WorldEx is designed to make subnational data discoverable using H3 indexing. Geocoders are implemented that are responsible for handling geospatial data types and converting them into H3 indices.
The project consists of a web application that allows for users to explore the available data at a given H3 cell. Users can search for names of places, which are processed using OpenStreetMap's Nominatim service.
APIs are implemented to allow users to interact with the indexed data.
First, install just. It is similar to make in that it's just a convenient wrapper around lengthier cli commands to setup the local environment.
to setup the simple password generator.
to setup aws environment variables (credentials, etc.) to access parquet files for dataset seeding later. This will create a ./secrets/aws.env file. At the time of writing however, you'll have to fill the aws key id and secret manually.
Ask the repo maintainers if you're part of the same team. Eventually, these parquet files should be made publicly accessible.
If you're on a Mac, you may have to run the following first to enable envsubst
brew install gettext
brew link --force gettext
to generate a password to the elastic user of the es elasticsearch service
to generate a postgres password and the environment files to be used by the api, db, and pgweb services as per the docker compose spec.
If you've previously generated a postgres password with this command and just want to regenerate the environment files, run create-envs
If you are on an ARM-based device (e.g. M1/M2 MacBooks), comment out the following line under the db service on docker-compose.yaml
dockerfile: arm64.Dockerfile
This PostGIS build will be more performant on your local machine.
to run the development environment on local. docker compose down to stop it.
Note that the cluster doesn't include the vite app by default. You'll have to run yarn to install the dependencies and
to run the dev server. We exclude the vite app since node_modules cannot be bind-mount if the dependencies were installed on a different architecture (e.g. mac host, linux container).
You may need to --force if you encounter issues with the lockfile.
If you're outside of development context that requires hot reloads, you can run the following to run/test actual builds of the vite app.
will run vite preview to preview the production build (not actually intended as a production server)
will run an nginx reverse proxy serving the vite app build.
See the API readme to finish setting the api (along with the database) service up.
We use this service to search against dataset metadata. After populating your database with the pre-indexed datasets, from the /api directory run
just run-script es_index_dataset
You can use https://elasticvue.com/ as a gui to the elasticsearch service/indices - the browser extension is sufficient. You will have to provide the uri http://localhost:9200 and the generated credentials to connect to the cluster.
is used to minimize nitpicks such as formatting during code review.
to install pre-commit. Or see https://pre-commit.com/ for installation instructions.
to then setup the pre-commit hooks.
pip install poetry
poetry install
poetry run python -m pytest