builtin/notes: add --allow-empty, to allow storing empty notes
Although the "git notes" man page advertises that we support binary-safe
notes addition (using the -C option), we currently do not support adding
the empty note (i.e. using the empty blob to annotate an object). Instead,
an empty note is always treated as an intent to remove the note
altogether.
Introduce the --allow-empty option to the add/append/edit subcommands,
to explicitly allow an empty note to be stored into the notes tree.
Also update the documentation, and add test cases for the new option.
Reported-by: James H. Fisher <jhf@trifork.com> Improved-by: Kyle J. McKay <mackyle@gmail.com> Improved-by: Junio C Hamano <gitster@pobox.com> Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
builtin/notes: split create_note() to clarify add vs. remove logic
create_note() has a non-trivial interface, and comprises three loosely
related parts:
1. launching the editor with the note contents, if needed
2. appending to an existing note, if append_only was given
3. adding or removing the resulting note, based on whether it's non-empty
Split it along those lines to make the logic clearer: The first part
goes into a new function - prepare_note_data(), with a simpler interface.
The second part is moved into append_edit(), which is the only user of
this code. Finally, the add vs. remove decision is moved into the callers
(add() and append_edit()), keeping the logic for writing the actual note
object in a separate function: write_note_data().
Suggested-by: Junio C Hamano <gitster@pobox.com> Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
In particular, git-fast-import and -export link to each
other, and gitremote-helpers links to existing remote
helpers, and vice versa. Also link to fast-import from the
remote helper spec, as this is relevant for remote helpers
using the fast-import format.
Signed-off-by: Max Horn <max@quendi.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
run-command: use void to declare that functions take no parameters
Explicitly declare that git_atexit_dispatch() and git_atexit_clear()
take no parameters instead of leaving their parameter list empty and
thus unspecified.
Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
In preparation for some needed refactoring, rename struct msg_arg to
struct note_data, and rename its instances from "msg" to "d" (also
removing some unnecessary parentheses). The 'msg_arg' name was
inherited from tag.c, but is not really a good name for the contents
of a note.
Also rename write_note_data() to copy_obj_to_fd(), which more aptly
describes what it actually does: Copying the contents of a git object
(given by its SHA1) into a given file descriptor.
Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t3301: verify that 'git notes' removes empty notes by default
Add test cases documenting the current behavior when trying to
add/append/edit empty notes. This is in preparation for adding
--allow-empty; to allow empty notes to be stored.
Improved-by: Eric Sunshine <sunshine@sunshineco.com> Improved-by: Junio C Hamano <gitster@pobox.com> Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
builtin/notes: fix premature failure when trying to add the empty blob
This fixes a small buglet when trying to explicitly add the empty blob
as a note object using the -c or -C option to git notes add/append.
Instead of failing with a nonsensical error message indicating that the
empty blob does not exist, we should rather behave as if an empty notes
message was given (e.g. using -m "" or -F /dev/null).
The next patch contains a test that verifies the fixed behavior.
Found-by: Eric Sunshine <sunshine@sunshineco.com> Signed-off-by: Johan Herland <johan@herland.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Convert users of struct child_process to using the managed argv_array
args instead of providing their own. This shortens the code a bit and
ensures that the allocated memory is released automatically after use.
Suggested-by: Jeff King <peff@peff.net> Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
trailer: use CHILD_PROCESS_INIT in apply_command()
Initialize the struct child_process variable cp at declaration time.
This is shorter, saves a function call and prevents using the variable
before initialization by mistake.
Suggested-by: Jeff King <peff@peff.net> Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Merge branch 'jc/conflict-hint' into cc/interpret-trailers-more
* jc/conflict-hint:
merge & sequencer: turn "Conflicts:" hint into a comment
builtin/commit.c: extract ignore_non_trailer() helper function
merge & sequencer: unify codepaths that write "Conflicts:" hint
builtin/merge.c: drop a parameter that is never used
git-tag.txt: Add a missing hyphen to `-s`
trailer: display a trailer without its trailing newline
Trailers passed to the parse_trailer() function often have
a trailing newline. When erroring out, we should display
the invalid trailer properly, that means without any
trailing newline.
Helped-by: Junio C Hamano <gitster@pobox.com> Helped-by: Jeff King <peff@peff.net> Signed-off-by: Christian Couder <chriscool@tuxfamily.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
t1410: fix breakage on case-insensitive filesystems
Two tests recently added to t1410 create branches "a" and
"a/b" to test d/f conflicts on reflogs. Earlier, unrelated
tests in that script create the path "A/B" in the working
tree. There's no conflict on a case-sensitive filesystem,
but on a case-insensitive one, "git log" will complain that
"a/b" is both a revision and a working tree path.
We could fix this by using a "--" to disambiguate, but we
are probably better off using names that are less confusing
to make it more clear that they are unrelated to the working
tree files. This patch turns "a/b" into "one/two".
Reported-by: Michael Blume <blume.mike@gmail.com> Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use libcurl's high-level API functions to implement git-imap-send
instead of the previous low-level OpenSSL-based functions.
Since version 7.30.0, libcurl's API has been able to communicate with
IMAP servers. Using those high-level functions instead of the current
ones would reduce imap-send.c by some 1200 lines of code. For now,
the old ones are wrapped in #ifdefs, and the new functions are enabled
by make if curl's version is >= 7.34.0, from which version on curl's
CURLOPT_LOGIN_OPTIONS (enabling IMAP authentication) parameter has been
available. The low-level functions will still be used for tunneling
into the server for now.
As I don't have access to that many IMAP servers, I haven't been able to
test the new code with a wide variety of parameter combinations. I did
test both secure and insecure (imaps:// and imap://) connections and
values of "PLAIN" and "LOGIN" for the authMethod.
In order to suppress a sparse warning about "using sizeof on a
function", we use the same solution used in commit 9371322a6
("sparse: suppress some "using sizeof on a function" warnings",
06-10-2013) which solved exactly this problem for the other commands
using libcurl.
Helped-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Signed-off-by: Bernhard Reiter <ockham@raz.or.at> Signed-off-by: Junio C Hamano <gitster@pobox.com>
git_connect: set ssh shell command in GIT_SSH_COMMAND
It may be impractical to install a wrapper script for GIT_SSH
when additional parameters need to be passed. Provide an alternative
way of specifying a shell command to be run, including command line
arguments, by means of the GIT_SSH_COMMAND environment variable,
which behaves like GIT_SSH but is passed to the shell.
The special circuitry to modify parameters in the case of using
PuTTY's plink/tortoiseplink is activated only when using GIT_SSH;
in the case of using GIT_SSH_COMMAND, it is deliberately left up to
the user to make any required parameters adaptation before calling
the underlying ssh implementation.
Signed-off-by: Thomas Quinot <thomas@quinot.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* rs/use-child-process-init-more:
bundle: split out ref writing from bundle_create
bundle: split out a helper function to compute and write prerequisites
bundle: split out a helper function to create pack data
use child_process_init() to initialize struct child_process variables
The option name "--store" was used early in development, but
never even made it into an applied patch, let alone a
released version of git. I forgot to update the matching
documentation at the time, though.
Noticed-by: Jesse Hopkins <jesse.hopkins@lmco.com> Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
While using diff-highlight with other tools, I have discovered that Python
ignores SIGPIPE by default. Unfortunately, this also means that tools
attempting to launch a pager under Python--and don't realize this is
happening--means that the subprocess inherits this setting. In this case, it
means diff-highlight will be launched with SIGPIPE being ignored. Let's work
with those broken scripts by restoring the default SIGPIPE handler.
Signed-off-by: John Szakmeister <john@szakmeister.net> Acked-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
ignore stale directories when checking reflog existence
When we update a ref, we have two rules for whether or not
we actually update the reflog:
1. If the reflog already exists, we will always append to
it.
2. If log_all_ref_updates is set, we will create a new
reflog file if necessary.
We do the existence check by trying to open the reflog file,
either with or without O_CREAT (depending on log_all_ref_updates).
If it fails, then we check errno to see what happened.
If we were not using O_CREAT and we got ENOENT, the file
doesn't exist, and we return success (there isn't a reflog
already, and we were not told to make a new one).
If we get EISDIR, then there is likely a stale directory
that needs to be removed (e.g., there used to be "foo/bar",
it was deleted, and the directory "foo" was left. Now we
want to create the ref "foo"). If O_CREAT is set, then we
catch this case, try to remove the directory, and retry our
open. So far so good.
But if we get EISDIR and O_CREAT is not set, then we treat
this as any other error, which is not right. Like ENOENT,
EISDIR is an indication that we do not have a reflog, and we
should silently return success (we were not told to create
it). Instead, the current code reports this as an error, and
we fail to update the ref at all.
Note that this is relatively unlikely to happen, as you
would have to have had reflogs turned on, and then later
turned them off (it could also happen due to a bug in fetch,
but that was fixed in the previous commit). However, it's
quite easy to fix: we just need to treat EISDIR like ENOENT
for the non-O_CREAT case, and silently return (note that
this early return means we can also simplify the O_CREAT
case).
Our new tests cover both cases (O_CREAT and non-O_CREAT).
The first one already worked, of course.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
When we start the git-fetch program, we call git_config to
load all config, but our callback only processes the
fetch.prune option; we do not chain to git_default_config at
all.
This means that we may not load some core configuration
which will have an effect. For instance, we do not load
core.logAllRefUpdates, which impacts whether or not we
create reflogs in a bare repository.
Note that I said "may" above. It gets even more exciting. If
we have to transfer actual objects as part of the fetch,
then we call fetch_pack as part of the same process. That
function loads its own config, which does chain to
git_default_config, impacting global variables which are
used by the rest of fetch. But if the fetch is a pure ref
update (e.g., a new ref which is a copy of an old one), we
skip fetch_pack entirely. So we get inconsistent results
depending on whether or not we have actual objects to
transfer or not!
Let's just load the core config at the start of fetch, so we
know we have it (we may also load it again as part of
fetch_pack, but that's OK; it's designed to be idempotent).
Our tests check both cases (with and without a pack). We
also check similar behavior for push for good measure, but
it already works as expected.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Locked paths can be saved in a linked list so that if something wrong
happens, *.lock are removed. For relative paths, this works fine if we
keep cwd the same, which is true 99% of time except:
- update-index and read-tree hold the lock on $GIT_DIR/index really
early, then later on may call setup_work_tree() to move cwd.
- Suppose a lock is being held (e.g. by "git add") then somewhere
down the line, somebody calls real_path (e.g. "link_alt_odb_entry"),
which temporarily moves cwd away and back.
During that time when cwd is moved (either permanently or temporarily)
and we decide to die(), attempts to remove relative *.lock will fail,
and the next operation will complain that some files are still locked.
Avoid this case by turning relative paths to absolute before storing
the path in "filename" field.
Reported-by: Yue Lin Ho <yuelinho777@gmail.com> Helped-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Helped-by: Johannes Sixt <j6t@kdbg.org> Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com> Adapted-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Merge branch 'for-junio' of git://bogomips.org/git-svn
* 'for-junio' of git://bogomips.org/git-svn:
git-svn: use SVN::Ra::get_dir2 when possible
git-svn: add space after "W:" prefix in warning
git-svn: (cleanup) remove editor param passing
git-svn: prepare SVN::Ra config pieces once
Git.pm: add specified name to tempfile template
git-svn: disable _rev_list memoization
git-svn: save a little memory as fetch progresses
git-svn: remove unnecessary DESTROY override
git-svn: reload RA every log-window-size
git-svn.txt: advertise pushurl with dcommit
git-svn: remove mergeinfo rev caching
git-svn: cache only mergeinfo revisions
git-svn: reduce check_cherry_pick cache overhead
git-svn: only look at the root path for svn:mergeinfo
git-svn: only look at the new parts of svn:mergeinfo
The bundle_create() function has a number of logical steps:
process the input, write the refs, and write the packfile.
Recent commits split the first and third into separate
sub-functions. It's worth splitting the middle step out,
too, if only because it makes the progression of the steps
more obvious.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
bundle: split out a helper function to compute and write prerequisites
The new helper compute_and_write_prerequistes() is ugly, but it
cannot be avoided. Ideally we should avoid a function that computes
and does I/O at the same time, but the prerequisites lines in the
output needs the human readable title only to help the recipient of
the bundle. The code copies them straight from the rev-list output
and immediately discards as no other internal computation needs that
information.
bundle: split out a helper function to create pack data
The create_bundle() function, while it does one single logical
thing, takes a rather large implementation to do so.
Let's start separating what it does into smaller steps to make it
easier to see what is going on. This is a first step to separate
out the actual pack-data generation, after the earlier part of the
function figures out which part of the history to place in the
bundle.
The callers of get_merge_bases() can choose to leave object flags
used during the merge-base traversal by passing cleanup=0 as a
parameter, but in practice a very few callers can afford to do so
(namely, "git merge-base"), as they need to compute merge base in
preparation for other processing of their own and they need to see
the object without contaminate flags.
Change the function signature of get_merge_bases_many() and
get_merge_bases() to drop the cleanup parameter, so that the
majority of the callers do not have to say ", 1" at the end.
Give a new get_merge_bases_many_dirty() API to support only a few
callers that know they do not need to spend cycles cleaning up the
object flags.
Unless there is a good reason to belieave that a particular
invocation of a get_merge_bases*() is the last one that cares about
the object flags the computation of merge bases leaves on the
objects, the "cleanup" parameter should always be true, and I do not
think there is one in this codepath.
cache-tree: avoid infinite loop on zero-entry tree
The loop in cache-tree's update_one iterates over all the
entries in the index. For each one, we find the cache-tree
subtree which represents our path (creating it if
necessary), and then recurse into update_one again. The
return value we get is the number of index entries that
belonged in that subtree. So for example, with entries:
a/one
a/two
b/one
We start by processing the first entry, "a/one". We would
find the subtree for "a" and recurse into update_one. That
would then handle "a/one" and "a/two", and return the value
2. The parent function then skips past the 2 handled
entries, and we continue by processing "b/one".
If the recursed-into update_one ever returns 0, then we make
no forward progress in our loop. We would process "a/one"
over and over, infinitely.
This should not happen normally. Any subtree we create must
have at least one path in it (the one that we are
processing!). However, we may also reuse a cache-tree entry
we found in the on-disk index. For the same reason, this
should also never have zero entries. However, certain buggy
versions of libgit2 could produce such bogus cache-tree
records. The libgit2 bug has since been fixed, but it does
not hurt to protect ourselves against bogus input coming
from the on-disk data structures.
Note that this is not a die("BUG") or assert, because it is
not an internal bug, but rather a corrupted on-disk
structure. It's possible that we could even recover from it
(by throwing out the bogus cache-tree entry), but it is not
worth the effort; the important thing is that we report an
error instead of looping infinitely.
Signed-off-by: Jeff King <peff@peff.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* git://ozlabs.org/~paulus/gitk:
gitk: Remove boilerplate for configuration variables
gitk: Show detached HEAD if --all is specified
gitk: Do not depend on Cygwin's "kill" command on Windows
If HEAD is detached, 'gitk --all' does not show it. This is inconvenient
for frontend program, and for example git log does show the detached HEAD.
gitk uses git rev-parse to find a list of branches to show.
Apparently, the command does not include detached HEAD to output if
--all argument is specified. This has been discussed in [1] and stated
as expected behavior. So rev-parse's parameters should be tuned in gitk.
gitk: Do not depend on Cygwin's "kill" command on Windows
Windows does not necessarily mean Cygwin, it could also be MSYS. The
latter ships with a version of "kill" that does not understand "-f".
In msysgit this was addressed by shipping Cygwin's version of kill.
Properly fix this by using the stock Windows "taskkill" command instead,
which is available since Windows XP Professional.
Signed-off-by: Sebastian Schuberth <sschuberth@gmail.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Memoizing these initialization functions saves some memory for
long fetches which require scanning many unwanted revisions
before any wanted revisions happen.
* dm/port2zos:
compat/bswap.h: detect endianness from XL C compiler macros
Makefile: reorder linker flags in the git executable rule
git-compat-util.h: support variadic macros with the XL C compiler
Tighten the logic to decide that an unreachable cruft is
sufficiently old by covering corner cases such as an ancient object
becoming reachable and then going unreachable again, in which case
its retention period should be prolonged.
* jk/prune-mtime: (28 commits)
drop add_object_array_with_mode
revision: remove definition of unused 'add_object' function
pack-objects: double-check options before discarding objects
repack: pack objects mentioned by the index
pack-objects: use argv_array
reachable: use revision machinery's --indexed-objects code
rev-list: add --indexed-objects option
rev-list: document --reflog option
t5516: test pushing a tag of an otherwise unreferenced blob
traverse_commit_list: support pending blobs/trees with paths
make add_object_array_with_context interface more sane
write_sha1_file: freshen existing objects
pack-objects: match prune logic for discarding objects
pack-objects: refactor unpack-unreachable expiration check
prune: keep objects reachable from recent objects
sha1_file: add for_each iterators for loose and packed objects
count-objects: use for_each_loose_file_in_objdir
count-objects: do not use xsize_t when counting object size
prune-packed: use for_each_loose_file_in_objdir
reachable: mark index blobs as SEEN
...
use child_process_init() to initialize struct child_process variables
Call child_process_init() instead of zeroing the memory of variables of
type struct child_process by hand before use because the former is both
clearer and shorter.
Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
receive-pack: avoid minor leak in case start_async() fails
If the asynchronous start of copy_to_sideband() fails, then any
env_array entries added to struct child_process proc by
prepare_push_cert_sha1() are leaked. Call the latter function only
after start_async() succeeded so that the allocated entries are
cleaned up automatically by start_command() or finish_command().
Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
merge & sequencer: turn "Conflicts:" hint into a comment
Just like other hints such as "Changes to be committed" we show in
the editor to remind the committer what paths were involved in the
resulting commit to help improving their log message, this section
is merely a reminder.
Traditionally, it was not made into comments primarily because it
has to be generated outside the wt-status infrastructure, and also
because it was meant as a bit stronger reminder than the others
(i.e. explaining how you resolved conflicts is much more important
than mentioning what you did to every paths involved in the commit).
But that still does not make this hint a part of the log message
proper, and not showing it as a comment is inviting mistakes.
Note that we still notice "Conflicts:" followed by list of indented
pathnames as an old-style cruft and insert a new Signed-off-by:
before it. This is so that "commit --amend -s" adds the new S-o-b
at the right place when used on an older commit.
builtin/commit.c: extract ignore_non_trailer() helper function
Extract a helper function from prepare_to_commit() to determine
where to place a new Signed-off-by: line, which is essentially the
true "end" of the log message, ignoring the trailing "Conflicts:"
line and everything below it.
The detection _should_ make sure the "Conflicts:" line it finds is
truly the conflict hint block by checking everything that follows is
a HT indented pathname to avoid false positive, but this logic will
be revamped in a later patch to ignore comments and blanks anyway,
so it is left as-is in this step.
Teach difftool to exit when a diff tool returns a non-zero exit
code when either --trust-exit-code is specified or
difftool.trustExitCode is true.
Forward exit codes from invoked diff tools to the caller when
--trust-exit-code is used.
Suggested-by: Adri Farr <14farresa@gmail.com> Helped-by: Johannes Sixt <j6t@kdbg.org> Signed-off-by: David Aguilar <davvid@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
grep: add color.grep.matchcontext and color.grep.matchselected
The config option color.grep.match can be used to specify the highlighting
color for matching strings. Add the options matchContext and matchSelected
to allow different colors to be specified for matching strings in the
context vs. in selected lines. This is similar to the ms and mc specifiers
in GNU grep's environment variable GREP_COLORS.
Tests are from Zoltan Klinger's earlier attempt to solve the same
issue in a different way.
Signed-off-by: Rene Scharfe <l.s.r@web.de> Signed-off-by: Junio C Hamano <gitster@pobox.com>
It took me a long time to notice the rider on the pack.threads
configuration option that it would multiple the memory consumption
by the number of CPUs in the machine. Clarify that the limit
applies per-thread.
Signed-off-by: Robert de Bath <rdebath@tvisiontech.co.uk> Signed-off-by: Junio C Hamano <gitster@pobox.com>
strbuf_add_commented_lines(): avoid SP-HT sequence in commented lines
The strbuf_add_commented_lines() function passes a pair of prefixes,
one to be used for a non-empty line, and the other for an empty
line, to underlying add_lines(). The former is set to a comment
char followed by a SP, while the latter is set to just the comment
char. This is designed to give a SP after the comment character,
e.g. "# <user text>\n", on a line with some text, and to avoid
emitting an unsightly "# \n" for an empty line.
Teach this machinery to also use the latter space-less prefix when
the payload line begins with a tab, to show e.g. "#\t<user text>\n";
otherwise we will end up showing "# \t<user text>\n" which is
similarly unsightly.