yera 0.1.0__py3-none-any.whl → 0.2.0__py3-none-any.whl

yera/agents/discovery.py

@@ -0,0 +1,124 @@
+"""Utilities for loading agents from files."""
+
+from __future__ import annotations
+
+import importlib.util
+import logging
+from pathlib import Path
+
+from yera.agents.decorator import AgentFunctionWrapper
+from yera.utils.path_utils import path_to_module_name
+
+
+def load_agent(
+    file_path: Path | str, agent_identifier: str | None = None
+) -> AgentFunctionWrapper:
+    """Load a Python file and extract the AgentFunctionWrapper instance with the given identifier.
+
+    Args:
+        file_path: Path to the Python file containing one or more agents.
+            Can be a Path object or string.
+        agent_identifier: Identifier of the agent to load. If None, the first agent is loaded.
+
+    Returns:
+        AgentFunctionWrapper instance.
+
+    Raises:
+        ValueError: If no agents are found in the file.
+        ValueError: If multiple agents are found in the file and no identifier is provided.
+        ValueError: If the specified agent identifier is not found in the file.
+        ValueError: If the specified agent identifier does not match the identifier of the loaded agent.
+    """
+    agents = load_agents(file_path)
+
+    if not agents:
+        raise ValueError(f"No agents found in {file_path}")
+
+    if agent_identifier is None:
+        if len(agents) > 1:
+            raise ValueError(
+                f"Multiple agents found in {file_path}, please specify an agent identifier"
+            )
+        return agents[0]
+
+    matching_agents = [
+        agent for agent in agents if agent.metadata.identifier == agent_identifier
+    ]
+    if not matching_agents:
+        raise ValueError(f"Agent with ID {agent_identifier} not found in {file_path}")
+    if len(matching_agents) > 1:
+        raise ValueError(
+            f"Multiple agents with ID {agent_identifier} found in {file_path}"
+        )
+    return matching_agents[0]
+
+
+def load_agents(file_path: Path | str) -> list[AgentFunctionWrapper]:
+    """Load a Python file and extract all AgentFunctionWrapper instances.
+
+    This helper looks for objects in the module that are instances of
+    AgentFunctionWrapper (created by the @yr.agent decorator).
+
+    Args:
+        file_path: Path to the Python file containing one or more agents.
+            Can be a Path object or string.
+
+    Returns:
+        Dictionary mapping agent identifiers to AgentFunctionWrapper instances.
+
+    Raises:
+        FileNotFoundError: If the file does not exist.
+    """
+    path = Path(file_path) if isinstance(file_path, str) else file_path
+    if not path.exists():
+        raise FileNotFoundError(f"Python file not found: {path}")
+
+    # Load the module from the given file path
+    module_name = path_to_module_name(path)
+    spec = importlib.util.spec_from_file_location(module_name, path)
+    if spec is None or spec.loader is None:
+        raise FileNotFoundError(f"Could not load Python module from: {path}")
+
+    module = importlib.util.module_from_spec(spec)
+    # Exec here is fine because we are only allowing users to run via the CLI locally,
+    # so there is no risk of arbitrary remote code execution.
+    # Deployment will require tracing, which will NOT exec the python directly.
+    spec.loader.exec_module(module)
+
+    # Collect all AgentFunctionWrapper instances defined in the module
+    return [
+        value
+        for value in vars(module).values()
+        if isinstance(value, AgentFunctionWrapper)
+    ]
+
+
+def discover_agents(directory: Path | str) -> list[AgentFunctionWrapper]:
+    """Discover all agents in a directory.
+
+    Scans the directory for .py files and loads all agents from each file.
+
+    Args:
+        directory: Directory to scan for agent files.
+
+    Returns:
+        List of AgentFunctionWrapper instances.
+
+    Raises:
+        FileNotFoundError: If the directory does not exist.
+    """
+    dir_path = Path(directory) if isinstance(directory, str) else directory
+    if not dir_path.exists():
+        raise FileNotFoundError(f"Directory not found: {dir_path}")
+    if not dir_path.is_dir():
+        raise ValueError(f"Path is not a directory: {dir_path}")
+
+    all_agents = []
+    for py_file in dir_path.rglob("*.py"):
+        try:
+            all_agents.extend(load_agents(py_file))
+        # Ignore PERF203 because I/O file loading is the real bottleneck.
+        except FileNotFoundError:  # noqa: PERF203
+            logging.warning(f"Failed to load agents from {py_file}")
+            continue
+    return all_agents

yera/agents/typing/coerce.py

@@ -0,0 +1,408 @@
+import io
+import json
+from datetime import date, datetime, time, timedelta
+from decimal import Decimal, InvalidOperation
+from enum import Enum
+from types import UnionType
+from typing import TYPE_CHECKING, Any, Literal, Union, get_args, get_origin
+
+if TYPE_CHECKING:
+    import pandas as pd
+
+from yera.agents.typing.utils import get_type_name
+from yera.agents.typing.validate import is_enum_type, is_struct_type
+from yera.dsl.struct import Struct
+
+from .validate import is_dataframe_type
+
+
+def coerce_set(value: Any, expected_type: Any) -> set:
+    if isinstance(value, str):
+        try:
+            value = json.loads(value)
+        except json.JSONDecodeError as e:
+            raise TypeError(f"Cannot parse set from JSON string: {e}") from None
+
+    if isinstance(value, set | list):
+        items = value
+    else:
+        raise TypeError(f"Expected set or list, got {type(value).__name__}")
+
+    type_args = get_args(expected_type)
+    if not type_args:
+        raise TypeError(
+            "Set type must specify element type (e.g., set[int], not plain set)"
+        )
+
+    element_type = type_args[0]
+    return {coerce_value(item, element_type) for i, item in enumerate(items)}
+
+
+def coerce_datetime(value: Any) -> datetime:
+    """Coerce to datetime."""
+    if isinstance(value, datetime):
+        return value
+    if isinstance(value, str):
+        # Try ISO format first
+        try:
+            return datetime.fromisoformat(value.replace("Z", "+00:00"))
+        except ValueError:
+            # Fallback to common formats
+            for fmt in [
+                "%Y-%m-%d %H:%M:%S",
+                "%Y-%m-%dT%H:%M:%S",
+                "%Y-%m-%d %H:%M:%S.%f",
+                "%Y-%m-%dT%H:%M:%S.%f",
+            ]:
+                try:
+                    return datetime.strptime(value, fmt)
+                except ValueError:  # noqa: PERF203
+                    continue
+            raise TypeError(f"Cannot parse datetime from string: {value!r}") from None
+    elif isinstance(value, int | float):
+        # Unix timestamp
+        try:
+            return datetime.fromtimestamp(value)
+        except (ValueError, OSError) as e:
+            raise TypeError(
+                f"Cannot convert timestamp {value} to datetime: {e}"
+            ) from None
+    else:
+        raise TypeError(f"Cannot convert {type(value).__name__} to datetime")
+
+
+def coerce_date(value: Any) -> date:
+    """Coerce to date."""
+    if isinstance(value, date) and not isinstance(value, datetime):
+        return value
+    if isinstance(value, datetime):
+        return value.date()
+    if isinstance(value, str):
+        try:
+            return date.fromisoformat(value)
+        except ValueError:
+            # Try parsing as datetime and extract date
+            try:
+                dt = datetime.fromisoformat(value.replace("Z", "+00:00"))
+                return dt.date()
+            except ValueError:
+                raise TypeError(f"Cannot parse date from string: {value!r}") from None
+    else:
+        raise TypeError(f"Cannot convert {type(value).__name__} to date")
+
+
+def coerce_time(value: Any) -> time:
+    """Coerce to time."""
+    if isinstance(value, time):
+        return value
+    if isinstance(value, str):
+        try:
+            return time.fromisoformat(value)
+        except ValueError:
+            raise TypeError(
+                f"Cannot convert {type(value).__name__} to time. "
+                f"Expected ISO format time."
+            ) from None
+    else:
+        raise TypeError(f"Cannot convert {type(value).__name__} to time")
+
+
+def coerce_timedelta(value: Any) -> timedelta:
+    """Coerce to timedelta."""
+    if isinstance(value, timedelta):
+        return value
+    if isinstance(value, str):
+        try:
+            return timedelta(seconds=float(value))
+        except ValueError:
+            raise TypeError(
+                f"Cannot convert timedelta from invalid string: {value!r}. "
+                f"Expected a float for number of seconds."
+            ) from None
+    else:
+        raise TypeError(f"Cannot convert {type(value).__name__} to timedelta")
+
+
+def coerce_decimal(value: Any) -> Decimal:
+    """Coerce to Decimal."""
+    if isinstance(value, Decimal):
+        return value
+    if isinstance(value, int | float | str):
+        try:
+            return Decimal(str(value))
+        except (ValueError, InvalidOperation) as e:
+            raise TypeError(f"Cannot convert {value!r} to Decimal: {e}") from None
+    else:
+        raise TypeError(f"Cannot convert {type(value).__name__} to Decimal")
+
+
+def coerce_enum(value: Any, expected_type: type[Enum]) -> Enum:
+    """Coerce to Enum type."""
+    if isinstance(value, expected_type):
+        return value
+    if isinstance(value, str):
+        # Try by name first
+        try:
+            return expected_type[value]
+        except KeyError:
+            valid_names = [e.name for e in expected_type]
+            raise TypeError(
+                f"'{value}' string value is not a valid {expected_type.__name__}. "
+                f"Valid names: {valid_names}."
+            ) from None
+    elif isinstance(value, int):
+        try:
+            return expected_type(value)
+        except ValueError:
+            valid_values = [e.value for e in expected_type]
+            raise TypeError(
+                f"{value} is not a valid {expected_type.__name__} value. "
+                f"Valid values: {valid_values}"
+            ) from None
+    else:
+        raise TypeError(
+            f"Cannot convert {type(value).__name__} to {expected_type.__name__}. "
+            f"Expected str or int."
+        )
+
+
+def serialise_dataframe(df: "pd.DataFrame") -> bytes:
+    """Serialize DataFrame to Parquet bytes."""
+    buffer = io.BytesIO()
+    df.to_parquet(buffer, engine="pyarrow")
+    return buffer.getvalue()
+
+
+def coerce_dataframe(value: Any) -> "pd.DataFrame":
+    """Coerce to pandas DataFrame."""
+    import pandas as pd
+
+    if isinstance(value, pd.DataFrame):
+        return value
+    if isinstance(value, bytes):
+        return pd.read_parquet(io.BytesIO(value), engine="pyarrow")
+    raise TypeError(
+        f"Cannot convert {type(value).__name__} to DataFrame. "
+        f"Must be a pandas DataFrame or parquet bytes."
+    )
+
+
+def coerce_none(value: Any) -> None:
+    """Coerce to None type."""
+    if value is None or value == "None":
+        return
+    raise TypeError(f"Expected None, got {type(value).__name__}")
+
+
+def coerce_bool(value: Any) -> bool:
+    """Coerce to bool with proper string handling."""
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        if value.lower() == "true":
+            return True
+        if value.lower() == "false":
+            return False
+        raise TypeError(
+            f"Cannot convert {value!r} to bool. "
+            f"Expected one of: true, false, 1, 0, yes, no, y, n, on, off (case-insensitive)"
+        )
+    if isinstance(value, int):
+        return bool(value)
+    raise TypeError(f"Cannot convert {type(value).__name__} to bool")
+
+
+def coerce_str(value: Any) -> str:
+    """Coerce to str with proper string handling."""
+    try:
+        return str(value)
+    except (ValueError, TypeError) as e:
+        raise TypeError(f"Cannot convert {value!r} to str. ") from None
+
+
+def coerce_int(value: Any) -> int:
+    """Coerce to int with proper string handling."""
+    try:
+        return int(value)
+    except (ValueError, TypeError) as e:
+        raise TypeError(f"Cannot convert {value!r} to int") from None
+
+
+def coerce_float(value: Any) -> float:
+    """Coerce to float with proper string handling."""
+    try:
+        return float(value)
+    except (ValueError, TypeError) as e:
+        raise TypeError(f"Cannot convert {value!r} to float") from None
+
+
+def coerce_literal(value: Any, expected_type: Any) -> Any:
+    """Coerce to Literal type."""
+    allowed_values = get_args(expected_type)
+
+    # First try exact match
+    if value in allowed_values:
+        return value
+
+    # Try coercing to match one of the literal values
+    for literal_value in allowed_values:
+        literal_type = type(literal_value)
+
+        # Skip None in literals for coercion attempts
+        if literal_value is None:
+            continue
+
+        try:
+            # Try to coerce the input value to the type of this literal
+            coerced = coerce_value(value, literal_type)
+            if coerced == literal_value:
+                return literal_value
+        except (TypeError, ValueError):
+            continue
+
+    # If no match found, raise error with helpful message
+    raise TypeError(
+        f"Value {value!r} is not one of the allowed literal values: {allowed_values}"
+    )
+
+
+def coerce_struct(value: Any, expected_type: type[Struct]) -> Any:
+    """Coerce to Struct (Pydantic model) type."""
+    if isinstance(value, expected_type):
+        return value
+    if isinstance(value, dict):
+        return expected_type.model_validate(value)
+    if isinstance(value, str):
+        return expected_type.model_validate_json(value)
+    raise TypeError(
+        f"Cannot convert {type(value).__name__} to {expected_type.__name__}. "
+        f"Expected dict or JSON string."
+    )
+
+
+def coerce_union(value: Any, expected_type: Any) -> Any:
+    """Coerce to Union type by trying each type in order."""
+    type_args = get_args(expected_type)
+    last_error = None
+
+    for arg_type in type_args:
+        try:
+            return coerce_value(value, arg_type)
+        except (TypeError, ValueError) as e:  # noqa: PERF203
+            last_error = e
+            continue
+
+    raise TypeError(
+        f"Cannot coerce value to any of {type_args}. Last error: {last_error}"
+    )
+
+
+def coerce_list(value: Any, expected_type: Any) -> list:
+    """Coerce to list[T] type."""
+    if isinstance(value, str):
+        try:
+            value = json.loads(value)
+        except json.JSONDecodeError as e:
+            raise TypeError(f"Cannot parse list from JSON string: {e}") from None
+
+    if not isinstance(value, list):
+        raise TypeError(f"Expected list, got {type(value).__name__}")
+
+    type_args = get_args(expected_type)
+    if not type_args:
+        raise TypeError(
+            "List type must specify element type (e.g., list[int], not plain list)"
+        )
+
+    element_type = type_args[0]
+    return [coerce_value(item, element_type) for i, item in enumerate(value)]
+
+
+def coerce_dict(value: Any, expected_type: Any) -> dict:
+    """Coerce to dict[K, V] type."""
+    # Handle string input (JSON)
+    if isinstance(value, str):
+        try:
+            value = json.loads(value)
+        except json.JSONDecodeError as e:
+            raise TypeError(f"Cannot parse dict from JSON string: {e}") from None
+
+    if not isinstance(value, dict):
+        raise TypeError(f"Expected dict, got {type(value).__name__}")
+
+    type_args = get_args(expected_type)
+    if not type_args:
+        raise TypeError(
+            "Dict type must specify key and value types (e.g., dict[str, int], "
+            "not plain dict)"
+        )
+
+    key_type, value_type = type_args[0], type_args[1]
+    return {
+        coerce_value(k, key_type): coerce_value(v, value_type) for k, v in value.items()
+    }
+
+
+def coerce_fallback(value: Any, expected_type: Any) -> Any:
+    """Fallback coercion - just check isinstance."""
+    raise TypeError(
+        f"Unsupported type {get_type_name(expected_type)} for coercion of value "
+        f"{value}. Type must be validated as allowed before coercion."
+    )
+
+
+PARAMETERISED_TYPES = (list, set, dict, Literal, Union, UnionType)
+
+
+def is_generic(t: type) -> bool:
+    return t in PARAMETERISED_TYPES
+
+
+TYPE_COERCERS = {
+    bool: coerce_bool,
+    int: coerce_int,
+    float: coerce_float,
+    str: coerce_str,
+    type(None): coerce_none,
+    datetime: coerce_datetime,
+    date: coerce_date,
+    time: coerce_time,
+    timedelta: coerce_timedelta,
+    Decimal: coerce_decimal,
+    list: coerce_list,
+    dict: coerce_dict,
+    set: coerce_set,
+    Literal: coerce_literal,
+    Union: coerce_union,
+    UnionType: coerce_union,
+}
+
+
+def coerce_value(value: Any, expected_type: Any) -> Any:
+    """Coerce a value to the expected type.
+
+    Delegates to specialized handlers based on type structure.
+    """
+    if expected_type is None:
+        return coerce_none(value)
+
+    origin = get_origin(expected_type)
+
+    if is_struct_type(expected_type):
+        return coerce_struct(value, expected_type)
+
+    if is_enum_type(expected_type):
+        return coerce_enum(value, expected_type)
+
+    if is_dataframe_type(expected_type):
+        return coerce_dataframe(value)
+
+    handler = TYPE_COERCERS.get(origin or expected_type)
+    if handler is None:
+        return coerce_fallback(value, expected_type)
+
+    if is_generic(origin or expected_type):
+        # noinspection PyArgumentList
+        return handler(value, expected_type)
+
+    return handler(value)

yera/agents/typing/utils.py

@@ -0,0 +1,19 @@
+from typing import get_args, get_origin
+
+
+def get_type_name(type_hint):
+    if type_hint is None or type_hint is type(None):
+        return "None"
+
+    origin = get_origin(type_hint)
+    if origin is not None:
+        args = get_args(type_hint)
+        if args:
+            arg_names = ", ".join(get_type_name(arg) for arg in args)
+            return f"{origin.__name__}[{arg_names}]"
+        return origin.__name__
+
+    if hasattr(type_hint, "__name__"):
+        return type_hint.__name__
+
+    return str(type_hint)