Skip to main content

Standardization

The Standardization Framework is a set of tools and guidelines that enable the user to query and obtain data in a consistent way across multiple providers.

Provider models can inherit from an already defined standard model. The standard models are also the metamodel names defined in the Router's function definition, and the Provider's fetcher_dict

Standardizing provider query parameters and response data enhances the user experience by overcoming things like:

  • Transparently defined schemas for the data and query parameters.
  • Outputs from multiple sources are comparable with each other and easily interchanged.
  • Consistent query parameters across all data sources for a function, or type of function.
    • Use symbol instead of ticker.
  • Output data that has conformed types, is validated, and will be JSON serializable.
    • NaN, NaT, "None", empty strings, should be NoneType (null).
    • Percent values should be represented as decimal points.
  • Parameter and fields are always lower_snake_case.

The standard models here are the contents of the repository code - /OpenBB/openbb_platform/core/openbb_core/provider/standard_models/ directory.

Every standard model consists of two classes, with each being a Pydantic model.

Any parameter or field can be assigned a custom field_validator, or the entire model can be passed through a model_validator on creation. See the page on validators for examples.

info
  • Function signatures prioritize the standard models, provider-specific implementations become extra_params.
  • Field aliases should be implemented at the model level - i.e., use __alias_dict__ = {"field": "mapped_field"}
  • Similarly, json_schema_extra, should be applied at the model level, __json_schema_extra__.
  • CamelCase to snake_case is handled automatically.