwise default to --mixed").
NOTE! This new mode is certainly not perfect. There's a few things to look
out for:
- if the index has unmerged entries, "--merge" will currently simply
refuse to reset ("you need to resolve your current index first").
You'll need to use "--hard" or similar in this case.
This is sad, because normally a unmerged index means that the working
tree file should have matched the source tree, so the correct action is
likely to make --merge reset such a path to the target (like --hard),
regardless of dirty state in-tree or in-index. But that's not how
read-tree has ever worked, so..
- "git checkout -m" actually knows how to do a three-way merge, rather
than refuse to update the working tree. So we do know how to do that,
and arguably that would be even nicer behavior.
At the same time it's also arguably true that there is a chance of loss
of state (ie you cannot get back to the original tree if the three-way
merge ends up resolving cleanly to no diff at all), so the "refuse to
do it" is in some respects the safer - but less user-friendly - option.
In other words, I think 'git reset --merge' could become a bit more
friendly, but this is already a big improvement. It allows you to undo a
recent commit without having to throw your current work away.
Yes, yes, with a dirty tree you could always do
git stash
git reset --hard
git stash apply
instead, but isn't "git reset --merge" a nice way to handle one particular
simple case?
Hmm? Maybe I'm the only one that does a lot of work with a dirty tree, and
sure, I can do other things like the "git stash" thing, or using "git
checkout" to actually create a new branch, and then playing games with
branch renaming etc to make it work like this one.
But I suspect others dislike how "git reset" works too. But see the
suggested improvements above.
git-gui: Teach start_push_anywhere_action{} to notice when remote is a mirror.
When the destination repository is a mirror, this function goofed by still
passing a refspec to git-push. Now it notices that the remote is a mirror
and holds the refspec.
Signed-off-by: Mark Burton <markb@ordern.com> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
gitk: Make line origin search update the busy status
Currently the 'show origin of this line' feature does not update the
status field of the gitk window, so it is not evident that any
processing is going on. It may seem at first that clicking the item
had no effect.
This commit adds calls to set and clear the busy status with an
appropriate title, similar to other search commands.
Signed-off-by: Alexander Gavrilov <angavrilov@gmail.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
gitk: Fix bug in accessing undefined "notflag" variable
As pointed out by Johannes Sixt and Alexander Gavrilov, commit 2958228430b63f2e38c55519d1f98d8d6d9e23f3 ("gitk: Fix switch statement
in parseviewargs") exposed a latent bug in that $notflag was never
initialized. Since it isn't used either, this removes it entirely.
gitk: Highlight only when search type is "containing:".
When the search type is "touching paths" or "adding/removing string",
it's not very useful to highlight instances of the search string in
the commit message, headline or author name, so this disables the
highlighting in those cases.
This was suggested by Mark Burton <markb@ordern.com>, but the
implementation is different to his patch, which tested $gdttype at
each place where $markingmatches was tested.
* git://git.kernel.org/pub/scm/gitk/gitk:
gitk: Avoid handling the Return key twice in Add Branch
gitk: Show local changes properly when we have a path limit
gitk: Fix switch statement in parseviewargs
gitk: Index line[hnd]tag arrays by id rather than row number
send-email: do not reverse the command line arguments
The loop picks elements from @ARGV one by one, sifts them into arguments
meant for format-patch and the script itself, and pushes them to @files
and @rev_list_opts arrays. Pick elements from @ARGV starting at the
beginning using shift, instead of at the end using pop, as push appends
them to the end of the array.
generate-cmdlist.sh: avoid selecting synopsis at wrong place
In "common" man pages there is luckily no "NAME" anywhere except at
beginning of documents. If there is another "NAME", sed could
mis-select it and lead to common-cmds.h corruption. So better nail it
at beginning of line, which would reduce corruption chance.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
gitweb: make gitweb_check_feature a boolean wrapper
The gitweb_get_feature() function retrieves the configuration parameters
for the feature (such as the list of snapshot formats or the list of
additional actions), but it is very often used to see if feature is
enabled (which is returned as the first element in the list).
Because accepting the returned list in the scalar context by mistake
yields the number of elements in the array, which is non-zero in all
cases, such a mistake would result in a bug for the latter use, with
disabled features appearing enabled. All existing callers that call the
function for this purpose assign the return value in the list context to
retrieve the first element, but that is only because we fixed careless
callers recently.
This adds gitweb_check_feature() as a wrapper to gitweb_get_feature() that
can be called safely in the scalar context to see if a feature is enabled
to reduce the risk of future bugs. Callers of "get" that use the call
only to see if the feature is enabled are updated to call this wrapper.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
gitweb: rename gitweb_check_feature to gitweb_get_feature
The function is about retrieving the configuration parameter list for the
feature. A more robust way to check if a feature is enabled will be
introduced in the next patch, and the function will be called
gitweb_check_feature.
gitweb_check_feature() function is to retrieve the configuration parameter
list and calling it in the scalar context does not give its first element
that tells if the feature is enabled. This fixes all the existing callers
to call the function correctly in the list context.
git add --intent-to-add: do not let an empty blob be committed by accident
Writing a tree out of an index with an "intent to add" entry is blocked.
This implies that you cannot "git commit" from such a state; however you
can still do "git commit -a" or "git commit $that_path".
git add --intent-to-add: fix removal of cached emptiness
This uses the extended index flag mechanism introduced earlier to mark
the entries added to the index via "git add -N" with CE_INTENT_TO_ADD.
The logic to detect an "intent to add" entry for the purpose of allowing
"git rm --cached $path" is tightened to check not just for a staged empty
blob, but with the CE_INTENT_TO_ADD bit. This protects an empty blob that
was explicitly added and then modified in the work tree from being dropped
with this sequence:
builtin-rm.c: explain and clarify the "local change" logic
Explain the logic to check local modification a bit more in the comment,
especially because the existing comment that talks about "git rm --cached"
was placed in a part that was not about "--cached" at all.
* rs/strbuf-expand:
remove the unused files interpolate.c and interpolate.h
daemon: deglobalize variable 'directory'
daemon: inline fill_in_extra_table_entries()
daemon: use strbuf_expand() instead of interpolate()
merge-recursive: use strbuf_expand() instead of interpolate()
add strbuf_expand_dict_cb(), a helper for simple cases
* mv/clone-strbuf:
builtin_clone: use strbuf in cmd_clone()
builtin-clone: use strbuf in clone_local() and copy_or_link_directory()
builtin-clone: use strbuf in guess_dir_name()
* ph/send-email:
git send-email: ask less questions when --compose is used.
git send-email: add --annotate option
git send-email: interpret unknown files as revision lists
git send-email: make the message file name more specific.
* maint:
sha1_file.c: resolve confusion EACCES vs EPERM
sha1_file: avoid bogus "file exists" error message
git checkout: don't warn about unborn branch if -f is already passed
bash: offer refs instead of filenames for 'git revert'
bash: remove dashed command leftovers
git-p4: fix keyword-expansion regex
fast-export: use an unsorted string list for extra_refs
Add new testcase to show fast-export does not always exports all tags
An earlier commit 916d081 (Nicer error messages in case saving an object
to db goes wrong, 2006-11-09) confused EACCES with EPERM, the latter of
which is an unlikely error from mkstemp().
This avoids the following misleading error message:
error: unable to create temporary sha1 filename ./objects/15: File exists
mkstemp can fail for many reasons, one of which, ENOENT, can occur if
the directory for the temp file doesn't exist. create_tmpfile tried to
handle this case by always trying to mkdir the directory, even if it
already existed. This caused errno to be clobbered, so one cannot tell
why mkstemp really failed, and it truncated the buffer to just the
directory name, resulting in the strange error message shown above.
Note that in both occasions that I've seen this failure, it has not been
due to a missing directory, or bad permissions, but some other, unknown
mkstemp failure mode that did not occur when I ran git again. This code
could perhaps be made more robust by retrying mkstemp, in case it was a
transient failure.
Signed-off-by: Joey Hess <joey@kitenet.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git checkout: don't warn about unborn branch if -f is already passed
I think it's unnecessary to warn that the checkout has been forced due to an
unborn current branch if -f has been explicitly passed. For one project, I am
using git-new-workdir to create workdirs from a bare repository whose HEAD is
set to an unborn branch, and this warning started to irritate me.
Signed-off-by: Matt McCutchen <matt@mattmccutchen.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
bash: offer refs instead of filenames for 'git revert'
The completion script for 'git revert' currently offers options and
filenames. However, 'git revert' doesn't take any filenames from the
command line, but a single commit. Therefore, it's more sane to offer
refs instead.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de> Acked-by: Shawn O. Pearce <spearce@spearce.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Sometimes it's handy to complete full refs, e.g. the user has some
refs outside of refs/{heads,remotes,tags} or the user wants to
complete some git command's special refs (like 'git show
refs/bisect/bad').
To do that, we check whether the ref to be completed starts with
'refs/' or is 'refs' (to reduce the risk of matching 'refs-'). If it
does, then we offer full refs for completion; otherwise everything
works as usual.
This way the impact on the common case is fairly small (hopefully not
many users have branches or tags starting with 'refs'), and in the
special case the cost of typing out 'refs' is bearable.
While at it, also remove the unused 'cmd' variable from '__git_refs'.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de> Acked-by: Shawn O. Pearce <spearce@spearce.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 5a625b07 (bash: remove fetch, push, pull dashed form leftovers,
2008-10-03) did that already, but there were still some git-cmd left
here and there.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de> Acked-by: Shawn O. Pearce <spearce@spearce.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
I figured the sections might as well be in some order, so I chose alphabetical
but with "core" at the beginning. This should help people add new variables
in the right places.
Signed-off-by: Matt McCutchen <matt@mattmccutchen.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t9129-git-svn-i18n-commitencoding: Make compare_svn_head_with() compatible with OSX sed
The sed call used in compare_svn_head_with() uses the + quantifier, which
is not supported in the OSX version of sed. It is replaced by the
equivalent \{1,\}.
Signed-off-by: Marcel Koeppen <git-dev@marzelpan.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The original intention of 72909be (Add diff-option --ext-diff, 2007-06-30)
was to optionally allow the use of external diff viewer in "git log"
family (while keeping them disabled by default). It exposed the "allow
external diff" bit to the UI, but forgot to adjust the "git diff" codepath
that was set up to always allow use of the external diff viewer.
Noticed by Nazri Ramliy; tests by René Scharfe squashed in.
Merge branch 'for-junio' of git://source.winehq.org/~julliard/git/git
* 'for-junio' of git://source.winehq.org/~julliard/git/git:
git.el: Allow to commit even if there are no marked files.
git.el: Add possibility to mark files directly in git-update-status-files.
git.el: Add an insert file command.
git.el: Never clear the status buffer, only update the files.
git.el: Fix git-amend-commit to support amending an initial commit.
git.el: Properly handle merge commits in git-amend-commit.
git.el: Simplify handling of merge heads in the commit log-edit buffer.
git.el: Remove the env parameter in git-call-process and git-call-process-string.
git.el: Improve error handling for commits.
bisect: teach "skip" to accept special arguments like "A..B"
The current "git bisect skip" syntax is "git bisect skip [<rev>...]"
so it's already possible to skip a range of revisions using
something like:
$ git bisect skip $(git rev-list A..B)
where A and B are the bounds of the range we want to skip.
This patch teaches "git bisect skip" to accept:
$ git bisect skip A..B
as an abbreviation for the former command.
This is done by checking each argument to see if it contains two
dots one after the other ('..'), and by expending it using
"git rev-list" if that is the case.
Note that this patch will not make "git bisect skip" accept all
that "git rev-list" accepts, as things like "^A B" for exemple
will not work. But things like "A B..C D E F.. ..G H...I" should
work as expected.
Signed-off-by: Christian Couder <chriscool@tuxfamily.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Remove the global variable 'directory' and pass it as a parameter of
the two functions that use it instead, (almost) restoring their
interface to how it was before 49ba83fb67d9e447b86953965ce5f949c6a93b81.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Having fill_in_extra_table_entries() as a separate function has no
advantage -- a function with no parameters and return values might as
well be an anonymous block of code. Its name still refers to the table
of interpolate() which has been removed earlier, so it's better to
inline it at its only call site.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>
add strbuf_expand_dict_cb(), a helper for simple cases
The new callback function strbuf_expand_dict_cb() can be used together
with strbuf_expand() if there is only a small number of placeholders
for static replacement texts. It expects its dictionary as an array of
placeholder+value pairs as context parameter, terminated by an entry
with the placeholder member set to NULL.
The new helper is intended to aid converting the remaining calls of
interpolate(). strbuf_expand() is smaller, more flexible and can be
used to go faster than interpolate(), so it should replace the latter.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>
fast-export: use an unsorted string list for extra_refs
The list extra_refs contains tags and the objects referenced by them,
so that they can be handled at the end. When a tag references a
commit, that commit is added to the list using the same name.
Also, the function handle_tags_and_duplicates() relies on the order
the items were added to extra_refs, so clearly we do not want to
use a sorted list here.
Noticed by Miklos Vajna.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Tested-by: Miklos Vajna <vmiklos@frugalware.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
This avoids the following misleading error message:
error: unable to create temporary sha1 filename ./objects/15: File exists
mkstemp can fail for many reasons, one of which, ENOENT, can occur if
the directory for the temp file doesn't exist. create_tmpfile tried to
handle this case by always trying to mkdir the directory, even if it
already existed. This caused errno to be clobbered, so one cannot tell
why mkstemp really failed, and it truncated the buffer to just the
directory name, resulting in the strange error message shown above.
Note that in both occasions that I've seen this failure, it has not been
due to a missing directory, or bad permissions, but some other, unknown
mkstemp failure mode that did not occur when I ran git again. This code
could perhaps be made more robust by retrying mkstemp, in case it was a
transient failure.
Signed-off-by: Joey Hess <joey@kitenet.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
compat/mingw.c: Teach mingw_rename() to replace read-only files
On POSIX, rename() can replace files that are not writable. On Windows,
however, read-only files cannot be replaced without additional efforts:
We have to make the destination writable first.
Since the situations where the destination is read-only are rare, we do not
make the destination writable on every invocation, but only if the first
try to rename a file failed with an "access denied" error.
Signed-off-by: Johannes Sixt <j6t@kdbg.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Fix deletion of last character in levenshtein distance
Without this change, "git tags" will not suggest "git tag"
(it will only suggest "git status"), and "git statusx" will
not suggest anything.
Signed-off-by: Samuel Tardieu <sam@rfc1149.net> Acked-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git.el: Simplify handling of merge heads in the commit log-edit buffer.
Use a single Merge: header instead of one Parent: header for each
parent, and don't list the current HEAD as a merged head. Support
symbolic references too.
Pass the verbose mode parameter to the underlying fetch command.
$ ./git remote -v update
Updating origin
From git://git.kernel.org/pub/scm/git/git
= [up to date] html -> origin/html
= [up to date] maint -> origin/maint
= [up to date] man -> origin/man
= [up to date] master -> origin/master
= [up to date] next -> origin/next
= [up to date] pu -> origin/pu
= [up to date] todo -> origin/todo
Signed-off-by: Cheng Renquan <crquan@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
gitk: Fix context menu items for generating diffs when in tree mode
Currently, if you invoke the "diff this -> selected" or "diff selected
-> this" and gitk is in "Tree" mode rather than "Patch" mode, the
diff display pane will just show the header but not the actual diff,
unless gitk has done the diff before and thus has the list of files
that differ. This was because the logic in gettreediffline that
checked whether we had moved on to doing something else checked the
mode (Tree or Patch) before checking whether the ids we're diffing
had changed.
This fixes it. The new logic in gettreediffline is slightly hacky
and relies on the fact that the Tree/Patch mode only applies when
we're looking at a single commit, not at the diff between two commits.
A similar change was made as part of commit 76f15947af7, that added
bindings to all dialogs, and this duplication causes mkbrgo to be
called twice, the second time after the window has been destroyed.
As a result, an error window appears when the code tries to access
widgets.
Signed-off-by: Alexander Gavrilov <angavrilov@gmail.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
gitk: Show local changes properly when we have a path limit
Since gitk looks for the HEAD commit to attach the fake commits for
local changes to, we can miss out on seeing the fake commits if we
have a path limit and the HEAD commit doesn't alter any of the files
in the path limit.
This fixes it by running
git rev-list -1 $head -- $paths
if we have a path limit, and taking the result of that as the commit
to attach the fake commits to. This means that we can be attaching
the fake commits to a different commit in each view, so we use a new
$viewmainhead($view) for that.
This also fixes a buglet where updatecommits would only fix up the
fake commits if the HEAD changed since the last call to updatecommits,
whereas it should fix them up if the HEAD has changed since this view
was last created or updated.
In Tcl, a comment in a switch command where a pattern would be expected
doesn't do what one would expect, so this moves the comments inside the
actions. Doing that shows up an extra "-" which this also removes.