caikit.core.signature_parsing.docstrings
========================================

.. py:module:: caikit.core.signature_parsing.docstrings

.. autoapi-nested-parse::

   This package handles all the gorp of finding types from docstrings given our custom
   conventions



Attributes
----------

.. autoapisummary::

   caikit.core.signature_parsing.docstrings.log


Functions
---------

.. autoapisummary::

   caikit.core.signature_parsing.docstrings.get_return_type
   caikit.core.signature_parsing.docstrings.is_optional
   caikit.core.signature_parsing.docstrings.get_arg_type
   caikit.core.signature_parsing.docstrings._get_candidate_type_names_from_docstring
   caikit.core.signature_parsing.docstrings._get_docstring_type
   caikit.core.signature_parsing.docstrings._extract_nested_type
   caikit.core.signature_parsing.docstrings._extract_type_from_pymodule


Module Contents
---------------

.. py:data:: log

.. py:function:: get_return_type(fn: Callable) -> Optional[Type]

   Grabs the return type off the docstring, if possible
   Args:
       fn: The function to get the return value of
           e.g. my_caikit_library.modules.classification.Transformer.run

   Returns:
       The return type of `fn`, if it can be parsed from the docstring. Otherwise, None


.. py:function:: is_optional(fn: Callable, arg_name: str) -> bool

   Checks if the `arg_name` param from `fn`s docstring is optional
   by checking if param description starts with "an optional"
   or "optional".

   Args:
       fn: The function to get the type of a parameter from
           e.g.  my_caikit_library.modules.classification.Transformer.run
       arg_name: The name of the parameter that we should try to get the type of
           e.g. "raw_document"


.. py:function:: get_arg_type(fn: Callable, arg_name: str) -> Optional[Type]

   Grabs the type of the `arg_name` param from `fn`s docstring, if possible
   Args:
       fn: The function to get the type of a parameter from
           e.g. my_caikit_library.modules.classification.Transformer.run
       arg_name: The name of the parameter that we should try to get the type of
           e.g. "raw_document"
   Returns:
       The return type of `fn`, if it can be parsed from the docstring. Otherwise, None


.. py:function:: _get_candidate_type_names_from_docstring(param: Optional[docstring_parser.common.DocstringParam]) -> Tuple[List[str], List[str]]

.. py:function:: _get_docstring_type(candidate_type_names: List[str]) -> Optional[Type]

   Given a parsed docstring parameter, look in all of the possible places
   for the actual type


.. py:function:: _extract_nested_type(type_name: str) -> Optional[Type]

.. py:function:: _extract_type_from_pymodule(py_module: Union[Type, Dict], type_name: str) -> Optional[Type]

   This walks down a type hierarchy to try to find the concrete type given an input string name

   Args:
       py_module (Type | Dict): A python module, or dictionary of modules, to
           start walking to find "type_name"
       type_name (str): The name of the type that we're trying to find. e.g.
           "caikit.core.data_model.ProducerId"

   Returns:
       Optional[Type]: The type of "type_name", or None if it cannot be found