谷歌浏览器的源码分析 26

Posted 2021-02-07 skiwnchh

 

消息的流通过程，是一个不同类相互交流的过程，如果不了解这个过程，根本就不知道这些类是怎么样相互协作的。由于上一次说到ViewHostMsg_RequestResource消息已经发送出来，它的处理过徎其实就是一般资源的消息处理过程，下面就来看看这个消息的处理过程，如下：<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

 

1.       base::MessagePumpWin::Run函数消息

 

2.       base::MessagePumpWin::RunWithDispatcher

 

3.       base::MessagePumpWin::DoRunLoop

 

4.       base::MessagePumpWin::WaitForWork

 

5.       base::MessagePumpWin::SignalWatcher

 

6.       IPC::Channel::OnObjectSignaled 通道的消息处理。

 

7.       IPC::Channel::ProcessIncomingMessages

 

8.       IPC::SyncChannel::SyncContext::OnMessageReceived

 

9.       IPC::ChannelProxy::Context::TryFilters

 

10.    ResourceMessageFilter::OnMessageReceived

 

11.    IPC::MessageWithTuple<Tuple2<int,ViewHostMsg_Resource_Request> >::Dispatch<ResourceMessageFilter,int,ViewHostMsg_Resource_Request const &>

 

12.    ResourceMessageFilter::OnRequestResource

 

13.    ResourceDispatcherHost::BeginRequest

 

 

 

消息通过上面12个函数的流转，到达到处理这个消息的函数ResourceDispatcherHost::BeginRequest，它在里面就会把消息变成一个网络下载请求，函数代码如下：

 

#001  void ResourceDispatcherHost::BeginRequest(

 

#002      Receiver* receiver,

 

#003      HANDLE render_process_handle,

 

#004      int render_process_host_id,

 

#005      int render_view_id,

 

#006      int request_id,

 

#007      const ViewHostMsg_Resource_Request& request_data,

 

#008      URLRequestContext* request_context,

 

#009      IPC::Message* sync_result) {

 

 

 

是否已经传送完成，或者关闭。

 

#010    if (is_shutdown_ ||

 

#011        !ShouldServiceRequest(render_process_host_id, request_data)) {

 

#012      // Tell the renderer that this request was disallowed.

 

#013      receiver->Send(new ViewMsg_Resource_RequestComplete(

 

#014          render_view_id,

 

#015          request_id,

 

#016          URLRequestStatus(URLRequestStatus::FAILED, net::ERR_ABORTED)));

 

#017      return;

 

#018    }

 

#019 

 

 

 

保证所有插件已经加载。

 

#020    // Ensure the Chrome plugins are loaded, as they may intercept network

 

#021    // requests.  Does nothing if they are already loaded.

 

#022    // TODO(mpcomplete): This takes 200 ms!  Investigate parallelizing this by

 

#023    // starting the load earlier in a BG thread.

 

#024    plugin_service_->LoadChromePlugins(this);

 

#025 

 

 

 

构造事件处理器。

 

#026    // Construct the event handler.

 

#027    scoped_refptr<EventHandler> handler;

 

#028    if (sync_result) {

 

#029      handler = new SyncEventHandler(receiver, request_data.url, sync_result);

 

#030    } else {

 

#031      handler = new AsyncEventHandler(receiver,

 

#032                                      render_process_host_id,

 

#033                                      render_view_id,

 

#034                                      render_process_handle,

 

#035                                      request_data.url,

 

#036                                      this);

 

#037    }

 

#038 

 

#039    if (HandleExternalProtocol(request_id, render_process_host_id, render_view_id,

 

#040                               request_data.url, request_data.resource_type,

 

#041                               handler)) {

 

#042      return;

 

#043    }

 

#044 

 

 

 

构造下载请求。

 

#045    // Construct the request.

 

#046    URLRequest* request = new URLRequest(request_data.url, this);

 

#047    request->set_method(request_data.method);

 

#048    request->set_policy_url(request_data.policy_url);

 

#049    request->set_referrer(request_data.referrer.spec());

 

#050    request->SetExtraRequestHeaders(request_data.headers);

 

#051    request->set_load_flags(request_data.load_flags);

 

#052    request->set_context(request_context);

 

#053    request->set_origin_pid(request_data.origin_pid);

 

#054 

 

 

 

设置上传数据。

 

#055    // Set upload data.

 

#056    uint64 upload_size = 0;

 

#057    if (!request_data.upload_content.empty()) {

 

#058      scoped_refptr<net::UploadData> upload = new net::UploadData();

 

#059      upload->set_elements(request_data.upload_content);  // Deep copy.

 

#060      request->set_upload(upload);

 

#061      upload_size = upload->GetContentLength();

 

#062    }

 

#063 

 

 

 

安装一个CrossSiteEventHandler事件处理器。

 

#064    // Install a CrossSiteEventHandler if this request is coming from a

 

#065    // RenderViewHost with a pending cross-site request.  We only check this for

 

#066    // MAIN_FRAME requests.

 

#067    // TODO(mpcomplete): remove "render_process_host_id != -1"

 

#068    //                   when http://b/viewIssue?id=1080959 is fixed.

 

#069    if (request_data.resource_type == ResourceType::MAIN_FRAME &&

 

#070        render_process_host_id != -1 &&

 

#071        Singleton<CrossSiteRequestManager>::get()->

 

#072            HasPendingCrossSiteRequest(render_process_host_id, render_view_id)) {

 

#073      // Wrap the event handler to be sure the current page‘s onunload handler

 

#074      // has a chance to run before we render the new page.

 

#075      handler = new CrossSiteEventHandler(handler,

 

#076                                          render_process_host_id,

 

#077                                          render_view_id,

 

#078                                          this);

 

#079    }

 

#080 

 

#081    if (safe_browsing_->enabled() &&

 

#082        safe_browsing_->CanCheckUrl(request_data.url)) {

 

#083      handler = new SafeBrowsingEventHandler(handler,

 

#084                                             render_process_host_id,

 

#085                                             render_view_id,

 

#086                                             request_data.url,

 

#087                                             request_data.resource_type,

 

#088                                             safe_browsing_,

 

#089                                             this);

 

#090    }

 

#091 

 

 

 

创建一个缓冲区处理。

 

#092    // Insert a buffered event handler before the actual one.

 

#093    handler = new BufferedEventHandler(handler, this, request);

 

#094 

 

#095    // Make extra info and read footer (contains request ID).

 

#096    ExtraRequestInfo* extra_info =

 

#097        new ExtraRequestInfo(handler,

 

#098                             request_id,

 

#099                             render_process_host_id,

 

#100                             render_view_id,

 

#101                             request_data.mixed_content,

 

#102                             request_data.resource_type,

 

#103                             upload_size);

 

#104    extra_info->allow_download =

 

#105        ResourceType::IsFrame(request_data.resource_type);

 

#106    request->set_user_data(extra_info);  // takes pointer ownership

 

#107 

 

 

 

开始调用内部处理请求函数。

 

#108    BeginRequestInternal(request, request_data.mixed_content);

 

#109  }

 

 

 

通过上面的分析，已经知道消息转换为一个请求任务URLRequest，这个任务就需要交给后面的工作进程来处理了，它是通过函数BeginRequestInternal来把任务进一步发送出去。

 

再分享一下我老师大神的人工智能教程吧。零基础！通俗易懂！风趣幽默！还带黄段子！希望你也加入到我们人工智能的队伍中来！https://blog.csdn.net/jiangjunshow