HttpClient 源码阅读

　　在项目中常用的HttpClient，与我们非常的亲密，为了能处理遇到的Http问题，我们应该去了解里面的机制和实现。　　

　　官方文档：http://hc.apache.org/httpcomponents-client-ga/tutorial/html/

　　Maven

        <!-- components.httpclient-->
        <dependency>
            <groupId>org.apache.httpcomponents</groupId>
            <artifactId>httpclient</artifactId>
            <version>4.5.5</version>
        </dependency>

　　HttpClient的看点主要是它的责任链设计、连接池机制、工具类的封装，个人觉得它设计得还比较优雅。

　　我们从一个简单的Http请求，就可以步步深入得去阅读它的源码。

        HttpGet get = new HttpGet("http://localhost:8080/hello/say");
        RequestConfig config = RequestConfig.custom()
                .setConnectTimeout(3000)
                .setSocketTimeout(3000)
                .build();
        get.setConfig(config);
        HttpClient client = HttpClientBuilder.create()
                .setMaxConnTotal(1<<6)
                .setMaxConnPerRoute(1<<3)
                .evictExpiredConnections()
                .build();
        HttpResponse response = client.execute(get);
        System.out.println(EntityUtils.toString(response.getEntity()));

　　首先，发起一个Get请求，它先要实例化请求方法，然后设置请求的配置（连接时间、读取时间）等，然后使用HttpClient的Builder实例化一个按照你配置的HttpClient，最后发起请求拿到响应，在解析响应的时候，它提供了EntityUtils帮你解析响应体。

　　当我们从它的execute方法进去的时候，它是一个抽象类CloseableHttpClient；它封装了各式各样发起请求的方法，可以说是走入核心之前的外部接口封装，它的真正功能在doExecute方法里面，这个方法是抽象的，取决与Builder在创建实例的时候，实现类或者内部类实现的这个方法。

　　 我们通过debug可以知道，它实际上是走入了InternalHttpClient中，这个时候已经走入了责任链的头部了，开始走入设计者想好的路径了；这个类的主要作用是封装内部调用使用的Request和Config，在它的return方法中，可以看出它将走入下一个execChain。

    protected CloseableHttpResponse doExecute(HttpHost target, HttpRequest request, HttpContext context) throws IOException, ClientProtocolException {
        Args.notNull(request, "HTTP request");
        HttpExecutionAware execAware = null;
        if(request instanceof HttpExecutionAware) {
            execAware = (HttpExecutionAware)request;
        }

        try {
            HttpRequestWrapper wrapper = HttpRequestWrapper.wrap(request, target);
            HttpClientContext localcontext = HttpClientContext.adapt((HttpContext)(context != null?context:new BasicHttpContext()));
            RequestConfig config = null;
            if(request instanceof Configurable) {
                config = ((Configurable)request).getConfig();
            }

            if(config == null) {
                HttpParams params = request.getParams();
                if(params instanceof HttpParamsNames) {
                    if(!((HttpParamsNames)params).getNames().isEmpty()) {
                        config = HttpClientParamConfig.getRequestConfig(params, this.defaultConfig);
                    }
                } else {
                    config = HttpClientParamConfig.getRequestConfig(params, this.defaultConfig);
                }
            }

            if(config != null) {
                localcontext.setRequestConfig(config);
            }

            this.setupContext(localcontext);
            HttpRoute route = this.determineRoute(target, wrapper, localcontext);
            return this.execChain.execute(route, wrapper, localcontext, execAware);
        } catch (HttpException var9) {
            throw new ClientProtocolException(var9);
        }
    }

　　 再往下走，它将走入RetryClient；这个类的主要作用是控制重试，当下层链出现超时的时候会进行重试。这个重试的次数是之前你创建实例的时候可以指定的，没有指定它也有默认值。

    @Override
    public CloseableHttpResponse execute(
            final HttpRoute route,
            final HttpRequestWrapper request,
            final HttpClientContext context,
            final HttpExecutionAware execAware) throws IOException, HttpException {
        Args.notNull(route, "HTTP route");
        Args.notNull(request, "HTTP request");
        Args.notNull(context, "HTTP context");
        final Header[] origheaders = request.getAllHeaders();
        for (int execCount = 1;; execCount++) {
            try {
                return this.requestExecutor.execute(route, request, context, execAware);
            } catch (final IOException ex) {
                if (execAware != null && execAware.isAborted()) {
                    this.log.debug("Request has been aborted");
                    throw ex;
                }
                if (retryHandler.retryRequest(ex, execCount, context)) {
                    if (this.log.isInfoEnabled()) {
                        this.log.info("I/O exception ("+ ex.getClass().getName() +
                                ") caught when processing request to "
                                + route +
                                ": "
                                + ex.getMessage());
                    }
                    if (this.log.isDebugEnabled()) {
                        this.log.debug(ex.getMessage(), ex);
                    }
                    if (!RequestEntityProxy.isRepeatable(request)) {
                        this.log.debug("Cannot retry non-repeatable request");
                        throw new NonRepeatableRequestException("Cannot retry request " +
                                "with a non-repeatable request entity", ex);
                    }
                    request.setHeaders(origheaders);
                    if (this.log.isInfoEnabled()) {
                        this.log.info("Retrying request to " + route);
                    }
                } else {
                    if (ex instanceof NoHttpResponseException) {
                        final NoHttpResponseException updatedex = new NoHttpResponseException(
                                route.getTargetHost().toHostString() + " failed to respond");
                        updatedex.setStackTrace(ex.getStackTrace());
                        throw updatedex;
                    } else {
                        throw ex;
                    }
                }
            }
        }
    }

　　 继续往下，它到达了MainClientExec类；这个类厉害了，控制着连接池的获取、socket连接的建立、链接的释放，可以说是HttpClient的核心了。它的下层调用是HttpRequestExecutor，里面控制着Http请求头的发送，Http Request Line的发送，以及响应的收集。至此，我们可以梳理一个路径。

　　 它使用了责任链进行拼装，并且每个链条上的抽象很干净，只做它负责的范围的工作。而它们这个链条的形成，是在builder里面组装的。

　　  我们把调用链抽象出来，结合builder，它的设计是比较优雅的。

package execChain;

public interface ExecutionChain {
    HttpResponse exec(String host,HttpRequest request,HttpContext context);
}

package execChain;

import java.io.IOException;

public interface HttpClient {
    HttpResponse execute(String route,HttpRequest request,HttpContext context)
            throws IOException;
}

package execChain;

import java.io.IOException;

public abstract class TopHttpClient implements HttpClient{
    protected abstract HttpResponse doExecute(String host,HttpRequest request,HttpContext context);

    @Override
    public HttpResponse execute(String route, HttpRequest request, HttpContext context) throws IOException {
        return doExecute(route,request,context);
    }
}

package execChain;

public class RetryExecClient extends TopHttpClient{

    private ExecutionChain requestExecChain;

    public RetryExecClient(ExecutionChain requestExecChain) {
        this.requestExecChain = requestExecChain;
    }

    @Override
    protected HttpResponse doExecute(String host, HttpRequest request, HttpContext context) {
        return this.requestExecChain.exec(host,request,context);
    }
}

package execChain;

public class MainExecClient implements ExecutionChain{

    @Override
    public HttpResponse exec(String host, HttpRequest request, HttpContext context) {
        System.out.println("host : "+host);
        return new HttpResponse();
    }
}

package execChain;

public class MyHttpClientBuilder {
    public static MyHttpClientBuilder create() {
        return new MyHttpClientBuilder();
    }
    public TopHttpClient build(){
        MainExecClient client = new MainExecClient();
        // append chain ....
        return new RetryExecClient(client);
    }
}

        TopHttpClient client = MyHttpClientBuilder.create().build();
        client.execute("http://www.baidu.com",new HttpRequest(),new HttpContext());

　　有关连接池的操作，是在PoolingHttpClientConnectionManager里边，它把对链接的申请称呼为lease，也就是租借的意思。

    @Override
    public ConnectionRequest requestConnection(
            final HttpRoute route,
            final Object state) {
        Args.notNull(route, "HTTP route");
        if (this.log.isDebugEnabled()) {
            this.log.debug("Connection request: " + format(route, state) + formatStats(route));
        }
        final Future<CPoolEntry> future = this.pool.lease(route, state, null);
        return new ConnectionRequest() {

            @Override
            public boolean cancel() {
                return future.cancel(true);
            }

            @Override
            public HttpClientConnection get(
                    final long timeout,
                    final TimeUnit tunit) throws InterruptedException, ExecutionException, ConnectionPoolTimeoutException {
                final HttpClientConnection conn = leaseConnection(future, timeout, tunit);
                if (conn.isOpen()) {
                    final HttpHost host;
                    if (route.getProxyHost() != null) {
                        host = route.getProxyHost();
                    } else {
                        host = route.getTargetHost();
                    }
                    final SocketConfig socketConfig = resolveSocketConfig(host);
                    conn.setSocketTimeout(socketConfig.getSoTimeout());
                }
                return conn;
            }

        };

    }

　　　我们观察到它的池子pool是CPool这个类，它的继承关系如下。

　　 发现大部分的逻辑实在AbstractConnPool里边，租借的时候走了这个方法org.apache.http.pool.AbstractConnPool#getPoolEntryBlocking。在这个方法中，我们发现它往一个RouteSpecificPool中通过route获取了一个池子。

　　然后进去看到它池子的定义，结合代码可以看出，它把链接的状态分为：已租借、可用链接、等待三种状态。梳理了一下lease方法，它是将avalible的链接拿出来放到lease中，如果可用队列没有链接，那它将创建一个并放入租借队列，这里如果它这个route的连接数超过了你设置的MaxPerRoute配置，那么它将会方法pending队列，并且await当前线程，直到有hold链接的线程调用了releaseConnection等方法才会被notify。

　　

　　　这个getPoolEntityBlocking方法，就是申请连接池链接的核心代码了。

    private E getPoolEntryBlocking(
            final T route, final Object state,
            final long timeout, final TimeUnit tunit,
            final Future<E> future) throws IOException, InterruptedException, TimeoutException {

        Date deadline = null;
        if (timeout > 0) {
            deadline = new Date (System.currentTimeMillis() + tunit.toMillis(timeout));
        }
        this.lock.lock();
        try {
            final RouteSpecificPool<T, C, E> pool = getPool(route);
            E entry;
            for (;;) {
                Asserts.check(!this.isShutDown, "Connection pool shut down");
                for (;;) {
                    entry = pool.getFree(state);
                    if (entry == null) {
                        break;
                    }
                    if (entry.isExpired(System.currentTimeMillis())) {
                        entry.close();
                    }
                    if (entry.isClosed()) {
                        this.available.remove(entry);
                        pool.free(entry, false);
                    } else {
                        break;
                    }
                }
                if (entry != null) {
                    this.available.remove(entry);
                    this.leased.add(entry);
                    onReuse(entry);
                    return entry;
                }

                // New connection is needed
                final int maxPerRoute = getMax(route);
                // Shrink the pool prior to allocating a new connection
                final int excess = Math.max(0, pool.getAllocatedCount() + 1 - maxPerRoute);
                if (excess > 0) {
                    for (int i = 0; i < excess; i++) {
                        final E lastUsed = pool.getLastUsed();
                        if (lastUsed == null) {
                            break;
                        }
                        lastUsed.close();
                        this.available.remove(lastUsed);
                        pool.remove(lastUsed);
                    }
                }

                if (pool.getAllocatedCount() < maxPerRoute) {
                    final int totalUsed = this.leased.size();
                    final int freeCapacity = Math.max(this.maxTotal - totalUsed, 0);
                    if (freeCapacity > 0) {
                        final int totalAvailable = this.available.size();
                        if (totalAvailable > freeCapacity - 1) {
                            if (!this.available.isEmpty()) {
                                final E lastUsed = this.available.removeLast();
                                lastUsed.close();
                                final RouteSpecificPool<T, C, E> otherpool = getPool(lastUsed.getRoute());
                                otherpool.remove(lastUsed);
                            }
                        }
                        final C conn = this.connFactory.create(route);
                        entry = pool.add(conn);
                        this.leased.add(entry);
                        return entry;
                    }
                }

                boolean success = false;
                try {
                    if (future.isCancelled()) {
                        throw new InterruptedException("Operation interrupted");
                    }
                    pool.queue(future);
                    this.pending.add(future);
                    if (deadline != null) {
                        success = this.condition.awaitUntil(deadline);
                    } else {
                        this.condition.await();
                        success = true;
                    }
                    if (future.isCancelled()) {
                        throw new InterruptedException("Operation interrupted");
                    }
                } finally {
                    // In case of 'success', we were woken up by the
                    // connection pool and should now have a connection
                    // waiting for us, or else we're shutting down.
                    // Just continue in the loop, both cases are checked.
                    pool.unqueue(future);
                    this.pending.remove(future);
                }
                // check for spurious wakeup vs. timeout
                if (!success && (deadline != null && deadline.getTime() <= System.currentTimeMillis())) {
                    break;
                }
            }
            throw new TimeoutException("Timeout waiting for connection");
        } finally {
            this.lock.unlock();
        }
    }

　　

　　回过头来，在MainClient里边的这一行，是建立Tcp链接的代码。

　　

　　　他们将链接的各种类型作为一种常量也可以说是一种枚举，然后通过while循环全部处理，这里边的状态跳转就不理了。

    /** Indicates that the route can not be established at all. */
    public final static int UNREACHABLE = -1;

    /** Indicates that the route is complete. */
    public final static int COMPLETE = 0;

    /** Step: open connection to target. */
    public final static int CONNECT_TARGET = 1;

    /** Step: open connection to proxy. */
    public final static int CONNECT_PROXY = 2;

    /** Step: tunnel through proxy to target. */
    public final static int TUNNEL_TARGET = 3;

    /** Step: tunnel through proxy to other proxy. */
    public final static int TUNNEL_PROXY = 4;

    /** Step: layer protocol (over tunnel). */
    public final static int LAYER_PROTOCOL = 5;

    /**
     * Establishes the target route.
     */
    void establishRoute(
            final AuthState proxyAuthState,
            final HttpClientConnection managedConn,
            final HttpRoute route,
            final HttpRequest request,
            final HttpClientContext context) throws HttpException, IOException {
        final RequestConfig config = context.getRequestConfig();
        final int timeout = config.getConnectTimeout();
        final RouteTracker tracker = new RouteTracker(route);
        int step;
        do {
            final HttpRoute fact = tracker.toRoute();
            step = this.routeDirector.nextStep(route, fact);

            switch (step) {

            case HttpRouteDirector.CONNECT_TARGET:
                this.connManager.connect(
                        managedConn,
                        route,
                        timeout > 0 ? timeout : 0,
                        context);
                tracker.connectTarget(route.isSecure());
                break;
            case HttpRouteDirector.CONNECT_PROXY:
                this.connManager.connect(
                        managedConn,
                        route,
                        timeout > 0 ? timeout : 0,
                        context);
                final HttpHost proxy  = route.getProxyHost();
                tracker.connectProxy(proxy, false);
                break;
            case HttpRouteDirector.TUNNEL_TARGET: {
                final boolean secure = createTunnelToTarget(
                        proxyAuthState, managedConn, route, request, context);
                this.log.debug("Tunnel to target created.");
                tracker.tunnelTarget(secure);
            }   break;

            case HttpRouteDirector.TUNNEL_PROXY: {
                // The most simple example for this case is a proxy chain
                // of two proxies, where P1 must be tunnelled to P2.
                // route: Source -> P1 -> P2 -> Target (3 hops)
                // fact:  Source -> P1 -> Target       (2 hops)
                final int hop = fact.getHopCount()-1; // the hop to establish
                final boolean secure = createTunnelToProxy(route, hop, context);
                this.log.debug("Tunnel to proxy created.");
                tracker.tunnelProxy(route.getHopTarget(hop), secure);
            }   break;

            case HttpRouteDirector.LAYER_PROTOCOL:
                this.connManager.upgrade(managedConn, route, context);
                tracker.layerProtocol(route.isSecure());
                break;

            case HttpRouteDirector.UNREACHABLE:
                throw new HttpException("Unable to establish route: " +
                        "planned = " + route + "; current = " + fact);
            case HttpRouteDirector.COMPLETE:
                this.connManager.routeComplete(managedConn, route, context);
                break;
            default:
                throw new IllegalStateException("Unknown step indicator "
                        + step + " from RouteDirector.");
            }

        } while (step > HttpRouteDirector.COMPLETE);
    }

　　　最近遇到了一个问题，设置了最大超时时间为5s，但是由于触发了某种条件，监控显示Http调用有长达12s也能成功的，因为阅读了源码，所以马上想到了是连接池的PerRoute设置得太小了，导致同一域名的请求超过了这个设置，引起了线程的await事件，所以Http层面的超时时间只能保证通信的超时，如果触发了线程排队，那么设置的超时时间并没有生效，而且在源码中也没找到能快速失败的路径，正在寻找这一问题的解决方案。