Rather than only clearing GIT_INDEX_FILE, take the list of environment
variables to clear from local_repo_env, appending the settings for
GIT_DIR and GIT_WORK_TREE.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
git-submodule used to take care of clearing GIT_DIR whenever it operated
on a submodule index or configuration, but forgot to unset GIT_WORK_TREE
or other repo-local variables. This would lead to failures e.g. when
GIT_WORK_TREE was set.
This only happened in very unusual contexts such as operating on the
main worktree from outside of it, but since "git-gui: set GIT_DIR and
GIT_WORK_TREE after setup" (a9fa11fe5b) such failures could also
be provoked by invoking an external tool such as "git submodule update"
from the Git Gui in a standard setup.
Solve by using the newly introduced clear_local_git_env() shell function
to ensure that all repo-local environment variables are unset.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Introduce an auxiliary function to clear all repo-local environment
variables. This should be invoked by any shell script that switches
repository during execution, to ensure that the environment is clean
and that things such as the git dir and worktree are set up correctly.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This prints the list of repo-local environment variables.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Move the list of GIT_* environment variables that are local to a
repository into a static list in environment.c, as it is also
useful elsewhere. Also add the missing GIT_CONFIG variable to the
list.
Make it easy to use the list both by NULL-termination and by size;
the latter (excluding the terminating NULL) is stored in the
local_repo_env_size define.
Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
These were written back when we always read objects from the standard
input. These days --revs and its friends can feed only the start and
end points and have the command internally enumerate the objects.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If the user runs
git config --global user.name Your Name
as suggested, user.name will be set to "Your". With this patch, the
suggested command will be
git config --global user.name "Your Name"
which will set user.name to "Your Name" and hopefully help users avoid
the former mistake.
Signed-off-by: Matt Kraai <kraai@ftbfs.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A simple "git shortlog" outside of a git repository stalls
waiting for an input. Check if that's the case by testing with
isatty() before read_from_stdin(), and warn the user like
"git commit" does in a similar case.
Signed-off-by: Michele Ballabio <barra_cuda@katamail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since "git fetch" learned "--all" and "--multiple" options, it has become
tempting for users to say "git pull --all". Even though it may fetch from
remotes that do not need to be fetched from for merging with the current
branch, it is handy.
"git fetch" however clears the list of fetched branches every time it
contacts a different remote. Unless the current branch is configured to
merge with a branch from a remote that happens to be the last in the list
of remotes that are contacted, "git pull" that fetches from multiple
remotes will not be able to find the branch it should be merging with.
Make "fetch" clear FETCH_HEAD (unless --append is given) and then append
the list of branches fetched to it (even when --append is not given). That
way, "pull" will be able to find the data for the branch being merged in
FETCH_HEAD no matter where the remote appears in the list of remotes to be
contacted by "git fetch".
Reported-by: Michael Lukashov
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If you have a branch.$X.merge config option, but no branch.$X.remote, and
your configuration tries to push tracking branches, git will segfault.
The problem is that even though branch->merge_nr is 1, you don't actually
have an upstream since there is no remote. Other callsites generally
check explicitly that branch->merge is not NULL, so let's do that here,
too.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When running a subfetch, the code propagated some options but not others.
Propagate --force, --update-head-ok and --keep options as well.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
All of these tests were bogus, as they created new directory and tried to
run "git pull" without even running "git init" in there. They were mucking
with the repository in $TEST_DIRECTORY.
While fixing it, modernize the style not to chdir around outside of
subshell. Otherwise a failed test will take us to an unexpected directory
and we need to chdir back to the test directory in each test, which is
ugly and error prone.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Note that in the documentation for git-pull, documentation for the
--progress option is displayed under the "Options related to fetching"
subtitle via fetch-options.txt.
Also, update the documentation of the -q/--quiet option for git-pull to
mention its effect on progress reporting during fetching.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Set transport->progress in transport.c::transport_set_verbosity() after
checking for the appropriate conditions (eg. --progress, isatty(2)),
and thereafter use it without having to check again.
The rules used are as follows (processing aborts when a rule is
satisfied):
1. Report progress, if force_progress is 1 (ie. --progress).
2. Don't report progress, if verbosity < 0 (ie. -q/--quiet).
3. Report progress if isatty(2) is 1.
This changes progress reporting behaviour such that if both --progress
and --quiet are specified, progress is reported.
In two areas, the logic to determine whether to *not* show progress is
changed to simply use the negation of transport->progress. This changes
behaviour in some ways (see previous paragraph for details).
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Remove the flags TRANSPORT_PUSH_QUIET and TRANSPORT_PUSH_VERBOSE; use
transport->verbose instead to determine verbosity for pushing.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
transport_set_verbosity() is now provided to transport users.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After 3f7a9b5 (Documentation/git-pull.txt: Add subtitles above included
option files, Thu Oct 22 2009), the -q/-v options were mentioned only
for the merge options section, giving the impression that git-fetch did
not take those arguments.
Follow 90e4311 (git-pull: do not mention --quiet and --verbose twice,
Mon Sep 7 2009) and hide -q/-v for merge options, while mentioning -q/-v
before the merge- and fetch-specific options.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After 5a518ad (clone: use --progress to force progress reporting),
-v/--verbose did not affect whether progress status was reported to
stderr, and users accustomed to using -v to do so since 21188b1
(Implement git clone -v) may be confused.
Mitigate such risks by stating -v does not affect progress in the
documentation.
Signed-off-by: Tay Ray Chuan <rctay89@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
gitk aliases either start with "!gitk", or look something like "!sh -c
FOO=bar gitk", IOW they contain the "gitk" word. With this patch the
completion script will recognize these cases and will offer gitk's
options.
Just like the earlier change improving on aliased command recognition,
this change can also be fooled easily by some complex aliases, but
users of such aliases could remedy it with custom completion
functions.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Shell command aliases can get rather complex, and the completion
script can not always determine correctly the git command invoked by
such an alias. For such cases users might want to provide custom
completion scripts the same way like for their custom commands made
possible by the previous patch.
The current completion script does not allow this, because if it
encounters an alias, then it will unconditionally perform completion
for the aliased git command (in case it can determine the aliased git
command, of course). With this patch the completion script will first
search for a completion function for the command given on the command
line, be it a git command, a custom git command of the user, or an
alias, and invoke that function to perform the completion. This has
no effect on git commands, because they can not be aliased anyway. If
it is an alias and there is a completion function for that alias (e.g.
_git_foo() for the alias 'foo'), then it will be invoked to perform
completion, allowing users to provide custom completion functions for
aliases. If such a completion function can not be found, only then
will the completion script check whether the command given on the
command line is an alias or not, and proceed as usual (i.e. find out
the aliased git command and provide completion for it).
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The bash completion script already provides support to complete
aliases, options and refs for aliases (if the alias can be traced back
to a supported git command by __git_aliased_command()), and the user's
custom git commands, but it does not support the options of the user's
custom git commands (of course; how could it know about the options of
a custom git command?). Users of such custom git commands could
extend git's bash completion script by writing functions to support
their commands, but they might have issues with it: they might not
have the rights to modify a system-wide git completion script, and
they will need to track and merge upstream changes in the future.
This patch addresses this by providing means for users to supply
custom completion scriplets for their custom git commands without
modifying the main git bash completion script.
Instead of having a huge hard-coded list of command-completion
function pairs (in _git()), the completion script will figure out
which completion function to call based on the command's name. That
is, when completing the options of 'git foo', the main completion
script will check whether the function '_git_foo' is declared, and if
declared, it will invoke that function to perform the completion. If
such a function is not declared, it will fall back to complete file
names. So, users will only need to provide this '_git_foo' completion
function in a separate file, source that file, and it will be used the
next time they press TAB after 'git foo '.
There are two git commands (stage and whatchanged), for which the
completion functions of other commands were used, therefore they
got their own completion function.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To support completion for aliases, the completion script tries to
figure out which git command is invoked by an alias. Its
implementation in __git_aliased_command() is rather straightforward:
it returns the first word from the alias. For simple aliases starting
with the git command (e.g. alias.last = cat-file commit HEAD) this
gives the right results. Unfortunately, it does not work with shell
command aliases, which can get rather complex, as illustrated by one
of Junio's aliases:
[alias]
lgm = "!sh -c 'GIT_NOTES_REF=refs/notes/amlog git log \"$@\" || :' -"
In this case the current implementation returns "!sh" as the aliased
git command, which is obviosly wrong.
The full parsing of a shell command alias like that in the completion
code is clearly unfeasible. However, we can easily improve on aliased
command recognition by eleminating stuff that is definitely not a git
command: shell commands (anything starting with '!'), command line
options (anything starting with '-'), environment variables (anything
with a '=' in it), and git itself. This way the above alias would be
handled correctly, and the completion script would correctly recognize
"log" as the aliased git command.
Of course, this solution is not perfect either, and could be fooled
easily. It's not hard to construct an alias, in which a word does not
match any of these filter patterns, but is still not a git command
(e.g. by setting an environment variable to a value which contains
spaces). It may even return false positives, when the output of a git
command is piped into an other git command, and the second gets the
command line options via $@, but options for the first one are
offered. However, the following patches will enable the user to
supply custom completion scripts for aliases, which can be used to
remedy these problematic cases.
Signed-off-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When git am does an automatic gc it doesn't clean up the rebase-apply
directory until after this has finished. This means that if the user
aborts the gc then future am or rebase operations will report that an
existing operation is in progress, which is undesirable and confusing.
Reported by Mark Brown <broonie@debian.org> through
http://bugs.debian.org/570966
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When RUNTIME_PREFIX is enabled, the installation prefix is derived by
trying a limited set of known locations where the git executable can
reside. If none of these is found, a warning is emitted.
When git is built in a directory that matches neither of these known names,
the warning would always be emitted when the uninstalled executable is run.
This is a problem on Windows, where gitk picks the uninstalled git when
invoked from the build directory and gets confused by the warning.
Print the warning only when GIT_TRACE is set.
Signed-off-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
According to relevant RFCs, in addition to alphanumerics, the following
characters are valid in URL scheme parts: '+', '-' and '.', but
currently only alphanumerics are allowed in remote helper names.
Allow those three characters in remote helper names (both 'foo://' and
'foo::' syntax).
Signed-off-by: Ilari Liusvaara <ilari.liusvaara@elisanet.fi>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
23a64c9e (conflict-marker-size: new attribute, 2010-01-16) introduced the
new attribute and also pass the conflict marker size as %L to merge driver
commands. This documents the substitution.
Signed-off-by: Bert Wesarg <bert.wesarg@googlemail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 1b22b6c897 made duplicated versions of encode_header() into a
common version called encode_in_pack_object_header(). There is however
a better location that sha1_file.c for such a function though, as
sha1_file.c contains nothing related to the creation of packs, and
it is quite populated already.
Also the comment that was moved to the header file should really remain
near the function as it covers implementation details and provides no
information about the actual function interface.
Signed-off-by: Nicolas Pitre <nico@fluxnic.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we look at a patch for adding hunks interactively, we
first split it into a header and a list of hunks. Some of
the header lines, such as mode changes and deletion, however,
become their own selectable hunks. Later when we reassemble
the patch, we simply concatenate the header and the selected
hunks. This leads to patches like this:
diff --git a/file b/file
index d95f3ad..0000000
--- a/file
+++ /dev/null
deleted file mode 100644
@@ -1 +0,0 @@
-content
Notice how the deletion comes _after_ the ---/+++ lines,
when it should come before.
In many cases, we can get away with this as git-apply
accepts the slightly bogus input. However, in the specific
case of a deletion line that is being applied via "apply
-R", this malformed patch triggers an assert in git-apply.
This comes up when discarding a deletion via "git checkout
-p".
Rather than try to make git-apply accept our odd input,
let's just reassemble the patch in the correct order.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We used to unnecessarily give the read permission to group and others,
regardless of the umask, which isn't serious because the objects are
still protected by their containing directory, but isn't necessary
either.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When reaching the end of git_mkstemps_mode, at least one call to open()
has been done, and errno has been set accordingly. Setting errno is
therefore not necessary, and actually harmfull since callers can't
distinguish e.g. permanent failure from ENOENT, which can just mean that
we need to create the containing directory.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We used to create 0600 files, and then use chmod to set the group and
other permission bits to the umask. This usually has the same effect
as a normal file creation with a umask.
But in the presence of ACLs, the group permission plays the role of
the ACL mask: the "g" bits of newly created files are chosen according
to default ACL mask of the directory, not according to the umask, and
doing a chmod() on these "g" bits affect the ACL's mask instead of
actual group permission.
In other words, creating files with 0600 and then doing a chmod to the
umask creates files which are unreadable by users allowed in the
default ACL. To create the files without breaking ACLs, we let the
umask do it's job at the file's creation time, and get rid of the
later chmod.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
gitmkstemps emulates the behavior of mkstemps, which is usually used
to create files in a shared directory like /tmp/, hence, it creates
files with permission 0600.
Add git_mkstemps_mode() that allows us to specify the desired mode, and
make git_mkstemps() a wrapper that always uses 0600 to call it. Later we
will use git_mkstemps_mode() when creating pack files.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This function used to be only a compatibility function, but we're
going to extend it and actually use it, so make it part of Git.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Right now, Git creates unreadable pack files on non-shared
repositories when the user has a umask of 077, even when the default
ACLs for the directory would give read/write access to a specific
user.
Loose object files are created world-readable, which doesn't break ACLs,
but isn't necessarily desirable.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This shrinks the top-level directory a bit, and makes it much more
pleasant to use auto-completion on the thing. Instead of
[torvalds@nehalem git]$ em buil<tab>
Display all 180 possibilities? (y or n)
[torvalds@nehalem git]$ em builtin-sh
builtin-shortlog.c builtin-show-branch.c builtin-show-ref.c
builtin-shortlog.o builtin-show-branch.o builtin-show-ref.o
[torvalds@nehalem git]$ em builtin-shor<tab>
builtin-shortlog.c builtin-shortlog.o
[torvalds@nehalem git]$ em builtin-shortlog.c
you get
[torvalds@nehalem git]$ em buil<tab> [type]
builtin/ builtin.h
[torvalds@nehalem git]$ em builtin [auto-completes to]
[torvalds@nehalem git]$ em builtin/sh<tab> [type]
shortlog.c shortlog.o show-branch.c show-branch.o show-ref.c show-ref.o
[torvalds@nehalem git]$ em builtin/sho [auto-completes to]
[torvalds@nehalem git]$ em builtin/shor<tab> [type]
shortlog.c shortlog.o
[torvalds@nehalem git]$ em builtin/shortlog.c
which doesn't seem all that different, but not having that annoying
break in "Display all 180 possibilities?" is quite a relief.
NOTE! If you do this in a clean tree (no object files etc), or using an
editor that has auto-completion rules that ignores '*.o' files, you
won't see that annoying 'Display all 180 possibilities?' message - it
will just show the choices instead. I think bash has some cut-off
around 100 choices or something.
So the reason I see this is that I'm using an odd editory, and thus
don't have the rules to cut down on auto-completion. But you can
simulate that by using 'ls' instead, or something similar.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Pagers that do not consume their input are dangerous: for example,
$ GIT_PAGER=: git log
$ echo $?
141
$
The only reason these tests were able to work before was that
'git log' would write to the pipe (and not fill it) before the
pager had time to terminate and close the pipe.
Fix it by using a program that consumes its input, namely wc (as
suggested by Johannes).
Reported-by: Johannes Sixt <j.sixt@viscovery.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We don't want the data being deflated and stored into loose objects
to be different from what we expect. While the deflated data is
protected by a CRC which is good enough for safe data retrieval
operations, we still want to be doubly sure that the source data used
at object creation time is still what we expected once that data has
been deflated and its CRC32 computed.
The most plausible data corruption may occur if the source file is
modified while Git is deflating and writing it out in a loose object.
Or Git itself could have a bug causing memory corruption. Or even bad
RAM could cause trouble. So it is best to make sure everything is
coherent and checksum protected from beginning to end.
To do so we compute the SHA1 of the data being deflated _after_ the
deflate operation has consumed that data, and make sure it matches
with the expected SHA1. This way we can rely on the CRC32 checked by
the inflate operation to provide a good indication that the data is still
coherent with its SHA1 hash. One pathological case we ignore is when
the data is modified before (or during) deflate call, but changed back
before it is hashed.
There is some overhead of course. Using 'git add' on a set of large files:
Before:
real 0m25.210s
user 0m23.783s
sys 0m1.408s
After:
real 0m26.537s
user 0m25.175s
sys 0m1.358s
The overhead is around 5% for full data coherency guarantee.
Signed-off-by: Nicolas Pitre <nico@fluxnic.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This patch adds two test cases for:
6977c25 git diff --quiet -w: check and report the status
Signed-off-by: Larry D'Anna <larry@elder-gods.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
