Great, thanks. I saw a few bugs playing around a bit more:
1) Consecutive escape codes not fontified:
b'\x12\x23' # Here only "\x12" would be fontified
2) Multi-line bytes literals fontified as string literals:
b'''\x12 \777 \1\23 \u1234''' # Here '\u1234" would be (incorrectly) fontified
"\1 \12 \123" # Here only "\123" would be fontified
I've attached a patch which fixes the above and adds a new test. The new test (tests different combinations of multi-line literals) makes up the majority of the diff.
Thanks, Laurence