I implemented a LRU-cache in C# and wanted to hear suggestions in making it more readable and maybe even more compact.
public class LRUCache<K, V>
{
public int Capacity { get; }
Dictionary<K, V> cache;
K[] keyRingBuffer;
int ringBufferIndex;
int LRUIndex => (ringBufferIndex + 1) % Capacity;
K LRU => keyRingBuffer[LRUIndex];
public LRUCache(int capacity)
{
Capacity = capacity;
cache = new Dictionary<K, V>(Capacity);
keyRingBuffer = new K[Capacity];
ringBufferIndex = 0;
}
public bool Contains(K key) => cache.ContainsKey(key);
public int Size => cache.Count;
public V Get(K key)
{
if (cache.TryGetValue(key, out var value))
{
keyRingBuffer[LRUIndex] = key;
ringBufferIndex = LRUIndex;
return value;
}
throw new Exception($"Element for key {key} not in cache");
}
public void Add(K key, V val)
{
if (cache.Count >= Capacity) cache.Remove(LRU);
ringBufferIndex = LRUIndex;
keyRingBuffer[ringBufferIndex] = key;
cache.Add(key, val);
}
}
I would love it to be more like -- how Robert Martin says -- "well written prose".
