GenericLoader

What it is

• A small Pydantic model that dynamically imports a Python module, retrieves a named callable (class or function), and instantiates/calls it using a provided configuration dict as **kwargs.

Public API

• class GenericLoader(BaseModel)
  • Fields:
    • python_module: str | None — fully qualified module path to import.
    • module_callable: str | None — name of the callable inside the module.
    • custom_config: Dict[str, Any] | None — keyword arguments to pass to the callable.
  • Methods:
    • load() -> Any
      • Imports python_module, fetches module_callable via getattr, and returns callable_obj(**custom_config).
      • Uses assert to require all three fields are non-None.

Configuration/Dependencies

• Dependencies:
  • Standard library: importlib
  • Third-party: pydantic.BaseModel
• Required configuration (must be set before calling load()):
  • python_module
  • module_callable
  • custom_config

Usage

from naas_abi_core.engine.engine_configuration.EngineConfiguration_GenericLoader import GenericLoader

loader = GenericLoader(
    python_module="collections",
    module_callable="Counter",
    custom_config={"iterable": "abca"},
)

obj = loader.load()
print(obj)  # Counter({'a': 2, 'b': 1, 'c': 1})

Caveats

• Missing python_module, module_callable, or custom_config triggers an AssertionError.
• Import and attribute failures will raise standard Python errors:
  • ModuleNotFoundError (module import fails)
  • AttributeError (callable not found)
  • TypeError (kwargs don’t match the callable signature)
• assert statements can be disabled with Python optimization (-O), which removes these checks.