Right now the parser requires to communicate in S1 mode, which is – albeit more readable – not as space efficient as possible. In particular when communicating with BTLE adapters, due to their inherently small bandwidths, every saved byte is a win.
It's not trivial to do this though since the current design relies on the space format to detect whether we have multi frames. Ideally we would feed the protocol a "normalized" form (no matter whether with or without spaces, we just have to decide on one) and take care about spaces or not on the parser level.
An added bonus would be to make the parser adaptive, i.e., to make it work with whatever input we feed, so that we would not have to configure to spaces-mode or non-spaces mode beforehand.
Right now the parser requires to communicate in
S1
mode, which is – albeit more readable – not as space efficient as possible. In particular when communicating with BTLE adapters, due to their inherently small bandwidths, every saved byte is a win.It's not trivial to do this though since the current design relies on the space format to detect whether we have multi frames. Ideally we would feed the protocol a "normalized" form (no matter whether with or without spaces, we just have to decide on one) and take care about spaces or not on the parser level.
An added bonus would be to make the parser adaptive, i.e., to make it work with whatever input we feed, so that we would not have to configure to spaces-mode or non-spaces mode beforehand.