HOCR python binary parser

[raphidoc]

The kaïtai python binary parser for HOCR takes ~3/4 min (on my machine), it would be good to speed up the process and remove dependency on python at the same time.

To do that we could use the cpp_stl version of the kaïtai parser, and use either:

1. Rcpp
2. cpp11

Which way to integrate C++ should be used?

My spec :
description: CPU
product: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz
vendor: Intel Corp.
physical id: 4
bus info: cpu@0
version: 6.158.10
serial: To Be Filled By O.E.M.
slot: U3E1
size: 4085MHz
capacity: 4500MHz
width: 64 bits
clock: 100MHz