definitions

agent_cover.instrumentation.definitions

Configuration classes for instrumentation targets.

Classes

TargetConfig dataclass

Represents the configuration for a target to be instrumented.

This dataclass holds information about a specific target, including the module and class name, methods to instrument, and optional parameters and version constraints.

Attributes:

Name	Type	Description
module	str	The module where the target class or function is located.
class_name	str	The name of the class or function to instrument.
methods	Union[Dict[str, str], List[str], None]	Methods to instrument, or None if not applicable. Can be a Dict (e.g., agent strategies) or a List (e.g., method names for prompts/tools). Defaults to None.
params	Dict[str, Any]	Additional parameters specific to the instrumentor (e.g., type="render" for promptflow). Defaults to an empty dictionary.
min_version	Optional[str]	Minimum version of the package required for instrumentation. Defaults to None.
max_version	Optional[str]	Maximum version of the package supported for instrumentation. Defaults to None.

Methods:

Name	Description
from_dict	Creates a TargetConfig instance from a dictionary.

Source code in src/agent_cover/instrumentation/definitions.py

@dataclass
class TargetConfig:
    """Represents the configuration for a target to be instrumented.

    This dataclass holds information about a specific target, including the module
    and class name, methods to instrument, and optional parameters and version
    constraints.

    Attributes:
        module (str): The module where the target class or function is located.
        class_name (str): The name of the class or function to instrument.
        methods (Union[Dict[str, str], List[str], None]): Methods to instrument,
            or None if not applicable. Can be a Dict (e.g., agent strategies)
            or a List (e.g., method names for prompts/tools). Defaults to None.
        params (Dict[str, Any]): Additional parameters specific to the
            instrumentor (e.g., type="render" for promptflow). Defaults to an
            empty dictionary.
        min_version (Optional[str]): Minimum version of the package required for
            instrumentation. Defaults to None.
        max_version (Optional[str]): Maximum version of the package supported
            for instrumentation. Defaults to None.

    Methods:
        from_dict: Creates a TargetConfig instance from a dictionary.
    """

    module: str
    class_name: str
    methods: Union[Dict[str, str], List[str], None] = None
    params: Dict[str, Any] = field(default_factory=dict)
    min_version: Optional[str] = None
    max_version: Optional[str] = None

    @classmethod
    def from_dict(cls, data: Union[Dict[str, Any], "TargetConfig"]) -> "TargetConfig":
        """Creates a TargetConfig instance from a dictionary.

        This class method allows creating a TargetConfig object from a dictionary,
        handling potential type conversions and default values.

        Args:
            data (Dict[str, Any] | TargetConfig): A dictionary containing the target
                configuration data or an existing TargetConfig object.

        Returns:
            TargetConfig: A TargetConfig object initialized from the data.
        """
        if isinstance(data, cls):
            return data

        if not isinstance(data, dict):
            raise ValueError(
                f"Invalid configuration format. Expected dict or TargetConfig, got {type(data)}"
            )

        return cls(
            module=data["module"],
            class_name=data["class_name"],
            methods=data.get("methods"),
            params=data.get("params", {}),
            min_version=data.get("min_version"),
            max_version=data.get("max_version"),
        )

Functions

from_dict(data) classmethod

Creates a TargetConfig instance from a dictionary.

This class method allows creating a TargetConfig object from a dictionary, handling potential type conversions and default values.

Parameters:

Name	Type	Description	Default
data	Dict[str, Any] | TargetConfig	A dictionary containing the target configuration data or an existing TargetConfig object.	required

Returns:

Name	Type	Description
TargetConfig	TargetConfig	A TargetConfig object initialized from the data.

Source code in src/agent_cover/instrumentation/definitions.py

@classmethod
def from_dict(cls, data: Union[Dict[str, Any], "TargetConfig"]) -> "TargetConfig":
    """Creates a TargetConfig instance from a dictionary.

    This class method allows creating a TargetConfig object from a dictionary,
    handling potential type conversions and default values.

    Args:
        data (Dict[str, Any] | TargetConfig): A dictionary containing the target
            configuration data or an existing TargetConfig object.

    Returns:
        TargetConfig: A TargetConfig object initialized from the data.
    """
    if isinstance(data, cls):
        return data

    if not isinstance(data, dict):
        raise ValueError(
            f"Invalid configuration format. Expected dict or TargetConfig, got {type(data)}"
        )

    return cls(
        module=data["module"],
        class_name=data["class_name"],
        methods=data.get("methods"),
        params=data.get("params", {}),
        min_version=data.get("min_version"),
        max_version=data.get("max_version"),
    )