Christian Couder [Tue, 18 Dec 2012 15:26:38 +0000 (16:26 +0100)]
Makefile: detect when PYTHON_PATH changes
When make is run, the python scripts are created from *.py files that
are changed to use the python given by PYTHON_PATH. And PYTHON_PATH
is set by default to /usr/bin/python on Linux.
However, next time make is run with a different value in PYTHON_PATH,
we failed to regenerate these scripts.
Signed-off-by: Christian Couder <redacted> Acked-by: Pete Wyckoff <redacted> Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Thu, 10 Jan 2013 22:04:19 +0000 (14:04 -0800)]
Merge branch 'nd/invalidate-i-t-a-cache-tree' into maint
* nd/invalidate-i-t-a-cache-tree:
cache-tree: invalidate i-t-a paths after generating trees
cache-tree: fix writing cache-tree when CE_REMOVE is present
cache-tree: replace "for" loops in update_one with "while" loops
cache-tree: remove dead i-t-a code in verify_cache()
Junio C Hamano [Thu, 10 Jan 2013 21:47:25 +0000 (13:47 -0800)]
Merge branch 'as/dir-c-cleanup'
Refactor and generally clean up the directory traversal API
implementation.
* as/dir-c-cleanup:
dir.c: rename free_excludes() to clear_exclude_list()
dir.c: refactor is_path_excluded()
dir.c: refactor is_excluded()
dir.c: refactor is_excluded_from_list()
dir.c: rename excluded() to is_excluded()
dir.c: rename excluded_from_list() to is_excluded_from_list()
dir.c: rename path_excluded() to is_path_excluded()
dir.c: rename cryptic 'which' variable to more consistent name
Improve documentation and comments regarding directory traversal API
api-directory-listing.txt: update to match code
Junio C Hamano [Thu, 10 Jan 2013 21:47:20 +0000 (13:47 -0800)]
Merge branch 'jk/config-uname'
Move the bits to set fallback default based on the platform from
the main Makefile to a separate file, so that it can be included in
Makefiles in subdirectories.
* jk/config-uname:
Makefile: hoist uname autodetection to config.mak.uname
Junio C Hamano [Thu, 10 Jan 2013 21:47:15 +0000 (13:47 -0800)]
Merge branch 'nd/wildmatch'
Allows pathname patterns in .gitignore and .gitattributes files
with double-asterisks "foo/**/bar" to match any number of directory
hierarchies.
* nd/wildmatch:
wildmatch: replace variable 'special' with better named ones
compat/fnmatch: respect NO_FNMATCH* even on glibc
wildmatch: fix "**" special case
t3070: Disable some failing fnmatch tests
test-wildmatch: avoid Windows path mangling
Support "**" wildcard in .gitignore and .gitattributes
wildmatch: make /**/ match zero or more directories
wildmatch: adjust "**" behavior
wildmatch: fix case-insensitive matching
wildmatch: remove static variable force_lower_case
wildmatch: make wildmatch's return value compatible with fnmatch
t3070: disable unreliable fnmatch tests
Integrate wildmatch to git
wildmatch: follow Git's coding convention
wildmatch: remove unnecessary functions
Import wildmatch from rsync
ctype: support iscntrl, ispunct, isxdigit and isprint
ctype: make sane_ctype[] const array
Junio C Hamano [Tue, 8 Jan 2013 08:02:49 +0000 (00:02 -0800)]
log --use-mailmap: optimize for cases without --author/--committer search
When we taught the commit_match() mechanism to pay attention to the
new --use-mailmap option, we started to unconditionally copy the
commit object to a temporary buffer, just in case we need the author
and committer lines updated via the mailmap mechanism, and rewrite
author and committer using the mailmap.
It turns out that this has a rather unpleasant performance
implications. In the linux kernel repository, running
$ git log --author='Junio C Hamano' --pretty=short >/dev/null
under /usr/bin/time, with and without --use-mailmap (the .mailmap
file is 118 entries long, the particular author does not appear in
it), cost (with warm cache):
The latter case is an unnecessary performance regression. We may
want to _show_ the result with mailmap applied, but we do not have
to copy and rewrite the author/committer of all commits we try to
match if we do not query for these fields.
Trivially optimize this performace regression by limiting the
rewrites for only when we are matching with author/committer fields.
Antoine Pelisse [Sat, 5 Jan 2013 21:26:40 +0000 (22:26 +0100)]
mailmap: simplify map_user() interface
Simplify map_user(), mostly to avoid copies of string buffers. It
also simplifies caller functions.
map_user() directly receive pointers and length from the commit buffer
as mail and name. If mapping of the user and mail can be done, the
pointer is updated to a new location. Lengths are also updated if
necessary.
The caller of map_user() can then copy the new email and name if
necessary.
Signed-off-by: Antoine Pelisse <redacted> Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Sat, 5 Jan 2013 21:26:39 +0000 (22:26 +0100)]
mailmap: remove email copy and length limitation
In map_user(), we have email pointer that points at the beginning of
an e-mail address, but the buffer is not terminated with a NUL after
the e-mail address. It typically has ">" after the address, and it
could have even more if it comes from author/committer line in a
commit object. Or it may not have ">" after it.
We used to copy the e-mail address proper into a temporary buffer
before asking the string-list API to find the e-mail address in the
mailmap, because string_list_lookup() function only takes a NUL
terminated full string.
Introduce a helper function lookup_prefix that takes the email
pointer and the length, and finds a matching entry in the string
list used for the mailmap, by doing the following:
- First ask string_list_find_insert_index() where in its sorted
list the e-mail address we have (including the possible trailing
junk ">...") would be inserted.
- It could find an exact match (e.g. we had a clean e-mail address
without any trailing junk). We can return the item in that case.
- Or it could return the index of an item that sorts after the
e-mail address we have.
- If we did not find an exact match against a clean e-mail address,
then the record we are looking for in the mailmap has to exist
before the index returned by the function (i.e. "email>junk"
always sorts later than "email"). Iterate, starting from that
index, down the map->items[] array until we find the exact record
we are looking for, or we see a record with a key that definitely
sorts earlier than the e-mail we are looking for (i.e. when we
are looking for "email" in "email>junk", a record in the mailmap
that begins with "emaik" strictly sorts before "email", if such a
key existed in the mailmap).
This, together with the earlier enhancement to support
case-insensitive sorting, allow us to remove an extra copy of email
buffer to downcase it.
A part of this is based on Antoine Pelisse's previous work.
Ralf Thielow [Thu, 10 Jan 2013 17:45:59 +0000 (18:45 +0100)]
commit: make default of "cleanup" option configurable
The default of the "cleanup" option in "git commit"
is not configurable. Users who don't want to use the
default have to pass this option on every commit since
there's no way to configure it. This commit introduces
a new config option "commit.cleanup" which can be used
to change the default of the "cleanup" option in
"git commit".
Signed-off-by: Ralf Thielow <redacted> Signed-off-by: Junio C Hamano <redacted>
Peter Eisentraut [Thu, 10 Jan 2013 12:29:58 +0000 (07:29 -0500)]
git-commit-tree(1): correct description of defaults
The old phrasing indicated that the EMAIL environment variable takes
precedence over the user.email configuration setting, but it is the
other way around.
Signed-off-by: Peter Eisentraut <redacted> Signed-off-by: Junio C Hamano <redacted>
John Keeping [Wed, 9 Jan 2013 19:45:29 +0000 (19:45 +0000)]
git-fast-import(1): reorganise options
The options in git-fast-import(1) are not currently arranged in a
logical order, which has caused the '--done' options to be documented
twice (commit 3266de10).
Rearrange them into logical groups under subheadings.
Suggested-by: Jonathan Nieder <redacted> Signed-off-by: John Keeping <redacted> Signed-off-by: Junio C Hamano <redacted>
John Keeping [Wed, 9 Jan 2013 19:44:38 +0000 (19:44 +0000)]
git-fast-import(1): combine documentation of --[no-]relative-marks
The descriptions of '--relative-marks' and '--no-relative-marks' make
more sense when read together instead of as two independent options.
Combine them into a single description block.
Signed-off-by: John Keeping <redacted> Reviewed-by: Jonathan Nieder <redacted> Signed-off-by: Junio C Hamano <redacted>
John Keeping [Wed, 9 Jan 2013 20:16:45 +0000 (20:16 +0000)]
git-shortlog(1): document behaviour of zero-width wrap
Commit 00d3947 (Teach --wrap to only indent without wrapping) added
special behaviour for a width of zero in the '-w' argument to
'git-shortlog' but this was not documented. Fix this.
Signed-off-by: John Keeping <redacted> Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Wed, 9 Jan 2013 16:27:09 +0000 (08:27 -0800)]
Merge branch 'nd/maint-branch-desc-doc'
Teach various forms of "format-patch" command line to identify what
branch the patches are taken from, so that the branch description
is picked up in more cases.
* nd/maint-branch-desc-doc:
format-patch: pick up branch description when no ref is specified
format-patch: pick up correct branch name from symbolic ref
t4014: a few more tests on cover letter using branch description
branch: delete branch description if it's empty
config.txt: a few lines about branch.<name>.description
"git merge" started calling prepare-commit-msg hook like "git
commit" does some time ago, but forgot to pay attention to the exit
status of the hook. t7505 may want a general clean-up but that is
a different topic.
Junio C Hamano [Wed, 9 Jan 2013 16:26:26 +0000 (08:26 -0800)]
Merge branch 'fc/remote-bzr'
New remote helper for bzr, with minimum fix squashed in.
* fc/remote-bzr:
remote-bzr: detect local repositories
remote-bzr: add support for older versions of bzr
remote-bzr: add support to push special modes
remote-bzr: add support for fecthing special modes
remote-bzr: add simple tests
remote-bzr: update working tree upon pushing
remote-bzr: add support for remote repositories
remote-bzr: add support for pushing
Add new remote-bzr transport helper
Junio C Hamano [Wed, 9 Jan 2013 16:26:20 +0000 (08:26 -0800)]
Merge branch 'jc/submittingpatches'
Streamline the document and update with a few e-mail addresses the
patches should be sent to.
* jc/submittingpatches:
SubmittingPatches: give list and maintainer addresses
SubmittingPatches: remove overlong checklist
SubmittingPatches: mention subsystems with dedicated repositories
SubmittingPatches: who am I and who cares?
The '--done' option to git-fast-import is documented twice in its manual
page. Combine the best bits of each description, keeping the location
of the instance that was added first.
Signed-off-by: John Keeping <redacted> Signed-off-by: Junio C Hamano <redacted>
René Scharfe [Tue, 8 Jan 2013 20:23:01 +0000 (21:23 +0100)]
t1402: work around shell quoting issue on NetBSD
The test fails for me on NetBSD 6.0.1 and reports:
ok 1 - ref name '' is invalid
ok 2 - ref name '/' is invalid
ok 3 - ref name '/' is invalid with options --allow-onelevel
ok 4 - ref name '/' is invalid with options --normalize
error: bug in the test script: not 2 or 3 parameters to test-expect-success
invalid_ref() constructs a test case description using its last argument,
but the shell seems to split it up into two pieces if it contains a
space. Minimal test case:
# on NetBSD with /bin/sh
$ a() { echo $#-$1-$2; }
$ t="x"; a "${t:+$t}"
1-x-
$ t="x y"; a "${t:+$t}"
2-x-y
$ t="x y"; a "${t:+x y}"
1-x y-
# and with bash
$ t="x y"; a "${t:+$t}"
1-x y-
$ t="x y"; a "${t:+x y}"
1-x y-
This may be a bug in the shell, but here's a simple workaround: Construct
the description string first and store it in a variable, and then use
that to call test_expect_success().
Signed-off-by: Rene Scharfe <redacted> Signed-off-by: Junio C Hamano <redacted>
upload-pack: only accept commits from "shallow" line
We only allow cuts at commits, not arbitrary objects. upload-pack will
fail eventually in register_shallow if a non-commit is given with a
generic error "Object %s is a %s, not a commit". Check it early and
give a more accurate error.
This should never show up in an ordinary session. It's for buggy
clients, or when the user manually edits .git/shallow.
Signed-off-by: Nguyễn Thái Ngọc Duy <redacted> Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 7 Jan 2013 20:24:55 +0000 (12:24 -0800)]
string-list: allow case-insensitive string list
Some string list needs to be searched case insensitively, and for
that to work correctly, the string needs to be sorted case
insensitively from the beginning.
Allow a custom comparison function to be defined on a string list
instance and use it throughout in place of strcmp().
Antoine Pelisse [Sun, 6 Jan 2013 22:09:39 +0000 (23:09 +0100)]
status: always report ignored tracked directories
When enumerating paths that are ignored, paths the index knows
about are not included in the result. The "index knows about"
check is done by consulting the name hash, not the actual
contents of the index:
- When core.ignorecase is false, directory names are not in the
name hash, and ignored ones are shown as ignored (directories
can never be tracked anyway).
- When core.ignorecase is true, however, the name hash keeps
track of the names of directories, in order to detect
additions of the paths under different cases. This causes
ignored directories to be mistakenly excluded when
enumerating ignored paths.
Stop excluding directories that are in the name hash when
looking for ignored files in dir_add_name(); the names that are
actually in the index are excluded much earlier in the callchain
in treat_file(), so this fix will not make them mistakenly
identified as ignored.
Signed-off-by: Antoine Pelisse <redacted> Reviewed-by: Jeff King <redacted> Signed-off-by: Junio C Hamano <redacted>
Ralf Thielow [Wed, 26 Dec 2012 15:46:52 +0000 (16:46 +0100)]
l10n: de.po: address the user formally
In the current German translation, the user was
addressed informally ("Du", "Dein") which is unusual
in German software. This commit changes the addressing
to be formal ("Sie", "Ihr").
Suggested-by: Christian Stimming <redacted> Suggested-by: Michael J Gruber <redacted> Signed-off-by: Ralf Thielow <redacted>
René Scharfe [Sun, 6 Jan 2013 17:59:39 +0000 (18:59 +0100)]
t5003: check if unzip supports symlinks
Only add a symlink to the repository if both the filesystem and
unzip support symlinks. To check the latter, add a ZIP file
containing a symlink, created like this with InfoZIP zip 3.0:
$ echo sample text >textfile
$ ln -s textfile symlink
$ zip -y infozip-symlinks.zip textfile symlink
If we can extract it successfully, we add a symlink to the test
repository for git archive --format=zip, or otherwise skip that
step. Users can see the skipped test and perhaps run it again
with a different unzip version.
Signed-off-by: Rene Scharfe <redacted> Signed-off-by: Junio C Hamano <redacted>
git-send-email: treat field names as case-insensitively
Field names like To:, Cc:, etc. are case-insensitive; use a
case-insensitive regexp to match them as such.
Previously, git-send-email would fail to pick-up the addresses when
in-body "fake" headers with different cases (e.g. lowercase "cc:")
are manually inserted to the messages it was asked to send, even
though the text will still show them.
Signed-off-by: Nickolai Zeldovich <redacted> Signed-off-by: Junio C Hamano <redacted>
René Scharfe [Sun, 6 Jan 2013 17:47:57 +0000 (18:47 +0100)]
t0024, t5000: clear variable UNZIP, use GIT_UNZIP instead
InfoZIP's unzip takes default parameters from the environment variable
UNZIP. Unset it in the test library and use GIT_UNZIP for specifying
alternate versions of the unzip command instead.
t0024 wasn't even using variable for the actual extraction. t5000
was, but when setting it to InfoZIP's unzip it would try to extract
from itself (because it treats the contents of $UNZIP as parameters),
which failed of course.
Signed-off-by: Rene Scharfe <redacted> Reviewed-by: Jonathan Nieder <redacted> Signed-off-by: Junio C Hamano <redacted>
Junio C Hamano [Mon, 7 Jan 2013 06:11:30 +0000 (22:11 -0800)]
Merge branch 'cc/no-gitk-build-dependency'
Remove leftover bits from an earlier change to move gitk in its own
subdirectory. Reimplementing the dependency tracking rules needs
to be done in gitk history separately.
* cc/no-gitk-build-dependency:
Makefile: replace "echo 1>..." with "echo >..."
Makefile: detect when PYTHON_PATH changes
Makefile: remove tracking of TCLTK_PATH
Junio C Hamano [Mon, 7 Jan 2013 06:11:16 +0000 (22:11 -0800)]
Merge branch 'jn/warn-on-inaccessible-loosen'
Deal with a situation where .config/git is a file and we notice
.config/git/config is not readable due to ENOTDIR, not ENOENT.
* jn/warn-on-inaccessible-loosen:
config: exit on error accessing any config file
doc: advertise GIT_CONFIG_NOSYSTEM
config: treat user and xdg config permission problems as errors
config, gitignore: failure to access with ENOTDIR is ok
Junio C Hamano [Mon, 7 Jan 2013 06:10:23 +0000 (22:10 -0800)]
Merge branch 'jc/apply-trailing-blank-removal'
Fix to update_pre_post_images() that did not take into account the
possibility that whitespace fix could shrink the preimage and
change the number of lines in it.
* jc/apply-trailing-blank-removal:
apply.c:update_pre_post_images(): the preimage can be truncated
Adam Spiers [Sun, 6 Jan 2013 16:58:10 +0000 (16:58 +0000)]
add.c: extract check_path_for_gitlink() from treat_gitlinks() for reuse
Extract the body of the for loop in treat_gitlinks() into a separate
check_path_for_gitlink() function so that it can be reused elsewhere.
This paves the way for a new check-ignore sub-command.
Also document treat_gitlinks().
Signed-off-by: Adam Spiers <redacted> Signed-off-by: Junio C Hamano <redacted>
Adam Spiers [Sun, 6 Jan 2013 16:58:09 +0000 (16:58 +0000)]
pathspec.c: rename newly public functions for clarity
Perform the following function renames to make it explicit that these
pathspec handling functions are for matching against the index, rather
than against a tree or the working directory.
Adam Spiers [Sun, 6 Jan 2013 16:58:05 +0000 (16:58 +0000)]
dir.c: provide clear_directory() for reclaiming dir_struct memory
By the end of a directory traversal, a dir_struct instance will
typically contains pointers to various data structures on the heap.
clear_directory() provides a convenient way to reclaim that memory.
Signed-off-by: Adam Spiers <redacted> Signed-off-by: Junio C Hamano <redacted>
Adam Spiers [Sun, 6 Jan 2013 16:58:04 +0000 (16:58 +0000)]
dir.c: keep track of where patterns came from
For exclude patterns read in from files, the filename is stored in the
exclude list, and the originating line number is stored in the
individual exclude (counting starting at 1).
For exclude patterns provided on the command line, a string describing
the source of the patterns is stored in the exclude list, and the
sequence number assigned to each exclude pattern is negative, with
counting starting at -1. So for example the 2nd pattern provided via
--exclude would be numbered -2. This allows any future consumers of
that data to easily distinguish between exclude patterns from files
vs. from the CLI.
Signed-off-by: Adam Spiers <redacted> Signed-off-by: Junio C Hamano <redacted>
Adam Spiers [Sun, 6 Jan 2013 16:58:03 +0000 (16:58 +0000)]
dir.c: use a single struct exclude_list per source of excludes
Previously each exclude_list could potentially contain patterns
from multiple sources. For example dir->exclude_list[EXC_FILE]
would typically contain patterns from .git/info/exclude and
core.excludesfile, and dir->exclude_list[EXC_DIRS] could contain
patterns from multiple per-directory .gitignore files during
directory traversal (i.e. when dir->exclude_stack was more than
one item deep).
We split these composite exclude_lists up into three groups of
exclude_lists (EXC_CMDL / EXC_DIRS / EXC_FILE as before), so that each
exclude_list now contains patterns from a single source. This will
allow us to cleanly track the origin of each pattern simply by adding
a src field to struct exclude_list, rather than to struct exclude,
which would make memory management of the source string tricky in the
EXC_DIRS case where its contents are dynamically generated.
Similarly, by moving the filebuf member from struct exclude_stack to
struct exclude_list, it allows us to track and subsequently free
memory buffers allocated during the parsing of all exclude files,
rather than only tracking buffers allocated for files in the EXC_DIRS
group.
Signed-off-by: Adam Spiers <redacted> Signed-off-by: Junio C Hamano <redacted>
Mark Levedahl [Sun, 6 Jan 2013 19:56:33 +0000 (11:56 -0800)]
Makefile: add comment on CYGWIN_V15_WIN32API
There is no documented, reliable, and future-proof method to
determine the installed w32api version on Cygwin. There are many
things that can be done that will work frequently, except when they
won't.
The only sane thing is to follow the guidance of the Cygwin
developers: the only supported configuration is that which the
current setup.exe produces, and in the case of problems, if the
installation is not up to date then updating is the first required
action.
Signed-off-by: Mark Levedahl <redacted> Signed-off-by: Junio C Hamano <redacted>
The documentation for the ALLOC_GROW API implicitly encouraged
developers to use "ary" as the variable name for the array which is
dynamically grown. However "ary" is an unusual abbreviation hardly
used anywhere else in the source tree, and it is also better to name
variables based on their contents not on their type.
Signed-off-by: Adam Spiers <redacted> Signed-off-by: Junio C Hamano <redacted>
René Scharfe [Sun, 6 Jan 2013 15:20:57 +0000 (16:20 +0100)]
archive-zip: write uncompressed size into header even with streaming
We record the uncompressed and compressed sizes and the CRC of streamed
files as zero in the local header of the file. The actual values are
recorded in an extra data descriptor after the file content, and in the
usual ZIP directory entry at the end of the archive.
While we know the compressed size and the CRC only after we processed
the contents, we actually know the uncompressed size right from the
start. And for files that we store uncompressed we also already know
their final size.
Do it like InfoZIP's zip and recored the known values, even though they
can be reconstructed using the ZIP directory and the data descriptors
alone. InfoZIP's unzip worked fine before, but NetBSD's version
actually depends on these fields.
The uncompressed size is already set by sha1_object_info(). We just
need to initialize the compressed size to zero or the uncompressed size
depending on the compression method (0 means storing). The CRC was
propertly initialized already.
Signed-off-by: Rene Scharfe <redacted> Signed-off-by: Junio C Hamano <redacted>
Johannes Sixt [Sun, 6 Jan 2013 14:07:43 +0000 (15:07 +0100)]
t6130-pathspec-noglob: Windows does not allow a file named "f*"
Windows disallows file names that contain a star. Arrange the test setup
to insert the file name "f*" in the repository without the corresponding
file in the worktree.
Signed-off-by: Johannes Sixt <redacted> Acked-by: Jeff King <redacted> Signed-off-by: Junio C Hamano <redacted>
git.99rst.org / git.git / log