lotus/composer
2020-06-23 10:02:05 -04:00
..
app make sure each group gets its own param objects 2020-06-23 10:02:05 -04:00
fixtures test runner working in docker 2020-06-23 10:02:05 -04:00
composer.ipynb new composition flow working e2e, some rough edges 2020-06-23 10:02:05 -04:00
composer.sh skip docker build if SKIP_BUILD is set 2020-06-23 10:02:05 -04:00
Dockerfile test runner working in docker 2020-06-23 10:02:05 -04:00
Makefile wip, jupyter in docker 2020-06-23 10:02:00 -04:00
README.md add readme 2020-06-23 10:02:05 -04:00
requirements.txt test runner working in docker 2020-06-23 10:02:05 -04:00

Testground Composer

This is a work-in-progress UI for configuring and running testground compositions.

The app code lives in ./app, and there's a thin Jupyter notebook shell in composer.ipynb.

Running

You can either run the app in docker, or in a local python virtualenv. Docker is recommended unless you're hacking on the code for Composer itself.

Running with docker

Run the ./composer.sh script to build a container with the latest source and run it. The first build will take a little while since it needs to build testground and fetch a bunch of python dependencies.

You can skip the build if you set SKIP_BUILD=true when running composer.sh, and you can rebuild manually with make docker.

The contents of $TESTGROUND_HOME/plans will be sync'd to a temporary directory and read-only mounted into the container.

After building and starting the container, the script will open a browser to the composer UI.

You should be able to load an existing composition or create a new one from one of the plans in $TESTGROUND_HOME/plans.

Right now docker only supports the standalone webapp UI; to run the UI in a Jupyter notebook, see below.

Running with local python

To run without docker, make a python3 virtual environment somewhere and activate it:

# make a virtualenv called "venv" in the current directory
python3 -m venv ./venv

# activate (bash/zsh):
source ./venv/bin/activate

# activate (fish):
source ./venv/bin/activate.fish

Then install the python dependencies:

pip install -r requirements.txt

And start the UI:

panel serve composer.ipynb

That will start the standalone webapp UI. If you want a Jupyter notebook instead, run:

jupyter notebook

and open composer.ipynb in the Jupyter file picker.