Older versions of pack-refs did not write peel lines for
refs outside of refs/tags. This meant that on reading the
pack-refs file, we might set the REF_KNOWS_PEELED flag for
such a ref, even though we do not know anything about its
peeled value.
The previous commit updated the writer to always peel, no
matter what the ref is. That means that packed-refs files
written by newer versions of git are fine to be read by both
old and new versions of git. However, we still have the
problem of reading packed-refs files written by older
versions of git, or by other implementations which have not
yet learned the same trick.
The simplest fix would be to always unset the
REF_KNOWS_PEELED flag for refs outside of refs/tags that do
not have a peel line (if it has a peel line, we know it is
valid, but we cannot assume a missing peel line means
anything). But that loses an important optimization, as
upload-pack should not need to load the object pointed to by
refs/heads/foo to determine that it is not a tag.
Instead, we add a "fully-peeled" trait to the packed-refs
file. If it is set, we know that we can trust a missing peel
line to mean that a ref cannot be peeled. Otherwise, we fall
back to assuming nothing.
[commit message and tests by Jeff King <peff@peff.net>]
Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t1507: Test that branchname@{upstream} is interpreted as branch
Syntax branchname@{upstream} should interpret its argument as a name of
a branch. Add the test to check that it doesn't try to interpret it as a
refname if the branch in question does not exist.
Signed-off-by: Kacper Kornet <draenog@pld-linux.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When fast-export wants to export a blob object, it first
calls parse_object to get a "struct object" and check
whether we have already shown the object. If we haven't
shown it, we then use read_sha1_file to pull it from disk
and write it out.
That means we load each blob from disk twice: once for
parse_object to find its type and check its sha1, and a
second time when we actually output it. We can drop this to
a single load by using lookup_object to check the SHOWN
flag, and then checking the signature on and outputting a
single buffer.
This provides modest speedups on git.git (best-of-five, "git
fast-export HEAD >/dev/null"):
[before] [after]
real 0m14.347s real 0m13.780s
user 0m14.084s user 0m13.620s
sys 0m0.208s sys 0m0.100s
and somewhat more on more blob-heavy repos (this is a
repository full of media files):
[before] [after]
real 0m52.236s real 0m44.451s
user 0m50.568s user 0m43.000s
sys 0m1.536s sys 0m1.284s
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The handle_object function is rather vaguely named; it only
operates on blobs, and its purpose is to export the blob to
the output stream. Let's call it "export_blob" to make it
more clear what it does.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we pack an annotated tag ref, we write not only the
sha1 of the tag object along with the ref, but also the sha1
obtained by peeling the tag. This lets readers of the
pack-refs file know the peeled value without having to
actually load the object, speeding up upload-pack's ref
advertisement.
The writer marks a packed-refs file with peeled refs using
the "peeled" trait at the top of the file. When the reader
sees this trait, it knows that each ref is either followed
by its peeled value, or it is not an annotated tag.
However, there is a mismatch between the assumptions of the
reader and writer. The writer will only peel refs under
refs/tags, but the reader does not know this; it will assume
a ref without a peeled value must not be a tag object. Thus
an annotated tag object placed outside of the refs/tags
hierarchy will not have its peeled value printed by
upload-pack.
The simplest way to fix this is to start writing peel values
for all refs. This matches what the reader expects for both
new and old versions of git.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
use parse_object_or_die instead of die("bad object")
Some call-sites do:
o = parse_object(sha1);
if (!o)
die("bad object %s", some_name);
We can now handle that as a one-liner, and get more
consistent output.
In the third case of this patch, it looks like we are losing
information, as the existing message also outputs the sha1
hex; however, parse_object will already have written a more
specific complaint about the sha1, so there is no point in
repeating it here.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Many call-sites of parse_object assume that they will get a
non-NULL return value; this is not the case if we encounter
an error while parsing the object.
This patch adds a wrapper function around parse_object that
handles dying automatically, and uses it anywhere we
immediately try to access the return value as a non-NULL
pointer (i.e., anywhere that we would currently segfault).
This wrapper may also be useful in other places. The most
obvious one is code like:
o = parse_object(sha1);
if (!o)
die(...);
However, these should not be mechanically converted to
parse_object_or_die, as the die message is sometimes
customized. Later patches can address these sites on a
case-by-case basis.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* maint:
rev-parse: clarify documentation of $name@{upstream} syntax
sha1_name: pass object name length to diagnose_invalid_sha1_path()
Makefile: keep LIB_H entries together and sorted
rev-parse: clarify documentation of $name@{upstream} syntax
"git rev-parse" interprets string in string@{upstream} as a name of
a branch not a ref. For example, refs/heads/master@{upstream} looks
for an upstream branch that is merged by git-pull to ref
refs/heads/refs/heads/master not to refs/heads/master.
However the documentation could mislead a user to believe that the
string is interpreted as ref.
Signed-off-by: Kacper Kornet <draenog@pld-linux.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
sha1_name: pass object name length to diagnose_invalid_sha1_path()
The only caller of diagnose_invalid_sha1_path() extracts a substring from
an object name by creating a NUL-terminated copy of the interesting part.
Add a length parameter to the function and thus avoid the need for an
allocation, thereby simplifying the code.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>
pull: Apply -q and -v options to rebase mode as well
git pull passed -q and -v only to git merge, but they can be useful for
git rebase as well, so pass them there, too.
In particular, using -q shuts up the "Already up-to-date." message.
Especially, a new test script runs the same "pull --rebase" twice to
make sure both cases are quiet, when it has something to fetch and
when it is already up to date.
Signed-off-by: Peter Eisentraut <peter@eisentraut.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Preallocate hash tables when the number of inserts are known in advance
This avoids unnecessary re-allocations and reinsertions. On webkit.git
(i.e. about 182k inserts to the name hash table), this reduces about
100ms out of 3s user time.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
As a follow-up to 60d24dd25 (Makefile: fold XDIFF_H and VCSSVN_H into
LIB_H), let the unconditional additions to LIB_H form a single sorted
list. Also drop the duplicate entry for xdiff/xdiff.h, which was easy
to spot after sorting.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>
upload-pack: load non-tip "want" objects from disk
It is a long-time security feature that upload-pack will not
serve any "want" lines that do not correspond to the tip of
one of our refs. Traditionally, this was enforced by
checking the objects in the in-memory hash; they should have
been loaded and received the OUR_REF flag during the
advertisement.
The stateless-rpc mode, however, has a race condition here:
one process advertises, and another receives the want lines,
so the refs may have changed in the interim. To address
this, commit 051e400 added a new verification mode; if the
object is not OUR_REF, we set a "has_non_tip" flag, and then
later verify that the requested objects are reachable from
our current tips.
However, we still die immediately when the object is not in
our in-memory hash, and at this point we should only have
loaded our tip objects. So the check_non_tip code path does
not ever actually trigger, as any non-tip objects would
have already caused us to die.
We can fix that by using parse_object instead of
lookup_object, which will load the object from disk if it
has not already been loaded.
We still need to check that parse_object does not return
NULL, though, as it is possible we do not have the object
at all. A more appropriate error message would be "no such
object" rather than "not our ref"; however, we do not want
to leak information about what objects are or are not in
the object database, so we continue to use the same "not
our ref" message that would be produced by an unreachable
object.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When upload-pack receives a "want" line from the client, it
adds it to an object array. We call lookup_object to find
the actual object, which will only check for objects already
in memory. This works because we are expecting to find
objects that we already loaded during the ref advertisement.
We use the resulting object structs for a variety of
purposes. Some of them care only about the object flags, but
others care about the type of the object (e.g.,
ok_to_give_up), or even feed them to the revision parser
(when --depth is used), which assumes that objects it
receives are fully parsed.
Once upon a time, this was OK; any object we loaded into
memory would also have been parsed. But since 435c833
(upload-pack: use peel_ref for ref advertisements,
2012-10-04), we try to avoid parsing objects during the ref
advertisement. This means that lookup_object may return an
object with a type of OBJ_NONE. The resulting mess depends
on the exact set of objects, but can include the revision
parser barfing, or the shallow code sending the wrong set of
objects.
This patch teaches upload-pack to parse each "want" object
as we receive it. We do not replace the lookup_object call
with parse_object, as the current code is careful not to let
just any object appear on a "want" line, but rather only one
we have previously advertised (whereas parse_object would
actually load any arbitrary object from disk).
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
upload-pack: drop lookup-before-parse optimization
When we receive a "have" line from the client, we want to
load the object pointed to by the sha1. However, we are
careful to do:
o = lookup_object(sha1);
if (!o || !o->parsed)
o = parse_object(sha1);
to avoid loading the object from disk if we have already
seen it. However, since ccdc603 (parse_object: try internal
cache before reading object db), parse_object already does
this optimization internally. We can just call parse_object
directly.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
index-pack: fix buffer overflow caused by translations
The translation of "completed with %d local objects" is put in a
48-byte buffer, which may be enough for English but not true for any
translations. Convert it to use strbuf (i.e. no hard limit on
translation length).
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
status: advise to consider use of -u when read_directory takes too long
Introduce advice.statusUoption to suggest considering use of -u to
strike different trade-off when it took more than 2 seconds to
enumerate untracked/ignored files.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git status: document trade-offs in choosing parameters to the -u option
In some repostories users experience that "git status" command takes
long time. The command spends some time searching the file system
for untracked files.
Explain the trade-off struck by the default choice of `normal` to
help users make an appropriate choice better, before talking about
the configuration variable.
setup.c: stop prefix_pathspec() from looping past the end of string
The code assumes that the string ends at either `)` or `,`, and does
not handle the case where strcspn() returns length due to end of
string. So specifying ":(top" as pathspec will cause the loop to go
past the end of string.
Signed-off-by: Andrew Wong <andrew.kw.w@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t2200: check that "add -u" limits itself to subdirectory
This behavior is due to change in the future, but let's test
it anyway. That helps make sure we do not accidentally
switch the behavior too soon while we are working in the
area, and it means that we can easily verify the change when
we do make it.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
to avoid the problem of being unable to save in some circumstances with
similar inputs.
Unfortunately this approach produces much worse results on differing
inputs. P4Merge really regards the blank file as the base, and once you
have just a couple of differences between the two branches you end up
with one a massive full-file conflict. The 3-way diff is not readable,
and you have to invoke "difftool MERGE_HEAD HEAD" manually to get a
useful view.
The original approach appears to have invoked special 2-way merge
behaviour in P4Merge that occurs only if the base filename is "" or
equal to the left input. You get a good visual comparison, and it does
not auto-resolve differences. (Normally if one branch matched the base,
it would autoresolve to the other branch).
But there appears to be no way of getting this 2-way behaviour and being
able to reliably save. Having base==left appears to be triggering other
assumptions. There are tricks the user can use to force the save icon
on, but it's not intuitive.
So we now follow a suggestion given in the original patch's discussion:
generate a virtual base, consisting of the lines common to the two
branches. This is the same as the technique used in resolve and octopus
merges, so we relocate that code to a shared function.
Note that if there are no differences at the same location, this
technique can lead to automatic resolution without conflict, combining
everything from the 2 files. As with the other merges using this
technique, we assume the user will inspect the result before saving.
Signed-off-by: Kevin Bracey <kevin@bracey.fi> Reviewed-by: David Aguilar <davvid@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Reverse LOCAL and REMOTE when invoking P4Merge as a mergetool, so that
the incoming branch is now in the left-hand, blue triangle pane, and the
current branch is in the right-hand, green circle pane.
This change makes use of P4Merge consistent with its built-in help, its
reference documentation, and Perforce itself. But most importantly, it
makes merge results clearer. P4Merge is not totally symmetrical between
left and right; despite changing a few text labels from "theirs/ours" to
"left/right" when invoked manually, it still retains its original
Perforce "theirs/ours" viewpoint.
Most obviously, in the result pane P4Merge shows changes that are common
to both branches in green. This is on the basis of the current branch
being green, as it is when invoked from Perforce; it means that lines in
the result are blue if and only if they are being changed by the merge,
making the resulting diff clearer.
Note that P4Merge now shows "ours" on the right for both diff and merge,
unlike other diff/mergetools, which always have REMOTE on the right.
But observe that REMOTE is the working tree (ie "ours") for a diff,
while it's another branch (ie "theirs") for a merge.
Ours and theirs are reversed for a rebase - see "git help rebase".
However, this does produce the desired "show the results of this commit"
effect in P4Merge - changes that remain in the rebased commit (in your
branch, but not in the new base) appear in blue; changes that do not
appear in the rebased commit (from the new base, or common to both) are
in green. If Perforce had rebase, they'd probably not swap ours/theirs,
but make P4Merge show common changes in blue, picking out our changes in
green. We can't do that, so this is next best.
Signed-off-by: Kevin Bracey <kevin@bracey.fi> Reviewed-by: David Aguilar <davvid@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
tag: --force does not have to warn when creating tags
"git tag --force" mentions what old tag object is being replaced
when it is used to update an existing tag, but it shows the same
message when creating a new one. Stop doing that, as it does not
add any information.
Add a test for this and also to ensure --force can replace tags at
all.
Signed-off-by: Phil Hord <hordp@cisco.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The generic chdir() helper sets the PWD environment
variable, as that is what is used by p4 to know its
current working directory. Normally the shell would
do this, but in git-p4, we must do it by hand.
However, when the path contains a symbolic link,
os.getcwd() will return the physical location. If the
p4 client specification includes symlinks, setting PWD
to the physical location causes p4 to think it is not
inside the client workspace. It complains, e.g.
Path /vol/bar/projects/foo/... is not under client root /p/foo
One workaround is to use AltRoots in the p4 client specification,
but it is cleaner to handle it directly in git-p4.
Other uses of chdir still require setting PWD to an
absolute path so p4 features like P4CONFIG work. See bf1d68f (git-p4: use absolute directory for PWD env
var, 2011-12-09).
[ pw: tweak patch and commit message ]
Thanks-to: John Keeping <john@keeping.me.uk> Signed-off-by: Pete Wyckoff <pw@padd.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git-completion.bash: zsh does not implement function redirection correctly
A recent change added functions whose entire standard error stream
is redirected to /dev/null using a construct that is valid POSIX.1
but is not widely used:
funcname () {
cd "$1" && run some command "$2"
} 2>/dev/null
Even though this file is "git-completion.bash", zsh completion
support dot-sources it (instead of asking bash to grok it like tcsh
completion does), and zsh does not implement this redirection
correctly.
With zsh, trying to complete an inexistant directory gave this:
git add no-such-dir/__git_ls_files_helper:cd:2: no such file or directory: no-such-dir/
Also these functions use "cd" to first go somewhere else before
running a command, but the location the caller wants them to go that
is given as an argument to them should not be affected by CDPATH
variable the users may have set for their interactive session.
To fix both of these, wrap the body of the function in a subshell,
unset CDPATH at the beginning of the subshell, and redirect the
standard error stream of the subshell to /dev/null.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr> Signed-off-by: Junio C Hamano <gitster@pobox.com>
add: update pathless 'add [-u|-A]' warning to reflect change of plan
We originally thought the transition would need a period where "git add
[-u|-A]" without pathspec would be forbidden, but the warning is big
enough to scare people and teach them not to use it (or, if so, to
understand the consequences).
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git-archive relies on get_pathspec to convert its argv into
a list of pathspecs. When get_pathspec is given an empty
argv list, it returns a single pathspec, the empty string,
to indicate that everything matches. When we feed this to
our path_exists function, we typically see that the pathspec
turns up at least one item in the tree, and we are happy.
But when our tree is empty, we erroneously think it is
because the pathspec is too limited, when in fact it is
simply that there is nothing to be found in the tree. This
is a weird corner case, but the correct behavior is almost
certainly to produce an empty archive, not to exit with an
error.
This patch teaches git-archive to create empty archives when
there is no pathspec given (we continue to complain if a
pathspec is given, since it obviously is not matched). It
also confirms that the tar and zip writers produce sane
output in this instance.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
We set up the $GIT_UNZIP variable and lazy prereq in
multiple places (and the next patch is about to add another
one). Let's factor it out to avoid repeating ourselves.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
shell: new no-interactive-login command to print a custom message
If I disable git-shell's interactive mode by removing the
~/git-shell-commands directory, attempts to ssh in to the service
produce a message intended for the administrator:
$ ssh git@myserver
fatal: Interactive git shell is not enabled.
hint: ~/git-shell-commands should exist and have read and execute access.
$
That is helpful for the new admin who is wondering "What? Why isn't
the git-shell I just set up working?", but once the site setup is
complete, it would be better to give the user a friendly hint that she
is on the right track, like GitHub does.
Hi <username>! You've successfully authenticated, but
GitHub does not provide shell access.
An appropriate greeting might even include more complex dynamic
information, like gitolite's list of repositories the user has access
to. Add support for a ~/git-shell-commands/no-interactive-login
command that generates an arbitrary greeting. When the user tries to
log in:
* If the file ~/git-shell-commands/no-interactive-login exists,
run no-interactive-login to let the server say what it likes,
then hang up.
* Otherwise, if ~/git-shell-commands/ is present, start an
interactive read-eval-print loop.
* Otherwise, print the usual configuration hint and hang up.
Reported-by: Ethan Reesor <firelizzard@gmail.com> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Improved-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The original git-shell(1) manpage emphasized that the shell supports
only git transport commands. As the shell gained features, that
emphasis and focus in the manual has been lost. Bring it back by
splitting the manpage into a few short sections and fleshing out each:
- SYNOPSIS, describing how the shell gets used in practice
- DESCRIPTION, which gives an overview of the purpose and guarantees
provided by this restricted shell
- COMMANDS, listing supported commands and restrictions on the
arguments they accept
- INTERACTIVE USE, describing the interactive mode
Also add a "see also" section with related reading.
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
format-patch: RFC 2047 says multi-octet character may not be split
Even though an earlier attempt (bafc478..41dd00bad) cleaned
up RFC 2047 encoding, pretty.c::add_rfc2047() still decides
where to split the output line by going through the input
one byte at a time, and potentially splits a character in
the middle. A subject line may end up showing like this:
".... fö?? bar". (instead of ".... föö bar".)
if split incorrectly.
RFC 2047, section 5 (3) explicitly forbids such beaviour
Each 'encoded-word' MUST represent an integral number of
characters. A multi-octet character may not be split across
adjacent 'encoded- word's.
Subject: =?UTF-8?q?....=20f=C3=B6=C3?= <-- NOTE ö is broken here
=?UTF-8?q?=B6=20bar?=
is not, because "ö" character UTF-8 encoding C3 B6 is split here across
adjacent encoded words.
To fix the problem, make the loop grab one _character_ at a time and
determine its output length to see where to break the output line. Note
that this version only knows about UTF-8, but the logic to grab one
character is abstracted out in mbs_chrlen() function to make it possible
to extend it to other encodings with the help of iconv in the future.
Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Split the backward-compatibility notes into two sections, the ones
that affect this release, and the other to describe changes meant
for Git 2.0. The latter gives a context to understand why the
changes for this release is necessary.
setup: suppress implicit "." work-tree for bare repos
If an explicit GIT_DIR is given without a working tree, we
implicitly assume that the current working directory should
be used as the working tree. E.g.,:
GIT_DIR=/some/repo.git git status
would compare against the cwd.
Unfortunately, we fool this rule for sub-invocations of git
by setting GIT_DIR internally ourselves. For example:
git init foo
cd foo/.git
git status ;# fails, as we expect
git config alias.st status
git status ;# does not fail, but should
What happens is that we run setup_git_directory when doing
alias lookup (since we need to see the config), set GIT_DIR
as a result, and then leave GIT_WORK_TREE blank (because we
do not have one). Then when we actually run the status
command, we do setup_git_directory again, which sees our
explicit GIT_DIR and uses the cwd as an implicit worktree.
It's tempting to argue that we should be suppressing that
second invocation of setup_git_directory, as it could use
the values we already found in memory. However, the problem
still exists for sub-processes (e.g., if "git status" were
an external command).
You can see another example with the "--bare" option, which
sets GIT_DIR explicitly. For example:
git init foo
cd foo/.git
git status ;# fails
git --bare status ;# does NOT fail
We need some way of telling sub-processes "even though
GIT_DIR is set, do not use cwd as an implicit working tree".
We could do it by putting a special token into
GIT_WORK_TREE, but the obvious choice (an empty string) has
some portability problems.
Instead, we add a new boolean variable, GIT_IMPLICIT_WORK_TREE,
which suppresses the use of cwd as a working tree when
GIT_DIR is set. We trigger the new variable when we know we
are in a bare setting.
The variable is left intentionally undocumented, as this is
an internal detail (for now, anyway). If somebody comes up
with a good alternate use for it, and once we are confident
we have shaken any bugs out of it, we can consider promoting
it further.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The GIT_PREFIX variable is set based on our location within
the working tree. It should therefore be cleared whenever
GIT_WORK_TREE is cleared.
In practice, this doesn't cause any bugs, because none of
the sub-programs we invoke with local_repo_env cleared
actually care about GIT_PREFIX. But this is the right thing
to do, and future proofs us against that assumption changing.
While we're at it, let's define a GIT_PREFIX_ENVIRONMENT
macro; this avoids repetition of the string literal, which
can help catch any spelling mistakes in the code.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git checkout -" is a short-hand for "git checkout @{-1}" and the
"@{nth}" notation for a negative number is to find nth previous
checkout in the reflog of the HEAD to determine the name of the
branch the user was on. We would want to find the nth most recent
reflog entry that matches "checkout: moving from X to Y" for this.
Unfortunately, reflog is implemented as an append-only file, and the
API to iterate over its entries, for_each_reflog_ent(), reads the
file in order, giving the entries from the oldest to newer. For the
purpose of finding nth most recent one, this API forces us to record
the last n entries in a rotating buffer and give the result out only
after we read everything. To optimize for a common case of finding
the nth most recent one for a small value of n, we also have a side
API for_each_recent_reflog_ent() that starts reading near the end of
the file, but it still has to read the entries in the "wrong" order.
The implementation of understanding @{-1} uses this interface.
This all becomes unnecessary if we add an API to let us iterate over
reflog entries in the reverse order, from the newest to older.
for_each_reflog_ent(): extract a helper to process a single entry
Split the logic that takes a single line of reflog entry in a
strbuf, parses the message, and calls the callback function out of
the loop into a separate helper function.
Documentation/git-push: clarify the description of defaults
We describe what gets pushed by default when the command line does
not give any <refspec> under the bullet point of <refspec>.
It is a bit unfriendly to expect users to read on <refspec> when
they are not giving any in the first place. "What gets pushed" is
determined by taking many factors (<refspec> argument being only one
of them) into account, and is a property of the entire command, not
an individual argument. Also we do not describe "Where the push
goes" when the command line does not say.
Give the description on "what gets pushed to where" upfront before
explaining individual arguments and options.
Also update the description of <refspec> to say what it is, what it
is used for, before explaining what shape it takes.
We keep a static array of variables that should be cleared
when invoking a sub-process on another repo. We statically
size the array with the LOCAL_REPO_ENV_SIZE macro so that
any readers do not have to count it themselves.
As it turns out, no readers actually use the macro, and it
creates a maintenance headache, as modifications to the
array need to happen in two places (one to add the new
element, and another to bump the size).
Since it's NULL-terminated, we can just drop the size macro
entirely. While we're at it, we'll clean up some comments
around it, and add a new mention of it at the top of the
list of environment variable macros. Even though
local_repo_env is right below that list, it's easy to miss,
and additions to that list should consider local_repo_env.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently this is cosmetic change - the merges are ignored, becuase the methods
(lookup_svn_merge, find_rev_before, find_rev_after) are failing on comparing text with number.
See http://www.open.collab.net/community/subversion/articles/merge-info.html
Extract:
The range r30430:30435 that was added to 1.5.x in this merge has a '*' suffix for 1.5.x\www.
This '*' is the marker for a non-inheritable mergeinfo range.
The '*' means that only the path on which the mergeinfo is explicitly set has had this range merged into it.
Signed-off-by: Jan Pesta <jan.pesta@certicon.cz> Signed-off-by: Eric Wong <normalperson@yhbt.net>
git p4 test: make sure P4CONFIG relative path works
This adds a test for the fix in bf1d68f (git-p4: use absolute
directory for PWD env var, 2011-12-09). It is necessary to
set PWD to an absolute path so that p4 can find files referenced
by non-absolute paths, like the value of the P4CONFIG environment
variable.
P4 does not open files directly; it builds a path by prepending
the contents of the PWD environment variable.
Signed-off-by: Pete Wyckoff <pw@padd.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
bundle: Fix "verify" output if history is complete
A more informative message for "complete" bundles was added in commit 8c3710fd3011 (tweak "bundle verify" of a complete history, 2012-06-04).
However, the prerequisites ref list is currently read *after* we
check if it equals zero, which means we never actually use the
number of prerequisite refs to decide when to print the newly
introduced message. The code incorrectly uses the number of
references recorded in the bundle instead.
Signed-off-by: Lukas Fleischer <git@cryptocrack.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The documentation of '-A' and '-u' is very confusing for someone who
doesn't already know what they do. Describe them with fewer words and
clearer parallelism to each other and to the behavior of plain 'add'.
Also mention the default <pathspec> for '-A' as well as '-u', because
it applies to both.
Signed-off-by: Greg Price <price@mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The new option "--follow-tags" tells "git push" to push annotated
tags that are missing from the other side and that can be reached by
the history that is otherwise pushed out.
For example, if you are using the "simple", "current", or "upstream"
push, you would ordinarily push the history leading to the commit at
your current HEAD and nothing else. With this option, you would
also push all annotated tags that can be reached from that commit to
the other side.
Similar to in_merge_bases(commit, other) that returns true when
commit is an ancestor (i.e. in the merge bases between the two) of
the other commit, in_merge_bases_many(commit, n_other, other[])
checks if commit is an ancestor of any of the other[] commits.
clear_commit_marks(struct commit *, unsigned) only can clear flag
bits starting from a single commit; introduce an API to allow
feeding an array of commits, so that flag bits can be cleared from
commits reachable from any of them with a single traversal.
git-completion.zsh: define __gitcomp_file compatibility function
Commit fea16b47b60 (Fri Jan 11 19:48:43 2013, Manlio Perillo,
git-completion.bash: add support for path completion), introduced a new
__gitcomp_file function that uses the bash builtin "compgen". The
function was redefined for ZSH in the deprecated section of
git-completion.bash, but not in the new git-completion.zsh script.
As a result, users of git-completion.zsh trying to complete "git add
fo<tab>" get an error:
git add fo__gitcomp_file:8: command not found: compgen
This patch adds the redefinition and removes the error.
Signed-off-by: Matthieu Moy <Matthieu.Moy@imag.fr> Signed-off-by: Junio C Hamano <gitster@pobox.com>
With "git submodule init" the user is able to tell git he cares about one
or more submodules and wants to have it populated on the next call to "git
submodule update". But currently there is no easy way he could tell git he
does not care about a submodule anymore and wants to get rid of his local
work tree (except he knows a lot about submodule internals and removes the
"submodule.$name.url" setting from .git/config together with the work tree
himself).
Help those users by providing a 'deinit' command. This removes the
whole submodule.<name> section from .git/config (either for the given
submodule(s) or for all those which have been initialized if '.' is used)
together with their work tree. Fail if the current work tree contains
modifications (unless forced), but don't complain when either the work
tree is already removed or no settings are found in .git/config.
Add tests and link the man pages of "git submodule deinit" and "git rm"
to assist the user in deciding whether removing or unregistering the
submodule is the right thing to do for him. Also add the deinit subcommand
to the completion list.
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
submodule update: when using recursion, show full path
Previously when using update with recursion, only the path for the
inner-most module was printed. Now the path is printed relative to
the directory the command was started from. This now matches the
behavior of submodule foreach.
Signed-off-by: William Entriken <github.com@phor.net> Acked-by: Jens Lehmann <Jens.Lehmann@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
CGit uses these symbols to output the correct HTML around graph
elements. Making these symbols private means that CGit cannot be
updated to use Git 1.8.0 or newer, so let's not do that.
On top of the revert, also add comments so that we avoid reintroducing
this problem in the future and suggest to those modifying this API
that they might want to discuss it with the CGit developers.
Signed-off-by: John Keeping <john@keeping.me.uk> Acked-by: Jason A. Donenfeld <Jason@zx2c4.com> Acked-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
A maildir does not technically record the order in which
items were placed into it. That means that when applying a
patch series from a maildir, we may get the patches in the
wrong order. We try to work around this by sorting the
filenames. Unfortunately, this may or may not work depending
on the naming scheme used by the writer of the maildir.
For instance, mutt will write:
${epoch_seconds}.${pid}_${seq}.${host}
where we have:
- epoch_seconds: timestamp at which entry was written
- pid: PID of writing process
- seq: a sequence number to ensure uniqueness of filenames
- host: hostname
None of the numbers are zero-padded. Therefore, when we sort
the names as byte strings, entries that cross a digit
boundary (e.g., 10) will sort out of order. In the case of
timestamps, it almost never matters (because we do not cross
a digit boundary in the epoch time very often these days).
But for the sequence number, a 10-patch series would be
ordered as 1, 10, 2, 3, etc.
To fix this, we can use a custom sort comparison function
which traverses each string, comparing chunks of digits
numerically, and otherwise doing a byte-for-byte comparison.
That would sort:
123.456_1.bar
123.456_2.bar
...
123.456_10.bar
according to the sequence number. Since maildir does not
define a filename format, this is really just a heuristic.
But it happens to work for mutt, and there is a reasonable
chance that it will work for other writers, too (at least as
well as a straight sort).
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Before 82dce99 (attr: more matching optimizations from .gitignore,
2012-10-15), .gitattributes did not have any special treatment of a
leading '!'. The docs, however, always said
The rules how the pattern matches paths are the same as in
`.gitignore` files; see linkgit:gitignore[5].
By those rules, leading '!' means pattern negation. So 82dce99
correctly determined that this kind of line makes no sense and should
be disallowed.
However, users who actually had a rule for files starting with a '!'
are in a bad position: before 82dce99 '!' matched that literal
character, so it is conceivable that users have .gitattributes with
such lines in them. After 82dce99 the unescaped version was
disallowed in such a way that git outright refuses to run(!) most
commands in the presence of such a .gitattributes. It therefore
becomes very hard to fix, let alone work with, such repositories.
Let's at least allow the users to fix their repos: change the fatal
error into a warning.
Reported-by: mathstuf@gmail.com Signed-off-by: Thomas Rast <trast@student.ethz.ch> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* wk/user-manual:
user-manual: Flesh out uncommitted changes and submodule updates
user-manual: Use request-pull to generate "please pull" text
user-manual: Reorganize the reroll sections, adding 'git rebase -i'
describe: --match=<pattern> must limit the refs even when used with --all
The logic to limit the refs used for describing with a matching pattern
with --match=<pattern> parameter was implemented incorrectly when --all
is in effect. It just demoted a ref that did not match the pattern to
lower priority---if there aren't other refs with higher priority
that describe the given commit, such an unmatching ref was still used.
When --match is used, reject refs that do not match the given
criteria, so that with or without --all, the output will only use
refs that match the pattern.
An earlier workaround designed to help people who list logical
directories that will not match what getcwd(3) returns in the
GIT_CEILING_DIRECTORIES had an adverse effect when it is slow to
stat and readlink a directory component of an element listed on it.
* mh/maint-ceil-absolute:
Provide a mechanism to turn off symlink resolution in ceiling paths
Add a credential() function which is an interface to the git
credential command. The code is heavily based on credential_*
functions in <contrib/mw-to-git/git-remote-mediawiki>.
Signed-off-by: Michal Nazarewicz <mina86@mina86.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
archive-zip: fix compressed size for stored export-subst files
Currently ZIP archive entries of files with export-subst attribute are
broken if they are stored uncompressed.
We get the size of a file from sha1_object_info(), but this number is
likely wrong for files whose contents are changed due to export-subst
placeholder expansion. We use sha1_file_to_archive() to get the
expanded file contents and size in that case. We proceed to use that
size for the uncompressed size field (good), but the compressed size
field is set based on the size from sha1_object_info() (bad).
This matters only for uncompressed files because for deflated files
we use the correct value after compression is done. And for files
without export-subst expansion the sizes from sha1_object_info() and
sha1_file_to_archive() are the same, so they are unaffected as well.
This patch fixes the issue by setting the compressed size based on the
uncompressed size only after we actually know the latter.
Also make use of the test file substfile1 to check for the breakage;
it was only stored verbatim so far. For that purpose, set the
attribute export-subst and replace its contents with the expected
expansion after committing.
Signed-off-by: Rene Scharfe <rene.scharfe@lsrfire.ath.cx> Signed-off-by: Junio C Hamano <gitster@pobox.com>