Supporting bit decoding and encoding

I'd like to specify the following:

BinaryStruct.decode([
  'L', :some_num,
  'b', :flag1,
  'b', :flag2,
  'b', :flag3,
  'L', :another_num,
])

Given that, I'd expect BinaryStruct.decode to extract a number, then on the next byte extract the first 3 bits as booleans stored in :flag1, :flag2, :flag3, then move ahead to the next byte and extract another number.

I'd probably be happy with having to align my definition to byte boundaries, so perhaps...

BinaryStruct.decode([
  'L', :some_num,
  'b', :flag1,
  'b', :flag2,
  'b', :flag3,
  'b5', :unused, # ideally optional
  'L', :another_num,
])

I would expect encode and size to handle similar structures.

Implementation wise, the 'b' identifier reads 1 bit then skips to the next byte, So, when building the decode string it will have to be aware of successive bits and then pack those into a b8.

Start from the BinaryStruct.decode, encode, and size entry points and inspect how format identifiers are currently parsed and packed. Define support for successive bit fields with optional padding and byte alignment, then verify that decoding, encoding, and size calculation agree for the structures shown in the issue.