扫清盲点，如何正确的从HttpClient 3.x系统升级到HttpClient 4.x

如果周期比较长的项目，或者这个项目开发人员换过了好几拨人，很有可能出现一些奇怪的问题，比如一个项目中出现了多种Spring注入bean的方式，不同版本的jar冲突等等

爬虫项目有的时候更是过犹不及，拿模拟登陆来说，开发人员的迭代，每个人的风格和技术各不相同，模拟登陆的方式也是五花八门，早在之前看到过一个项目的源码，其中使用HttpClient也是各种风格，虽然官方已经强烈建议使用HttpClient 4.x之后版本，但是在该项目中均存在HttpClient 3.x的和HttpClient 4.x的创建Httpclient对象。造成风格不一，维护繁杂，而新进项目的同学又很难进行区分不易上手，是有可能继续使用旧代码造成不良循环。

本文针对以上的情况进行了详细讲述：

1.   关于HttpClient 3.x 和 HttpClient 4.x项目的历史

如下图所示：

原Commons HttpClient project(HttpClient 3.x) 是 Jakarta Commons 项目下的一个子项目，在2007年Commons项目独立，Commons摒弃了Httpclient 3.x，不再对其维护；而在早些年2005年Jakarta成立了子项目HttpComponents项目用于开发HttpClients 3.x的替代项目HttpClient 4.x。就在Commons项目脱离Jakarta项目的同年2007年，HttpComponents也脱离Jakarta项目成为Apache顶级项目，进行开发维护HttpClient 4.x及后续版本。

从Httpclient历史中可以看出早在2005年Apache就有了要取代3.x的打算，成立了单独项目HttpComponents，并在两年后发布4.x版本取代了3.x。

Apache官方公告链接：https://hc.apache.org/httpclient-3.x/

2.   区分HttpClient 3.x 和 HttpClient 4.x 的jar包

如下图所示，如果是HttpClient 3.x 在项目中需要引入 import org.apache.commons.httpclient。而HttpClient 4.x 则需要引入 import org.apache.http.client.HttpClient。

另外关于3.x和4.x的jar包依赖，在上图中可以看出，3.x的jar依赖于commons-logging和commons-codec。而4.x依赖于httpcore、commons-logging和commons-codec。从下图的maven依赖中也可以清晰的看出区分。

3.   HttpClient 3.x 升级到 4.x 的参照表总结如下：

	Commons HttpClient 3.x	HttpComponents HttpClient 4.x
import	import org.apache.commons.httpclient.Cred;font-family:Consolas, SimSun, Helvetica, sans-serif;font-size:13px;entials; import org.apache.commons.httpclient.Header; import org.apache.commons.httpclient.HostConfiguration; import org.apache.commons.httpclient.HttpClient; import org.apache.commons.httpclient. MultiThreadedHttpConnectionManager; import org.apache.commons.httpclient. UsernamePasswordCredentials; import org.apache.commons.httpclient.auth.AuthScope; import org.apache.commons.httpclient.methods.DeleteMethod; import org.apache.commons.httpclient.methods.GetMethod; import org.apache.commons.httpclient.methods.PostMethod;	import org.apache.http.Header; import org.apache.http.HttpEntity; import org.apache.http.HttpHost; import org.apache.http.HttpResponse; import org.apache.http.auth.AuthScope; import org.apache.http.auth.Credentials; import org.apache.http.auth.UsernamePasswordCredentials; import org.apache.http.client.HttpClient; import org.apache.http.client.methods.HttpGet; import org.apache.http.client.methods.HttpPost; import org.apache.http.conn.params.ConnRoutePNames; import org.apache.http.entity.ContentType; import org.apache.http.entity.StringEntity; import org.apache.http.impl.client.DefaultHttpClient; import org.apache.http.util.EntityUtils;
HttpClient	HttpClient client = new HttpClient(connectionManager);	DefaultHttpClient client = new DefaultHttpClient(); // 4.3版本前 // 4.3版本后的创建方式见表下方代码
proxy configuration	client.getHostConfiguration().setProxy( proxyHost, proxyPort); client.getState().setAuthenticationPreemptive( true); Credentials cred = new  UsernamePasswordCredentials(proxyUser, proxyPassword); client.getState().setProxyCredentials(AuthScope.ANY_HOST, proxyHost, cred)	client.getCredentialsProvider().setCredentials(  new AuthScope(proxyHost, proxyPort),  new UsernamePasswordCredentials(proxyUser, proxyPassword)); HttpHost proxy = new HttpHost(proxyHost proxyPort); client.getParams().setParameter( ConnRoutePNames.DEFAULT_PROXY, proxy);
GET	GetMethod getMethod = new GetMethod(); getMethod.setPath(url.getFile());	HttpGet httpGet = new HttpGet(url.getFile());
GET Header	getMethod.setRequestHeader(key, value);	httpGet.addHeader(key, value);
Execute GET	client.executeMethod(getMethod);	HttpHost targetHost = new HttpHost(url.getHost(), url.getPort(), url.getProtocol()); HttpResponse httpResponse = client.execute( targetHost, httpGet);
POST	PostMethod post = new PostMethod(); post.setPath(url.getFile());	HttpPost post = new HttpPost(url.getFile());
POST Header	post.setRequestHeader(key, (String) headers.get(key));	post.addHeader(key, (String) headers.get(key));
POST Body	post.setRequestBody(message);	StringEntity messageEntity = new StringEntity( message, ContentType.create("text/plain", "UTF-8")); post.setEntity(messageEntity);
Execute POST	client.executeMethod(post);	HttpHost targetHost = new HttpHost(url.getHost(), url.getPort(), url.getProtocol()); HttpResponse httpResponse = client.execute(targetHost, post);
Response Header	Header[] rspHeaders = post.getResponseHeaders();	Header[] rspHeaders = httpResponse.getAllHeaders();
Response Body	String responseMsg = post.getResponseBodyAsString()	StringBuffer buffer = new StringBuffer(); BufferedReader reader = new BufferedReader(new InputStreamReader(httpResponse.getEntity().getContent())); String dataLine = null; while((dataLine = reader.readLine()) != null){     buffer.append(dataLine); } String responseMsg =    buffer.toString();

在上表中需要注意的是：如果是HttpClient 4.3之后的版本，创建HttpClient对象的方法有如下几种：

// 一般默认的创建HC对象的方法，实际上底层实现为:HttpClientBuilder.create().build()
CloseableHttpClient httpClient = HttpClients.createDefault(); 
 
// HttpClientBuilder 创建HC对象的各种方式
CloseableHttpClient closeableHttpClient = new HTTPClientBuilder().build();
 
CloseableHttpClient closeableHttpClient = new HTTPClientBuilder().withCookiesStore().build();
 
CloseableHttpClient closeableHttpClient = new HTTPClientBuilder().withGlobalConfig().build();
 
CloseableHttpClient closeableHttpClient = new HTTPClientBuilder().withSllContext().build();
 
CloseableHttpClient closeableHttpClient = new HTTPClientBuilder().withCookiesStore()
    .withGlobalConfig()
    .withSllContext()
    .build();

4.   HttpClient 3.x 和 4.x 废弃API一览：

在每个版本的HttpClient中都有Deprecated list（废弃API一览），下面的官方连接中可以方便查询到官方不建议使用的API，这些API如果继续使用会和JDK的API一样出现删除线。

3.x 版本 https://hc.apache.org/httpclient-3.x/apidocs/deprecated-list.html
4.x 版本 https://hc.apache.org/httpcomponents-client-ga/httpclient/apidocs/deprecated-list.html

5.   HttpClient 3.x 和 4.x 的常量变化一览：

在无论是3.x还是4.x的版本中，默认都定义了常量文件，里面提供了默认的状态码，协议头等等的常量，这样一些常用的就不需要自己再次定义了，可以直接使用HttpClient的常量类直接来使用，具体的详细常量参照下面的链接：

3.x 版本 https://hc.apache.org/httpclient-3.x/apidocs/constant-values.html
4.x 版本 https://hc.apache.org/httpcomponents-client-ga/httpclient/apidocs/constant-values.html

6.   HttpClient 3.x 和 4.x 超时Timeout设置变化：

HttpClient 3.x

HttpClient client = new HttpClient();
client.setConnectionTimeout(5000);  
client.setTimeout(5000);
 
HttpClient httpClient= new HttpClient();  
httpClient.getHttpConnectionManager().getParams().setConnectionTimeout(5000);

HttpClient 4.x到HttpClient4.3以下

HttpClient httpClient=new DefaultHttpClient();
httpClient.getParams().setParameter(CoreConnectionPNames.CONNECTION_TIMEOUT,5000);
httpClient.getParams().setParameter(CoreConnectionPNames.SO_TIMEOUT,5000);

HttpClient4.3以上

CloseableHttpClient httpClient = HttpClients.createDefault();
HttpGet httpGet = new HttpGet("https://www.csdn.net/");
RequestConfig requestConfig = RequestConfig.custom().setSocketTimeout(5000).setConnectTimeout(5000).build();
httpGet.setConfig(requestConfig);
httpClient.execute(httpGet);

7.   对于HttpClient 4.5.x 之后的版本废弃API版本替代策略（针对SSL相关API）

HttpClient4.5之后依然有很多的变化，其中一些API也废弃掉了。
下文针对 4.x 以后 SSL证书验证相关废弃（@Deprecated）API提供了最新的替换策略。

HttpClient 4.5.x 之后 Deprecated 废弃API 的替代对应策略：https://blog.csdn.net/dietime1943/article/details/86705539

8.   其他一些有用的链接：

HttpClient官方指南教程：https://hc.apache.org/httpcomponents-client-ga/tutorial/html/index.html

What does setDefaultMaxPerRoute and setMaxTotal mean in HttpClient?
https://stackoverflow.com/questions/30689995/what-does-setdefaultmaxperroute-and-setmaxtotal-mean-in-httpclient

注：本文原创由`bluetata`发布于blog.csdn.net、转载请务必注明出处。