Wrapper for virtualenv, pip, and wheel
Project description
A wrapper for virtualenv, pip, and wheel that caches wheels to speed up virtualenv creation.
Features:
Create and maintain a cache of wheels.
Create a virtualenv and install packages from the wheel cache in a single command.
Add to the wheel cache as new packages are built.
Download the virtualenv source code when not already installed or out of date.
Note that this package should be considered alpha/unstable!
Requirements
Python 2.7
Unix-like systems (ie, not Windows)
Examples
Create a wheel cache (~/.mkvenv by default):
% mkvenv init % mkvenv list-wheels # Wheels in /Users/nhoffman/.mkvenv/2.7.7/
Now create a new virtualenv named test-env and install some packages. As each package is installed, a wheel is first built and saved to the cache, along with wheels for any dependencies:
% cat requirements.txt flake8 % mkvenv install --venv test-env -r requirements.txt % test-env/bin/pip freeze flake8==2.3.0 mccabe==0.3 pep8==1.5.7 pyflakes==0.8.1 wsgiref==0.1.2 % mkvenv list-wheels # Wheels in /Users/nhoffman/.mkvenv/2.7.7/ = flake8-2.3.0-py2.py3-none-any.whl mccabe-0.3-py2.py3-none-any.whl pep8-1.5.7-py2.py3-none-any.whl pyflakes-0.8.1-py2.py3-none-any.whl
Subsequent requests to install any of these packages will use the cached wheels.
Installation from the wheel cache can speed things up a lot when packages require compilation:
% cat scientific.txt numpy scipy pandas biopython seqmagick % mkvenv install --venv sci-env -r scientific.txt 620.66s user 36.41s system 75% cpu 14:28.52 total % mkvenv install --venv another-env -r scientific.txt 10.87s user 4.82s system 18% cpu 1:24.65 total
Installation
Installation is most easily performed from PyPi using pip:
pip install mkvenv
Alternatively, obtain the source code from either PyPi (https://pypi.python.org/pypi/mkvenv) or GitHub (https://github.com/nhoffman/mkvenv) and install by running either python setup.py install or pip install . from within the package directory. Installation provides a script named mkvenv as an entry point. mkvenv.py may also be executed directly from the top level of the package directory.
Alternatively, note that the mkvenv script is implemented as a single python file that can be invoked directly as a script. This script can be downloaded and used to create a virtualenv on a system on which the virtualenv package is not available:
wget https://raw.githubusercontent.com/nhoffman/mkvenv/master/mkvenv/mkvenv.py python mkvenv.py
It may also be useful to distribute mkvenv.py along with other projects to facilitate creation of execution environments.
Execution
Run mkvenv -h for a list of subcommands and common options, or mkvenv <subcommand> -h for help on a subcommand. Note that common options must be provided before the name of the subcommand (eg, mkvenv -v wheelstreet -r requirements.txt)
Known Bugs
There’s a known bug in pip using python 2.7 on OS X (https://github.com/pypa/pip/issues/1964 - the issue is not specific to this project) that results in an error on installation from PyPI with the message “AssertionError: Multiple .dist-info directories” after a previous installation. The solution is to delete any residual pip build directories:
find /private -name 'mkvenv' -exec rm -r "{}" \; 2> /dev/null
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file mkvenv-0.2.3.tar.gz
.
File metadata
- Download URL: mkvenv-0.2.3.tar.gz
- Upload date:
- Size: 7.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 094884554a7904b4855fcdf1e04a210975807a9365565a2bad1c090fee52f3a4 |
|
MD5 | 9d5c3f22ebb6f3fd3c01a17072ec3bbf |
|
BLAKE2b-256 | 10b98b3bea50488f0c59ba60508360dbe99d77e9903848d6eba8b58a8a88fadb |
File details
Details for the file mkvenv-0.2.3-py2-none-any.whl
.
File metadata
- Download URL: mkvenv-0.2.3-py2-none-any.whl
- Upload date:
- Size: 7.3 kB
- Tags: Python 2
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 306186c85c2b351c62e11e42de64da133430e5fe13e8aab3aca802cd069c1bbf |
|
MD5 | 1ee5e5033bce5d2f2e4b975164ab7203 |
|
BLAKE2b-256 | 6b4e61dd9aecabde7b81f09b3844c9235f460e798dda08c432c179a6081c54c8 |