appenv

Self-contained bootstrapping/updating of Python applications deployed through shared repositories.

The following examples use the ducker package to illustrate how to use appenv. ducker and appenv are not related at all.

Bootstrapping an application / project

Use curl -sL https://github.com/flyingcircusio/appenv/raw/master/bootstrap | sh for bootstrapping a new project.

$ curl -sL https://github.com/flyingcircusio/appenv/raw/master/bootstrap | sh
Let's create a new appenv project.

What should the command be named? ducker <return>
What is the main dependency as found on PyPI? [ducker] <return>
Where should we create this? [/private/tmp/ducker] <return>

Creating appenv setup in /private/tmp/ducker ...

Done. You can now `cd ducker` and call `./ducker` to bootstrap and run it.

$ cd ducker
$ ./ducker
Running unclean installation from requirements.txt
Ensuring unclean install ...
Please initiate a query.
Ducker (? for help) q

Freezing requirements for repeatable builds

Using frozen requirements makes the builds repeatable for you and your team and also speeds up subsequent invocations:

$ ./appenv update-lockfile
Updating lockfile
Installing packages ...

$ time ./ducker wikipedia
Installing ducker ...
./ducker wikipedia  2.91s user 0.99s system 88% cpu 4.407 total

$ time ./ducker wikpedia
./ducker wikipedia  0.22s user 0.11s system 90% cpu 0.371 total

Using a specific version of Python for your application

appenv tries to use the best Python version available. It bootstraps with the Python 3 interpreter available in your PATH as python3 and then can either detect the newest Python or select the best python of your choice.

Two disable the automatic detection of the newest version and provide a list of acceptable Python versions (tried in the order you list them) add the following line to your requirements.txt file:

# appenv-python-preference: 3.6,3.9,3.8

The best version that is found on the system will be used to re-spawn appenv and then also used to manage the virtual environments for your application.

AppEnv itself is tested against Python 3.6+.

Learning more about appenv

$ ./appenv --help
usage: appenv [-h] {update-lockfile,init,reset,prepare,python,run} ...

positional arguments:
  {update-lockfile,init,reset,prepare,python,run}
    update-lockfile     Update the lock file.
    init                Create a new appenv project.
    reset               Reset the environment.
    prepare             Prepare the venv.
    python              Spawn the embedded Python interpreter REPL
    run                 Run a script from the bin/ directory of the virtual env.

options:
  -h, --help            show this help message and exit

Testing

If you want to contribute, please install tox and run it.

$ tox