def test_petfind_orgname():
# Single word organization should not need slicing
org1 = 'BARC'
org1_pets, _ = find_pets(pf, org_name=org1)
# Casing should not matter
org2 = 'barc'
org2_pets, _ = find_pets(pf, org_name=org2)
# This org will require slicing due to the API not finding the org based on the full name
org3 = 'Adopt A Pup Animal Rescue'
org3_pets, _ = find_pets(pf, org_name=org3)
assert isinstance(org1_pets, DataFrame)
assert isinstance(org2_pets, DataFrame)
assert isinstance(org3_pets, DataFrame)
assert org1_pets.shape == org2_pets.shape
def test_default_coords():
# Originally NaN values for rows with BC as a province/state
pets, _ = find_pets(pf, location='Vancouver, BC', distance=100)
print(pets[[
'contact.address.lat', 'contact.address.long', 'contact.address.state'
]])
assert isinstance(pets, DataFrame)
assert pets[isnull(pets['contact.address.lat'])].shape[0] == 0
def test_default_image():
# This should have a couple defaults in it but due to the way Petfinder selects the pets, you may get a
# different set each time so if you don't see default replacing the NaN and just all links then you may want to
# rerun this test again until you see it
pets, _ = find_pets(pf, location='Baltimore, MD')
print(pets['primary_photo_cropped.small'])
assert isinstance(pets, DataFrame)
assert pets[isnull(pets['primary_photo_cropped.small'])].shape[0] == 0
def test_petfind_location():
# City, State
loc1 = 'Baltimore, MD'
loc1_pets, _ = find_pets(pf, location=loc1, distance=25)
# Postal Code
loc2 = '21250'
loc2_pets, _ = find_pets(pf, location=loc2, distance=25)
# lat, long
loc3 = '39.258, -76.713'
loc3_pets, _ = find_pets(pf, location=loc3, distance=25, lat_long=True)
assert isinstance(loc1_pets, DataFrame)
assert Series(['contact.address.address', 'contact.address.lat', 'contact.address.long']). \
isin(loc1_pets.columns).all()
assert isinstance(loc2_pets, DataFrame)
assert Series(['contact.address.address', 'contact.address.lat', 'contact.address.long']). \
isin(loc2_pets.columns).all()
assert isinstance(loc3_pets, DataFrame)
assert Series(['contact.address.address', 'contact.address.lat', 'contact.address.long']). \
isin(loc3_pets.columns).all()
def test_petfind_json_location():
search_json_path = os.path.join(
os.path.dirname((os.path.dirname(os.path.abspath(__file__)))),
'tests/search.json')
with open(search_json_path, 'r') as fp:
search = json.load(fp)
pets, _ = find_pets(pf, location=search['location'], distance=25)
assert isinstance(search, dict)
assert isinstance(pets, DataFrame)
def test_petfind_complex():
pets, _ = find_pets(pf,
location='Baltimore, MD',
animal_type='dog',
breed='beagle',
distance=50,
name='Bayla',
age='young',
size='small',
gender='female',
coat='short')
assert isinstance(pets, DataFrame) and pets['name'][0] == 'Bayla'
def test_geocode():
locator = Nominatim(user_agent='myGeocoder')
pets, _ = find_pets(pf,
animal_type='dog',
location='Baltimore, MD',
distance=25)
pets['contact.address.address'] = pets['contact.address.city'].fillna('') + ', ' + \
pets['contact.address.state'].fillna('') + ', ' + \
pets['contact.address.postcode'].fillna('')
unique_addresses = pets['contact.address.address'].unique()
location_list = np.array([locator.geocode(x) for x in unique_addresses],
dtype=object)
lat_long_tup = np.array(location_list[:, 1])
for i in range(len(unique_addresses)):
pets.loc[pets['contact.address.address'] == unique_addresses[i], ['contact.address.lat',
'contact.address.long']] = \
lat_long_tup[i][0], lat_long_tup[i][1]
print(pets[[
'contact.address.address', 'contact.address.lat',
'contact.address.long'
]])
assert isinstance(pets, DataFrame)
def test_petfind_goodwith():
pets, _ = find_pets(pf, good_with=['cat', 'dog'])
assert all(pets['environment.cats']) and all(pets['environment.dogs'])
def test_petfind():
pets, _ = find_pets(pf)
assert isinstance(pets, DataFrame) and pets.shape[0] == MAX_PETS
def find_pets_with_cache(pf: Petfinder,
location=None,
animal_type=None,
breed=None,
size=None,
gender=None,
age=None,
color=None,
coat=None,
org_name=None,
distance=None,
name=None,
good_with=[],
house_trained=None,
special_needs=None,
sort=None):
# if cache is empty fill with query
if cache.get('default') is None:
pets, _ = find_pets(pf, location, animal_type, breed, size, gender,
age, color, coat, org_name, distance, name,
good_with, house_trained, special_needs, sort)
# if 'primary_photo_cropped' in pets.columns:
# del pets['primary_photo_cropped']
cache.set('default', pets)
else:
pets = cache.get('default')
# search cache for characteristics
if animal_type is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['animal_type'] == animal_type]
if breed is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['breeds'] == breed]
if size is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['size'] == size]
if gender is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['gender'] == gender]
if age is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['age'] == age]
if color is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['colors'] == color]
if coat is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['coat'] == coat]
if name is not None and pets.shape[0] >= num_results:
pets = pets.loc[pets['name'] == name]
# location search
if location is not None and pets.shape[0] >= num_results:
curr_location = locator.geocode(location)
curr_latlong = (curr_location.latitude, curr_location.longitude)
# remove results with no latlong
pets = pets.loc[(np.isnan(pets['contact.address.lat']) is False) &
(np.isnan(pets['contact.address.long']) is False)]
# pets = pets.loc[((geodesic(curr_latlong, (pets['contact.address.lat'],
# pets['contact.address.long'])).miles) < distance)]
# there's probably a better way but this works
pets['new'] = 0
for index in pets.index:
pets.loc[index, 'new'] = geodesic(
curr_latlong,
(pets.loc[index, 'contact.address.lat'],
pets.loc[index, 'contact.address.long'])).miles
pets = pets.loc[pets['new'] < distance]
del pets['new']
# if there are not enough results in the cache, run query and cache results
if pets.shape[0] < num_results:
pets, _ = find_pets(pf, location, animal_type, breed, size, gender,
age, color, coat, org_name, distance, name,
good_with, house_trained, special_needs, sort)
temp = cache.get('default')
# if 'primary_photo_cropped' in pets.columns:
# del pets['primary_photo_cropped']
if 'primary_photo_cropped' in temp.columns:
del temp['primary_photo_cropped']
pets = pets.append(temp, ignore_index=True)
pets = pets.drop_duplicates(
subset=['animal_type', 'gender', 'age', 'coat', 'name'])
cache.set('default', pets)
else:
# if more results than wanted, drop bottom values until there are proper number of results
while pets.shape[0] > num_results:
pets.drop(pets.index[num_results])
return pets