changing predefined recognizers to use the config file

[RoeyBC]

Change Description

This PR is going to accomplish the following:

1. Refactor load_predefined_recognizers to load the settings from the default yml file
2. Add warning in case recognizer is listed with a language not supported by the registry
3. raise an exception in case there's discrepancy between the supported language in the registry and the one listed in the analyzer
4. Edit the docs accordingly I guess :D

Checklist

• I have reviewed the contribution guidelines
• I have signed the CLA (if required)
• My code includes unit tests
• All unit tests and lint checks pass locally
• My PR contains documentation updates / additions if required

[RoeyBC]

when refactored the codebase, added documentation & testing and you've got less text, you know it's a good change :D

[omri374]

I know this is working, but it feels like a weakspot. Is it complex to extract this to an external module that both this class and RecognizerRegistryProvoder uses?

[omri374]

Worth checking for duplications. Isn't the nlp provider added twice?

[RoeyBC]

recognizer_registry doesn't add the nlp recognizer, so there's no duplication. In fact when the analyzer get initialized without a registry we do the following:

        if not registry:
            logger.info("registry not provided, creating default.")
            provider = RecognizerRegistryProvider(
                registry_configuration={"supported_languages": self.supported_languages}
            )
            registry = provider.create_recognizer_registry()
            registry.add_nlp_recognizer(nlp_engine=self.nlp_engine)

Which is basically this function now :D

[omri374]

Wouldn't it add it if it's in the yaml?