Hi Luiz,
On Fri, Mar 01, 2013, Luiz Augusto von Dentz wrote:
> On Fri, Mar 1, 2013 at 1:20 PM, Johan Hedberg <johan.hedberg@xxxxxxxxx> wrote:
> > Hi Luiz,
> >
> > On Thu, Feb 28, 2013, Luiz Augusto von Dentz wrote:
> >> +static const char *type_to_string(uint8_t type)
> >> +{
> >> + switch (type) {
> >> + case PLAYER_ITEM_TYPE_AUDIO:
> >> + return "audio";
> >> + case PLAYER_ITEM_TYPE_VIDEO:
> >> + return "video";
> >> + case PLAYER_ITEM_TYPE_FOLDER:
> >> + return "folder";
> >> + }
> >> +
> >> + return NULL;
> >> +}
> >> +
> >>
> >> +static const char *folder_type_to_string(uint8_t type)
> >> +{
> >> + switch (type) {
> >> + case PLAYER_FOLDER_TYPE_MIXED:
> >> + return "mixed";
> >> + case PLAYER_FOLDER_TYPE_TITLES:
> >> + return "titles";
> >> + case PLAYER_FOLDER_TYPE_ALBUMS:
> >> + return "albums";
> >> + case PLAYER_FOLDER_TYPE_ARTISTS:
> >> + return "artists";
> >> + case PLAYER_FOLDER_TYPE_GENRES:
> >> + return "genres";
> >> + case PLAYER_FOLDER_TYPE_PLAYLISTS:
> >> + return "playlists";
> >> + case PLAYER_FOLDER_TYPE_YEARS:
> >> + return "years";
> >> + }
> >> +
> >> + return NULL;
> >> +}
> >> diff --git a/profiles/audio/player.h b/profiles/audio/player.h
> >> index 28689c5..9888f56 100644
> >> --- a/profiles/audio/player.h
> >> +++ b/profiles/audio/player.h
> >> @@ -23,6 +23,20 @@
> >> *
> >> */
> >>
> >> +#define PLAYER_ITEM_TYPE_AUDIO 0x00
> >> +#define PLAYER_ITEM_TYPE_VIDEO 0x01
> >> +#define PLAYER_ITEM_TYPE_FOLDER 0x02
> >> +#define PLAYER_ITEM_TYPE_INVALID 0xff
> >> +
> >> +#define PLAYER_FOLDER_TYPE_MIXED 0x00
> >> +#define PLAYER_FOLDER_TYPE_TITLES 0x01
> >> +#define PLAYER_FOLDER_TYPE_ALBUMS 0x02
> >> +#define PLAYER_FOLDER_TYPE_ARTISTS 0x03
> >> +#define PLAYER_FOLDER_TYPE_GENRES 0x04
> >> +#define PLAYER_FOLDER_TYPE_PLAYLISTS 0x05
> >> +#define PLAYER_FOLDER_TYPE_YEARS 0x06
> >> +#define PLAYER_FOLDER_TYPE_INVALID 0xff
> >
> > Why not enums? That'd let the compiler verify that all values are
> > covered in switch statements and since this is an internal API and not a
> > protocol detail there really isn't a need to use a fixed size integer
> > (uint8_t).
>
> Well I was expecting it should follow the AVRCP spec, otherwise we are
> always going to have to do 2 conversions: AVRCP <-> enum <-> string.
There is no AVRCP correspondence at least for the item type where you've
molded two protocol concepts together into one value. Also, using an
enum doesn't mean that you can't use protocol values.
Johan
--
To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
