EggsPacker is a Python module that allows you to put in a single directory a Python egg and all its dependencies as long as it is able to find them in a Python index.
It is almost equivalent as running "easy_install -zmaxd basket somepackage", with 2 important differences :
1. The python version and target platform can be changed from the system ones. This allow to prepare all the necessary eggs for a py2.5 win32 distribution on a py2.7 linux (for example).
2. EggsPacker will only work with eggs, no sources. So it requires an index that have all the dependencies available as eggs for the correct target platform / python version.
The implementation rely as much as possible on pkg_resources to resolve the requirements and read the metadata, so there should be no wrong version packed although I of course give no warranty of any kind.
Reading the package index on the other hand was tested only on a an index builded by http://pypi.python.org/pypi/basketweaver/, so expect issues if hitting directly the pypi index.
Usage: packeggs [options] pkgname [pkgname [pkgname...]]
-h, --help show this help message and exit
-i INDEX, --index=INDEX
A Pypi compatible index URL. Default to pypi
-t TARGETDIR, --targetdir=TARGETDIR
Target directory. Default: "basket".
Target python version.
--platform=PLATFORM Target platform
--unzip=UNZIP Unzip the asked egg if present in the dependencies.
The special value "auto" will guess what to do with
with not-zip-safe flag, and "all" will force unzipping
of all the eggs. Can be specified several times.
The following command will get into the 'basket' directory the ANicePackage-0.1-py2.5 egg and all its dependencies for the win32 platform, while unzipping the eggs not marked as zip-safe plus forcely unzipping ANicePackage:
packeggs -i http://yourserver/your/package/index --python-version 2.5 --platform win32 --targetdir basket ANicePackage==0.1 --unzip auto --unzip ANicePackage