Add JSON generation

[danellecline]

Add JSON generation to simplify the steps for processing with pypam.

Reference repo:
https://github.com/mbari-org/soundscape-wavdata
https://github.com/ioos/soundcoop/

Tasks

Beta Give feedback

1. Create branch for this work (jsongen)
2. Copy modules from soundscape-wavdata into appropriate modules
3. Add entrypoint for calling the metadata generation
4. Add pytest
5. Add simple notebook to demonstrate local JSON generation

[danellecline]

@ryjombari, just FYI, in case you want to follow along.

[danellecline]

@carueda - where would you like me to put the modules? It should be different than the .nc metadata in metadata to avoid confusion.
Carlos and D.Cline spoke on Slack about this. This will live in src/metadata.

[ryjombari]

Thanks for the jumpstart @danellecline!

[ryjombari]

Relaying a thought from communications with Samara @ NOAA: Tooling this demonstration to gather metadata and read data directly from the Google Cloud Platform would pave the way for PBP to be the engine for a lot of NOAA passive acoustic data. If I understand, MinIO can be used to move data from GCP to the processing instance, and PBP can move to using MinIO? How about the gathering of file metadata?

[danellecline]

Thanks @ryjombari. @carueda I'm' wondering if 1) does Minio replace boto3, or do they stand side-by-side, and 2) does boto3 support GCP. I know that minio is S3 compliant and can be used for both local and cloud storage operations - I am using it now for video processing pipelines successfully this way.

[carueda]

I haven't looked into minio with enough detail to tell for sure, but I think it supports AWS, GCP, and probably others. If so, we could replace the boto3 dependency with minio and adjust the appropriate places dealing with file downloads/uploads (s3 and gs URIs)

[carueda]

As part of #6, I note that Minio seems to only be for s3 compliant endpoints, and I started using the google-cloud-storage package for handling gs://.