Functions to parse YAML output to KGX-compliant CSV

[serenalotreck]

Added two functions to src/ontogpt/io/csv_wrapper.py to convert the YAML output to KGX-compliant CSVs as per #343.I left the original two functions untouched, as I wasn't sure how to integrate my changes into them.

This code relies on the assumptions that:

1. User has followed the naming conventions in the schema re: camel case & underscored versions of the entity and relation types (i.e. that there is a class called EntityContainingDocument that contains pluralized, underscored versions of the camelcase names for all entities and relations defined in the rest of the classes section).
2. OntoGPT's extracted_object will always have entities as lists of strings and relations as lists of dictionaries.

[caufieldjh]

Big thanks @serenalotreck !
I'll finish integrating this and add the options to the CLI.

[caufieldjh]

One detail - I'm going to make it such that the root class in the schema doesn't have to be EntityContainingDocument - it will just use the root class or accept an argument to specify a different class.

[caufieldjh]

OK, basic functionality is all there, though it's a bit rough.
The parse_yaml_predictions function will take a path to YAML output and the corresponding schema, yielding dataframes.
Specifying -O kgx along with a ontogpt extract will do the same but output node and edge files (as nodes.tsv and edges.tsv) to the working directory.