README

Hierarchical (Geospatial) Topic Aggregation in Pub/Sub Message Brokers
----------------------------------------------------------------------

HTAbroker is a prototype for propagating and aggregating metrics
submitted to pub/sub topics along geospatial hierarchies. It supports
multiple indexing systems such as NUTS, geohash, H3 - some
computationally, some data-driven, and some in hybrid mode.

Run HTAbroker as CLI tool using either simulation or attached to the
Apache Pulsar broker. Similar to the concept of FaaS, Pulsar allows
deploying light-weight functions which are processed in an event-driven
way; HTAbroker makes use of this concept to inject hierarchical
subscription routes. For convenience, the system is also containerised.

Run from code:
 term0% # install deps, e.g. pip3 install pulsar-client
 term1% ./pulsar-dockerised.sh
 term2% python3 htabroker.py

Run as containers:
 term0% docker-compose run htabroker

Once in the interactive shell, marked with ")))" in HTAbroker,
everything should be self-explaining. You can follow the examples from
the 'doc' file to get started.

(Note: Function deployment is currently not working in the containerised
version.)