just simple example/proof of concept
in file df_firestore_stream.py
set GCP project and BUCKET (if you want to run on Dataflow)
run_local
variable in the file controls whether pipeline works locally or on Dataflow
to run pipeline just execute python df_firestore_stream.py
publish.py
just publishes messages to PubSub, for testing purposes. in variable PROJECT
set GCP project
At the moment this doesn't work with Apache Beam 2.13.0 but works with 2.12.0. Python 2.7 was used