summaryrefslogtreecommitdiff
path: root/lib
Commit message (Collapse)AuthorAgeFilesLines
...
* Corrected a problem with commits_between returning None.Michael Trier2009-01-241-2/+2
| | | | | This was caused by calling reverse on the list, which reverses the list in place and returns None. We really need tests for this.
* Fixed the gzip archive creation using python gzip moduleEric Jonas2009-01-241-1/+8
| | | | (cherry picked from commit 160bd2bb74d88c110874c21635e24cc74ef58459)
* fixed Commit.stats retrieval for parentless commits in bare reposMichael Trier2008-12-161-4/+4
| | | | (cherry picked from commit 88852ed7bcde4f4b18c1ae8b6fba7f3fab8e9bf5)
* Fixed a bug with branch names omitting path components.Michael Trier2008-12-161-1/+6
| | | | | | | | | Git allows branches to be named and organized using path components, e.g using a branch called "refactoring/feature1", which gets stored under refs/heads/refactoring/feature1. The previous code omitted everything but the last path component giving the name "feature1" instead of "refactoring/feature1" for the branch. This changeset fixes that. (cherry picked from commit dc4738bc53e580754e47037e26c7eec3047aeb69)
* Corrected conflict with pep8 changes from apollo13.Michael Trier2008-12-152-10/+10
|\
| * pep008!Florian Apolloner2008-12-152-9/+9
| |
* | Added in new properties Diff.renamed, Diff.rename_from, and Diff.rename_toMichael Trier2008-12-152-5/+10
| |
* | Make message the full message and add a summary propertyPaul Sowden2008-12-151-2/+6
| | | | | | | | | | Adds a summary property to the Commit object which returns just the first line of the commit message and makes the message property contain the entire commit message (previously the message property only contained the first line of the commit message). This breaks backwards compatibility a little in that the message property now contains a different value but previously there was no way to access the entire commit message from the Commit object and this is in keeping with git vocabulary, where message generally refers to the entire commit message. (cherry picked from commit 9d2962d8306c894d4cca55bab551677b92d96352)
* | add support for parsing rename info to the diff parserPaul Sowden2008-12-151-3/+7
| | | | | | | | | | Currently the parsed headers are just ignored but I'll add them to the Diff object in a future commit (cherry picked from commit 711b655f29b42821c51be8e592143c7db31ed140)
* | implement a far far faster diff parserPaul Sowden2008-12-151-42/+19
| | | | | | | | | | The old diff parser in list_from_string took a large amount of time to parse long diffs, on one of my repositories it took over 3 minutes to parse the initial commit. The new parser uses a single regexp to match the header of a diff, and iterates over the each individual diff by splitting the entire string by the diff seperator, attempting to match the header for each individual diff. With the new parser parsing the same repository is almost instant, woohoo! (cherry picked from commit 5b6b27f153bdc30380bea12a528ef483571dd57a)
* | add a path parameter to most commit methodsPaul Sowden2008-12-152-14/+35
| | | | | | | | | | The path parameter allows you to specify a path to constrain queries by. This changes potentially breaks backwards compatibility for the Repo.commits and Repo.commits_since methods as it moves the positional arguments. (cherry picked from commit cc8a20e78da4864060bd0c9279633009bc10d871)
* | remove ambiguity between paths and treeishsPaul Sowden2008-12-152-10/+9
|/ | | | | When calling commands that accept treeish and path arguments and there is a path with the same name as a treeish git cowardly refuses to pick one and asks for the command to use the unambiguous syntax where '--' seperates the treeish from the paths. Add '--' to the git commands to indicate that the argument is a treeish and not a path. (cherry picked from commit a9a57fa93a2b121ab9b17fcd6062b9a9c9740883)
* fixed git.Tree.__len__ argumentsk00012008-12-151-1/+1
| | | | (cherry picked from commit 657444d75d7b2d6913c2d491841b9bd70c8bd289)
* Replace GitPython with git in repr() outputs.Steve Frécinaux2008-09-177-41/+41
| | | | | The imported module is called git (as in "import git"), so it's less confusing to do so than to call everything GitPython.something.
* Merge branch 'master' of git://gitorious.org/git-python/dokais-cloneMichael Trier2008-09-141-0/+31
|\
| * Added a read-only Repo.active_branch property which returns the name of the ↵Kai Lautaportti2008-09-121-0/+14
| | | | | | | | currently active branch.
| * Cosmetic fixes to be consistent with the overall coding style.Kai Lautaportti2008-09-121-4/+5
| |
| * Added a read-only property Repo.is_dirty which reflects the status of theKai Lautaportti2008-09-121-0/+16
| | | | | | | | | | | | | | working directory. A working directory is dirty if it has any uncommitted changes (in the working directory or in the index). Bare repositories are by nature always clean.
* | Merge branch 'master' of git://gitorious.org/git-python/nudMichael Trier2008-09-144-96/+108
|\ \ | |/ |/|
| * Allow modifying the project descriptionSteve Frécinaux2008-09-071-14/+13
| | | | | | | | | | | | | | | | Do this: >>> repo.description = "Foo Bar" >>> repo.description 'Foo Bar'
| * Make daemon export a property of git.RepoSteve Frécinaux2008-09-071-19/+16
| | | | | | | | | | | | | | Now you can do this: >>> exported = repo.daemon_export >>> repo.daemon_export = True
| * Do not use **kwargs for constructors.Steve Frécinaux2008-09-063-33/+35
| | | | | | | | | | It is better to have an explicit list of variables for the constructors, be it only to avoid mispelled arguments.
| * Make Tree.content_from_string a static method.Steve Frécinaux2008-09-061-1/+2
| | | | | | | | It doesn't use an object's private contents, so let's go...
| * Implement dict protocol for trees.Steve Frécinaux2008-09-061-5/+31
| | | | | | | | | | It is rather intuitive to consider trees as a dict of objects (like a directory could be seen as a dict of files).
| * Initialize trees completely in tree.__bake__().Steve Frécinaux2008-09-052-21/+9
| | | | | | | | | | | | | | | | | | | | | | This is a simplification of the tree baking code. As a matter of consequency, Tree.construct() and tree.construct_initialize() have been killed, and repo.tree() has lost the "paths" argument. This is not a problem since one can just have the same result with: dict(k, o for k, o in tree.items() if k in paths)
| * Use a dictionnary for tree contentsSteve Frécinaux2008-09-051-6/+5
| | | | | | | | | | | | | | It seems more natural to use a dictionnary for directories, since we usually want to access them by name, and entry order is not relevant. Also, finding a particular blob given its name is O(1) instead of O(N).
* | Merge branch 'master' of git://gitorious.org/git-python/nudFlorian Apolloner2008-09-057-25/+25
|\ \ | |/
| * Fix unneeded dict unpacking.Steve Frécinaux2008-09-057-25/+25
| |
* | commit: handle --bisect-all output in Commit.list_from_stringDavid Aguilar2008-09-021-4/+4
|/ | | | | | | | | | | | Rui Abreu Ferrerira pointed out that "git rev-list --bisect-all" returns a slightly different format which we can easily accomodate by changing the way we parse rev-list output. http://groups.google.com/group/git-python/browse_thread/thread/aed1d5c4b31d5027 This resolves the issue mentioned in that thread. Signed-off-by: David Aguilar <davvid@gmail.com>
* Merge branch 'master' of git@gitorious.org:git-python/mainlineFlorian Apolloner2008-08-211-1/+1
|\
| * fixed it so init actually creates a bare repo. Thank you Daniel Watkins.Michael Trier2008-08-181-1/+1
| |
* | fixed unneeded list unpackingFlorian Apolloner2008-08-211-3/+3
|/
* use shell=True in windows (git.exe needs to be on %PATH%)Florian Apolloner2008-08-111-3/+8
| | | | One bug remaining: git on windows is returning status 0 for `git this-does-not-exist`, so no GitCommandError is raised.
* Fix use of bare repositories.Jelmer Vernooij2008-07-281-1/+2
| | | | (cherry picked from commit 80f8a18a69e307f6ea0bfeb14791fc75e239faa2)
* Remove debug statement.Jelmer Vernooij2008-07-281-1/+0
| | | | (cherry picked from commit 1f9e1bacb65fe503ad23efba7f99ca6346fa648c)
* Avoid stripping newlines in blob data.Jelmer Vernooij2008-07-281-1/+1
| | | | (cherry picked from commit ccca12ee26e40fb4c4df2d77154ed496144569b9)
* Removed method_missing since it was only used in one place.Michael Trier2008-07-172-29/+7
|
* fixed a silly typoFlorian Apolloner2008-07-171-1/+1
|
* Added license information to all files.Michael Trier2008-07-1615-1/+93
|
* fixed ↵0.1.4-preFlorian Apolloner2008-07-081-5/+9
| | | | http://groups.google.com/group/git-python/browse_thread/thread/62b972d2345c74c2?hl=en# and added lines to the per file stats
* fixed ↵Florian Apolloner2008-06-254-90/+52
| | | | http://groups.google.com/group/git-python/browse_thread/thread/b8f3580abf31f9db?hl=en# and passed Git a working_tree again (sort of).
* added os.path.expanduser to the repo initializationFlorian Apolloner2008-06-211-0/+1
|
* continued yesterdays work on the execute stuffFlorian Apolloner2008-06-211-24/+16
|
* made with_exceptions=True default (don't look before you leak ;)) and ↵Florian Apolloner2008-06-202-34/+33
| | | | | | | | removed the combined output of stderr and stdout. Also renamed with_status to extended_output. The method_missing function needs to be modified, as it does a kwargs.pop(xxx, None); which resulted in with_excpetions=None -> False all the time... Test should follow tomorrow.
* Merge branch 'master' of git://gitorious.org/git-python/david into bisectFlorian Apolloner2008-06-202-40/+93
|\
| * cmd: rename with_keep_cwd to keep_cwdDavid Aguilar2008-06-151-7/+9
| | | | | | | | | | | | | | | | Having execute() use a different directory is an important piece of API information so I added more documentation about it and renamed the flag to just "keep_cwd" since that's shorter and simpler. Signed-off-by: David Aguilar <davvid@gmail.com>
| * Merge commit 'sverre/master' into davidDavid Aguilar2008-06-151-2/+23
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * commit 'sverre/master': Added a with_keep_cwd option Improved the GIT_PYTHON_TRACE=full output format Added /dist to .gitignore Conflicts: lib/git/cmd.py Signed-off-by: David Aguilar <davvid@gmail.com>
| | * Added a with_keep_cwd optionSverre Rabbelier2008-06-131-1/+13
| | | | | | | | | | | | | | | | | | | | | When executing commands, if the with_keep_cwd option is specified, the current working directory will be set to os.getcwd() instead of the directory containing the .git directory.
| | * Improved the GIT_PYTHON_TRACE=full output formatSverre Rabbelier2008-06-131-1/+10
| | | | | | | | | | | | | | | | | | It now also shows stderr if there was any on it, and only shows stdout if there was any output. Also added a '->' between the command and the return value as a visual clue.
| * | cmd: better support for bare repositoriesDavid Aguilar2008-06-152-3/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In order to avoid the expense of parsing .git/config just to know whether or not a repository is bare at __init__ time, we just pass an optional flag to Git.__init__(): bare_repo with a default value of False. Repo.init_bare() was updated to pass this flag. We could have an optional Git.read_bare_status() function that does the expensive lookup. Then, users can optionally call it at runtime if they really need to know whether or not a repository is bare. That seems like a decent tradeoff between speed, correctness, and common use cases. Signed-off-by: David Aguilar <davvid@gmail.com>
View Lorry Controller Status