pyfcach 0.2.14

Best production cache, use async - better than sync

PyFCach - High-Performance Python Caching Library

A comprehensive, multi-strategy caching solution for Python applications with enterprise-grade performance and flexibility.

Features

Performance Optimized

· Optimized cache operations with minimal overhead · Thread-safe implementations with fine-grained locking · Batch operations for improved throughput · Memory-efficient storage with configurable limits

Multiple Eviction Strategies

· LRU (Least Recently Used) - Classic LRU implementation · MRU (Most Recently Used) - For temporal locality patterns · LFU (Least Frequently Used) - Frequency-based eviction · TTL (Time-to-Live) - Automatic expiration with background cleanup · ARC (Adaptive Replacement Cache) - Self-tuning adaptive algorithm

Advanced Capabilities

· Synchronous and asynchronous API support · Memory usage tracking and limits · Performance monitoring and statistics · Decorator-based caching for functions · Cached properties with TTL support · Global cache management and statistics

Performance Benchmarks

Based on comprehensive testing, PyFCach demonstrates exceptional performance:

Core Operations Performance

BASIC TEST — HighPerformanceCache strategy=lru
set() 20,000 items    | 195,501 ops/sec
get() 20,000 items    | 265,736 ops/sec  
delete() 10,000 items | 499,259 ops/sec

BASIC TEST — HighPerformanceCache strategy=mru
set() 20,000 items    | 195,587 ops/sec
get() 20,000 items    | 242,919 ops/sec
delete() 10,000 items | 521,393 ops/sec

BASIC TEST — HighPerformanceCache strategy=lfu  
set() 20,000 items    | 180,350 ops/sec
get() 20,000 items    | 275,304 ops/sec
delete() 10,000 items | 490,736 ops/sec

Advanced Features Performance

BATCH OPERATIONS TEST
batch_set 10,000 items | 59,774 concurrent ops/sec
batch_get 10,000 items | 166,465 concurrent ops/sec

CONCURRENCY STRESS TEST (5 Threads)
concurrent_operations 30,000 items | 28,747 concurrent ops/sec

CACHE DECORATOR PERFORMANCE
First run (CACHE MISSES) 1,000 calls | 17 ops/sec
Second run (CACHE HITS) 1,000 calls  | 155 ops/sec

CACHED PROPERTY TEST
cached_property access (50 gets) | 661 ops/sec

Installation

pip install pyfcach

Quick Start

Basic Usage

from pyfcach import cache, CacheStrategy, get_global_stats

@cache(maxsize=1000, strategy=CacheStrategy.LRU)
def expensive_operation(x: int, y: int) -> int:
    return x * y + expensive_calculation(x, y)

# The result is automatically cached
result1 = expensive_operation(10, 20)  # Computed
result2 = expensive_operation(10, 20)  # From cache

Async Support

from pyfcach import async_cache

@async_cache(maxsize=500, ttl=3600)  # 1 hour TTL
async def fetch_user_data(user_id: int):
    return await database.fetch_user(user_id)

# Usage in async context
user_data = await fetch_user_data(123)

Advanced Configuration

from pyfcach import cache, CacheStrategy

@cache(
    maxsize=1000,
    strategy=CacheStrategy.ARC,  # Adaptive replacement
    memory_limit=100*1024*1024,  # 100MB memory limit
    ttl=300  # 5 minute expiration
)
def memory_intensive_operation(data: bytes):
    return process_large_data(data)

Cache Strategies

LRU (Least Recently Used)

@cache(strategy="lru", maxsize=1000)
def lru_cached_function():
    pass

LFU (Least Frequently Used)

@cache(strategy="lfu", maxsize=1000)
def lfu_cached_function():
    pass

TTL (Time-to-Live)

@cache(strategy="ttl", ttl=60)  # 60 seconds
def ttl_cached_function():
    pass

ARC (Adaptive Replacement)

@cache(strategy="arc", maxsize=1000)
def arc_cached_function():
    pass

Advanced Features

Cached Properties

from pyfcach import cached_property, AsyncCachedProperty

class DataProcessor:
    @cached_property(ttl=300)  # 5 minute cache
    def processed_data(self):
        return expensive_processing()
    
    @AsyncCachedProperty(ttl=600)  # 10 minute cache
    async def async_data(self):
        return await async_expensive_processing()

Batch Operations

from pyfcach import batch_set_operations, batch_get_operations

# Batch setting
keys_values = [("key1", "value1"), ("key2", "value2"), ...]
batch_set_operations(cache_instance, keys_values)

# Batch getting
keys = ["key1", "key2", "key3"]
results = batch_get_operations(cache_instance, keys)

Memory Management

# Set memory limits
@cache(memory_limit=50*1024*1024)  # 50MB limit
def memory_aware_function(data):
    return process_data(data)

Performance Monitoring

# Get cache statistics
cache_info = expensive_operation.cache_info()
print(f"Hit rate: {cache_info.hits / (cache_info.hits + cache_info.misses):.2%}")

# Global statistics
stats = get_global_stats()
print(f"Total caches: {stats['total_caches']}")
print(f"Global hit rate: {stats['global_hit_rate']:.2%}")

API Reference

Core Classes

· CacheBase: Base class for all cache implementations · HighPerformanceCache: General-purpose cache supporting LRU and MRU strategies · TTLCache: Time-based expiration cache with background cleanup · OptimizedLFUCache: Frequency-based eviction cache · ARCCache: Adaptive Replacement Cache - self-tuning between LRU and LFU · AsyncCache: Asynchronous wrapper for all cache strategies

Decorators

@cache

Main caching decorator for synchronous functions.

Parameters:

· maxsize: Maximum number of entries · ttl: Time-to-live in seconds · strategy: Eviction strategy (LRU, LFU, MRU, TTL, ARC) · memory_limit: Maximum memory usage in bytes · typed: Type-aware key generation

@async_cache

Caching decorator for asynchronous functions.

Properties

· cached_property: Cached property descriptor for synchronous methods · AsyncCachedProperty: Cached property descriptor for asynchronous methods

Utility Functions

· cache_clear_all(): Clear all registered caches · get_cache_info(): Get info for all caches · get_global_stats(): Get aggregated statistics · shutdown_all_executors(): Cleanup background executors

Examples

Web Application Caching

from pyfcach import cache, async_cache
from flask import Flask

app = Flask(__name__)

@cache(maxsize=1000, ttl=300)
def get_template_data(template_id):
    return database.get_template(template_id)

@async_cache(maxsize=500, strategy="lfu")
async def get_user_session(user_id):
    return await redis.get(f"session:{user_id}")

Data Processing Pipeline

from pyfcach import cache, cached_property

class DataAnalyzer:
    def __init__(self, dataset):
        self.dataset = dataset
    
    @cached_property(ttl=3600)
    def statistical_summary(self):
        return self._compute_summary()
    
    @cache(maxsize=100, strategy="arc")
    def correlation_matrix(self, columns):
        return self._compute_correlation(columns)

Real-time Analytics

from pyfcach import async_cache, AsyncCachedProperty

class AnalyticsService:
    @AsyncCachedProperty(ttl=60)
    async def real_time_metrics(self):
        return await self.fetch_live_metrics()
    
    @async_cache(maxsize=1000, strategy="ttl", ttl=30)
    async def get_user_analytics(self, user_id, time_range):
        return await self.compute_analytics(user_id, time_range)

Performance Characteristics

Based on benchmark results:

· High Throughput: Up to 500,000+ operations per second for basic operations · Excellent Concurrency: 18,000+ concurrent operations with 5 threads · Efficient Batching: 166,000+ operations per second for batch gets · Low Latency: Sub-millisecond response times for cache hits · Memory Efficient: Configurable memory limits and efficient storage

Performance Tips

1. Strategy Selection: Use ARC for mixed access patterns, TTL for time-sensitive data
2. Size Configuration: Balance memory usage and hit rates with appropriate maxsize
3. Batch Operations: Use batch methods for bulk operations to maximize throughput
4. Memory Management: Set memory limits when caching large objects
5. Monitoring: Regularly check cache statistics to optimize configuration

License

Apache-2.0