Convert channel counts to size_t.

talk/media/webrtc/fakewebrtcvoiceengine.h

Index: talk/media/webrtc/fakewebrtcvoiceengine.h
diff --git a/talk/media/webrtc/fakewebrtcvoiceengine.h b/talk/media/webrtc/fakewebrtcvoiceengine.h
index bf22a290b89673cc245ff05d717ac94d161433e3..5fc2f485190c14b9ca5a2211f2f7f40b273e71dc 100644
--- a/talk/media/webrtc/fakewebrtcvoiceengine.h
+++ b/talk/media/webrtc/fakewebrtcvoiceengine.h
@@ -77,9 +77,9 @@ class FakeAudioProcessing : public webrtc::AudioProcessing {
   WEBRTC_STUB_CONST(input_sample_rate_hz, ());
   WEBRTC_STUB_CONST(proc_sample_rate_hz, ());
   WEBRTC_STUB_CONST(proc_split_sample_rate_hz, ());
-  WEBRTC_STUB_CONST(num_input_channels, ());
-  WEBRTC_STUB_CONST(num_output_channels, ());
-  WEBRTC_STUB_CONST(num_reverse_channels, ());
+  size_t num_input_channels() const override { return 0; }

[perkj_webrtc, 2016/01/07 09:32:32]

I am not sure I understand the purpose of WEBRTC_STUB_CONST.. Should there be a
macro for size_t as well to keep the same pattern?
solenberg - I guess that is your decision if you know own the voice engine
implementation.

[Peter Kasting, 2016/01/07 09:47:50]

AFAICT it's just to reduce verbosity.  I don't see much benefit to adding
another macro form for "const function which returns size_t" when that would
literally just be used here.

Frankly, given the style guide's direction to avoid macros when possible, I
think all the macros here should be removed and replaced with normal function
declarations like we use almost everywhere else, but I'm not going to try to
write that change and push it through.

[the sun, 2016/01/07 12:27:59]

Don't worry; this is going away. Don't create any new macros.

+  size_t num_output_channels() const override { return 0; }
+  size_t num_reverse_channels() const override { return 0; }
   WEBRTC_VOID_STUB(set_output_will_be_muted, (bool muted));
   WEBRTC_STUB(ProcessStream, (webrtc::AudioFrame* frame));
   WEBRTC_STUB(ProcessStream, (