Flask server that runs ISIS commands

Environment

Any environment variables without a default are required for the app to start

Name	Description	Default
APP_LOG_LEVEL	A log level name (see Python docs)	"info"
APP_PORT	The port that the app should run on	8080
APP_SMTP_SERVER	The SMTP server for emailing pipeline output	localhost
APP_SMTP_PORT	The port for APP_SMTP_SERVER	25
APP_SMTP_USERNAME	The username APP_SMTP_SERVER	None
APP_SMTP_PASSWORD	The password APP_SMTP_USERNAME	None
S3_SERVER	The URL of the S3 server that the app should use to store intermediate data
S3_BUCKET	The name of the S3 bucket that the app should use
S3_ACCESS_KEY	The access key for S3_BUCKET
S3_SECRET_KEY	The secret key for S3_BUCKET

These variables can be set directly from the CLI or in a .env file at the root of the app:

APP_LOG_LEVEL=debug
APP_PORT=8080

APP_SMTP_SERVER=...
APP_SMTP_PORT=...
APP_SMTP_USERNAME=...
APP_SMTP_PASSWORD=...

S3_SERVER=...
S3_BUCKET=...
S3_ACCESS_KEY=...
S3_SECRET_KEY=...

Development

Install GDAL

This project requires GDAL >= 3.1.0 be installed globally, since GDAL and ISIS cannot currently be installed in the same conda environment.

On Ubuntu:

# apt-get install -y python3-dev
# add-apt-repository ppa:ubuntugis/ppa
# apt-get update
# apt-get install -y gdal-bin

Create the project environment

1. Install miniconda3
2. Install the project's dependencies into .python

   $ conda env create -p .python -f environment.yml

3. Activate the new environment in .python

   $ conda activate ./.python 

4. Install the development dependencies

   $ pip install -r requirements.dev.txt

5. Start the server, it'll automatically pick up any changes to the code and restart itself and run on port 8000 by default.

   $ gunicorn -c gunicorn.conf.py --reload isis_server:app

If you make updates to environment.yml, you can update the .python directory to match by running:

$ conda env update -p ./.python -f environment.yml --prune
$ pip install -r requirements.dev.txt

To run tests:

$ pytest

To test a command (I00831002RDR.cub available via LFS in data):

$ curl -s -X POST \
    -H 'Content-Type: application/json' \
    -d '{"cmd": "lowpass", "input_file": "I00831002RDR.cub", "args": []}' \
    localhost:8000 | jq
{
  "output": "I00831002RDR.lowpass.cub"
}