def convert(url, mapping):
'''Convert given URL into affiliate URL based on mapping.
mapping maps domains with URL query paramater/value pairs to add or replace
in given URL. Parameter values must be specified as lists.
Example mapping for various amazon domains:
{
'amazon.com': {'tag': ['affurl-20']},
'amazon.de': {'tag': ['affurl-21']},
'amazon.co.uk': {'tag': ['afflink-21']}
}
'''
new_url = urlsplit(url)
if not new_url.netloc:
return None # rather raise an Exception?
# Parse_domain returns a tuple like ('co.uk', 'amazon', 'www').
domain = '.'.join(domain_parser.parse_domain(new_url.netloc)[:2][::-1])
# Leave URLs with no matching domain as they are.
if domain not in mapping:
return url
# Add new and replace existing query paramters with given ones.
query = parse_qs(new_url.query)
params = mapping[domain]
query.update(params)
# Concatenate and unsplit tuples to create a URL string.
return urlunsplit(new_url[:3] + (urlencode(query, True), ) + new_url[4:])
def test_google(self):
"""Is google.com properly parsed?"""
assert domain_parser.parse_domain('http://www.google.com') == (
'com', 'google', 'www')
def test_internationalized_domain_name(self):
"""Is 'маил.гоогле.рф', which is entirely composed of non-latin characters, parsed properly?"""
# Should always pass when run with Python 3.
assert domain_parser.parse_domain('http://маил.гоогле.рф') == (
'рф', 'гоогле', 'маил')
def test_secure_scheme(self):
"""Is 'https://www.google.com', which include 'https' instead of 'http', parsed properly?"""
assert domain_parser.parse_domain('https://www.google.com') == (
'com', 'google', 'www')
def test_no_scheme(self):
"""Is 'www.google.com', which doesn't include the scheme ('http'), parsed properly?"""
assert domain_parser.parse_domain('www.google.com') == ('com',
'google',
'www')
def test_guardian(self):
"""Is 'co.uk', which is wildcarded in the TLD list, parsed properly?"""
assert domain_parser.parse_domain('http://www.guardian.co.uk') == (
'co.uk', 'guardian', 'www')
def test_more_levels():
"""Is two levels sub-domain properly parsed?"""
assert domain_parser.parse_domain('www.staging.google.com') == (
'com', 'google', 'www.staging')
def test_no_subdomain():
"""Is twitter.com properly parsed?"""
assert domain_parser.parse_domain('twitter.com') == ('com', 'twitter',
'')
def test_google(self):
"""Is google.com properly parsed?"""
assert domain_parser.parse_domain(
'http://www.google.com') == ('com', 'google', 'www')
def test_secure_scheme(self):
"""Is 'https://www.google.com', which include 'https' instead of 'http', parsed properly?"""
assert domain_parser.parse_domain(
'https://www.google.com') == ('com', 'google', 'www')
def test_no_scheme(self):
"""Is 'www.google.com', which doesn't include the scheme ('http'), parsed properly?"""
assert domain_parser.parse_domain(
'www.google.com') == ('com', 'google', 'www')
def test_guardian(self):
"""Is 'co.uk', which is wildcarded in the TLD list, parsed properly?"""
assert domain_parser.parse_domain(
'http://www.guardian.co.uk') == ('co.uk', 'guardian', 'www')
def test_internationalized_domain_name(self):
"""Is 'маил.гоогле.рф', which is entirely composed of non-latin characters, parsed properly?"""
# Should always pass when run with Python 3.
assert domain_parser.parse_domain(
'http://маил.гоогле.рф') == ('рф', 'гоогле', 'маил')
