|
aea8a638
|
2011-05-29T18:00:35
|
|
Implemented read_unique_short_oid method for loose backend.
|
|
6c8ca697
|
2011-05-29T17:57:25
|
|
Fixed some error messages related to searching objects from a short oid. Fixed forgot to check that prefix length is greater than minimum prefix length in read_unique_short_oid method from pack backend.
|
|
da03c9f3
|
2011-05-29T11:45:34
|
|
Changed return value of git_oid_match to be consistent with the other compare methods (0 means oids match). Added method to compare prefixes of hex formatted oids.
|
|
7d74cd44
|
2011-05-29T11:46:42
|
|
Deleted unused method git_cached_object_match, since we do not explore the cache when searching objects from a short oid.
|
|
ecd6fdf1
|
2011-05-27T18:49:09
|
|
Added a read_unique_short_oid method to backends, to make it possible to find objects from sha1 prefixes in the future. Default implementations throw GIT_ENOTIMPLEMENTED for strict prefixes (i.e. length < GIT_OID_HEXSZ).
|
|
ac2b94ad
|
2011-05-28T21:24:25
|
|
Added a GIT_OID_MINPREFIXLEN constant to define the minimum length allowed for oid prefixes (set to 4, like in git). Consequently updated some object lookup methods and their documentation.
|
|
dd453c4d
|
2011-05-27T22:46:41
|
|
Added git.git sha1 lookup method to replace simple binary search in pack backend.
Implemented find_unique_short_oid for pack backend, based on git sha1 lookup method;
finding an object given its full oid is just a particular case of searching
the unique object matching an oid prefix (short oid).
Added git_odb_read_unique_short_oid, which iterates over all the backends to
find and read the unique object matching the given oid prefix.
Added a git_object_lookup_short_oid method to find the unique object in
the repository matching a given oid prefix : it generalizes git_object_lookup
which now does nothing but calls git_object_lookup_short_oid.
|
|
53c0bd81
|
2011-05-27T22:37:10
|
|
Added error for ambiguous oid prefixes. Added methods to compare the first nth hexadecimal characters (i.e. packets of 4 bits) of OIDs.
|
|
4a51e998
|
2011-06-01T12:57:47
|
|
Merge pull request #224 from glesserd/tagparsing
Fix tag and signature parsing
|
|
dd8a2070
|
2011-06-01T12:54:15
|
|
Merge pull request #215 from schu/typos
Fix typos
|
|
a7fdce62
|
2011-06-01T12:53:16
|
|
Merge pull request #223 from carlosmn/valgrind
Plug a leak in the index unmerged entries vector
|
|
9db04160
|
2011-06-01T12:52:50
|
|
Merge pull request #222 from carlosmn/config-bugfix
Config bugfix
|
|
50b7334e
|
2011-06-01T09:58:21
|
|
Merge pull request #206 from nulltoken/topic/is-bare
Add git_repository_is_bare() accessor
|
|
a02fc2cd
|
2011-05-24T15:24:45
|
|
index: correctly parse invalidated TREE extensions
A TREE extension with an entry count of -1 means that it was
invalidated and we should ignore it. Do so instead of returning an
error.
This fixes issue #202
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
cdd9fd47
|
2011-05-24T14:55:34
|
|
Allow read_tree_internal to return an error code
There are two reasons why read_tree_internal might return a NULL
tree. The first one is a corrupt index, but the second one is an
invalidated TREE extension. Up to now, its only way to communicate
with its caller was through the return value being NULL or not.
Allow read_tree_internal to report its exit status independently from
the tree pointer.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
786ad84f
|
2011-06-01T18:51:54
|
|
index: Cleanup tree parsing
|
|
f7e59c4d
|
2011-06-01T18:34:21
|
|
index: Change the memory management for repo indexes
The `git_repository_index` call now returns a brand new index that must
be manually free'd.
|
|
fbfc7580
|
2011-05-31T18:38:59
|
|
Fix tag and signature parsing
Before this commit, malformed tag and signature were considered as
valid by the parser. See the test t3800-mktag.sh of git to see example
of malformed tag and signature.
|
|
c7e6e958
|
2011-05-31T17:44:55
|
|
Don't try to parse an empty config file
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
5892277c
|
2011-05-31T15:16:25
|
|
Config parse header ext: don't allow text after closing quote
Nothing is allowed betwen the closing quotation mark and the ] so
return an error if there is.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
7288d8b6
|
2011-05-31T15:11:49
|
|
Parse section header ext: don't leak on error
Also free the subsection if we find too many quotes
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
7bc9e2aa
|
2011-05-31T15:06:22
|
|
Guard against double-freeing the current section
If parse_section_header{,_ext} return an error, current_section
doesn't get allocated. Set it to NULL after freeing so we don't try to
free it again.
This fixes part 2-2 of Issue #210.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
f2abee47
|
2011-05-29T01:24:09
|
|
cfg_readline: really ignore empty lines
Simplify cfg_readline and at the same time fix it so that it does
really ignore empty lines.
This fixes point 2-1 of Issue #210
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
a5f43b95
|
2011-05-31T13:48:44
|
|
Config file open: don't free memory that doesn't belong to us
On error, it would free the configuration object even though it didn't
own that memory, which would cause a double-free.
This fixes the first part of Issue #210
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
71da57ae
|
2011-05-31T16:49:15
|
|
Plug a leak in the index unmerged entries vector
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
f5a86fef
|
2011-05-29T17:44:14
|
|
gitfo_isdir: fix error message
Signed-off-by: schu <schu-github@schulog.org>
|
|
fa9bcd81
|
2011-05-24T21:48:07
|
|
Add git_repository_is_bare() accessor
|
|
3a1c4310
|
2011-05-24T18:55:35
|
|
Fix compilation warnings in MSVC
This allows to successfully build libgit2 with waf on Windows.
|
|
8146fe7c
|
2011-05-23T21:41:13
|
|
index: Fix unused error messages
|
|
d320c52d
|
2011-05-19T15:59:18
|
|
index.c: Move to new error handling mechanism
|
|
ec9edd56
|
2011-05-19T15:46:36
|
|
config_file.c: Move to new error handling mechanism
|
|
a6d647d2
|
2011-05-19T15:37:10
|
|
win32/pthread.c: Move to new error handling mechanism
|
|
7c8b597e
|
2011-05-19T15:36:09
|
|
win32/map.c: Move to new error handling mechanism
|
|
450ac186
|
2011-05-19T15:29:22
|
|
unix/map.c: Move to new error handling mechanism
|
|
57435a6d
|
2011-05-19T15:24:24
|
|
backends/sqlite.c: Move to new error handling mechanism
|
|
43a0bc01
|
2011-05-19T15:17:06
|
|
backends/hiredis.c: Move to new error handling mechanism
|
|
33b1d19e
|
2011-05-19T15:13:40
|
|
util.c: Move to new error handling mechanism
|
|
bc06a4ee
|
2011-05-19T15:10:54
|
|
tree.c: Move to new error handling mechanism
|
|
6e34ee61
|
2011-05-19T15:07:41
|
|
tag.c: Move to new error handling mechanism
|
|
00556496
|
2011-05-19T14:59:10
|
|
signature.c: Move to new error handling mechanism
|
|
c0cd9d50
|
2011-05-19T14:57:41
|
|
revwalk.c: Move to new error handling mechanism
|
|
4f664a1b
|
2011-05-19T14:51:16
|
|
repository.c: Move to new error handling mechanism
|
|
0d5d5190
|
2011-05-23T21:37:48
|
|
refs: Reword errors
|
|
5bdf7b9f
|
2011-05-19T14:36:52
|
|
refs.c: Move to new error handling mechanism
|
|
bea54842
|
2011-05-19T13:50:46
|
|
oid.c: Move to new error handling mechanism
|
|
f84d9819
|
2011-05-23T21:14:58
|
|
odb_pack: Reword errors
|
|
267d539f
|
2011-05-19T13:38:12
|
|
odb_pack.c: Move to new error handling mechanism
|
|
60e1b49a
|
2011-05-23T21:12:18
|
|
odb_loose: Reword errors
|
|
dfb12cd5
|
2011-05-19T12:28:46
|
|
odb_loose.c: Move to new error handling mechanism
|
|
f93f8ec5
|
2011-05-19T12:27:43
|
|
odb_loose.c: Return GIT_ENOMEM when allocation fails
When trying to inflate a buffer, a GIT_ERROR was returned when malloc()
failed.
Fix this to return GIT_ENOMEM.
|
|
1e85d1aa
|
2011-05-23T21:09:07
|
|
odb: Reword errors
|
|
d3d5d86d
|
2011-05-18T12:35:08
|
|
odb.c: Move to new error handling mechanism
|
|
6623f5c9
|
2011-05-23T21:05:43
|
|
object: Reword errors
|
|
75eb97fe
|
2011-05-18T12:19:48
|
|
object.c: Move to new error handling mechanism
|
|
0be7f000
|
2011-05-23T21:04:09
|
|
hashtable: Reword errors
|
|
da533495
|
2011-05-18T12:12:16
|
|
hashtable.c: Move to new error handling mechanism
|
|
d3ca89fe
|
2011-05-23T21:03:01
|
|
fileops: Cleanup
|
|
cc2ac058
|
2011-05-17T18:02:08
|
|
fileops.c: Move to new error handling mechanism
|
|
f6328611
|
2011-05-23T21:00:19
|
|
filebuf: Reword errors
|
|
374db5f9
|
2011-05-17T17:43:00
|
|
filebuf.c: Move to new error handling mechanism
|
|
68384a27
|
2011-05-17T17:24:28
|
|
config.c: Fix format string error
|
|
4152a44b
|
2011-05-17T17:22:33
|
|
delta-apply.c: Move to new error handling mechanism
|
|
21dcb75a
|
2011-05-17T17:21:00
|
|
delta-apply.c: Use GIT_ENOMEM instead of GIT_ERROR when allocating
git__delta_apply used to return with a generic GIT_ERROR when allocating
memory for the delta failed.
Fix this to return GIT_ENOMEM.
|
|
5c36f6db
|
2011-05-23T20:43:19
|
|
config: Cleanup
|
|
4c7a5e9e
|
2011-05-23T20:36:06
|
|
commit: Reword errors
|
|
726c6fd2
|
2011-05-17T17:00:37
|
|
commit.c: Move to new error handling mechanism
|
|
adef9922
|
2011-05-23T20:33:20
|
|
blob: Reword errors
|
|
91369100
|
2011-05-17T16:50:49
|
|
blob.c: Move to new error handling mechanism
|
|
b0b527e0
|
2011-05-20T03:20:12
|
|
config: Cleanup & renaming of the external API
"git_config_backend" have been renamed to "git_config_file", which
implements a generic interface to access a configuration file -- be it
either on disk, from a DB or whatever mumbojumbo.
I think this makes more sense.
|
|
8adbf2ed
|
2011-05-20T02:58:33
|
|
Rewrite `git_config_open_global`
We have a lot of utility methods that make path building trivial. Use
them!
|
|
274f2c21
|
2011-05-19T14:18:57
|
|
Merge pull request #193 from carlosmn/config
A couple of config improvements
|
|
44d16d6f
|
2011-05-19T11:21:22
|
|
Merge pull request #198 from sschuberth/fix-typos
Fix a few minor typos in comments and error messages
|
|
f4e2aca2
|
2011-05-19T20:38:17
|
|
index: Fix issues in the unmerged entries API
|
|
8133afef
|
2011-05-19T09:26:01
|
|
Fix a few minor typos in comments and error messages
Regarding "initialize" vs. "initialise", www.dict.cc says the first is American
English whereas the latter in British English. For consistency, we should
stick to American English.
|
|
7b134cfe
|
2011-05-18T17:04:12
|
|
Merge pull request #141 from jpfender/unmerged
index: Add API for unmerged entries
|
|
7d6274ba
|
2011-05-18T17:00:20
|
|
Merge pull request #197 from schu/errors
Move odb.c to the new error handling
|
|
12de98c1
|
2011-05-18T18:00:34
|
|
Move odb.c to the new error handling
Add missing free in git_odb_new().
Signed-off-by: schu <schu-github@schulog.org>
|
|
765fdf4a
|
2011-05-18T17:19:38
|
|
Use "__inline" instead of "inline" with MSVC
MSVC supports "inline" only in C++ code, not in C code.
|
|
07234716
|
2011-05-18T16:46:56
|
|
Define str(n)casecmp for MSVC
MSVC has _str(n)icmp instead.
|
|
6421c49a
|
2011-05-17T18:55:52
|
|
Fix variable name normalization
When I changed it over to use different strings for the variable and
the name, cvar_name_normalize was left behind. Fix this and rename to
cvar_normalize_name to reflect the incompatible change.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
9d27fd3b
|
2011-05-17T16:51:37
|
|
index.c: Fix typo
git__rethrow was missing an underscore.
|
|
c90bfec7
|
2011-05-17T16:18:56
|
|
Move index.c to new error handling mechanism
|
|
050e8877
|
2011-05-17T15:31:05
|
|
Merge branch 'development' into unmerged
|
|
32234541
|
2011-05-17T14:18:42
|
|
Implement git_config_open_global
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
0da2c700
|
2011-05-17T15:09:30
|
|
utils: Move git__str[n]tolower
|
|
128d3731
|
2011-05-17T15:05:41
|
|
config_file: Fix compilation
|
|
6adcb5f3
|
2011-05-17T15:04:18
|
|
utils: Implement git__strndup
|
|
b3d94069
|
2011-05-17T14:51:42
|
|
repository: Properly free the index on close
|
|
29dca088
|
2011-05-17T13:38:19
|
|
Move config to the new error methods
Take this opportunity to fix an instance of returning
GIT_EOBJCORRUPTED when malloc failed.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
|
|
94711cad
|
2011-05-17T12:12:59
|
|
Merge upstream/development
|
|
335d6c99
|
2011-05-17T01:46:07
|
|
cache: Drop cuckoo hashing
Now we use a simple closed-addressing cache. Cuckoo hashing was creating
too many issues with race conditions. Fuck that.
Let's see what happens performance wise, we may have to roll back or
come up with another way to implement an efficient multi-threaded cache.
|
|
3de79280
|
2011-05-17T00:51:52
|
|
cache: Fix deadlock
Do not try to adquire the same node lock twice when the cuckoo hashing
resolves to the same node.
|
|
5ca2f580
|
2011-05-15T23:48:05
|
|
Do not set error message on `GIT_EREVWALKOVER`
This is not really an error, just a special return code to mark the end
of an iteration.
|
|
7cadd1f6
|
2011-05-15T23:46:22
|
|
Check error code from `git_cache_init`
|
|
4edf3e09
|
2011-05-15T23:45:24
|
|
Return success code on `git_cache_init`
|
|
3fe2e770
|
2011-05-15T13:34:43
|
|
Merge pull request #184 from nulltoken/repo-error-handling
Updated fileops.c and repository.c to new error handling mechanism
|
|
71747bca
|
2011-05-15T20:07:54
|
|
fix git_otype typo when calling `git_odb_read_header`.
|
|
81201a4c
|
2011-05-15T06:57:34
|
|
Move cache.c to the new error handling
|
|
3abe3bba
|
2011-05-14T16:05:33
|
|
Move repository.c to the new error handling
|
|
77c3999c
|
2011-05-14T14:40:56
|
|
Move fileops.c to the new error handling
|