backend/caching/decorators.py

"""Cache decorators for easy integration with existing functions.

This module provides decorators to automatically cache function results
with configurable TTL strategies and cache key generation.
"""

import functools
import hashlib
import inspect
import json
import logging
from typing import Any, Callable, Optional, TypeVar

from backend.caching.factory import get_cache
from backend.caching.redis_cache import CacheDataType, TTLStrategy

logger = logging.getLogger(__name__)

T = TypeVar("T")


def cached(
    namespace: str = "default",
    data_type: CacheDataType = CacheDataType.ANALYSIS_RESULTS,
    ttl: Optional[int] = None,
    key_prefix: Optional[str] = None,
    include_self: bool = False,
) -> Callable[[Callable[..., T]], Callable[..., T]]:
    """Decorator to cache function results.

    Args:
        namespace: Cache namespace for key generation.
        data_type: Type of data being cached (for TTL strategy).
        ttl: Custom TTL override (seconds).
        key_prefix: Optional prefix for cache key.
        include_self: Include 'self' in cache key for instance methods.

    Returns:
        Decorated function with caching.

    Example:
        >>> @cached(namespace="market", data_type=CacheDataType.MARKET_DATA)
        >>> def get_stock_price(ticker: str) -> float:
        ...     # Expensive API call
        ...     return fetch_price_from_api(ticker)
        >>>
        >>> # First call - fetches from API and caches
        >>> price1 = get_stock_price("AAPL")
        >>>
        >>> # Second call - returns cached value
        >>> price2 = get_stock_price("AAPL")
    """

    def decorator(func: Callable[..., T]) -> Callable[..., T]:
        @functools.wraps(func)
        def wrapper(*args, **kwargs) -> T:
            cache = get_cache()

            # Generate cache key from function arguments
            cache_key = _generate_cache_key(
                namespace,
                func.__name__,
                args,
                kwargs,
                key_prefix=key_prefix,
                include_self=include_self,
            )

            # Try to get from cache
            cached_value = cache.get(cache_key)
            if cached_value is not None:
                logger.debug(f"Cache HIT for {cache_key}")
                return cached_value

            logger.debug(f"Cache MISS for {cache_key}")

            # Call function and cache result
            result = func(*args, **kwargs)

            # Get TTL for this data type
            cache_ttl = ttl if ttl is not None else TTLStrategy.get_ttl(data_type)

            # Cache the result
            cache.set(cache_key, result, ttl=cache_ttl)

            return result

        # Add cache control methods to the wrapper
        wrapper.cache_key = lambda *args, **kwargs: _generate_cache_key(
            namespace,
            func.__name__,
            args,
            kwargs,
            key_prefix=key_prefix,
            include_self=include_self,
        )
        wrapper.invalidate = lambda *args, **kwargs: get_cache().delete(
            wrapper.cache_key(*args, **kwargs)
        )

        return wrapper

    return decorator


def cached_async(
    namespace: str = "default",
    data_type: CacheDataType = CacheDataType.ANALYSIS_RESULTS,
    ttl: Optional[int] = None,
    key_prefix: Optional[str] = None,
    include_self: bool = False,
) -> Callable[[Callable[..., Any]], Callable[..., Any]]:
    """Decorator to cache async function results.

    Args:
        namespace: Cache namespace for key generation.
        data_type: Type of data being cached (for TTL strategy).
        ttl: Custom TTL override (seconds).
        key_prefix: Optional prefix for cache key.
        include_self: Include 'self' in cache key for instance methods.

    Returns:
        Decorated async function with caching.

    Example:
        >>> @cached_async(namespace="portfolio", data_type=CacheDataType.PORTFOLIO_METRICS)
        >>> async def analyse_portfolio(portfolio_id: str) -> dict:
        ...     # Expensive async analysis
        ...     return await perform_analysis(portfolio_id)
        >>>
        >>> # First call - performs analysis and caches
        >>> result1 = await analyse_portfolio("portfolio-123")
        >>>
        >>> # Second call - returns cached value
        >>> result2 = await analyse_portfolio("portfolio-123")
    """

    def decorator(func: Callable[..., Any]) -> Callable[..., Any]:
        @functools.wraps(func)
        async def wrapper(*args, **kwargs) -> Any:
            cache = get_cache()

            # Generate cache key from function arguments
            cache_key = _generate_cache_key(
                namespace,
                func.__name__,
                args,
                kwargs,
                key_prefix=key_prefix,
                include_self=include_self,
            )

            # Try to get from cache
            cached_value = cache.get(cache_key)
            if cached_value is not None:
                logger.debug(f"Cache HIT for {cache_key}")
                return cached_value

            logger.debug(f"Cache MISS for {cache_key}")

            # Call async function and cache result
            result = await func(*args, **kwargs)

            # Get TTL for this data type
            cache_ttl = ttl if ttl is not None else TTLStrategy.get_ttl(data_type)

            # Cache the result
            cache.set(cache_key, result, ttl=cache_ttl)

            return result

        # Add cache control methods to the wrapper
        wrapper.cache_key = lambda *args, **kwargs: _generate_cache_key(
            namespace,
            func.__name__,
            args,
            kwargs,
            key_prefix=key_prefix,
            include_self=include_self,
        )
        wrapper.invalidate = lambda *args, **kwargs: get_cache().delete(
            wrapper.cache_key(*args, **kwargs)
        )

        return wrapper

    return decorator


def cache_invalidate(namespace: str, pattern: str = "*") -> int:
    """Invalidate cache entries matching a pattern.

    Args:
        namespace: Cache namespace.
        pattern: Pattern to match (default: all in namespace).

    Returns:
        Number of entries invalidated.

    Example:
        >>> # Invalidate all portfolio cache entries
        >>> cache_invalidate("portfolio", "portfolio-123:*")
        >>>
        >>> # Invalidate all market data
        >>> cache_invalidate("market")
    """
    cache = get_cache()
    full_pattern = f"{namespace}:{pattern}"
    deleted = cache.delete_pattern(full_pattern)
    logger.info(f"Invalidated {deleted} cache entries matching {full_pattern}")
    return deleted


def _generate_cache_key(
    namespace: str,
    func_name: str,
    args: tuple,
    kwargs: dict,
    key_prefix: Optional[str] = None,
    include_self: bool = False,
) -> str:
    """Generate deterministic cache key from function arguments.

    Args:
        namespace: Cache namespace.
        func_name: Function name.
        args: Positional arguments.
        kwargs: Keyword arguments.
        key_prefix: Optional prefix.
        include_self: Include 'self' parameter for instance methods.

    Returns:
        Generated cache key.
    """
    # Filter out 'self' from args unless explicitly included
    filtered_args = args
    if not include_self and args and hasattr(args[0], "__dict__"):
        # Likely an instance method - skip first arg (self)
        filtered_args = args[1:]

    # Create a deterministic representation of arguments
    key_data = {
        "args": [_serialise_arg(arg) for arg in filtered_args],
        "kwargs": {k: _serialise_arg(v) for k, v in sorted(kwargs.items())},
    }

    # Generate hash of arguments
    key_json = json.dumps(key_data, sort_keys=True)
    args_hash = hashlib.md5(key_json.encode()).hexdigest()[:16]

    # Build cache key
    parts = [namespace, func_name, args_hash]
    if key_prefix:
        parts.insert(0, key_prefix)

    return ":".join(parts)


def _serialise_arg(arg: Any) -> Any:
    """Serialise argument for cache key generation.

    Args:
        arg: Argument to serialise.

    Returns:
        Serialisable representation.
    """
    # Handle common types
    if isinstance(arg, (str, int, float, bool, type(None))):
        return arg

    # Handle lists and tuples
    if isinstance(arg, (list, tuple)):
        return [_serialise_arg(item) for item in arg]

    # Handle dictionaries
    if isinstance(arg, dict):
        return {k: _serialise_arg(v) for k, v in sorted(arg.items())}

    # Handle Pydantic models
    if hasattr(arg, "model_dump"):
        return arg.model_dump()

    # Handle objects with __dict__
    if hasattr(arg, "__dict__"):
        return str(arg.__dict__)

    # Fallback to string representation
    return str(arg)