Make as a C module (with Cython)

[suminb]

Our preliminary test shows promising results. Building base62 with Cython results in great improvement in performance.

Test code

import random

import base62


for _ in range(1000000):
    value = random.randint(0, 0xffffffff) 
    encoded = base62.encode(value)
    assert base62.decode(encoded) == value

Test with native Python (3.5)

(pybase62) ➜  base62 git:(develop) ✗ time python test.py
python test.py  8.41s user 0.00s system 99% cpu 8.420 total

Test with Cython

(pybase62) ➜  base62 git:(develop) ✗ time python test.py
python test.py  6.63s user 0.00s system 99% cpu 6.636 total

That's approximately 21% of speed improvement. More thorough tests shall be conducted in the near future.

[amcgregor]

For comparison, running on a 2.7GHz i7 MacBook Pro:

Python 3.6.2

# time python test.py
python tmp.py  9.79s user 0.03s system 99% cpu 9.835 total

PyPy 5.8.0 (≅ CPython 3.5.3)

# time pypy3 test.py
pypy3 tmp.py  2.52s user 0.04s system 82% cpu 3.116 total

[suminb]

Thank you for your test results. PyPy is surely a better option for most cases. However, I'm using base62 for my work and we are unfortunately relying on some PyPy-incompatible packages :-(

I would recommend PyPy (or perhaps Pyston?) for anyone looking for performance improvement for base62.