README.md

Cachet Tier

Core cache tier abstractions for building cache backends.

This crate defines the CacheTier trait that all cache implementations must satisfy, along with CacheEntry for storing values with metadata and Error types for fallible operations.

Overview

The cache tier abstraction separates storage concerns from caching features. Implement CacheTier for your storage backend, then use cachet to add telemetry, TTL, multi-tier fallback, and other features on top.

Implementing a Cache Tier

Implement all required methods of CacheTier:

use std::collections::HashMap;
use std::sync::RwLock;

use cachet_tier::{CacheEntry, CacheTier, Error};

struct SimpleCache<K, V>(RwLock<HashMap<K, CacheEntry<V>>>);

impl<K, V> CacheTier<K, V> for SimpleCache<K, V>
where
    K: Clone + Eq + std::hash::Hash + Send + Sync,
    V: Clone + Send + Sync,
{
    async fn get(&self, key: &K) -> Result<Option<CacheEntry<V>>, Error> {
        Ok(self.0.read().unwrap().get(key).cloned())
    }

    async fn insert(&self, key: K, entry: CacheEntry<V>) -> Result<(), Error> {
        self.0.write().unwrap().insert(key.clone(), entry);
        Ok(())
    }

    async fn invalidate(&self, key: &K) -> Result<(), Error> {
        self.0.write().unwrap().remove(key);
        Ok(())
    }

    async fn clear(&self) -> Result<(), Error> {
        self.0.write().unwrap().clear();
        Ok(())
    }
}

Dynamic Dispatch

DynamicCache wraps any CacheTier in a type-erased container. This is useful for multi-tier caches with heterogeneous storage backends.

---

_{This crate was developed as part of The Oxidizer Project. Browse this crate's source code.}