Add get_cmd_id_offsets helper function

[pipermerriam]

extracted from #684

What was wrong?

For a given set of protocols we need to compute the offsets for their command ids

How was it fixed?

Added a helper function for computing these offset values.

To-Do

• Clean up commit history

• Add entry to the release notes

Cute Animal Picture

[lithp]

Looks good to me!

[lithp]

It could be nice to include a link here to https://github.com/ethereum/devp2p/blob/master/rlpx.md#message-id-based-multiplexing, as explanation for why cmd length is 16 even though there are only four commands.

[lithp]

Now that I know what it does the code as-is makes sense but the easiest way for me to figure out what it was doing was to play with the implementation. Dropping what I came up here with for inspiration, I think a comment explaining why the [:-1] exists could be valuable.

def get_cmd_offsets(protocol_types: Tuple[Type[ProtocolAPI], ...]) -> Tuple[int, ...]:
    protocol_lengths = (protocol_type.cmd_length for protocol_type in protocol_types)
    offsets = accumulate(operator.add, protocol_lengths, P2P_PROTOCOL_COMMAND_LENGTH)
    return tuple(take(len(protocol_types), offsets))  # don't compute more offsets than necessary```

[lithp]

Also, I don't know how you plan on using this method, but it might be slightly cleaner and more general if, instead of hard-coding P2P_PROTOCOL_COMMAND_LENGTH, it was always passed P2PProtocol as its first argument.