- import pkgs.lib by default, there's no reason why not
- Define `isPy*` helper functions for all python versions
- Define `pythonName` that returns unique python name based on `isPy*`
helpers
- Fix `python2.7` version detection
- Fix pythonPackages importing(with) for `python2.7` only packages
Before we used `easy_install` command to handle installation
in one shot, now this is split into two phases:
- buildPhase: python setup.py build
- installPhase: python setup.py install
Each of those commands have the ability to pass extra
parameters through buildPythonPackage parameters as
`setupPyInstallFlags` and `setupPyBuildFlags`.
Phases now correctly execute post/pre hooks.
In configurePhase we inject setuptools dependency before distutils
is imported to apply monkeypatching by setuptools that is needed
for special features to apply.
We don't have to reorder default phases anymore, as test
phase comes after build and that works.
I rewrote offineDistutils into distutils-cfg with a bit cleaner
syntax and ability to specify extraCfg to the config file.
Plone packages are failing and garbas said he will adopt them to
the new functions. The rest of the packages I fixed and these commits
shouldn't break any package (according to my testings) and they introduce
16 new jobs and fix 38 that were broken before.
This release contains only minor fixes, which are:
* Fix AttributeError bugs in known_hosts file (re)loading.
* Note ecdsa requirement in README.
Full changelog with credits and issue IDs can be found at:
https://github.com/paramiko/paramiko/blob/master/NEWS
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Qt and CUPS are not supported on Darwin and are dependencies.
Note: this makes ipython the same as ipythonLight on Darwin, but
ipython is used as an input for other packages (ipdb and ipdplugin)
and it is reasonable to assume that users on other platforms may
choose ipythonLight.
Matplotlib is currently using hacked together build / install phases
which mimick this functionality of buildPythonPackage. This patch
changes the expression to remove the duplicated build scripting.
pycollada is a python library for reading and writing collada documents.
collada is an open file format (XML based) for 3D applications. Filename
extension is .dae (digital asset exchange).
This only happens during setup.py build and generates the corresponding
_pb2.py files from the .proto files, as they're no longer included in
the source distribution starting from 2.5.0.
Thanks to @iElectric for reporting this build failure:
http://hydra.nixos.org/build/6897112
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Changes:
0.9.0 (2013-10-05)
Support python 3.2 and 3.3 (thanks to @masayuko)
Drop supports for python 2.4 and 2.5
Replace dependency: PIL -> Pillow
(The test phase still fails, so continue having it disabled.)
Changes:
1.0.0 (2013-10-05)
Support python 3.2 and 3.3 (thanks to @masayuko)
Drop supports for python 2.4 and 2.5
Replace dependency: PIL -> Pillow
(The test phase still fails, so continue having it disabled.)
Changes:
0.5.1 (2013-10-22)
Fix bugs
0.5.0 (2013-10-05)
Support python 3.2 and 3.3 (thanks to @masayuko)
Drop supports for python 2.4 and 2.5
Replace dependency: PIL -> Pillow
(The test phase still fails for one test, so continue having it disabled.)
Changes:
1.3.2 (2013-11-19)
Fix bugs
1.3.1 (2013-10-22)
Fix bugs
1.3.0 (2013-10-05)
Support python 3.2 and 3.3 (thanks to @masayuko)
Drop supports for python 2.4 and 2.5
Replace dependency: PIL -> Pillow
(The test phase still fails for one test, so continue having it disabled.)
Changes:
* Added context manager support to Cursor
* Added padding for driver bugs writing an extra byte
* Cursor.executemany now accepts an iterator or generator.
* Compilation improvements for FreeBSD, Cygwin, and OS/X
* Use SQL_DATA_AT_EXEC instead of SQL_DATA_LEN_AT_EXEC when possible for driver compatibility.
* Row objects can now be pickled.
