Handle media access permission request

android_webview/android_webview_tests.gypi

Index: android_webview/android_webview_tests.gypi
diff --git a/android_webview/android_webview_tests.gypi b/android_webview/android_webview_tests.gypi
index 9ead6fe42b9bcd2a28ca0f367526adba48bc408a..c387bac7bdabeff37b0623d9cdad2ce57073794b 100644
--- a/android_webview/android_webview_tests.gypi
+++ b/android_webview/android_webview_tests.gypi
@@ -89,6 +89,7 @@
         '../ui/base/ui_base.gyp:ui_base_jni_headers',
         'android_webview_common',
         'android_webview_unittests_jni',
+        'libwebviewchromium',

[mkosiba (inactive), 2014/04/23 18:04:40]

I'm surprised it doesn't crash. - you've just introduced a 2nd entry point to
the test binary.

what error do you get without this? All the webview code the tests depend on
should already be part of android_webview_common.

[michaelbai, 2014/04/24 00:56:43]

oops, it should be removed, I were hoping it could avoid call RegisterJni. there
is no crash, why you think it is add 2nd entry point, were you talking about
Jni_onload?

On 2014/04/23 18:04:40, mkosiba wrote:

[mkosiba (inactive), 2014/04/24 09:24:58]

yes, testing/android/native_test_launcher.cc defines JNI_OnLoad which I thought
would be shadowed by the one from
android_webview/lib/main/webview_entry_point.cc

       ],
       'include_dirs': [
         '..',
@@ -101,6 +102,8 @@
         'browser/net/input_stream_reader_unittest.cc',
         'lib/main/webview_tests.cc',
         'native/input_stream_unittest.cc',
+        'native/permission/media_access_permission_request_unittest.cc',
+        'native/permission/permission_request_handler_unittest.cc',
         'native/state_serializer_unittest.cc',
       ],
     },