audio policy: enhancing type conversion helper

This patch updates the conversion helper class (string to/from policy
common types) by adding direction agnostic helper for channels and
formats.
It also adds the support of device category and stream type parsing.

Change-Id: I75d5a77d78bc3aeb2389b424bfb41febbf020f1a
Signed-off-by: François Gaffie <francois.gaffie@intel.com>
diff --git a/services/audiopolicy/common/include/policy.h b/services/audiopolicy/common/include/policy.h
index 4b73e3c..1a1cbb2 100755
--- a/services/audiopolicy/common/include/policy.h
+++ b/services/audiopolicy/common/include/policy.h
@@ -18,6 +18,10 @@
 
 #include <system/audio.h>
 
+static const uint32_t gDynamicRate = 0;
+static const audio_format_t gDynamicFormat = AUDIO_FORMAT_DEFAULT;
+static const audio_channel_mask_t gDynamicChannelMask = AUDIO_CHANNEL_NONE;
+
 // For mixed output and inputs, the policy will use max mixer sampling rates.
 // Do not limit sampling rate otherwise
 #define MAX_MIXER_SAMPLING_RATE 192000