Volley使用了线程池来作为基础结构，主要分为主线程，cache线程和network线程。
主线程和cache线程都只有一个，而NetworkDispatcher线程可以有多个，这样能解决比并行问题。如下图:
其中左下角是NetworkDispatcher线程，大致步骤是：
1.不断从请求队列中取出请求
request = mQueue.take();
2.发起网络请求
NetworkResponse networkResponse = mNetwork.performRequest(request);
3.把这个networkResponse转化为期望的数据类型，比如Response<String>,Response<Json>,Response<Bitmap>
Response<?> response = request.parseNetworkResponse(networkResponse);
4.将网络响应加入缓存
mCache.put(request.getCacheKey(), response.cacheEntry);
下面是NetworkDispatcher线程的主要代码：
@Override
public void run() {
Process.setThreadPriority(Process.THREAD_PRIORITY_BACKGROUND);
Request request;
while (true) {
try {
// Take a request from the queue.
request = mQueue.take();//1.从请求队列中取出一个网络请求，mQueue是BlockingQueue<Request>的实现类
} catch (InterruptedException e) {
// We may have been interrupted because it was time to quit.
if (mQuit) {
return;
}
continue;
}
try {
request.addMarker("network-queue-take");
// If the request was cancelled already, do not perform the
// network request.
if (request.isCanceled()) {
request.finish("network-discard-cancelled");
continue;
}
// Tag the request (if API >= 14)
if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.ICE_CREAM_SANDWICH) {
TrafficStats.setThreadStatsTag(request.getTrafficStatsTag());
}
// Perform the network request.
NetworkResponse networkResponse = mNetwork.performRequest(request);//2.发起网络请求
request.addMarker("network-http-complete");
// If the server returned 304 AND we delivered a response already,
// we're done -- don't deliver a second identical response.
if (networkResponse.notModified && request.hasHadResponseDelivered()) {
request.finish("not-modified");
continue;
}
// Parse the response here on the worker thread.
Response<?> response = request.parseNetworkResponse(networkResponse);//3.把这个networkResponse转化为期望的数据类型
request.addMarker("network-parse-complete");
// Write to cache if applicable.
// TODO: Only update cache metadata instead of entire record for 304s.
if (request.shouldCache() && response.cacheEntry != null) {
mCache.put(request.getCacheKey(), response.cacheEntry);//4.将网络响应加入缓存
request.addMarker("network-cache-written");
}
// Post the response back.
request.markDelivered();
mDelivery.postResponse(request, response);
} catch (VolleyError volleyError) {
parseAndDeliverNetworkError(request, volleyError);
} catch (Exception e) {
VolleyLog.e(e, "Unhandled exception %s", e.toString());
mDelivery.postError(request, new VolleyError(e));
}
}
}
要理解Volley的并行性实现，必需理解PriorityBlockingQueue并发类。注意到我们在NetworkDispatcher循环中并没有使用显式的同步（使用Lock或者使用synchronize），因为PriorityBlockingQueue的实现是线程安全的。在使用显式的wait()和notifyAll()时存在的类和类之间的耦合可以因此消除，因为每个类只和BlockingQueue通信。这个知识点可以参考《Java编程思想》P713.
BlockingQueue implementations are thread-safe. All queuing methods achieve their effects atomically using internal locks or other forms of concurrency control.
所以可以有多个NetworkDispatcher线程同时从请求队列PriorityBlockingQueue中取出请求而不会产生线程冲突，那就是Volley支持多线程下载图片的方式。
实际上，BlockingQueue接口是用于解决生产者-消费者问题的。
class Producer implements Runnable {
private final BlockingQueue queue;
Producer(BlockingQueue q) { queue = q; }
public void run() {
try {
while (true) { queue.put(produce()); }
} catch (InterruptedException ex) { ... handle ...}
}
Object produce() { ... }
}
class Consumer implements Runnable {
private final BlockingQueue queue;
Consumer(BlockingQueue q) { queue = q; }
public void run() {
try {
while (true) { consume(queue.take()); }
} catch (InterruptedException ex) { ... handle ...}
}
void consume(Object x) { ... }
}
class Setup {
void main() {
BlockingQueue q = new SomeQueueImplementation();
Producer p = new Producer(q);
Consumer c1 = new Consumer(q);
Consumer c2 = new Consumer(q);
new Thread(p).start();
new Thread(c1).start();
new Thread(c2).start();
}
}}
所以在Volley中，request就是产品，主线程是生产者，NetworkDispatcher线程是消费者。
另外注意到，NetworkDispatcher的实现其实是策略设计模式：
/** The queue of requests to service. */
private final BlockingQueue<Request> mQueue;
/** The network interface for processing requests. */
private final Network mNetwork;
/** The cache to write to. */
private final Cache mCache;
/** For posting responses and errors. */
private final ResponseDelivery mDelivery;
/** Used for telling us to die. */
private volatile boolean mQuit = false;
/**
* Creates a new network dispatcher thread. You must call {@link #start()}
* in order to begin processing.
*
* @param queue Queue of incoming requests for triage
* @param network Network interface to use for performing requests
* @param cache Cache interface to use for writing responses to cache
* @param delivery Delivery interface to use for posting responses
*/
public NetworkDispatcher(BlockingQueue<Request> queue,
Network network, Cache cache,
ResponseDelivery delivery) {
mQueue = queue;
mNetwork = network;
mCache = cache;
mDelivery = delivery;
}
NetworkDispatcher构造函数的几个参数都是接口，而run方法则使用这些策略类方法实现了算法的主体流程，具体实现有些留给了开发者，有些则是框架实现。比如ImageCache作为一级缓存的Cache方法留给了开发者实现，由开发者控制具体的缓存策略，当然Volley建议我们使用LRUCache作为L1缓存的实现。
最后，NetworkDispatcher的数组则构成了RequestQueue类中线程池，由RequestQueue统一启动和停止：
/**
* Creates the worker pool. Processing will not begin until {@link #start()} is called.
*
* @param cache A Cache to use for persisting responses to disk
* @param network A Network interface for performing HTTP requests
* @param threadPoolSize Number of network dispatcher threads to create
* @param delivery A ResponseDelivery interface for posting responses and errors
*/
public RequestQueue(Cache cache, Network network, int threadPoolSize,
ResponseDelivery delivery) {
mCache = cache;
mNetwork = network;
mDispatchers = new NetworkDispatcher[threadPoolSize];
mDelivery = delivery;
}
/**
* Starts the dispatchers in this queue.
*/
public void start() {
stop(); // Make sure any currently running dispatchers are stopped.
// Create the cache dispatcher and start it.
mCacheDispatcher = new CacheDispatcher(mCacheQueue, mNetworkQueue, mCache, mDelivery);
mCacheDispatcher.start();
// Create network dispatchers (and corresponding threads) up to the pool size.
for (int i = 0; i < mDispatchers.length; i++) {
NetworkDispatcher networkDispatcher = new NetworkDispatcher(mNetworkQueue, mNetwork,
mCache, mDelivery);
mDispatchers[i] = networkDispatcher;
networkDispatcher.start();
}
}
关于volley的网络请求部分可以看博客：http://www.cnblogs.com/bvin/p/3291611.html
网络请求中有几个转换解析请求获取响应结果的地方：
1.HttpStack接口的performRequest()方法
public HttpResponse performRequest(Request<?> request, Map<String, String> additionalHeaders)
2.Network接口的performRequest()方法
public NetworkResponse performRequest(Request<?> request)
3.Request类的parseNetworkResponse()抽象方法
abstract protected Response<T> parseNetworkResponse(NetworkResponse response);
可以很鲜明得看出第一个是对原生Http请求的解析，解析出来是一个Apach的HttpResponse 实例，这个结果就是通过上述两个HttpStack的实现类HttpClientStack和HurlStack执行获取的获取的。
而第二个解析出来是框架自定的NetworkResponse，这是通过Network的实现类BasicNetwork来获取的。
第三个就是第二个得出来NetworkResponse解析成用户期望Response<T> 了，这个Response和Request是对应的，有String型，json型还有Image型的。然后在通过ResponseDelivery把解析好的结果发送到主线程。
从Request到Response就是这样一步步走来的。
再详细看下上面第二个方法performRequest()，这个方法是NetworkDispatcher中主要步骤的第二步，其中真正网络请求的工作委托给mHttpStack实现，通过HttpStack实现类就是上图的两个HttpClientStack和HurlStack执行。这两个类的区别在于HttpClientStack是直接用HttpClient的execute()方法执行一个Http请求，而HurlStack就是直接用URL.openConnection()进行连接，这种方式在2.3以上是不能用的，所以分开了这两种方式。
/**
* @title performRequest执行各种Request请求并以NetworkResponse的形式返回结果
* @param Request
* @return NetworkResponse
* @throws VolleyError
* 定义：{@link Network#performRequest(Request)}
* 被调：{@link NetworkDispatcher#run()}
*
*/
@Override//NetworkDispatcher的run()方法中调用
public NetworkResponse performRequest(Request<?> request) throws VolleyError {
long requestStart = SystemClock.elapsedRealtime();//开始请求时间
while (true) {
HttpResponse httpResponse = null;//apache的请求结果
byte[] responseContents = null;//请求的内容
Map<String, String> responseHeaders = new HashMap<String, String>();//响应结果头部信息
try {
// Gather headers.
Map<String, String> headers = new HashMap<String, String>();//保存缓存数据
addCacheHeaders(headers, request.getCacheEntry());//先获取缓存数据
httpResponse = mHttpStack.performRequest(request, headers);//去调用mHttpStack的实现方法执行请求
StatusLine statusLine = httpResponse.getStatusLine();//获取http状态线
int statusCode = statusLine.getStatusCode();//获取状态码
responseHeaders = convertHeaders(httpResponse.getAllHeaders());
// Handle cache validation.//处理缓存验证
if (statusCode == HttpStatus.SC_NOT_MODIFIED) {//返回缓存数据
return new NetworkResponse(HttpStatus.SC_NOT_MODIFIED,
request.getCacheEntry().data, responseHeaders, true);
}
//把HttpEntity转化为byte[]数据
responseContents = entityToBytes(httpResponse.getEntity());
// if the request is slow, log it.//如果请求很慢，就打印出来看一下
long requestLifetime = SystemClock.elapsedRealtime() - requestStart;
logSlowRequests(requestLifetime, request, responseContents, statusLine);//打印
//连接正常但是返回无内容，抛出IO异常
if (statusCode != HttpStatus.SC_OK && statusCode != HttpStatus.SC_NO_CONTENT) {
throw new IOException();
}
return new NetworkResponse(statusCode, responseContents, responseHeaders, false);
} catch (SocketTimeoutException e) {//读取超时，重试
attemptRetryOnException("socket", request, new TimeoutError());
} catch (ConnectTimeoutException e) {//连接超时，重试
attemptRetryOnException("connection", request, new TimeoutError());
} catch (MalformedURLException e) {//Bad URL
throw new RuntimeException("Bad URL " + request.getUrl(), e);
} catch (IOException e) {//IO异常
int statusCode = 0;
NetworkResponse networkResponse = null;
if (httpResponse != null) {
statusCode = httpResponse.getStatusLine().getStatusCode();
} else {//如果没有返回httpResponse，就说明没连接
throw new NoConnectionError(e);
}
VolleyLog.e("Unexpected response code %d for %s", statusCode, request.getUrl());
if (responseContents != null) {//返回数据不为空
networkResponse = new NetworkResponse(statusCode, responseContents,
responseHeaders, false);//创建响应体
if (statusCode == HttpStatus.SC_UNAUTHORIZED ||
statusCode == HttpStatus.SC_FORBIDDEN) {//认证失败异常，重试
attemptRetryOnException("auth",
request, new AuthFailureError(networkResponse));
} else {//服务器异常
// TODO: Only throw ServerError for 5xx status codes.
throw new ServerError(networkResponse);//只有状态码为5XX才抛出服务器异常
}
} else {//网络异常
throw new NetworkError(networkResponse);
}
}
}
}
A：先是通过mHttpStack把请求执行并且获取它的响应结果，根据HttpStatus做出各种判断。
B：然后再把httpResponse的Entity转化为ByteArray，并处理各种发生的异常。
C：最后的过程是这样的：通过Volley创建一个RequestQueue请求队列，当这个队列开始运作的时候会启动NetworkDispatcher这个工作线程，而BasicNetwork的performRequest()的方法就在NetworkDispatcher线程run()方法中调用，然后通过mHttpStack的performRequest()方法获取一个networkResponse，在NetworkDispatcher线程把这个networkResponse转化为期望的数据类型，比如Response<String>,Response<Json>,Response<Bitmap>。