summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2014-07-10Tweak makefile test runnerGravatar Jon Bernard 1-1/+1
2014-07-10Update gitignore to include .tox directoryGravatar Jon Bernard 1-0/+1
2014-07-10Add tox configuration fileGravatar Jon Bernard 1-0/+6
2014-07-10Remove executable bit from test_basic.pyGravatar Jon Bernard 1-0/+0
2014-07-08Ignore build directory for wheelsGravatar Jon Bernard 1-0/+1
2014-07-08Push test packages to correct serverGravatar Jon Bernard 1-2/+2
2014-07-08Fix typo in setup.py classifiersGravatar Jon Bernard 1-1/+1
2014-07-07Add unit test for adding a package file via CLIGravatar Jon Bernard 1-0/+24
Re #20
2014-07-07Remove accidental tabs in unit testsGravatar Jon Bernard 1-6/+6
Python 2 will overlook this and continue executing, but python 3 will issue an error and exit.
2014-07-07Cleanup setup.pyGravatar Jon Bernard 1-27/+17
2014-07-07Use entry point for bin stubGravatar Jon Bernard 1-1/+4
The script in bin/ remains for running directly from the source directory.
2014-07-07Parse version string without importing source codeGravatar Jon Bernard 1-9/+7
It's probably better to execute as little code as possible from setup.py, so this approach seems cleaner to me.
2014-07-07Remove test and publish functions from setup.pyGravatar Jon Bernard 1-10/+0
2014-07-07Add setup.cfgGravatar Jon Bernard 1-0/+2
2014-07-07Simplify MANIFEST.inGravatar Jon Bernard 1-1/+1
2014-07-07Update travis configuration to use make testGravatar Jon Bernard 1-4/+4
2014-07-07Add Makefile for testing and publishingGravatar Jon Bernard 1-0/+12
2014-07-07Move tests into a directoryGravatar Jon Bernard 1-0/+0
I'm hoping to add more tests for specific features and break then into separate files so things are a bit easier to manage.
2014-05-29Add Alexandre Viau to list of authorsGravatar Jon Bernard 1-2/+2
Thanks for the contributions! They are much appreciated.
2014-05-29Correct logic in package prefix testGravatar Alexandre Viau 1-2/+2
In reference to #33.
2014-05-27Add test for syncing packages with a non-default prefixGravatar Jon Bernard 1-0/+28
This test currently fails because of a bug. In reference to #33.
2014-01-01Update copyright yearGravatar Jon Bernard 1-1/+1
2014-01-01Update gitignore to exclude setuptools cruftGravatar Jon Bernard 1-0/+2
2014-01-01Update history and version number for 0.6.3v0.6.3Gravatar Jon Bernard 2-1/+7
2014-01-01Prevent adding toplevel package directoriesGravatar Jon Bernard 1-0/+3
This change prevents users from adding toplevel package directories via the add operation in the cli. Until add is package-aware, allowing this behaviour can lead to loss of configuration files if the --force option is later used to sync. Adding packages already-defined packages will be possible in the future, but this patch is necessary to prevent users from reaching an invalid state. Re #20, #27
2014-01-01Add unit test for adding a missing packageGravatar Jon Bernard 1-0/+23
This fails because the add operation is not aware of package settings, and therefore adds a package dotfiles as it normally would - as a toplevel symlink. The problem arises when a user attempts to sync and the expected symlink layout is not as it should be. With the force option, this can lead to deleted files with no way to recover. Re #20, #27
2014-01-01Update gitignore to include .venvGravatar Jon Bernard 1-0/+1
2013-12-19Fidget with the badges for the last timeGravatar Jon Bernard 1-5/+5
2013-12-17Move waffle and travis badges to contribute sectionGravatar Jon Bernard 1-6/+6
2013-12-17Remove python 2.5 from travis-ci configurationGravatar Jon Bernard 1-1/+0
2013-12-17Only replace existing dotfiles on remove operationGravatar Jon Bernard 1-1/+7
It is possible that the dotifle is removed from the repository before the remove is issued. In this case only the symlink is removed. Fixes #25
2013-12-17Verify symlink removal in unit testGravatar Jon Bernard 1-0/+4
Re: #25
2013-12-17Add unit test for removing a dotfile with a missing targetGravatar Jon Bernard 1-0/+21
If the remove operation is issued on a dotfile pointing to a target that's already been removed from the repository, then the symlink should just be removed quietly. Re: #25
2013-11-09Fix history formatting for 0.6.2Gravatar Jon Bernard 1-3/+3
2013-11-09Update history and bump version numberv0.6.2Gravatar Jon Bernard 2-1/+8
2013-11-09Add waffle.io badgeGravatar Jon Bernard 1-0/+3
2013-11-09Add Jay Sweeney to AUTHORSGravatar Jon Bernard 1-0/+1
2013-11-09A set makes more sense (and is faster) than list hereGravatar Jay Sweeney 1-1/+1
2013-11-09Python 3.x compat: map is lazy in 3.xGravatar Jay Sweeney 1-1/+1
Closes: #21
2013-11-09Fixes #22: Ensure --force option is handed onGravatar Jay Sweeney 2-1/+12
2013-11-07Fix unit tests to exit with the correct return codeGravatar Jon Bernard 1-5/+1
2013-11-07Bump version number and update HISTORYv0.6.1Gravatar Jon Bernard 2-1/+8
2013-11-07Add Jesús García Crespo to AUTHORSGravatar Jon Bernard 1-0/+1
2013-11-07Update contribution documentationGravatar Jon Bernard 1-3/+3
Remove mention of the develop branch since it's going away in favor of a much simpler branching model (feature branches off master). The use of git-flow was a great learning experience, but is really too much for projects such as this.
2013-11-07Update travis badge for master branch onlyGravatar Jon Bernard 1-1/+1
I'm abandoning git-flow for a much simpler branching model: https://gist.github.com/jbenet/ee6c9ac48068889b0912 Since the master branch must always be deployable, it makes sense to show the travis badge for the master branch only.
2013-11-07Merge branch 'feature/single-sync'Gravatar Jon Bernard 5-15/+116
2013-11-07Update move documentation to be more conciseGravatar Jon Bernard 1-2/+3
2013-11-07Update sync documentation to indicate optional file argumentGravatar Jon Bernard 1-1/+1
2013-11-07Implement single-sync for specified filesGravatar Jon Bernard 1-2/+7
This commit allows the user to sync only one (or a subset) of the available dotfiles in the repository.
2013-11-07Restrict the sync files to what user specifiedGravatar Jon Bernard 1-1/+5
If the user doesn't specify a set of files to sync then the entire set of dotfiles is assumed. There is an obvious bug here, as user-passed strings are not yet converted to dotfile objects.