Features/4 yaml to json pipeline

[radishmouse]

Closes #4

This PR:

• adds dependencies to requirements.txt
• copies the .yamllint from https://github.com/thunderbird/thunderbird-notes/tree/master
• creates a NotificationSchema class that validates incoming yaml, produces equivalent json
• adds a python script that converts a folder of yaml files using NotificationSchema
  • uses argparse to accept input and output directory names
• adds a python file that generates a JSON schema file using NotificationSchema
  • uses argparse to accept output file name

[radishmouse]

@MelissaAutumn I might have gone a little too far for this PR, but I created the NotificationSchema using pydantic and included type and value restrictions.

I'm wondering: would this change (using pydantic to validate the incoming yaml and produce the json) basically close #1 and #3?

• If the pydantic model can validate the yaml and produce the corresponding JSON, then maybe this PR supersedes #1? (Because the pydantic model is basically a translation of the JSON schema.)
• And if the requirements/constraints of the JSON schema are enforced by the pydantic model, the model should always produce valid JSON that conforms to that schema, which takes care of #3.

Not that we shouldn't do a final check of all of the JSON produced, but what do we check it against? If we have a separate JSON schema file, there are now two sources of truth for what is/isn't valid data. It seems like it would be best if there were only a single source of truth. (And the pydantic model seems like the more robust of the two.)

Please let me know if I'm way off base here! (Also would love to hear your thoughts @malini)

[MelissaAutumn]

I'll look at this more in-depth tomorrow, but I believe you can generate a json schema file from a pydantic model https://docs.pydantic.dev/latest/concepts/json_schema/

Dump a copy, and check to see if it differs in any significant way from our current schema.

We can have a github action later on to auto-generate this per model change. (Manual is fine for now.)

[radishmouse]

I'll look at this more in-depth tomorrow, but I believe you can generate a json schema file from a pydantic model https://docs.pydantic.dev/latest/concepts/json_schema/

Dump a copy, and check to see if it differs in any significant way from our current schema.

We can have a github action later on to auto-generate this per model change. (Manual is fine for now.)

Looking at them side by side, the generated one is pretty close to the one submitted for #1. The only thing not included in the generated schema is the "if CTA exists, URL must exist" requirement.

Also, the generated schema doesn't define min/max constraints for "percent_chance", so I'll add a @field_validator to account for that.

[radishmouse]

@MelissaAutumn I've made a few more changes:

• added the field validator for percent_chance
• created a schema generation method
• created a separate python script file for generating a JSON schema
• renamed the conversion script to convert-yaml.py

[radishmouse]

@MelissaAutumn really awesome advice! I think it's ready for another pass.