gossamr / covid

Covid visualization

covid-19 visualizer

Architecture

• webserver - a webserver using fastapi to render templates.

Development

Using docker (the easiest)

docker-compose up  # possibly with --build

and go to localhost:8000 (main server)

locally/manually

Manually using poetry:

$ poetry install
$ poetry shell

then start up the fastapi server:

$ cd src/server
$ uvicorn main:app --reload

Using conda environments

1. Install miniconda (e.g. Windows)
2. open terminal in the clonned repository and run:

$ conda create -n covid python=3.8  # creates a new conda environment with python 3.8
$ conda activate covid  # activates the conda environment
$ pip install poetry  # installs poetry inside this environment
$ poetry install  # installs the project dependencies (you must be in the root of the cloned repository)

and then you should be able to follow locally/manually way of triggering the server.

Getting inside the container

$ docker-compose run --entrypoint bash server

can be used to e.g. install deps:

$ docker-compose run --entrypoint poetry server add pyarrow

tests

poetry run pytest tests

or inside the container

linting

install the needed packages locally with

npm i

Then run the linting tests with

npm test

To automatically lint your files run

prettier --write "{src/server/static/**, src/server/templates/**}"

Deployment

We use Github Actions. The pipeline is specified in .github/workflows/pythonapp.yml

• On a merge to staging branch, the code is deployed to the staging environment.
• On a merge to master branch, the code is deployed to the production environment.