More control when specifying formats.

[scott-griffiths]

So for example a trailing r to indicate that the bitstring should be bit-wise reversed before the interpretation is used:

a, b = s.readlist('2*u4r') # Read two 4-bit bitstrings, reverse them and interpret as unsigned integers.

It would also have to work for unpack, but that should be fine.

How about ordinary interpretations?:

a = s[0:4].ur

which I don't hate. so there would be floatr, uintr, intr, bytesr, uer, sier, intner, float16r, float:64r, hexr, octr, binr, etc. etc.

And for setting:

s.binr = '0001' # So s.bin = 1000`

Also could possibly allow LSB0/MSB0 to be specified to override whatever the default has been set to be. How would that work?

a = s.read('u4msb0')

Fine, now s.pos = 4. So if we do another read but instead use lsb0:

b = s.read('u4lsb0')

where does it read from? Reading from pos=4 makes no sense as that is now relative to the other end of the bitstring. If we read from the same position then it will re-read the same bits which is a bit pointless. Overall I think that a bitstring's pos is invalidated by a change of LSB0/MSB0 so they can't be combined in this way or used in reads.

They could be used in unpack though as that deals with the whole bitstring.

a, b = s.unpack('2*4u', lsb0=True)

So this overrides the default in much the same was as bytealigned works.

This doesn't need to change interpretations as they don't care if it's LSB0 or MSB0.

Is there a way to override the default in slices? Instead of s[0:4] we actually could say s[4:0] but that's a big breaking change to change that meaning! And it would mean that s[0:4] would always be the left-most bits irrespective of the LSB0 flag. So to switch you'd need to recode all your slice indices - not very helpful.