]> git.proxmox.com Git - libgit2.git/log
libgit2.git
11 years agoFix compilation on OpenBSD
Carlos Martín Nieto [Sun, 20 Jan 2013 00:57:32 +0000 (01:57 +0100)]
Fix compilation on OpenBSD

11 years agoMerge pull request #1260 from phkelley/development
Philip Kelley [Fri, 18 Jan 2013 20:05:10 +0000 (12:05 -0800)]
Merge pull request #1260 from phkelley/development

Fix really bad error handling in git_smart__negotiate_fetch

11 years agoFix really bad error handling in git_smart__negotiate_fetch
Philip Kelley [Fri, 18 Jan 2013 19:51:46 +0000 (14:51 -0500)]
Fix really bad error handling in git_smart__negotiate_fetch

11 years agoMerge pull request #1258 from zcbenz/fix_index_remove_test
Vicent Martí [Fri, 18 Jan 2013 06:09:57 +0000 (22:09 -0800)]
Merge pull request #1258 from zcbenz/fix_index_remove_test

Fix linking error caused by ddcb28a41f3774e26fc6ae0a7174a5565e4749ce.

11 years agoFix linking error caused by ddcb28a41f3774e26fc6ae0a7174a5565e4749ce.
Zhao Cheng [Fri, 18 Jan 2013 05:22:55 +0000 (13:22 +0800)]
Fix linking error caused by ddcb28a41f3774e26fc6ae0a7174a5565e4749ce.

11 years agoMerge pull request #1239 from ethomson/index_remove
Vicent Martí [Fri, 18 Jan 2013 00:56:57 +0000 (16:56 -0800)]
Merge pull request #1239 from ethomson/index_remove

add an index_remove_bypath that removes conflicts

11 years agoMerge pull request #1256 from arrbee/asciify-test-data
Vicent Martí [Fri, 18 Jan 2013 00:33:40 +0000 (16:33 -0800)]
Merge pull request #1256 from arrbee/asciify-test-data

Move all non-ascii test data to raw hex

11 years agoMerge pull request #1257 from ethomson/crlf_test_fix
Vicent Martí [Fri, 18 Jan 2013 00:32:46 +0000 (16:32 -0800)]
Merge pull request #1257 from ethomson/crlf_test_fix

add a git config, don't run crlf tests on non-win32

11 years agoadd a git config, don't run crlf tests on non-win32
Edward Thomson [Fri, 18 Jan 2013 00:18:44 +0000 (18:18 -0600)]
add a git config, don't run crlf tests on non-win32

11 years agoMove all non-ascii test data to raw hex
Russell Belfer [Thu, 17 Jan 2013 23:47:10 +0000 (15:47 -0800)]
Move all non-ascii test data to raw hex

This takes all of the characters in core::env and makes them use
hex sequences instead of keeping tricky character data inline in
the test.

11 years agoMerge pull request #1244 from carlosmn/pack-evict
Vicent Martí [Thu, 17 Jan 2013 21:45:52 +0000 (13:45 -0800)]
Merge pull request #1244 from carlosmn/pack-evict

A comparison of eviction algorithms for the delta base cache

11 years agoMerge pull request #1247 from sba1/dont-segfault-if-transport-doesnt-support-push
Ben Straub [Thu, 17 Jan 2013 21:37:32 +0000 (13:37 -0800)]
Merge pull request #1247 from sba1/dont-segfault-if-transport-doesnt-support-push

Don't segfault if transport doesn't support push.

11 years agoMerge pull request #1255 from arrbee/fix-signed-commit-header-parsing
Vicent Martí [Thu, 17 Jan 2013 21:36:33 +0000 (13:36 -0800)]
Merge pull request #1255 from arrbee/fix-signed-commit-header-parsing

Add skipping of unknown commit headers

11 years agoMerge pull request #1254 from ethomson/index_filtered_size
Vicent Martí [Thu, 17 Jan 2013 21:34:20 +0000 (13:34 -0800)]
Merge pull request #1254 from ethomson/index_filtered_size

cache should contain on-disk (filtered) file size

11 years agoTest that pushs properly fail for transports that don't provide a push implementation.
Sebastian Bauer [Thu, 17 Jan 2013 21:27:04 +0000 (22:27 +0100)]
Test that pushs properly fail for transports that don't provide a push implementation.

11 years agoAdd skipping of unknown commit headers
Russell Belfer [Thu, 17 Jan 2013 21:19:09 +0000 (13:19 -0800)]
Add skipping of unknown commit headers

This moves the check for the "encoding" header into a loop which
is just scanning for non-required headers at the end of a commit
header.  That loop will skip unrecognized lines (including header
continuation lines) until a terminating completely blank line is
found, and only then does it move to reading the commit message.

11 years agocache should contain on-disk (filtered) file size
Edward Thomson [Thu, 17 Jan 2013 19:11:57 +0000 (13:11 -0600)]
cache should contain on-disk (filtered) file size

11 years agoDon't segfault if transport doesn't support push.
Sebastian Bauer [Tue, 15 Jan 2013 20:39:37 +0000 (21:39 +0100)]
Don't segfault if transport doesn't support push.

Instead, set an more informative error message.

11 years agoMerge pull request #1211 from arrbee/fix-icase-status-file
Vicent Martí [Wed, 16 Jan 2013 23:52:58 +0000 (15:52 -0800)]
Merge pull request #1211 from arrbee/fix-icase-status-file

Fix case insensitivity issues in git_status_file

11 years agoMerge pull request #1248 from sba1/doc-fix
Vicent Martí [Wed, 16 Jan 2013 22:20:41 +0000 (14:20 -0800)]
Merge pull request #1248 from sba1/doc-fix

Some doc improvements in transport.h

11 years agoMerge pull request #1245 from nulltoken/topic/git_branch_tracking_name
Vicent Martí [Wed, 16 Jan 2013 22:17:48 +0000 (14:17 -0800)]
Merge pull request #1245 from nulltoken/topic/git_branch_tracking_name

Introduce git_branch_tracking_name()

11 years agobranch: Introduce git_branch_tracking_name()
nulltoken [Mon, 14 Jan 2013 13:22:11 +0000 (14:22 +0100)]
branch: Introduce git_branch_tracking_name()

11 years agoFix indentations
nulltoken [Mon, 14 Jan 2013 11:09:10 +0000 (12:09 +0100)]
Fix indentations

11 years agoSome doc improvements
Sebastian Bauer [Wed, 16 Jan 2013 11:25:28 +0000 (12:25 +0100)]
Some doc improvements

11 years agoShortcut spool and sort for empty iterator
Russell Belfer [Thu, 10 Jan 2013 19:11:42 +0000 (11:11 -0800)]
Shortcut spool and sort for empty iterator

11 years agoSupport case insensitive tree iterators and status
Russell Belfer [Thu, 10 Jan 2013 00:07:54 +0000 (16:07 -0800)]
Support case insensitive tree iterators and status

This makes tree iterators directly support case insensitivity by
using a secondary index that can be sorted by icase.  Also, this
fixes the ambiguity check in the git_status_file API to also be
case insensitive.  Lastly, this adds new test cases for case
insensitive range boundary checking for all types of iterators.

With this change, it should be possible to deprecate the spool
and sort iterator, but I haven't done that yet.

11 years agoAdd git_tree_entry_cmp and git_tree_entry_icmp
Russell Belfer [Thu, 10 Jan 2013 00:03:35 +0000 (16:03 -0800)]
Add git_tree_entry_cmp and git_tree_entry_icmp

This adds a new external API git_tree_entry_cmp and a new internal
API git_tree_entry_icmp for sorting tree entries.  The case
insensitive one is internal only because general users should
never be seeing case-insensitively sorted trees.

11 years agoAdd git_path_icmp to case-insensitive path cmp
Russell Belfer [Thu, 10 Jan 2013 00:02:42 +0000 (16:02 -0800)]
Add git_path_icmp to case-insensitive path cmp

This adds git_path_icmp to complement git_path_cmp.

11 years agoAdd payload "_r" versions of bsearch and tsort
Russell Belfer [Thu, 10 Jan 2013 00:00:16 +0000 (16:00 -0800)]
Add payload "_r" versions of bsearch and tsort

git__bsearch and git__tsort did not pass a payload through to the
comparison function.  This makes it impossible to implement sorted
lists where the sort order depends on external data (e.g. building
a secondary sort order for the entries in a tree).  This commit
adds git__bsearch_r and git__tsort_r versions that pass a third
parameter to the cmp function of a user payload.

11 years agoTest for ignore_case ranges on workdir iterator
Russell Belfer [Tue, 8 Jan 2013 23:56:11 +0000 (15:56 -0800)]
Test for ignore_case ranges on workdir iterator

This adds a test that confirms that the working directory iterator
can actually correctly process ranges of files case insensitively
with proper sorting and proper boundaries.

11 years agoUpdate iterator API with flags for ignore_case
Russell Belfer [Tue, 8 Jan 2013 23:53:13 +0000 (15:53 -0800)]
Update iterator API with flags for ignore_case

This changes the iterator API so that flags can be passed in to
the constructor functions to control the ignore_case behavior.
At this point, the flags are not supported on tree iterators (i.e.
there is no functional change over the old API), but the API
changes are all made to accomodate this.

By the way, I went with a flags parameter because in the future
I have a couple of other ideas for iterator flags that will make
it easier to fix some diff/status/checkout bugs.

11 years agoFix err msg for ambiguous path in git_status_file
Russell Belfer [Tue, 8 Jan 2013 21:45:06 +0000 (13:45 -0800)]
Fix err msg for ambiguous path in git_status_file

Returning GIT_EAMBIGUOUS from inside the status callback gets
overridden with GIT_EUSER.  `git_status_file` accounted for this
via the callback payload, but was allowing the error message to
be cleared.  Move the `giterr_set` call outside the callback to
where the EUSER case was being dealt with.

11 years agoMinor iterator API cleanups
Russell Belfer [Tue, 8 Jan 2013 21:39:15 +0000 (13:39 -0800)]
Minor iterator API cleanups

In preparation for further iterator changes, this cleans up a few
small things in the iterator API:

* removed the git_iterator_for_repo_index_range API
* made git_iterator_free not be inlined
* minor param name and test function name tweaks

11 years agoSimplify git_diff__paired_foreach icase handling
Russell Belfer [Tue, 8 Jan 2013 21:27:25 +0000 (13:27 -0800)]
Simplify git_diff__paired_foreach icase handling

11 years agostatus: Enhance git_status_file() test coverage
nulltoken [Sat, 5 Jan 2013 19:58:25 +0000 (20:58 +0100)]
status: Enhance git_status_file() test coverage

11 years agoMerge pull request #1238 from nulltoken/fix/checkout-index
Russell Belfer [Tue, 15 Jan 2013 17:46:50 +0000 (09:46 -0800)]
Merge pull request #1238 from nulltoken/fix/checkout-index

checkout: Teach checkout to cope with orphaned Head

11 years agoMerge pull request #1242 from sba1/init-with-template-fix
Russell Belfer [Mon, 14 Jan 2013 18:05:35 +0000 (10:05 -0800)]
Merge pull request #1242 from sba1/init-with-template-fix

Some fixes for external template support

11 years agopack: evict all of the pages at once
Carlos Martín Nieto [Mon, 14 Jan 2013 16:42:12 +0000 (17:42 +0100)]
pack: evict all of the pages at once

Somewhat surprisingly, this can increase the speed considerably, as we
don't bother trying to decide what to evict, and the most used entries
are quickly back into the cache.

11 years agopack: evict objects from the cache in groups of eight
Carlos Martín Nieto [Mon, 14 Jan 2013 15:39:22 +0000 (16:39 +0100)]
pack: evict objects from the cache in groups of eight

This drops the cache eviction below libcrypto and zlib in the perf
output. The number has been chosen empirically.

11 years agoAltered the description of the template.
Sebastian Bauer [Sun, 13 Jan 2013 11:21:52 +0000 (12:21 +0100)]
Altered the description of the template.

Before, it was identical to the default template making it difficult
to check, if the proper template was copied an external template
test.

11 years agoNow checks in the template test whether the description file has
Sebastian Bauer [Sun, 13 Jan 2013 11:23:30 +0000 (12:23 +0100)]
Now checks in the template test whether the description file has
been properly copied.

This is a minimal effort to test whether the template really has
been used when creating an repo with external templates.

11 years agoDo not use GIT_CPDIR_CHMOD flag when copying the template.
Sebastian Bauer [Sun, 13 Jan 2013 10:30:05 +0000 (11:30 +0100)]
Do not use GIT_CPDIR_CHMOD flag when copying the template.

This is an intermin solution. While this essentially disables the
--shared flag feature, previously external templates did not work
at all. This change fixes the previously corrected, and since
then failing, repo_init__extended_with_template() test.

The problem is now documented in the source code comments.

11 years agoFix Travis compilation warnings
nulltoken [Sat, 12 Jan 2013 18:37:14 +0000 (19:37 +0100)]
Fix Travis compilation warnings

11 years agocheckout: Teach checkout to cope with orphaned Head
nulltoken [Sat, 12 Jan 2013 18:27:31 +0000 (19:27 +0100)]
checkout: Teach checkout to cope with orphaned Head

Fix #1236

11 years agoadd an index_remove_bypath that removes conflicts, renamed add_from_workdir to match
Edward Thomson [Sat, 12 Jan 2013 19:47:56 +0000 (13:47 -0600)]
add an index_remove_bypath that removes conflicts, renamed add_from_workdir to match

11 years agoClear local error variable after invoking giterr_clear().
Sebastian Bauer [Sat, 12 Jan 2013 19:02:00 +0000 (20:02 +0100)]
Clear local error variable after invoking giterr_clear().

11 years agopack: fixes to the cache
Carlos Martín Nieto [Sat, 12 Jan 2013 18:31:07 +0000 (19:31 +0100)]
pack: fixes to the cache

The offset should be git_off_t, and we should check the return value
of the mutex lock function.

11 years agotests: plug leaks
Carlos Martín Nieto [Sat, 12 Jan 2013 18:01:45 +0000 (19:01 +0100)]
tests: plug leaks

11 years agoindexer: properly free the packfile resources
Carlos Martín Nieto [Sat, 12 Jan 2013 17:38:19 +0000 (18:38 +0100)]
indexer: properly free the packfile resources

The indexer needs to call the packfile's free function so it takes care of
freeing the caches.

We still need to close the mwf descriptor manually so we can rename the
packfile into its final name on Windows.

11 years agoAdded flag GIT_REPOSITORY_INIT_EXTERNAL_TEMPLATE to test_repo_init__extended_with_tem...
Sebastian Bauer [Sat, 12 Jan 2013 17:44:50 +0000 (18:44 +0100)]
Added flag GIT_REPOSITORY_INIT_EXTERNAL_TEMPLATE to test_repo_init__extended_with_template().

Otherwise the template functionallity is not tested (as a TODO we
also shall test that the specified template really got copied).

11 years agoMerge pull request #1222 from scunz/clone_branch
Vicent Martí [Sat, 12 Jan 2013 10:14:14 +0000 (02:14 -0800)]
Merge pull request #1222 from scunz/clone_branch

Switch to specified branch during clone

11 years agoDefault git_clone_options' checkout strategy to GIT_CHECKOUT_SAFE_CREATE
Sascha Cunz [Sat, 12 Jan 2013 04:51:00 +0000 (05:51 +0100)]
Default git_clone_options' checkout strategy to GIT_CHECKOUT_SAFE_CREATE

11 years agobuf: Is this the function you were looking for?
Vicent Marti [Sat, 12 Jan 2013 00:40:35 +0000 (01:40 +0100)]
buf: Is this the function you were looking for?

11 years agoKill vestigal dangling-remote code
Ben Straub [Sat, 12 Jan 2013 04:23:46 +0000 (20:23 -0800)]
Kill vestigal dangling-remote code

Fixes #1232

11 years agoMerge pull request #1215 from phkelley/binaryunicode
Philip Kelley [Sat, 12 Jan 2013 01:16:55 +0000 (17:16 -0800)]
Merge pull request #1215 from phkelley/binaryunicode

Add a failing test for CRLF filters

11 years agoFix up binaryunicode test repo
Philip Kelley [Sat, 12 Jan 2013 00:00:19 +0000 (19:00 -0500)]
Fix up binaryunicode test repo

11 years agoTests: Add test for check out of given branch during clone
Sascha Cunz [Fri, 11 Jan 2013 23:08:48 +0000 (00:08 +0100)]
Tests: Add test for check out of given branch during clone

11 years agoTest: Cleanup some cleaning code
Sascha Cunz [Fri, 11 Jan 2013 23:07:44 +0000 (00:07 +0100)]
Test: Cleanup some cleaning code

11 years agoCheck for binary blobs in checkout
Russell Belfer [Thu, 10 Jan 2013 23:59:36 +0000 (15:59 -0800)]
Check for binary blobs in checkout

This adds a git_buf_text_is_binary check to blobs before applying
filters when the blob data is being written to disk.

11 years agoMerge pull request #1230 from arrbee/match-core-git-diff-binary-detection
Vicent Martí [Fri, 11 Jan 2013 19:35:09 +0000 (11:35 -0800)]
Merge pull request #1230 from arrbee/match-core-git-diff-binary-detection

Match binary file check of core git in diff

11 years agoMerge pull request #1229 from arrbee/fix-diff-patch-line-numbers
Vicent Martí [Fri, 11 Jan 2013 19:34:13 +0000 (11:34 -0800)]
Merge pull request #1229 from arrbee/fix-diff-patch-line-numbers

Fix diff patch line number calculation

11 years agoMatch binary file check of core git in diff
Russell Belfer [Fri, 11 Jan 2013 19:24:26 +0000 (11:24 -0800)]
Match binary file check of core git in diff

Core git just looks for NUL bytes in files when deciding about
is-binary inside diff (although it uses a better algorithm in
checkout, when deciding if CRLF conversion should be done).
Libgit2 was using the better algorithm in both places, but that
is causing some confusion.  For now, this makes diff just look
for NUL bytes to decide if a file is binary by content in diff.

11 years agoFix diff patch line number calculation
Russell Belfer [Fri, 11 Jan 2013 19:20:44 +0000 (11:20 -0800)]
Fix diff patch line number calculation

This was just wrong.  Added a test that verifying patch line
numbers even for hunks further into a file and then fixed the
algorithm.  I needed to add a little extra state into the patch
so that I could track old and new file numbers independently,
but it should be okay.

11 years agoRevert "pack: packfile_free -> git_packfile_free and use it in the indexers"
Carlos Martín Nieto [Fri, 11 Jan 2013 19:15:06 +0000 (20:15 +0100)]
Revert "pack: packfile_free -> git_packfile_free and use it in the indexers"

This reverts commit f289f886cb81bb570bed747053d5ebf8aba6bef7, which
makes the tests fail on Windows. Revert until we can figure out a
solution.

11 years agoMerge pull request #1227 from nulltoken/topic/emergeconflict
Vicent Martí [Fri, 11 Jan 2013 18:59:31 +0000 (10:59 -0800)]
Merge pull request #1227 from nulltoken/topic/emergeconflict

Introduce EMERGECONFLICT

11 years agoFix MSVC Clar compilation warnings
nulltoken [Fri, 11 Jan 2013 14:39:14 +0000 (15:39 +0100)]
Fix MSVC Clar compilation warnings

11 years agoFix MSVC compilation warnings
nulltoken [Fri, 11 Jan 2013 13:40:09 +0000 (14:40 +0100)]
Fix MSVC compilation warnings

11 years agocheckout: Deploy EMERGECONFLICT usage
nulltoken [Thu, 10 Jan 2013 22:27:13 +0000 (23:27 +0100)]
checkout: Deploy EMERGECONFLICT usage

11 years agoerrors: Introduce EMERGECONFLICT error code
nulltoken [Thu, 10 Jan 2013 11:11:49 +0000 (12:11 +0100)]
errors: Introduce EMERGECONFLICT error code

11 years agoFix indentations
nulltoken [Thu, 10 Jan 2013 11:32:20 +0000 (12:32 +0100)]
Fix indentations

11 years agopack: That declaration
Vicent Marti [Fri, 11 Jan 2013 17:21:09 +0000 (18:21 +0100)]
pack: That declaration

11 years agoregex: Proper define for this thing
Vicent Marti [Fri, 11 Jan 2013 17:19:52 +0000 (18:19 +0100)]
regex: Proper define for this thing

11 years agoMerge pull request #1228 from carlosmn/delta-base
Vicent Martí [Fri, 11 Jan 2013 17:12:21 +0000 (09:12 -0800)]
Merge pull request #1228 from carlosmn/delta-base

Introduce a delta base cache

11 years agoMerge pull request #1224 from sba1/some-regex-warning-fixes
Vicent Martí [Fri, 11 Jan 2013 16:52:31 +0000 (08:52 -0800)]
Merge pull request #1224 from sba1/some-regex-warning-fixes

Some regex warning fixes

11 years agopack: packfile_free -> git_packfile_free and use it in the indexers
Carlos Martín Nieto [Fri, 11 Jan 2013 16:24:52 +0000 (17:24 +0100)]
pack: packfile_free -> git_packfile_free and use it in the indexers

It turns out the indexers have been ignoring the pack's free function
and leaking data. Plug that.

11 years agopack: limit the amount of memory the base delta cache can use
Carlos Martín Nieto [Fri, 21 Dec 2012 12:46:48 +0000 (13:46 +0100)]
pack: limit the amount of memory the base delta cache can use

Currently limited to 16MB (like git) and to objects up to 1MB in
size.

11 years agopack: abstract out the cache into its own functions
Carlos Martín Nieto [Fri, 21 Dec 2012 09:59:10 +0000 (10:59 +0100)]
pack: abstract out the cache into its own functions

11 years agoMerge pull request #1226 from nulltoken/fix/refspec_free
Vicent Martí [Fri, 11 Jan 2013 16:26:26 +0000 (08:26 -0800)]
Merge pull request #1226 from nulltoken/fix/refspec_free

refspec: prevent git_refspec__free() from segfaulting

11 years agopack: refcount entries and add a mutex around cache access
Carlos Martín Nieto [Thu, 20 Dec 2012 06:55:51 +0000 (07:55 +0100)]
pack: refcount entries and add a mutex around cache access

11 years agopack: introduce a delta base cache
Carlos Martín Nieto [Wed, 19 Dec 2012 15:48:12 +0000 (16:48 +0100)]
pack: introduce a delta base cache

Many delta bases are re-used. Cache them to avoid inflating the same
data repeatedly.

This version doesn't limit the amount of entries to store, so it can
end up using a considerable amound of memory.

11 years agotests: plug a couple of leaks
Carlos Martín Nieto [Fri, 11 Jan 2013 15:54:57 +0000 (16:54 +0100)]
tests: plug a couple of leaks

11 years agorefspec: prevent git_refspec__free() from segfaulting
nulltoken [Fri, 11 Jan 2013 15:14:17 +0000 (16:14 +0100)]
refspec: prevent git_refspec__free() from segfaulting

Fix libgit2/libgit2sharp#247

11 years agoregex: Fixed warnings about unused parameter values.
Sebastian Bauer [Fri, 11 Jan 2013 09:47:44 +0000 (10:47 +0100)]
regex: Fixed warnings about unused parameter values.

There are different solutions to the problem. In this change, we
define an UNUSED macro that maps to __attribute__((unused)) when
compiling with gcc. Otherwise it is a NOOP. We apply this macro
in all function headers for each parameter value that is not used
within the function body.

The change is local to regex.

11 years agoregex: Fixed several warnings about signed/unsigned conversions.
Sebastian Bauer [Fri, 11 Jan 2013 09:25:51 +0000 (10:25 +0100)]
regex: Fixed several warnings about signed/unsigned conversions.

11 years agoAdd missing git_buf_free
Sascha Cunz [Fri, 11 Jan 2013 02:07:50 +0000 (03:07 +0100)]
Add missing git_buf_free

11 years agoImplement analog for 'git checkout --branch xxx ...'
Sascha Cunz [Fri, 11 Jan 2013 01:45:55 +0000 (02:45 +0100)]
Implement analog for 'git checkout --branch xxx ...'

11 years agoFix possible free'ing of unitialized pointer in error case
Sascha Cunz [Fri, 11 Jan 2013 01:18:27 +0000 (02:18 +0100)]
Fix possible free'ing of unitialized pointer in error case

11 years agoMerge pull request #1221 from arrbee/checkout-without-pathspec-match
Vicent Martí [Thu, 10 Jan 2013 23:18:49 +0000 (15:18 -0800)]
Merge pull request #1221 from arrbee/checkout-without-pathspec-match

Add GIT_CHECKOUT_DISABLE_PATHSPEC_MATCH

11 years agoAdd GIT_CHECKOUT_DISABLE_PATHSPEC_MATCH
Russell Belfer [Thu, 10 Jan 2013 23:15:37 +0000 (15:15 -0800)]
Add GIT_CHECKOUT_DISABLE_PATHSPEC_MATCH

This adds an option to checkout a la the diff option to turn off
fnmatch evaluation for pathspec entries.  This can be useful to
make sure your "pattern" in really interpretted as an exact file
match only.

11 years agoMerge pull request #1206 from ben/stock-auth
Vicent Martí [Thu, 10 Jan 2013 19:24:09 +0000 (11:24 -0800)]
Merge pull request #1206 from ben/stock-auth

Expose stock user/pass credential utility

11 years agoMerge pull request #1220 from ethomson/reuc_empty_sides
Vicent Martí [Thu, 10 Jan 2013 18:04:53 +0000 (10:04 -0800)]
Merge pull request #1220 from ethomson/reuc_empty_sides

REUC needs to handle empty sides

11 years agoREUC needs to handle empty sides
Edward Thomson [Thu, 10 Jan 2013 17:56:02 +0000 (11:56 -0600)]
REUC needs to handle empty sides

11 years agoMerge pull request #1182 from libgit2/odb-file-refresh
Vicent Martí [Thu, 10 Jan 2013 16:15:10 +0000 (08:15 -0800)]
Merge pull request #1182 from libgit2/odb-file-refresh

Sane refresh logic for #1180

11 years agoMerge pull request #1218 from sba1/amiga.2
Vicent Martí [Thu, 10 Jan 2013 16:13:02 +0000 (08:13 -0800)]
Merge pull request #1218 from sba1/amiga.2

Libgit2 for AmigaOS4

11 years agoodb: Refresh on `exists` query too
Vicent Marti [Thu, 10 Jan 2013 14:43:08 +0000 (15:43 +0100)]
odb: Refresh on `exists` query too

11 years agodat errorcode
Vicent Marti [Fri, 4 Jan 2013 16:42:41 +0000 (17:42 +0100)]
dat errorcode

11 years agoSane refresh logic
Vicent Marti [Thu, 3 Jan 2013 19:36:26 +0000 (20:36 +0100)]
Sane refresh logic

All the ODB backends have a specific refresh interface. When reading an
object, first we attempt every single backend: if the read fails, then
we refresh all the backends and retry the read one more time to see if
the object has appeared.

11 years agoMerge pull request #1219 from sba1/fetch-example-return-fix
Vicent Martí [Thu, 10 Jan 2013 14:01:00 +0000 (06:01 -0800)]
Merge pull request #1219 from sba1/fetch-example-return-fix

Don't call pthread_exit() in the callback of the fetch example

11 years agoDon't call pthread_exit() in the callback.
Sebastian Bauer [Thu, 10 Jan 2013 11:39:17 +0000 (12:39 +0100)]
Don't call pthread_exit() in the callback.

Compilers that are not aware that pthread_exit() does not return
issue a warning when compiling the present code. This change
exchanges the call to pthread_exit() with a simple return
statement. According to the pthread specification this is
equivalent.

11 years agoIgnore build-amiga
Sebastian Bauer [Fri, 4 Jan 2013 22:49:28 +0000 (23:49 +0100)]
Ignore build-amiga