def testIssue49(self):
'''
country recognition
'''
show = self.debug
texts = [
'United Kingdom', 'UK', 'Great Britain', 'GB', 'United States'
]
expected = [
"United Kingdom", "United Kingdom", "United Kingdom",
"United Kingdom", "United States of America"
]
if show:
print("lookup with geograpy.get_geoPlace_context")
for text in texts:
countries = geograpy.get_geoPlace_context(text=text).countries
if show:
print(f"{text}:{countries}")
if show:
print("lookup with PlaceContext")
for i, text in enumerate(texts):
pc = PlaceContext([text])
pc.set_countries()
if show:
print(f"{text}:{pc.countries}")
self.assertEqual([expected[i]], pc.countries)
def get_place_context(url=None, text=None):
e = Extractor(url=url, text=text)
e.find_entities()
pc = PlaceContext(e.places)
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
return pc
def testGetRegionNames(self):
'''
test getting region names
'''
pc = PlaceContext(place_names=["Berlin"])
regions = pc.getRegions("Germany")
self.assertEqual(16, len(regions))
for region in regions:
if self.debug:
print(region)
self.assertTrue(region.iso.startswith("DE"))
regionNames = pc.get_region_names("Germany")
self.assertEqual(16, len(regionNames))
if self.debug:
print(regionNames)
def get_place_context(url=None, text=None):
e = Extractor(url=url, text=text)
e.find_entities()
pc = PlaceContext(e.places, e.people, e.organs)
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
return pc
# url = 'http://www.bbc.com/news/world-us-canada-39821789'
# places = get_place_context(url=url)
# len(places)
def testIssue25(self):
'''
https://github.com/somnathrakshit/geograpy3/issues/25
'''
pc = PlaceContext(
place_names=["Bulgaria", "Croatia", "Czech Republic", "Hungary"])
if self.debug:
print(pc.countries)
def get_place_context(url=None, text=None, labels=Labels.default, debug=False):
'''
Get a place context for a given text with information
about country, region, city and other
based on NLTK Named Entities in the label set Geographic(GPE),
Person(PERSON) and Organization(ORGANIZATION).
Args:
url(String): the url to read text from (if any)
text(String): the text to analyze
debug(boolean): if True show debug information
Returns:
pc:
PlaceContext: the place context
'''
e = Extractor(url=url, text=text, debug=debug)
e.find_entities(labels=labels)
pc = PlaceContext(e.places)
pc.setAll()
return pc
def testPlaces(self):
'''
test places
'''
pc = PlaceContext(['Ngong', 'Nairobi', 'Kenya'],setAll=False)
pc.setAll()
if self.debug:
print (pc)
self.assertEqual(1,len(pc.countries))
self.assertEqual("Kenya",pc.countries[0])
assert len(pc.cities) == 1
# assert len(pc.other) == 1
# assert 'Ngong' in pc.other
assert pc.cities_for_name('Nairobi')[0][4] == 'Kenya'
assert pc.regions_for_name('Ohio')[0][4] == 'United States'
pc = PlaceContext(['Mumbai'])
if self.debug:
print(pc)
def testPlaces(self):
'''
test places
'''
pc = PlaceContext(['Ngong', 'Nairobi', 'Kenya'], setAll=False)
pc.setAll()
if self.debug:
print(pc)
self.assertEqual(1, len(pc.countries))
self.assertEqual("Kenya", pc.countries[0])
self.assertEqual(2, len(pc.cities))
cityNames = ['Nairobi', 'Ohio', 'Amsterdam']
countries = ['Kenya', 'United States', 'Netherlands']
for index, cityName in enumerate(cityNames):
cities = pc.cities_for_name(cityName)
country = cities[0].country
self.assertEqual(countries[index], country.name)
pc = PlaceContext(['Mumbai'])
if self.debug:
print(pc)
def testPlaces(self):
'''
test places
'''
pc = PlaceContext(['Ngong', 'Nairobi', 'Kenya'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
if self.debug:
print("countries=%s" % pc.countries)
print("cities=%s" % pc.cities)
assert len(pc.countries) == 3
assert len(pc.cities) == 1
# assert len(pc.other) == 1
# assert 'Ngong' in pc.other
assert pc.cities_for_name('Nairobi')[0][4] == 'Kenya'
assert pc.regions_for_name('Ohio')[0][4] == 'United States'
pc = PlaceContext(['Mumbai'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
if self.debug:
print(pc)
stop_regions = {
'city': [
'sign', 'home', 'shop', 'us', 'about', 'unknown', 'industry', 'other',
'welcome', 'contact', 'jobs', 'job', 'staff', 'chesapeake', 'find',
'laboratory', 'services', 'email', 'register', 'company', 'locate',
'zip', 'close', 'else', 'city', 'make', 'model', 'yes', 'login',
'location', 'are', 'share', 'log', 'hours', 'remember', 'force',
'cart', 'our', 'telephone', 'posts', 'those', 'sort', 'many', 'even'
],
'country': ['at', 'by', 'to'],
'state': ['follow', 'at', 'by', 'to', 'and'],
'city_district': [],
'suburb': []
}
keys = ['suburb', 'city_district', 'city', 'state', 'country']
pc = PlaceContext('')
db_locations = {}
secondary_keys = [
'house_number', 'road', 'unit', 'po_box', 'postcode', 'suburb'
]
ignored = [
'house', 'near', 'category', 'level', 'island', 'country_region',
'world_region'
]
def get_key(dict_keys, loc):
loc_key = ''
for _key in dict_keys:
if _key in loc:
loc_key += loc[_key]
def test():
pc = PlaceContext(['Ngong', 'Nairobi', 'Kenya'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
assert len(pc.countries) == 1
assert len(pc.cities) == 1
assert len(pc.other) == 1
assert 'Ngong' in pc.other
assert pc.cities_for_name('Nairobi')[0][4] == 'Kenya'
assert pc.regions_for_name('Ohio')[0][4] == 'United States'
pc = PlaceContext(['Aleppo', 'Syria'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
assert 'Aleppo' in pc.cities
def test():
pc = PlaceContext(['Ngong', 'Nairobi', 'Kenya'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
assert len(pc.countries) == 1
assert len(pc.cities) == 1
assert len(pc.other) == 1
assert 'Ngong' in pc.other
assert pc.cities_for_name('Nairobi')[0][4] == 'Kenya'
assert pc.regions_for_name('Ohio')[0][4] == 'United States'
pc = PlaceContext(['Aleppo', 'Syria'])
pc.set_countries()
pc.set_regions()
pc.set_cities()
pc.set_other()
assert 'Aleppo' in pc.cities