Intended to be used for companies or organizations that need a private PyPi. Originally based on simplepypi, however this solution requires a compute engine instance which is quite expensive for the task.
This application runs on Google App Engine (and uses webapp2 instead of tornado) and stores the uploaded packages on Google Cloud Storage. It supports
- Uploading of packages
- Downloading by package (or package and version)
- A / page that is navigatable with a web browser
- /pypi/ listing
- Basic ACL through http auth with optional account roles (read/write)
- Prevent overwriting a package with the same name & version
Unsupported are:
- Package registration
Before deploying, install the required dependencies to the libs folder:
pip install -t libs -r requirements-app.txt
Next, have a look at config.json, you can set up multiple accounts for the http auth. The passwords are assumed to be hashed with sha1, you can obtain the hash as follows:
from webapp2_extras import security
security.hash_password('password', method='sha1')
Now, you are ready to deploy to google app engine. Depending on how you set up your projects, this may require adjusting app.yaml, or setting the right project for gcloud. Then:
gcloud app deploy app.yaml
will launch your package index
Uploading a python package is quite straightforward. First, add your private package index to ~/.pypirc
:
[distutils]
index-servers =
pypi
local
[local]
username = username
password = password
repository = https://project.appspot.com
[pypi]
repository=https://pypi.python.org/pypi
username=your_pypi_username
password=your_pypi_password
The later reflects a configuration for uploading on the public pypi. Now go to the directory containing setup.py and upload your package to the private package index by specifying the local configuration:
python setup.py sdist upload -r local
When installing a package (e.g. dummy), specify the url to the package index:
pip install -i https://username:password@project.appspot.com/pypi dummy
To run the testsuite, we recommend to use the nox tool:
pip install nox-automation
Next, to run the tests:
nox