Distriploy

This tool can assist you to perform deployment of release artifacts.

Usage

After its installation, distriploy is typically called from your repo's root folder using:

distriploy

or:

python -m distriploy

The repo should have been configured prior to that, see configuration.

Run it with --help to get to know the options.

Prior to running distriploy, you might want to create a custom tag. It will be used to name the release. E.g.:

git tag -s r$(date +%Y%m%d)

Installation

distriploy is available on pypi:

python -m pip install --user distriploy

or you can grab a release or clone the repository and use it.

Configuration

An enabled repository has a .distriploy.yml file in its root folder. See this repo's for inspiration.

The file is YAML, it contains an object with the following members:

• release: release information object

• release.method: release method. Supported values are:

  • github: the project has a github repo.

    Additional info may be provided if so:

    • release.remote: string, optional (defaults to origin)

      The local git repo remote corresponding to github.

    • postrelease.add_mirror_urls: bool (defaults to false)

      Whether to add to the release description a list of mirrors. Limited use because if github is down, the list is not accessible.

    • Environment variable GITHUB_TOKEN, must be populated by user, contains a token string created from https://github.com/settings/tokens

      Set it up with eg.:

      read GITHUB_TOKEN
      export GITHUB_TOKEN

      Or add it to an environment file not under revision control.

• mirrors: mirroring information object, containing members whose key is a mirror handle.

  Each entry contains:

  mirrors.${mirror}.method: mirroring upload method. Supported values (unsupported are ignored, with a warning) are (TODO):

  • osf: upload to https://osf.io

    Use requires installation with osf option, or subsequent installation of the osfclient dependency.

    Additional info may be provided if so:

    • project: string, OSF project key

    • folder: string, OSF file folder

    • name: string, optional (defaults to release artifact name)

      Useful if using OSF's revision system.

    • Environment variables OSF_USERNAME and OSF_PASSWORD.

      Set it up with eg.:

      read OSF_USERNAME OSF_PASSWORD
      export OSF_USERNAME OSF_PASSWORD

      Or add it to an environment file not under revision control.

  • rsync: upload with rsync

    Additional info may be provided if so:

    • remote: string, rsync destination
    • public: string, public URI prefix, accessible once upload is completed

  • academictorrents: create torrent file, and upload to https://academictorrents.com

    The torrent file will be generated in cwd, named like the release artifact and added`.torrent` extension.

    Additional info may be provided if so:

    • params: object, academictorrents upload params, to be POST'ed

    • Environment variables ACADEMICTORRENTS_USERNAME and ACADEMICTORRENTS_PASSWORD, must be populated by user, they come from uid & pass in https://academictorrents.com/about.php#apikeys

      Set it up with eg.:

      read ACADEMICTORRENTS_USERNAME ACADEMICTORRENTS_PASSWORD
      export ACADEMICTORRENTS_USERNAME ACADEMICTORRENTS_PASSWORD

      Or add it to an environment file not under revision control.

TODO

• get some use and feedback

License

MIT.

Releasing

read version # eg. 0.14
sed -i -e 's/^version = .*/version = "'${version}'"/g' setup.py
git add setup.py
git commit -m "preparation for v${version}"
git tag --annotate --message "" v${version}
git push; git push --tags
rm -rf dist; python setup.py sdist && twine upload dist/* -r pypi
python -m distriploy