Make SpeechRecognition per RenderFrame instead of per RenderView.

content/public/browser/speech_recognition_session_context.h

 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef CONTENT_PUBLIC_BROWSER_SPEECH_RECOGNITION_SESSION_CONTEXT_H_
 #define CONTENT_PUBLIC_BROWSER_SPEECH_RECOGNITION_SESSION_CONTEXT_H_
 
 #include <string>
 
 #include "content/common/content_export.h"
 #include "content/public/common/media_stream_request.h"
 #include "ui/gfx/rect.h"
 
 namespace content {
 
 // The context information required by clients of the SpeechRecognitionManager
 // and its delegates for mapping the recognition session to other browser
 // elements involved with it (e.g., the page element that requested the
 // recognition). The manager keeps this struct attached to the recognition
 // session during all the session lifetime, making its contents available to
 // clients (In this regard, see SpeechRecognitionManager::GetSessionContext and
 // SpeechRecognitionManager::LookupSessionByContext methods).
 struct CONTENT_EXPORT SpeechRecognitionSessionContext {
   SpeechRecognitionSessionContext();
   ~SpeechRecognitionSessionContext();
 
   int render_process_id;
-  int render_view_id;
   int render_frame_id;
 
-  // Browser plugin guest's render view id, if this context represents a speech
+  // Browser plugin guest's render frame id, if this context represents a speech
   // recognition request from an embedder on behalf of the guest. This is used
   // for input tag where speech bubble is to be shown.
   //
   // TODO(lazyboy): Right now showing bubble from guest does not work, we fall
   // back to embedder instead, fix this and use
-  // embedder_render_process_id/embedder_render_view_id similar to Web Speech
+  // embedder_render_process_id/embedder_render_frame_id similar to Web Speech
   // API below.
-  int guest_render_view_id;
+  int guest_render_frame_id;
 
-  // The pair (|embedder_render_process_id|, |embedder_render_view_id|)
+  // The pair (|embedder_render_process_id|, |embedder_render_frame_id|)
   // represents a Browser plugin guest's embedder. This is filled in if the
   // session is from a guest Web Speech API. We use these to check if the
   // embedder (app) is permitted to use audio.
   int embedder_render_process_id;
-  int embedder_render_view_id;
+  int embedder_render_frame_id;
 
   int request_id;
 
   // A texual description of the context (website, extension name) that is
   // requesting recognition, for prompting security notifications to the user.
   std::string context_name;
 
   // The label for the permission request, it is used for request abortion.
   std::string label;
 
   // A list of devices being used by the recognition session.
   MediaStreamDevices devices;
 };
 
 }  // namespace content
 
 #endif  // CONTENT_PUBLIC_BROWSER_SPEECH_RECOGNITION_SESSION_CONTEXT_H_