Currently, the message parsing functions require a contiguous block of memory (a &[u8]) to parse a message, but often, when MIDI messages arrive in a stream that all have the same status byte, only the first message will have its status byte set, and the same "running status" byte is assumed for following messages until it is changed. This means that consumers of this crate either need to do their own separate parsing for following messages, or need to allocate a Vec or stack array to parse the messages with the running status.
It'd be nice if there were a parse function that took a &[u8] and a separate status byte.
Currently, the message parsing functions require a contiguous block of memory (a &[u8]) to parse a message, but often, when MIDI messages arrive in a stream that all have the same status byte, only the first message will have its status byte set, and the same "running status" byte is assumed for following messages until it is changed. This means that consumers of this crate either need to do their own separate parsing for following messages, or need to allocate a Vec or stack array to parse the messages with the running status.
It'd be nice if there were a parse function that took a &[u8] and a separate status byte.