Add workaround for type change in webrtc api.

content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc

Index: content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc
diff --git a/content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc b/content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc
index 30e0125c40f634fa245f98f9b0bb6a5a29f78194..936506459e041b03c1ce58edf6ec10bdb4721ef2 100644
--- a/content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc
+++ b/content/renderer/media/webrtc/webrtc_video_capturer_adapter_unittest.cc
@@ -105,7 +105,10 @@ class WebRtcVideoCapturerAdapterTest
 
     // Request smaller scale to make sure scaling normally kicks in.
     rtc::VideoSinkWants wants;
-    wants.max_pixel_count = rtc::Optional<int>(kInputWidth * kInputHeight / 2);
+    // TODO(sprang): Remove this type hack when webrtc has updated the sink
+    // wants api. https://codereview.webrtc.org/2781433002/
+    using MaxPixelCountType = decltype(wants.max_pixel_count);
+    wants.max_pixel_count = MaxPixelCountType(kInputWidth * kInputHeight / 2);
     adapter_->AddOrUpdateSink(this, wants);
 
     adapter_->OnFrameCaptured(frame);