Added AudioBuffer/AudioChunk abstractions for speech recognition and improved speech_recognizer_impl_unittest.

audio_encoder  - Introduced AudioBuffer class in order to hide the current string-based implementation (which involved a lot of dirty and distributed casts) and make room for future implementations based on a circular buffer.

speech_recognizer_impl_unittest
 - Created MockAudioManager class, in order to avoid using the true audio manager on trybots, which could lead to errors accessing the audio device.

BUG=116954
TEST=speech_recognizer_impl_uinittest should never raise errors related to the audio driver (e.g, device in use, no microphone attached, etc).


Review URL: http://codereview.chromium.org/9646031

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@126512 0039d316-1c4b-4281-b951-d872f2087c98

1 files changed, 3 insertions, 1 deletions

diff --git a/content/browser/speech/speech_recognition_request.h b/content/browser/speech/speech_recognition_request.h
index b6ce077..2f29e0a 100644
--- a/content/browser/speech/speech_recognition_request.h
+++ b/content/browser/speech/speech_recognition_request.h
@@ -27,6 +27,8 @@ class URLRequestContextGetter;
 
 namespace speech {
 
+class AudioChunk;
+
 // Provides a simple interface for sending recorded speech data to the server
 // and get back recognition results.
 class SpeechRecognitionRequest : public content::URLFetcherDelegate {
@@ -61,7 +63,7 @@ class SpeechRecognitionRequest : public content::URLFetcherDelegate {
                             const std::string& content_type);
 
   // Send a single chunk of audio immediately to the server.
-  CONTENT_EXPORT void UploadAudioChunk(const std::string& audio_data,
+  CONTENT_EXPORT void UploadAudioChunk(const AudioChunk& audio_chunk,
                                        bool is_last_chunk);
 
   CONTENT_EXPORT bool HasPendingRequest() { return url_fetcher_ != NULL; }