Files
node/cache/lru/sized_cache.go
zeekay 7cf87b230b node-unified is the node
The module already called itself github.com/luxfi/node while containing no node
in Go: cmd/luxd was twenty-one lines that printed the size of a few types, and
the Go column of the matrix measured a VM plugin. A repo that fetches the thing
it claims to be has not replaced it, so the Go sources move in — 1515 files,
the daemon and everything under it — and the build reaches outside this module
for nothing.

Rust and C++ sit beside it unchanged and are still measured against it by the
differential. That is the point of one repo: the reference implementation and
the two ports are the same checkout, so parity is a thing you can run rather
than a thing you assert.

Verified: builds, and the closure carries no lux-private — checked at build time
rather than believed. The daemon starts twelve chains on a local network
(P X C Q Z A B G K M F, plus D through its plugin) and reports healthy.
2026-09-07 16:15:25 -07:00

141 lines
3.0 KiB
Go

// Copyright (C) 2019-2025, Lux Industries Inc. All rights reserved.
// See the file LICENSE for licensing terms.
package lru
import (
"sync"
"github.com/luxfi/container/linked"
"github.com/luxfi/node/cache"
"github.com/luxfi/util"
)
var _ cache.Cacher[struct{}, any] = (*SizedCache[struct{}, any])(nil)
// sizedElement is used to store the element with its size, so we don't
// calculate the size multiple times.
//
// This ensures that any inconsistencies returned by the size function can not
// corrupt the cache.
type sizedElement[V any] struct {
value V
size int
}
// SizedCache is a key value store with bounded size. If the size is attempted
// to be exceeded, then elements are removed from the cache until the bound is
// honored, based on evicting the least recently used value.
type SizedCache[K comparable, V any] struct {
lock sync.Mutex
elements *linked.Hashmap[K, *sizedElement[V]]
maxSize int
currentSize int
size func(K, V) int
}
func NewSizedCache[K comparable, V any](maxSize int, size func(K, V) int) *SizedCache[K, V] {
return &SizedCache[K, V]{
elements: linked.NewHashmap[K, *sizedElement[V]](),
maxSize: maxSize,
size: size,
}
}
func (c *SizedCache[K, V]) Put(key K, value V) {
c.lock.Lock()
defer c.lock.Unlock()
c.put(key, value)
}
func (c *SizedCache[K, V]) Get(key K) (V, bool) {
c.lock.Lock()
defer c.lock.Unlock()
return c.get(key)
}
func (c *SizedCache[K, V]) Evict(key K) {
c.lock.Lock()
defer c.lock.Unlock()
c.evict(key)
}
func (c *SizedCache[K, V]) Flush() {
c.lock.Lock()
defer c.lock.Unlock()
c.flush()
}
func (c *SizedCache[_, _]) Len() int {
c.lock.Lock()
defer c.lock.Unlock()
return c.len()
}
func (c *SizedCache[_, _]) PortionFilled() float64 {
c.lock.Lock()
defer c.lock.Unlock()
return c.portionFilled()
}
func (c *SizedCache[K, V]) put(key K, value V) {
newEntrySize := c.size(key, value)
if newEntrySize > c.maxSize {
c.flush()
return
}
if oldElement, ok := c.elements.Get(key); ok {
c.currentSize -= oldElement.size
}
// Remove elements until the size of elements in the cache <= [c.maxSize].
for c.currentSize > c.maxSize-newEntrySize {
oldestKey, oldestElement, _ := c.elements.Oldest()
c.elements.Delete(oldestKey)
c.currentSize -= oldestElement.size
}
c.elements.Put(key, &sizedElement[V]{
value: value,
size: newEntrySize,
})
c.currentSize += newEntrySize
}
func (c *SizedCache[K, V]) get(key K) (V, bool) {
element, ok := c.elements.Get(key)
if !ok {
return utils.Zero[V](), false
}
c.elements.Put(key, element) // Mark [k] as MRU.
return element.value, true
}
func (c *SizedCache[K, _]) evict(key K) {
if element, ok := c.elements.Get(key); ok {
c.elements.Delete(key)
c.currentSize -= element.size
}
}
func (c *SizedCache[K, V]) flush() {
c.elements.Clear()
c.currentSize = 0
}
func (c *SizedCache[_, _]) len() int {
return c.elements.Len()
}
func (c *SizedCache[_, _]) portionFilled() float64 {
return float64(c.currentSize) / float64(c.maxSize)
}