文档章节

LruCache的理解

新年
 新年
发布于 2015/02/15 10:33
字数 1639
阅读 36
收藏 2

package android.util;  
 
import java.util.LinkedHashMap;  
import java.util.Map;  
 
/**
 * Cache保存一个强引用来限制内容数量,每当Item被访问的时候,此Item就会移动到队列的头部。
 * 当cache已满的时候加入新的item时,在队列尾部的item会被回收。
 * 如果你cache的某个值需要明确释放,重写entryRemoved()
 * 如果key相对应的item丢掉啦,重写create().这简化了调用代码,即使丢失了也总会返回。
 * 默认cache大小是测量的item的数量,重写sizeof计算不同item的
 *  大小。
 * <pre>   {@code  
 *   int cacheSize = 4 * 1024 * 1024; // 4MiB
 *   LruCache<String, Bitmap> bitmapCache = new LruCache<String, Bitmap>(cacheSize) {
 *       protected int sizeOf(String key, Bitmap value) {
 *           return value.getByteCount();
 *       }
 *   }}</pre>
 *
 * <p>This class is thread-safe. Perform multiple cache operations atomically by
 * synchronizing on the cache: <pre>   {@code  
 *   synchronized (cache) {
 *     if (cache.get(key) == null) {
 *         cache.put(key, value);
 *     }
 *   }}</pre>
 *
 * <p>This class does not allow null to be used as a key or value. A return
 * value of null from {@link  #get}, {@link  #put} or {@link  #remove} is
 * unambiguous: the key was not in the cache.
 * 不允许key或者value为null
 *  当get(),put(),remove()返回值为null时,key相应的项不在cache中
 */  
public class LruCache<K, V> {  
    private final LinkedHashMap<K, V> map;  
 
    /** Size of this cache in units. Not necessarily the number of elements. */  
    private int size; //已经存储的大小
    private int maxSize; //规定的最大存储空间
 
    private int putCount;  //put的次数
    private int createCount;  //create的次数
    private int evictionCount;  //回收的次数
    private int hitCount;  //命中的次数
    private int missCount;  //丢失的次数
 
    /**
     * @param maxSize for caches that do not override {@link #sizeOf}, this is
     *     the maximum number of entries in the cache. For all other caches,
     *     this is the maximum sum of the sizes of the entries in this cache.
     */  
    public LruCache(int maxSize) {  
        if (maxSize <= 0) {  
            throw new IllegalArgumentException("maxSize <= 0");  
        }  
        this.maxSize = maxSize;  
        this.map = new LinkedHashMap<K, V>(0, 0.75f, true);  
    }  
 
    /**
     * Returns the value for {@code key} if it exists in the cache or can be
     * created by {@code #create}. If a value was returned, it is moved to the
     * head of the queue. This returns null if a value is not cached and cannot
     * be created. 通过key返回相应的item,或者创建返回相应的item。相应的item会移动到队列的头部,
     * 如果item的value没有被cache或者不能被创建,则返回null。
     */  
    public final V get(K key) {  
        if (key == null) {  
            throw new NullPointerException("key == null");  
        }  
 
        V mapValue;  
        synchronized (this) {  
            mapValue = map.get(key);  
            if (mapValue != null) {  
                hitCount++;  //命中
                return mapValue;  
            }  
            missCount++;  //丢失
        }  
 
        /*
         * Attempt to create a value. This may take a long time, and the map
         * may be different when create() returns. If a conflicting value was
         * added to the map while create() was working, we leave that value in
         * the map and release the created value.
         * 如果丢失了就试图创建一个item
         */  
 
        V createdValue = create(key);  
        if (createdValue == null) {  
            return null;  
        }  
 
        synchronized (this) {  
            createCount++;//创建++  
            mapValue = map.put(key, createdValue);  
 
            if (mapValue != null) {  
                // There was a conflict so undo that last put  
                //如果前面存在oldValue,那么撤销put()
                map.put(key, mapValue);  
            } else {  
                size += safeSizeOf(key, createdValue);  
            }  
        }  
 
        if (mapValue != null) {  
            entryRemoved(false, key, createdValue, mapValue);  
            return mapValue;  
        } else {  
            trimToSize(maxSize);  
            return createdValue;  
        }  
    }  
 
    /**
     * Caches {@code value} for {@code key}. The value is moved to the head of
     * the queue.
     *
     * @return the previous value mapped by {@code key}.
     */  
    public final V put(K key, V value) {  
        if (key == null || value == null) {  
            throw new NullPointerException("key == null || value == null");  
        }  
 
        V previous;  
        synchronized (this) {  
            putCount++;  
            size += safeSizeOf(key, value);  
            previous = map.put(key, value);  
            if (previous != null) {  //返回的先前的value值
                size -= safeSizeOf(key, previous);  
            }  
        }  
 
        if (previous != null) {  
            entryRemoved(false, key, previous, value);  
        }  
 
        trimToSize(maxSize);  
        return previous;  
    }  
 
    /**
     * @param maxSize the maximum size of the cache before returning. May be -1
     *     to evict even 0-sized elements.
     *  清空cache空间
     */  
    private void trimToSize(int maxSize) {  
        while (true) {  
            K key;  
            V value;  
            synchronized (this) {  
                if (size < 0 || (map.isEmpty() && size != 0)) {  
                    throw new IllegalStateException(getClass().getName()  
                            + ".sizeOf() is reporting inconsistent results!");  
                }  
 
                if (size <= maxSize) {  
                    break;  
                }  
 
                Map.Entry<K, V> toEvict = map.eldest();  
                if (toEvict == null) {  
                    break;  
                }  
 
                key = toEvict.getKey();  
                value = toEvict.getValue();  
                map.remove(key);  
                size -= safeSizeOf(key, value);  
                evictionCount++;  
            }  
 
            entryRemoved(true, key, value, null);  
        }  
    }  
 
    /**
     * Removes the entry for {@code key} if it exists.
     * 删除key相应的cache项,返回相应的value
     * @return the previous value mapped by {@code key}.
     */  
    public final V remove(K key) {  
        if (key == null) {  
            throw new NullPointerException("key == null");  
        }  
 
        V previous;  
        synchronized (this) {  
            previous = map.remove(key);  
            if (previous != null) {  
                size -= safeSizeOf(key, previous);  
            }  
        }  
 
        if (previous != null) {  
            entryRemoved(false, key, previous, null);  
        }  
 
        return previous;  
    }  
 
    /**
     * Called for entries that have been evicted or removed. This method is
     * invoked when a value is evicted to make space, removed by a call to
     * {@link #remove}, or replaced by a call to {@link #put}. The default
     * implementation does nothing.
     * 当item被回收或者删掉时调用。改方法当value被回收释放存储空间时被remove调用,
     * 或者替换item值时put调用,默认实现什么都没做。
     * <p>The method is called without synchronization: other threads may
     * access the cache while this method is executing.
     *
     * @param evicted true if the entry is being removed to make space, false
     *     if the removal was caused by a {@link #put} or {@link #remove}.
     * true---为释放空间被删除;false---put或remove导致
     * @param newValue the new value for {@code key}, if it exists. If non-null,
     *     this removal was caused by a {@link #put}. Otherwise it was caused by
     *     an eviction or a {@link #remove}.
     */  
    protected void entryRemoved(boolean evicted, K key, V oldValue, V newValue) {}  
 
    /**
     * Called after a cache miss to compute a value for the corresponding key.
     * Returns the computed value or null if no value can be computed. The
     * default implementation returns null.
     * 当某Item丢失时会调用到,返回计算的相应的value或者null
     * <p>The method is called without synchronization: other threads may
     * access the cache while this method is executing.
     *
     * <p>If a value for {@code key} exists in the cache when this method
     * returns, the created value will be released with {@link #entryRemoved}
     * and discarded. This can occur when multiple threads request the same key
     * at the same time (causing multiple values to be created), or when one
     * thread calls {@link #put} while another is creating a value for the same
     * key.
     */  
    protected V create(K key) {  
        return null;  
    }  
 
    private int safeSizeOf(K key, V value) {  
        int result = sizeOf(key, value);  
        if (result < 0) {  
            throw new IllegalStateException("Negative size: " + key + "=" + value);  
        }  
        return result;  
    }  
 
    /**
     * Returns the size of the entry for {@code key} and {@code value} in
     * user-defined units.  The default implementation returns 1 so that size
     * is the number of entries and max size is the maximum number of entries.
     * 返回用户定义的item的大小,默认返回1代表item的数量,最大size就是最大item值
     * <p>An entry's size must not change while it is in the cache.
     */  
    protected int sizeOf(K key, V value) {  
        return 1;  
    }  
 
    /**
     * Clear the cache, calling {@link #entryRemoved} on each removed entry.
     * 清空cacke
     */  
    public final void evictAll() {  
        trimToSize(-1); // -1 will evict 0-sized elements  
    }  
 
    /**
     * For caches that do not override {@link #sizeOf}, this returns the number
     * of entries in the cache. For all other caches, this returns the sum of
     * the sizes of the entries in this cache.
     */  
    public synchronized final int size() {  
        return size;  
    }  
 
    /**
     * For caches that do not override {@link #sizeOf}, this returns the maximum
     * number of entries in the cache. For all other caches, this returns the
     * maximum sum of the sizes of the entries in this cache.
     */  
    public synchronized final int maxSize() {  
        return maxSize;  
    }  
 
    /**
     * Returns the number of times {@link #get} returned a value that was
     * already present in the cache.
     */  
    public synchronized final int hitCount() {  
        return hitCount;  
    }  
 
    /**
     * Returns the number of times {@link #get} returned null or required a new
     * value to be created.
     */  
    public synchronized final int missCount() {  
        return missCount;  
    }  
 
    /**
     * Returns the number of times {@link #create(Object)} returned a value.
     */  
    public synchronized final int createCount() {  
        return createCount;  
    }  
 
    /**
     * Returns the number of times {@link #put} was called.
     */  
    public synchronized final int putCount() {  
        return putCount;  
    }  
 
    /**
     * Returns the number of values that have been evicted.
     * 返回被回收的数量
     */  
    public synchronized final int evictionCount() {  
        return evictionCount;  
    }  
 
    /**
     * Returns a copy of the current contents of the cache, ordered from least
     * recently accessed to most recently accessed. 返回当前cache的副本,从最近最少访问到最多访问
     */  
    public synchronized final Map<K, V> snapshot() {  
        return new LinkedHashMap<K, V>(map);  
    }  
 
    @Override public synchronized final String toString() {  
        int accesses = hitCount + missCount;  
        int hitPercent = accesses != 0 ? (100 * hitCount / accesses) : 0;  
        return String.format("LruCache[maxSize=%d,hits=%d,misses=%d,hitRate=%d%%]",  
                maxSize, hitCount, missCount, hitPercent);  
    }  
}  

© 著作权归作者所有

共有 人打赏支持
新年
粉丝 7
博文 83
码字总数 45935
作品 0
海淀
程序员
Android内存优化之内存缓存

什么是缓存? 缓存技术原理就是把用户访问的所有对象看作一个全集,经过算法标记哪些是用户经常访问的对象,把这些对象放到一个集合里,这个集合是全集一个子集,下一次用户再访问的时候会先...

今晚打猴子
2015/08/21
0
0
安卓图片的异步请求及使用LruCache缓存和手机内存两层存储图片,避免重新加载页面带来的重新请求

看到网友的一片技术博客讲解了LruCache的使用,我把它加到了我的项目中,但是加入断点发现,列表上下滑动时,确实可以不用重新加载图片,但是重新打开这个activity或者重新启动应用,LruCach...

bluecoffee
2015/04/03
0
0
为什么Android官方废弃SoftRefrerence软引用和WeakReference弱引用,而拥抱LruCache?

为什么Android官方废弃SoftRefrerence软引用和WeakReference弱引用,而拥抱LruCache? 一些具有Java背景的研发者喜欢使用软引用(SoftRefrerence)和弱引用(WeakReference)来缓存Java对象和数据...

开开心心过
06/09
0
0
Android源码解析——LruCache

我认为在写涉及到数据结构或算法的实现类的源码解析博客时,不应该急于讲它的使用或马上展开对源码的解析,而是要先交待一下这个数据结构或算法的资料,了解它的设计,再从它的设计出发去讲如...

浩码农
2016/05/13
0
0
Android LruCache 缓存图片

使用图片缓存技术 在 你应用程序的UI界面加载一张图片是一件很简单的事情,但是当你需要在界面上加载一大堆图片的时候,情况就变得复杂起来。在很多情况下,(比如使用 ListView, GridView 或...

娶到笨笨
2014/04/23
0
0

没有更多内容

加载失败,请刷新页面

加载更多

Hbase 概述及特点

1、Hbase概述 HBase是一种构建在HDFS之上的分布式、面向列的存储系统。在需要实时读写、随机访问超大规模数据集时,可以使用HBase。 尽管已经有许多数据存储和访问的策略和实现方法,但事实上...

PeakFang-BOK
19分钟前
0
0
TortoiseGit(乌龟git)保存用户名密码的方法

windows下比较比较好用的git客户端有2种: 1. msysgit + TortoiseGit(乌龟git) 2. GitHub for Windows github的windows版也用过一段时间,但还是不太习惯。所以目前仍然青睐与msysgit+乌龟g...

simpower
38分钟前
0
0
Java并发编程:volatile关键字解析

volatile这个关键字可能很多朋友都听说过,或许也都用过。在Java 5之前,它是一个备受争议的关键字,因为在程序中使用它往往会导致出人意料的结果。在Java 5之后,volatile关键字才得以重获生...

engeue
55分钟前
2
0
通过ajax访问远程天气预报服务

http://www.webxml.com.cn/zh_cn/index.aspx 更改wsdl文件 打开文件将15行,51行,101行去掉 然后把文件复制到c盘 然后在桌面上面就生成了文件 将文件打成jar包 package cn.it.ws.weather;...

江戸川
今天
1
0
聊聊storm的tickTuple

序 本文主要研究一下storm的tickTuple 实例 TickWordCountBolt public class TickWordCountBolt extends BaseBasicBolt { private static final Logger LOGGER = LoggerFactory.getLogg......

go4it
今天
1
0

没有更多内容

加载失败,请刷新页面

加载更多

返回顶部
顶部