简介
okhttp是Android中应用最广的http网络请求框架。结构优雅,性能强大。我们通过阅读它,对网络库的架构进行学习。本篇主要阅读okhttp任务队列的结构,了解okhttp的任务调度。
基本结构
网络请求是一个典型的生产/消费模型。我们的每一个请求都会加入一个缓冲队列,然后在合适的时候进行消费。
我们以异步请求为例:
OKHttpClient
提供方法生产Call,Dispatcher
负责管理、分发和执行AsyncCall
。
下面,我们来细致地看一下代码。
详细代码
基本调用
还是以异步任务为例,一次典型的okhttp的调用:
String url = "http://wwww.baidu.com";
OkHttpClient okHttpClient = new OkHttpClient();
final Request request = new Request.Builder()
.url(url)
.get()//默认就是GET请求,可以不写
.build();
Call call = okHttpClient.newCall(request);
call.enqueue(new Callback() {
@Override
public void onFailure(Call call, IOException e) {
Log.d(TAG, "onFailure: ");
}
@Override
public void onResponse(Call call, Response response) throws IOException {
Log.d(TAG, "onResponse: " + response.body().string());
}
});
我们可以看到,这个过程大致分为两部,生成一个新的call,将call加入队列。
Request
Request是okhttp中的http请求封装类,它管理了http协议中的参数,如Header,RequestBody等
class Request internal constructor(
internal val url: HttpUrl,
builder: Builder
) {
internal val method: String = builder.method
internal val headers: Headers = builder.headers.build()
internal val body: RequestBody? = builder.body
internal val tags: Map, Any> = Util.immutableMap(builder.tags)
......
}
OKHttpClient
OKHttpClient是okhttp中总管类,管理了okhttp几乎所有抽象模块和各种配置。
open class OkHttpClient internal constructor(
builder: Builder
) : Cloneable, Call.Factory, WebSocket.Factory {
private val dispatcher: Dispatcher = builder.dispatcher
private val proxy: Proxy? = builder.proxy
private val protocols: List = builder.protocols
private val connectionSpecs: List = builder.connectionSpecs
private val interceptors: List =
Util.immutableList(builder.interceptors)
private val networkInterceptors: List =
Util.immutableList(builder.networkInterceptors)
private val eventListenerFactory: EventListener.Factory = builder.eventListenerFactory
private val proxySelector: ProxySelector = builder.proxySelector
private val cookieJar: CookieJar = builder.cookieJar
private val cache: Cache? = builder.cache
private val socketFactory: SocketFactory = builder.socketFactory
private val sslSocketFactory: SSLSocketFactory?
private val hostnameVerifier: HostnameVerifier = builder.hostnameVerifier
private val certificatePinner: CertificatePinner
private val proxyAuthenticator: Authenticator = builder.proxyAuthenticator
private val authenticator: Authenticator = builder.authenticator
private val connectionPool: ConnectionPool = builder.connectionPool
private val dns: Dns = builder.dns
private val followSslRedirects: Boolean = builder.followSslRedirects
private val followRedirects: Boolean = builder.followRedirects
private val retryOnConnectionFailure: Boolean = builder.retryOnConnectionFailure
private val callTimeout: Int = builder.callTimeout
private val connectTimeout: Int = builder.connectTimeout
private val readTimeout: Int = builder.readTimeout
private val writeTimeout: Int = builder.writeTimeout
private val pingInterval: Int = builder.pingInterval
private val internalCache: InternalCache? = builder.internalCache
private val certificateChainCleaner: CertificateChainCleaner?
......
/** Prepares the [request] to be executed at some point in the future. */
override fun newCall(request: Request): Call {
return RealCall.newRealCall(this, request, false /* for web socket */)
}
}
RealCall
RealCall是okhttp中任务的封装类
companion object {
fun newRealCall(
client: OkHttpClient,
originalRequest: Request,
forWebSocket: Boolean
): RealCall {
// Safely publish the Call instance to the EventListener.
return RealCall(client, originalRequest, forWebSocket).apply {
transmitter = Transmitter(client, this)
}
}
}
internal class RealCall private constructor(
val client: OkHttpClient,
/** The application's original request unadulterated by redirects or auth headers. */
val originalRequest: Request,
val forWebSocket: Boolean
) : Call {
/**
* There is a cycle between the [Call] and [Transmitter] that makes this awkward.
* This is set after immediately after creating the call instance.
*/
private lateinit var transmitter: Transmitter
// Guarded by this.
var executed: Boolean = false
@Synchronized override fun isExecuted(): Boolean = executed
override fun isCanceled(): Boolean = transmitter.isCanceled
override fun request(): Request = originalRequest
override fun execute(): Response {
synchronized(this) {
check(!executed) { "Already Executed" }
executed = true
}
transmitter.timeoutEnter()
transmitter.callStart()
try {
client.dispatcher().executed(this)
return getResponseWithInterceptorChain()
} finally {
client.dispatcher().finished(this)
}
}
override fun enqueue(responseCallback: Callback) {
synchronized(this) {
check(!executed) { "Already Executed" }
executed = true
}
transmitter.callStart()
client.dispatcher().enqueue(AsyncCall(responseCallback))
}
override fun cancel() {
transmitter.cancel()
}
...
}
在异步任务的例子中,我们通常使用enqueue
将我们的请求加入请求队列。
client.dispatcher().enqueue(AsyncCall(responseCallback))
这句话中,我们可以看到,我们新建了AsyncCall
加入到OKHttpClient的dispatcher的队列中。
我们先看AsyncCall封装了什么
internal inner class AsyncCall(
private val responseCallback: Callback
) : NamedRunnable("OkHttp %s", redactedUrl()) {
@Volatile private var callsPerHost = AtomicInteger(0)
fun callsPerHost(): AtomicInteger = callsPerHost
fun reuseCallsPerHostFrom(other: AsyncCall) {
this.callsPerHost = other.callsPerHost
}
fun host(): String = originalRequest.url().host()
fun request(): Request = originalRequest
fun get(): RealCall = this@RealCall
/**
* Attempt to enqueue this async call on `executorService`. This will attempt to clean up
* if the executor has been shut down by reporting the call as failed.
*/
fun executeOn(executorService: ExecutorService) {
assert(!Thread.holdsLock(client.dispatcher()))
var success = false
try {
executorService.execute(this)
success = true
} catch (e: RejectedExecutionException) {
val ioException = InterruptedIOException("executor rejected")
ioException.initCause(e)
transmitter.noMoreExchanges(ioException)
responseCallback.onFailure(this@RealCall, ioException)
} finally {
if (!success) {
client.dispatcher().finished(this) // This call is no longer running!
}
}
}
override fun execute() {
var signalledCallback = false
transmitter.timeoutEnter()
try {
val response = getResponseWithInterceptorChain()
signalledCallback = true
responseCallback.onResponse(this@RealCall, response)
} catch (e: IOException) {
if (signalledCallback) {
// Do not signal the callback twice!
Platform.get().log(INFO, "Callback failure for ${toLoggableString()}", e)
} else {
responseCallback.onFailure(this@RealCall, e)
}
} finally {
client.dispatcher().finished(this)
}
}
}
我们可以看到,AsyncCall主要提供了executeOn
和execute
两个方法。
executeOn
executeOn
是AsyncCall自身的方法。它的主要内容是对异常的处理。如果一切正常,就用传入的ExecutorService
执行当前AsyncCall。
execute
而execute
是实现的Runnable的run方法。
public abstract class NamedRunnable implements Runnable {
protected final String name;
public NamedRunnable(String format, Object... args) {
this.name = Util.format(format, args);
}
@Override public final void run() {
String oldName = Thread.currentThread().getName();
Thread.currentThread().setName(name);
try {
execute();
} finally {
Thread.currentThread().setName(oldName);
}
}
protected abstract void execute();
}
我们可以看到,这个方法,只是为了给线程起名。
当AsyncCall在线程池中执行时,execute
方法就会被调用。execute
方法的主要逻辑是:
getResponseWithInterceptorChain
,它的实现,我们再不展开,属于okhttp网络能力了。
铺垫了上面这么多,我们终于可以看看okhttp的任务队列如何设计的了。
Dispatcher
private var idleCallback: Runnable? = null
/** Executes calls. Created lazily. */
private var executorService: ExecutorService? = null
/** Ready async calls in the order they'll be run. */
private val readyAsyncCalls = ArrayDeque()
/** Running asynchronous calls. Includes canceled calls that haven't finished yet. */
private val runningAsyncCalls = ArrayDeque()
/** Running synchronous calls. Includes canceled calls that haven't finished yet. */
private val runningSyncCalls = ArrayDeque()
internal fun enqueue(call: AsyncCall) {
synchronized(this) {
readyAsyncCalls.add(call)
// Mutate the AsyncCall so that it shares the AtomicInteger of an existing running call to
// the same host.
if (!call.get().forWebSocket) {
val existingCall = findExistingCallWithHost(call.host())
if (existingCall != null) call.reuseCallsPerHostFrom(existingCall)
}
}
promoteAndExecute()
}
这一段主要是AsyncCall的callsPerHost复用逻辑,注释说明比较清晰。如果有相同的AsyncCall存在于runningAsyncCalls中,则会增加callsPerHost。接下来,看一下promoteAndExecute
的实现:
private fun promoteAndExecute(): Boolean {
assert(!Thread.holdsLock(this))
val executableCalls = ArrayList()
val isRunning: Boolean
synchronized(this) {
val i = readyAsyncCalls.iterator()
while (i.hasNext()) {
val asyncCall = i.next()
if (runningAsyncCalls.size >= this.maxRequests) break // Max capacity.
if (asyncCall.callsPerHost().get() >= this.maxRequestsPerHost) continue // Host max capacity.
i.remove()
asyncCall.callsPerHost().incrementAndGet()
executableCalls.add(asyncCall)
runningAsyncCalls.add(asyncCall)
}
isRunning = runningCallsCount() > 0
}
for (i in 0 until executableCalls.size) {
val asyncCall = executableCalls[i]
asyncCall.executeOn(executorService())
}
return isRunning
}
promoteAndExecute
方法,是将readyAsyncCalls
队列中的任务,在最大任务数没有超标的情况下,移入runningAsyncCalls
队列中。并对涉及转移的方法,调用executeOn
方法。executeOn
被调用后,就会执行到,RealCall中AsyncCall中的execute
方法了。
小结
以上就是okhttp在请求和任务上的基本结构,还没有涉及到具体的网络请求。核心类是RealCall和Dispatcher,通过ArrayDeque和AsyncCall完成任务的调度。
如有问题,欢迎指正。