Module cache 

Query result caching for Thread pipeline

This module provides LRU caching for frequently accessed query results, reducing database round-trips and improving response times.

Features

• Async-first: Built on moka’s async cache for tokio compatibility
• Type-safe: Generic caching with compile-time type checking
• TTL support: Configurable time-to-live for cache entries
• Statistics: Track cache hit/miss rates for monitoring
• Size limits: Automatic eviction when cache exceeds capacity

Usage

use thread_flow::cache::{QueryCache, CacheConfig};
use thread_services::conversion::Fingerprint;

// Create cache with 1000 entry limit, 5 minute TTL
let cache = QueryCache::new(CacheConfig {
    max_capacity: 1000,
    ttl_seconds: 300,
});

// Cache symbol query results
let fingerprint = compute_content_fingerprint("fn main() {}");
cache.insert(fingerprint, symbols).await;

// Retrieve from cache
if let Some(symbols) = cache.get(&fingerprint).await {
    // Cache hit - saved D1 query!
}

Performance Impact

Scenario	Without Cache	With Cache	Savings
Symbol lookup	50-100ms (D1)	<1µs (memory)	99.9%
Metadata query	20-50ms (D1)	<1µs (memory)	99.9%
Re-analysis (90% hit)	100ms total	10ms total	90%

Structs

CacheConfig

Configuration for query result cache

CacheStats

Cache statistics for monitoring

QueryCache

No-op cache for when caching feature is disabled