url.py

from urllib.parse import urlparse, urlunparse

"""
url helpers
"""

def isurl(o):
    return o.startswith('http://') \
        or o.startswith('https://')

def isabsolute(url):
    p = urlparse(url)
    return p.scheme is not '' and p.netloc is not ''

def reconciliate_path(np, op, sep='/'):
    if np.startswith(sep):
        return np
    else:
        lnp = np.split(sep)
        lop = op.split(sep)
        rnp = lop[:]
        for e in lnp:
            if e == '.':
                pass
            elif e == '..':
                if len(rnp) > 1:
                    rnp.pop()
            else:
                rnp.append(e)
        # crude fix
        r = sep.join(rnp)
        return r if r.startswith(sep) else sep + r
            
def normalize(new, old):
    """turns an href into an absolute url if possible"""
    if isabsolute(new):
        return new
    else:
        np = urlparse(new)
        op = urlparse(old)
        rp = reconciliate_path(np.path, op.path)
        nu = f'{op.scheme}://{op.netloc}{rp}{np.params}{np.query}{np.fragment}'
        return nu