To get set up, clone and enter the repo.
git clone https://github.com/os-climate/physrisk-api.git
cd physrisk-api
We recommend using pdm for a consistent working environment. Install via, e.g.:
pip install pdm
For ease of using Jupyter notebooks (e.g. in VS Code) the config can be used:
pdm config venv.with_pip True
The command:
pdm install
will create a virtual environment (typically .venv folder in the project folder) and install the dependencies. We recommend that the IDE workspace uses this virtual environment when developing.
When adding a package for use in new or improved functionality,
pdm add <package-name>. Or, when adding something helpful for
testing or development, pdm add -dG <group> <package-name>.
Patches may be contributed via pull requests to https://github.com/os-climate/physrisk-api.
All changes must pass the automated test suite, along with various static checks.
Enabling automatic formatting via pre-commit is also recommended:
pre-commit install
or
pre-commit run --all-files
To ensure compliance with automated checks, developers may wish to run formatting and test steps which can be done via:
pdm run lint
pdm run test
For those using VS Code, configure tests ('Python: Configure Tests') to use 'pytest' to allow running of tests within the IDE.
This is a useful clarification of the forking workflow:
https://gist.github.com/Chaser324/ce0505fbed06b947d962
https://www.atlassian.com/git/tutorials/comparing-workflows/forking-workflow
An image can be built and run via:
docker build . -t physrisk-api-image
docker run -d --name physrisk-api -p 8081:8081 physrisk-api-image
The API should then be accessible via:
http://localhost:8081
and docs via:
http://localhost:8081/docs
An example is also given of using Docker Compose to use nginx as a reverse proxy. To run the application locally, run either of the following commands;
docker-compose up
# or
podman-compose up
This is just an example and details of how the image are used in an Application are mostly beyond the scope of this repo. There is a useful introduction here:
https://fastapi.tiangolo.com/deployment/docker/.
The example 'physrisk' project is hosted on the OpenShift Container platform. This is the 'One Load Balancer - Multiple Worker Containers' case where replication is done at the cluster level, which explains why the Dockerfile configures a single worker process.
https://physrisk-ui-physrisk.apps.odh-cl2.apps.os-climate.org/
https://physrisk-api-physrisk.apps.odh-cl2.apps.os-climate.org/docs
Currently, this image is built and released automatically to quay.io and manually (/automatically for a test instance) deployed onto OpenShift.
We recommend the approach of https://fastapi.tiangolo.com/tutorial/debugging/. That is, the "Python Debugger: Current File" option can be selected, running file .\src\physrisk_api\app\main.py.