dcpy user configuration

I figure for a first pass, use pyproject.toml or another type of config file, but since we use it pyproject.toml already this seems like a decent path. But this issue is meant to be a bit of a brainstorming space

sqlfluff has a lot of nice logic around how you might go about this just in terms of reading in a config, especially around looking for different files at different paths - see here

But we could do something simpler to start - just assume that pyproject.toml is in the current folder or something for now. Basically, for ingest-related stuff, I'm thinking something like this

[tools.dcpy.lifecycle.connectors]
recipes.datasets = { conn_type = "s3", prefix = "datasets" }
ingest = { conn_type = "ingest_datastore", storage = "recipes.datasets" }

and then we'd have something in python

config = config_from_toml("pyproject.toml")
...
connectors = ConnectorRegistry()
...
for connector in config["lifecycle"]["connectors"]:
    connectors.register(Connector(**connector))

this is clearly a little ugly as is, but something along these lines.

Start by reviewing the proposed pyproject.toml structure and the linked sqlfluff config loader for approaches to locating configuration files. Then inspect the existing connector and ingest-related entry points to determine how configuration could reach ConnectorRegistry. Done should include an agreed configuration shape and a defined loading and connector-registration flow.