aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2016-06-20Merge branch 'master' into fix-non-ascii-chars-in-status-linesSebastian Thiel
2016-06-20Merge pull request #475 from warsaw/issue470Sebastian Thiel
Fix issue #470
2016-06-15Fix issue #470Barry Warsaw
2016-06-14Fix for parsing non-ASCII chars in status linesVincent Driessen
2016-06-14Python 2.6 compatVincent Driessen
2016-06-14doc(changes): inform about new APISebastian Thiel
Relates to #446
2016-06-14fix(test_docs): skip master-dependent assertionSebastian Thiel
It usually fails on branches, which doesn't help assessing PRs.
2016-06-14Merge pull request #446 from guyzmo/updating_remote_url_handlingSebastian Thiel
Adding support for git remote set-url/get-url API to Remote
2016-06-14Merge pull request #467 from ↵Sebastian Thiel
gitpython-developers/fix-dont-choke-on-invalid-unicode-paths Don't choke on (legitimately) invalidly encoded Unicode paths
2016-06-14fix(flake): misc whitespace fixesSebastian Thiel
2016-06-13fix(cmd): allow any kind of status messageSebastian Thiel
I see no need in verifying the status code. It's enough to just get the error.
2016-06-13doc(README): add basic contribution guidelinesSebastian Thiel
The main point is that from now on, tests are required to add new features. If the fix is minor enough, not having a test is probably alright. That distinction is not represented in the contribution guide as more tests are better - people should prefer to have a test whenever they contribute anything. My motivation to finally do this is the sad realization that I grow too unconfident about the quality of some contributions without having tests that proof they are valid. It's not enough to not break anything that exists, as the current test-suite is certainly not perfect either.
2016-06-13fix(misc): various cleanupSebastian Thiel
Just went through all changes and adjusted them to the best of my abilities. As there are no tests to claim otherwise, I believe this is correct enough. However, it becomes evident that it's no longer possible to just make changes without backing them with a respective test.
2016-06-13Merge branch 'pr-cmd-raise-with-stderr-on-error' of ↵Sebastian Thiel
https://github.com/barry-scott/GitPython into barry-scott-pr-cmd-raise-with-stderr-on-error
2016-06-13fix(repo): prevent error messages from being swallowedSebastian Thiel
This issue must have rosen from `to_progress_instance()` being inserted in a spot where `None` was a legit value. Fixes #462
2016-06-08Switching the `urls` property to use `git remote show` instead of `git ↵Guyzmo
remote get-url` `get-url` is a new API that is not widely available yet (introduced in git 2.7.0), and provokes failure on travis. Signed-off-by: Guyzmo <guyzmo+github@m0g.net>
2016-06-08Adding support for git remote set-url/get-url API to RemoteGuyzmo
Both commands enable handling of a little known feature of git, which is to support multiple URL for one remote. You can add multiple url using the `set_url` subcommand of `git remote`. As listing them is also handy, there's a nice method to do it, using `get_url`. * adding set_url method that maps to the git remote set-url command¶ * can be used to set an URL, or replace an URL with optional positional arg¶ * can be used to add, delete URL with kwargs (matching set-url options)¶ * adding add_url, delete_url methods that wraps around set_url for conveniency¶ * adding urls property that yields an iterator over the setup urls for a remote¶ * adding a test suite that checks all use case scenarii of this added API.¶ Signed-off-by: Guyzmo <guyzmo+github@m0g.net>
2016-06-06Don't choke on (legitimately) invalidly encoded Unicode pathsVincent Driessen
2016-06-06Can get a str object from stream.read rather then bytes.Barry Scott
Convert to the expected bytes.
2016-06-06log all the output from stdout and stderr for debugging process failuresBarry Scott
2016-06-06Merge remote-tracking branch 'upstream/master' into ↵Barry Scott
pr-cmd-raise-with-stderr-on-error
2016-06-02Merge pull request #460 from ddanier/autointerrupt_del_no_os_killSebastian Thiel
Make sure os is not even partly destroyed
2016-06-02fix(test): remove unused importSebastian Thiel
2016-06-01Make sure os is not even partly destroyedDavid Danier
2016-06-01Merge pull request #459 from andy-maier/fix-457-typerror-on-py26Vincent Driessen
Fixes #457: 'TypeError: decode() takes no keyword arguments' on Python 2.6
2016-06-01Fixed 'TypeError: decode() takes no keyword arguments' on Python 2.6Andreas Maier
2016-06-01fix(test): do not skip test on travisSebastian Thiel
Please exclude the particular assertion instead. Related to https://github.com/gitpython-developers/GitPython/commit/a3f24f64a20d1e09917288f67fd21969f4444acd#commitcomment-17691581
2016-06-01chore(compat): state py2.6 support officiallySebastian Thiel
More information in the respective issue. Fixes #453
2016-05-30Bump for new versionVincent Driessen
2016-05-30This is 2.0.52.0.5Vincent Driessen
2016-05-30Fix regexVincent Driessen
This catches the case where the matched line contains "(" or ")" characters.
2016-05-30This is 2.0.42.0.4Vincent Driessen
2016-05-30Update changelogVincent Driessen
2016-05-30Merge pull request #456 from ↵Sebastian Thiel
gitpython-developers/fix-for-invalid-data-in-commits Add test case as example of Git commit with invalid data
2016-05-30Merge pull request #454 from ↵Sebastian Thiel
gitpython-developers/fix-octal-escaped-path-parser-bug Fix bug in diff parser output
2016-05-30Make sure that stderr is converted to bytesBarry Scott
remove stderr for a wait() that is not the GitPython wrapper.
2016-05-30Ignore invalid data when decoding commit objectsVincent Driessen
Previously, GitPython chokes on this while decoding. Rather than choking, instead accept the error and replace the invalid bytes by the � (\x80) char.
2016-05-30Merge pull request #455 from gitpython-developers/fix-ci-testsVincent Driessen
Skip test that always fails on Travis CI
2016-05-30Add test case as example of Git commit with invalid dataVincent Driessen
This is a real commit from the microjs.com open source project, see https://github.com/madrobby/microjs.com/commit/7e8457c17850d0991763941213dcb403d80f39f8, which is declared to be encoded in UTF-8, but contains invalid bytes. This makes GitPython choke on it while decoding. Rather than choking, this should instead accept the error and replace the invalid bytes by the � (\x80) char.
2016-05-30Fix flake8 complaintsBarry Scott
2016-05-30Skip test that always fails on Travis CIfix-ci-testsVincent Driessen
2016-05-30Fix bug in diff parser outputVincent Driessen
The diff --patch parser was missing some edge case where Git would encode non-ASCII chars in path names as octals, but these weren't decoded properly. \360\237\222\251.txt Decoded via utf-8, that will return: 💩.txt
2016-05-29Return stderr lines from a pull() call that failsBarry Scott
2016-05-29Return all the stderr messge after an error is detected for pull()Barry Scott
2016-05-29chore(remote): better super-class call syntaxSebastian Thiel
Python :) !! Related to #451
2016-05-29chore(flake8): whitespace ...Sebastian Thiel
Related to #451 Signed-off-by: Sebastian Thiel <byronimo@gmail.com>
2016-05-29Merge pull request #451 from barry-scott/pr-fix-callable-remote-progressSebastian Thiel
Fix traceback because _seen_ops is not initialised
2016-05-29Fix traceback because _seen_ops is not initialisedBarry Scott
must call the base class __init__
2016-05-29fix(remote): improve version checkSebastian Thiel
Make version check much more readable, and fix it at the same time. The previous implementation would assume progress is supported just by looking at the patch-level for instance. A quick check of the git sources seems to indicate the --progress flag exists in v1.7 of the git command-line already. Fixes #449
2016-05-29fix(remote): Add CallableRemoteProgressSebastian Thiel
That way, the base type doesn't need any adjustment. Related to #450