Make the code faster by compiling regex on global context and reusing named collection.

[jaysonsantos]

Running the following code as benchmark

import timeit
import giturlparse

URLS = [
    '[email protected]:jaysonsantos/abc.git',
    'https://github.com/jaysonsantos/abc.git'
]

print(timeit.timeit(
    """
parsed = [giturlparse.parse(url) for url in URLS]
""",
    number=100_000,
    globals={
        'URLS': URLS,
        'giturlparse': giturlparse
    }))

on master it returned 20.684334864999997 and on my branch 1.4601109110000001

[retr0h]

Thanks @jaysonsantos you rock! \m/