def test_location_is_not_matched_if_it_is_not_flat_address(self):
mocked_address_provider = MockedAddressProvider(streets=[{
"official":
"Szeroka",
"colloquial": [],
}, {
"official":
"Karmelicka",
"colloquial": [],
}],
places=[{
"official":
"Ikea",
"colloquial": [],
}])
ctx_analyser = NearbyLocationContext(
introducers={'w sąsiedztwie'},
conjunctions={'i'},
address_provider=mocked_address_provider)
extractor = AddressExtractor(mocked_address_provider,
excluded_contexts=[ctx_analyser])
found_address = extractor(
"Mieszkanie znajduje się na ulicy Karmelickiej. W sąsiedztwie ul. Szeroka i Ikea"
)
self.assertIn("Karmelicka",
[str(match.location) for match in found_address.street])
self.assertEqual(1, len(found_address.all_addresses))
def test_address_extractor_correctly_recognize_location_type(self):
mocked_address_provider = MockedAddressProvider(streets=[{
"official":
"Stanisława",
"colloquial": [],
}],
estates=[{
"official":
"Grzegorza",
"colloquial": [],
}],
districts=[{
"official":
"Piotra",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor(
"blah blah Piotra blah Grzegorza blah Stanisława")
self.assertEqual(len(found_address.street), 1)
self.assertIn("Stanisława",
[str(match.location) for match in found_address.street])
self.assertEqual(len(found_address.estate), 1)
self.assertIn("Grzegorza",
[str(match.location) for match in found_address.estate])
self.assertEqual(len(found_address.district), 1)
self.assertIn(
"Piotra",
[str(match.location) for match in found_address.district])
def test_only_longest_location_from_overlapping_matches_is_returned(self):
with self.subTest():
mocked_address_provider = MockedAddressProvider(streets=[{
"official":
"Zygmunta Starego",
"colloquial": [],
}, {
"official":
"Stare Podgórze",
"colloquial": [],
}], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor(
"\nDo wynajęcia 1-pokojowe funkcjonalne mieszkanie w spokojnej, dobrze skomunikowanej"
" okolicy - Stare Podgórze przy ulicy Zamoyskiego, bardzo dobry dojazd do każdej części miasta."
)
names_of_matched_locations = [
str(match.location) for match in found_address.all
]
self.assertIn("Stare Podgórze", names_of_matched_locations)
self.assertNotIn("Zygmunta Starego", names_of_matched_locations)
with self.subTest():
mocked_address_provider = MockedAddressProvider(
streets=[{
"official": "Bronowicka",
"colloquial": [],
}],
places=[{
"official": "Galeria Bronowicka",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("Galeria Bronowicka")
names_of_matched_locations = [
str(match.location) for match in found_address.all
]
self.assertIn("Galeria Bronowicka", names_of_matched_locations)
self.assertNotIn("Bronowicka", names_of_matched_locations)
def test_address_extractor_performs_morphological_comparison(self):
mocked_address_provider = MockedAddressProvider(
streets=[{
"official": "Stanisława",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("Stanisławowi")
self.assertIn("Stanisława",
[str(match.location) for match in found_address.street])
def test_address_extractor_correctly_compares_names(self):
streets = [{
"official": "Tadeusza Kościuszki",
"colloquial": [],
}]
extractor = AddressExtractor(MockedAddressProvider(streets=streets))
found_address = extractor("Kościuszki")
self.assertIn("Tadeusza Kościuszki",
[str(match.location) for match in found_address.street])
extractor = AddressExtractor(MockedAddressProvider(streets=streets))
found_address = extractor("Tadeusza Kościuszki")
self.assertIn("Tadeusza Kościuszki",
[str(match.location) for match in found_address.street])
extractor = AddressExtractor(MockedAddressProvider(streets=streets))
found_address = extractor("Tadeusza")
self.assertNotIn(
"Tadeusza Kościuszki",
[str(match.location) for match in found_address.street])
def test_extract_address_with_unit_number(self):
mocked_address_provider = MockedAddressProvider(
streets=[{
"official": "Jana Zamoyskiego",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("Zamoyskiego 15")
self.assertIn("Jana Zamoyskiego 15",
[str(match.location) for match in found_address.street])
def test_case_does_not_matter_phrase_in_text_is_all_upper_case(self):
mocked_address_provider = MockedAddressProvider(
streets=[{
"official": "Śliczna",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("mieszkanie przy ulicy ŚLICZNEJ")
self.assertIn("Śliczna",
[str(match.location) for match in found_address.street])
def test_case_matters(self):
mocked_address_provider = MockedAddressProvider(
streets=[{
"official": "Śliczna",
"colloquial": [],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor(
"Oferuję do wynajęcia śliczne mieszkanie 4-pokojowe")
self.assertEqual(0, len(found_address.all))
def test_address_extractor_returns_official_name_if_colloquial_name_matched(
self):
mocked_address_provider = MockedAddressProvider(
estates=[{
"official": "Osiedle Na Kozłówce",
"colloquial": ["Kozłówek"],
}])
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("Kozłówek")
self.assertIn("Osiedle Na Kozłówce",
[match.location for match in found_address.estate])
def test_osiedle_street_is_not_matched_to_osiedle_location_prefix(self):
mocked_address_provider = MockedAddressProvider(streets=[
{
"official": "Osiedle",
"colloquial": [],
},
], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("Duże osiedle.")
self.assertNotIn("Osiedle",
[match.location for match in found_address.all])
def test_duplications_are_merged(self):
mocked_address_provider = MockedAddressProvider(districts=[
{
"official": "Nowa Huta",
"colloquial": [],
},
], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor(
'Dzielnica Nowa Huta. Mieszkanie się na Nowej Hucie')
self.assertEqual(1, len(found_address.all))
def test_actual_all_uppercase_bug(self):
mocked_address_provider = MockedAddressProvider(streets=[
{
"official": "Czyżyny",
"colloquial": [],
},
], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor('CZYŻYNY')
self.assertIn("Czyżyny",
[str(match.location) for match in found_address.all])
self.assertEqual(1, len(found_address.all))
def test_zl_is_not_matched_to_zlota_street(self):
mocked_address_provider = MockedAddressProvider(streets=[
{
"official": "Złota",
"colloquial": [],
},
], )
extractor = AddressExtractor(mocked_address_provider,
excluded_contexts=[PriceContext()])
found_address = extractor(
'czynsz najmu : 1600 zł + 553 ZŁ czynsz administracyjny + media .')
self.assertNotIn("Złota",
[match.location for match in found_address.all])
def test_Krakow_city_is_not_recognized_as_Kraka_street(self):
mocked_address_provider = MockedAddressProvider(streets=[{
"official":
"Kraka",
"colloquial": [],
}], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor("miasto Kraków")
self.assertEqual(0, len(found_address.all))
found_address = extractor("w Krakowie")
self.assertEqual(0, len(found_address.all))
def test_temp(self): # TODO remove
import logging
logging.root.setLevel(logging.NOTSET)
all_test_cases = self._load_regression_cases()
flat = all_test_cases[46]
flat['ignore_extra_matches'] = False
extractor = AddressExtractor(
address_provider,
excluded_contexts=[
FirstWordOfSentenceContext(),
NearbyLocationContext(address_provider=address_provider)
])
found_address = extractor(flat['title'] + '.\n' + flat['description'])
self._compare_address_results(flat, found_address)
def runner(flat):
try:
extractor = AddressExtractor(
address_provider,
excluded_contexts=[
FirstWordOfSentenceContext(),
NearbyLocationContext(
address_provider=address_provider),
PriceContext()
])
found_address = extractor(flat['title'] + '.\n' +
flat['description'])
return flat, found_address
except Exception as e:
trace = traceback.format_exc()
return None, Exception(str(e) + '\n' + trace)
def test_word_is_not_interpreted_as_location_if_it_is_first_word_of_a_sentence(
self):
mocked_address_provider = MockedAddressProvider(streets=[{
"official":
"Piękna",
"colloquial": [],
}], )
extractor = AddressExtractor(
mocked_address_provider,
excluded_contexts=[FirstWordOfSentenceContext()])
found_address = extractor("Jakieś zdanie. Piękna okolica.")
self.assertEqual(0, len(found_address.all))
found_address = extractor("Jakieś zdanie. Lokalizacja - Piękna 13")
self.assertNotEqual(0, len(found_address.all))
def does_introducer_refer_to_tested_location(self, introducer_subject):
address_extractor = AddressExtractor(self.address_provider)
matches = address_extractor(' '.join(introducer_subject))
found_addresses = chain(matches.street, matches.estate,
matches.district, matches.place)
match_slices = (address.match_slice_position
for address in found_addresses)
is_the_word_an_address_part_or_conjunction = [
False
] * len(introducer_subject)
# locations
for beg, end in match_slices:
for i in range(beg, end):
is_the_word_an_address_part_or_conjunction[i] = True
# conjunctions
for i in range(len(introducer_subject)):
if not is_the_word_an_address_part_or_conjunction[i]:
if introducer_subject[i] in self.conjunctions:
is_the_word_an_address_part_or_conjunction[i] = True
# location types
for i in range(len(introducer_subject)):
if not is_the_word_an_address_part_or_conjunction[i]:
if introducer_subject[i].lower(
) in self.location_type_prefixes:
is_the_word_an_address_part_or_conjunction[i] = True
try:
if introducer_subject[i + 1] == '.':
is_the_word_an_address_part_or_conjunction[
i + 1] = True
except IndexError:
pass
# newline character
for i in range(len(introducer_subject)):
if not is_the_word_an_address_part_or_conjunction[i]:
if introducer_subject[i] == '\n':
is_the_word_an_address_part_or_conjunction[i] = True
return all(is_the_word_an_address_part_or_conjunction)
def test_street_duplications_are_merged(self):
mocked_address_provider = MockedAddressProvider(streets=[
{
"official": "Mogilska",
"colloquial": [],
},
], )
extractor = AddressExtractor(mocked_address_provider)
found_address = extractor(
'Mieszkanie przy ulicy Mogilskiej. Adres Mogilska 66')
self.assertIn("Mogilska 66",
[str(match.location) for match in found_address.all])
self.assertEqual(1, len(found_address.all))
found_address = extractor('Mieszkanie przy ulicy Mogilskiej')
self.assertIn("Mogilska",
[str(match.location) for match in found_address.all])
self.assertEqual(1, len(found_address.all))
"Mistrzejowice",
"Swoszowice",
"Ruczaj",
'Eliasza Radzikowskiego',
'Aleja 29 Listopada',
'Dobrego Pasterza',
'Żabiniec',
'Jana Sobieskiego',
'Białoprądnicka',
])
extractors = [
AddressExtractor(
address_provider,
excluded_contexts=[
FirstWordOfSentenceContext(),
NearbyLocationContext(address_provider=address_provider),
PriceContext()
]),
InterconnectingRoomExtractor(),
KitchenetteExtractor(),
BachelorPadExtractor()
]
two_room_cfg = {'price_low': 800, 'price_high': 1500, 'room': 2}
three_room_cfg = {'price_low': 1000, 'price_high': 2200, 'room': 3}
mgr = ScrappingManager(check_interval_in_seconds=1 * 60 * 60,
filters=[
without_kitchenette,
