Switch LiveAudio to source provider solution.

content/renderer/media/webrtc_local_audio_source_provider.h

+// Copyright 2013 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_RENDERER_MEDIA_WEBRTC_LOCAL_AUDIO_SOURCE_PROVIDER_H_
+#define CONTENT_RENDERER_MEDIA_WEBRTC_LOCAL_AUDIO_SOURCE_PROVIDER_H_
+
+#include "base/memory/scoped_ptr.h"
+#include "base/synchronization/lock.h"
+#include "base/threading/thread_checker.h"
+#include "base/time/time.h"
+#include "content/common/content_export.h"
+#include "media/base/audio_converter.h"
+#include "third_party/WebKit/public/platform/WebVector.h"
+#include "third_party/WebKit/public/web/WebAudioSourceProvider.h"
+
+namespace media {
+class AudioBus;
+class AudioConverter;
+class AudioFifo;
+class AudioParameters;
+}
+
+namespace WebKit {
+class WebAudioSourceProviderClient;
+}
+
+namespace content {
+
+// WebRtcLocalAudioSourceProvider provides a bridge between classes:
+//     WebRtcAudioCapturer ---> WebKit::WebAudioSourceProvider
+//
+// WebRtcLocalAudioSourceProvider works as a sink to the WebRtcAudiocapturer
+// and store the capture data to a FIFO. When the media stream is connected to
+// WebAudio as a source provider, WebAudio will periodically call
+// provideInput() to get the data from the FIFO.
+//
+// All calls are protected by a lock.
+class CONTENT_EXPORT WebRtcLocalAudioSourceProvider
+    : NON_EXPORTED_BASE(public media::AudioConverter::InputCallback),
+      NON_EXPORTED_BASE(public WebKit::WebAudioSourceProvider) {
+ public:
+  WebRtcLocalAudioSourceProvider();
+  virtual ~WebRtcLocalAudioSourceProvider();
+
+  // Initialize function for the souce provider. This can be called multiple
+  // times if the source format has changed.

[tommi (sloooow) - chröme, 2013/09/12 11:02:16]

Multiple times for the same source?  what would be examples of when that would
happen?

[no longer working on chromium, 2013/09/12 19:14:56]

For example, Initialize() can be called again when the capturer is setting to
peer connection mode, where the capturer will use 10ms as buffer size then the
source_param will be changed.

+  void Initialize(const media::AudioParameters& source_params);
+
+  // Called by the WebRtcAudioCapturer to deliever captured data into fifo on
+  // the capture audio thread.
+  void DeliverData(media::AudioBus* audio_source,
+                   int audio_delay_milliseconds,
+                   int volume,
+                   bool key_pressed);
+
+  // Called by the WebAudioCapturerSource to get the audio processing params.
+  // This function is triggered by provideInput() on the WebAudio audio thread,
+  // so it has been under the protection of |lock_|.
+  void GetAudioProcessingParams(int* delay_ms, int* volume, bool* key_pressed);
+
+  // WebKit::WebAudioSourceProvider implementation.
+  virtual void setClient(WebKit::WebAudioSourceProviderClient* client) OVERRIDE;
+  virtual void provideInput(const WebKit::WebVector<float*>& audio_data,
+                            size_t number_of_frames) OVERRIDE;
+
+  // media::AudioConverter::Inputcallback implementation.
+  // This function is triggered by provideInput()on the WebAudio audio thread,
+  // so it has been under the protection of |lock_|.
+  virtual double ProvideInput(media::AudioBus* audio_bus,
+                              base::TimeDelta buffer_delay) OVERRIDE;
+
+  // Method to allow the unittests to inject its own sink parameters to avoid
+  // query the hardware.
+  void SetSinkParamsForTesting(const media::AudioParameters& sink_params);

[tommi (sloooow) - chröme, 2013/09/12 11:02:16]

Can you add:
TODO(xians,tommi): Remove and instead offer a way to inject the sink parameters
so that the implementation doesn't rely on the global default hardware config
but instead gets the parameters directly from the sink (WebAudio in this case). 
Ideally the unit test should be able to use that same mechanism to inject the
sink parameters for testing.

[no longer working on chromium, 2013/09/12 19:14:56]

Done.
I would love to inject a new interface to webkit::audiosourceprovider to do
this. but let me do it in a separate CL.

+
+ private:
+  // Used to DCHECK that we are called on the correct thread.
+  base::ThreadChecker thread_checker_;
+
+  scoped_ptr<media::AudioConverter> audio_converter_;
+  scoped_ptr<media::AudioFifo> fifo_;
+  scoped_ptr<media::AudioBus> bus_wrapper_;
+  int audio_delay_ms_;
+  int volume_;
+  bool key_pressed_;
+  bool is_enabled_;
+  media::AudioParameters source_params_;
+  media::AudioParameters sink_params_;
+
+  // Protects all the member variables above.
+  base::Lock lock_;
+
+  // Used to report the correct delay to |webaudio_source_|.
+  base::TimeTicks last_fill_;
+
+  DISALLOW_COPY_AND_ASSIGN(WebRtcLocalAudioSourceProvider);
+};
+
+}  // namespace content
+
+#endif  // CONTENT_RENDERER_MEDIA_WEBRTC_LOCAL_AUDIO_SOURCE_PROVIDER_H_