class TestHelloWorldThings(unittest.TestCase):
# Todo very important to test hard cases. Huge cases, edge cases, corner cases--try to break it.
def setUp(self):
# Blank out the test JSON files:
data_map = { # todo DRY, this is repeated in every model interface's tests module
"user_data": "test/testing_mockUserDB.json",
"datespot_data": "test/testing_mockDatespotDB.json",
"match_data": "test/testing_mockMatchData.json",
"review_data": "test/testing_mockReviewData.json",
"message_data": "test/testing_mockMessageData.json",
"chat_data": "test/testing_mockChatData.json"
}
with open(TEST_JSON_DB_NAME, 'w') as fobj:
json.dump(data_map, fobj)
fobj.seek(0)
# make sure all the test-mock JSONs exist
for filename in data_map:
with open(data_map[filename], 'w') as fobj:
json.dump({}, fobj)
fobj.seek(0)
# Make a mock restaurant
self.terrezanos_location = (40.72289821341384, -73.97993915779077)
self.terrezanos_name = "Terrezano's"
self.terrezanos_traits = [
"italian", "wine", "pasta", "NOT FROM PIZZA HUT", "authentic",
"warehouse"
]
self.terrezanos_price_range = 2
self.terrezanos_hours = [
[14, 22], [14, 21], [14, 21], [14, 21], [14, 23], [14,
23], [14, 20]
] # ints in [0..23] representing hours, for now
terrezanos_data = {
"location": self.terrezanos_location,
"name": self.terrezanos_name,
"traits": self.terrezanos_traits,
"price_range": self.terrezanos_price_range,
"hours": self.terrezanos_hours,
}
# Make mock text
self.mock_text_positive_relevant = "This was a wonderful place to go on a date. I had the pasta. It was authentic and not from Pizza Hut."
self.expected_sentiment = 0.1906 # todo hardcoded
self.expected_relevance = round(
1 / len(self.mock_text_positive_relevant),
SENTIMENT_DECIMAL_PLACES) # i.e. "date" appears once.
# Connect to the database with the mock data set
self.db = DatabaseAPI(json_map_filename=TEST_JSON_DB_NAME)
args_data = {
"object_model_name": "datespot",
"object_data": terrezanos_data
}
self.terrezanos_id = self.db.post_object(args_data)
# Instantiate mock Review object
self.review_obj = models.Review(datespot_id=self.terrezanos_id,
text=self.mock_text_positive_relevant)
def test_init(self):
self.assertIsInstance(self.review_obj, models.Review)
def test_tokenize(self):
"""Does the internal tokenize method tokenize a multi-sentence text into an array of sentences as expected?"""
expected_sentences = [
"This was a wonderful place to go on a date.", "I had the pasta.",
"It was authentic and not from Pizza Hut."
]
self.review_obj._tokenize()
for i in range(len(expected_sentences)):
self.assertEqual(expected_sentences[i],
self.review_obj._sentences[i])
# todo test the length of each
def test_analyze_sentiment(self):
self.review_obj._analyze_sentiment()
self.assertAlmostEqual(self.expected_sentiment,
self.review_obj._sentiment)
def test_public_sentiment_attribute(self):
"""
Can the sentiment be accessed as expected via the object's public sentiment attribute?
"""
self.assertAlmostEqual(self.expected_sentiment,
self.review_obj.sentiment)
def test_analyze_relevance(self):
self.review_obj._analyze_relevance()
self.assertAlmostEqual(self.expected_relevance,
self.review_obj._analyze_relevance())
def test_public_relevance_attribute(
self): # todo tbd if any external code ever needs this
self.assertAlmostEqual(self.expected_relevance,
self.review_obj.relevance)
class TestHelloWorldThings(unittest.TestCase):
def setUp(self):
# Todo: Very hard to test the model in isolation here. Because chats require messages, and current architecture is that
# messages are created directly into a chat. The model itself doesn't persist any data, so it's impossible for the
# create_message method to find the correct chat to update.
# Same un-DRY boilerplate to configure the testing DB:
data_map = { # todo DRY, this is repeated in every model interface's tests module
"user_data": "test/testing_mockUserDB.json",
"datespot_data": "test/testing_mockDatespotDB.json",
"match_data": "test/testing_mockMatchData.json",
"review_data": "test/testing_mockReviewData.json",
"message_data": "test/testing_mockMessageData.json",
"chat_data": "test/testing_mockChatData.json"
}
with open(TEST_JSON_DB_NAME, 'w') as fobj:
json.dump(data_map, fobj)
fobj.seek(0)
# make sure all the test-mock JSONs exist and start as "{}":
for filename in data_map:
with open(data_map[filename], 'w') as fobj:
json.dump({}, fobj)
fobj.seek(0)
# Mock DB
self.db = DatabaseAPI(json_map_filename=TEST_JSON_DB_NAME)
self.message_data = model_interfaces.MessageModelInterface(json_map_filename=TEST_JSON_DB_NAME)
self.chat_data = model_interfaces.ChatModelInterface(json_map_filename=TEST_JSON_DB_NAME)
# Mock users
self.akatosh_name = "Akatosh"
self.akatosh_location = (40.73517750328247, -74.00683227856715)
self.akatosh_id = "1"
akatosh_data = {
"name": self.akatosh_name,
"current_location": self.akatosh_location,
"force_key": self.akatosh_id
}
self.db.post_object({"object_model_name": "user", "object_data": akatosh_data}) # Don't need to store the key returned by this, forced it to "1"
self.stendarr_name = "Stendarr"
self.stendarr_location = (40.74769591216627, -73.99447266003756)
self.stendarr_id = "2"
stendarr_json = {
"name": self.stendarr_name,
"current_location": self.stendarr_location,
"force_key": self.stendarr_id
}
self.db.post_object({"object_model_name": "user", "object_data": stendarr_json})
self.talos_name = "Talos"
self.talos_location = (40.76346250260515, -73.98013893542904)
self.talos_id = "3"
talos_json = {
"name": self.talos_name,
"current_location": self.talos_location,
"force_key": self.talos_id
}
self.db.post_object({"object_model_name": "user", "object_data": talos_json})
# Instantiate chat object with two participants
# Messages can't be instantiated without a Chat ID
# Todo and that chat ID must come from the DB. So must initiate a chat in the DB.
self.chat_start_time = time.time()
self.chat_json = {
"start_time": self.chat_start_time,
"participant_ids": [self.akatosh_id, self.stendarr_id]
}
self.chat_id = self.db.post_object({"object_model_name": "chat", "object_data": self.chat_json})
# Mock messages
self.first_timestamp = time.time()
self.first_message_text = "Lord Akatosh lends you his might. When your own strength fails you, trust in the Nine."
self.first_message_json = {
"time_sent": self.first_timestamp,
"sender_id": self.akatosh_id,
"chat_id": self.chat_id,
"text": self.first_message_text
}
self.first_message_id = self.db.post_object({"object_model_name": "message", "object_data": self.first_message_json})
self.first_message_obj = self.message_data.lookup_obj(self.first_message_id)
self.first_message_sentiment = self.first_message_obj.sentiment
# Second message in same chat:
self.second_timestamp = time.time()
self.second_message_text = "K, thanks for letting me know!"
self.second_message_json = {
"time_sent": self.second_timestamp,
"sender_id": self.stendarr_id,
"chat_id": self.chat_id,
"text": self.second_message_text
}
self.second_message_id = self.db.post_object({"object_model_name": "message", "object_data": self.second_message_json})
self.second_message_obj = self.message_data.lookup_obj(self.second_message_id)
self.second_message_sentiment = self.second_message_obj.sentiment
# create_message should append the message to the chat
# Fetch the chat object at the end, to create one with the messages appended
self.chat_obj = self.chat_data.lookup_obj(self.chat_id)
def test_init(self):
self.assertIsInstance(self.chat_obj, models.Chat)
def test_eq(self):
"""Does the custom __eq__() behave as expected?"""
self.assertTrue(self.chat_obj == self.chat_obj)
def test_message_id_order(self):
"""Are both test messages in the Chat, and is the first message before the second?"""
self.assertEqual([message.id for message in self.chat_obj.messages], [self.first_message_id, self.second_message_id]) # todo hacky/obfuscating to have the list comp here
def test_average_sentiment(self):
"""Does the average sentiment match the value expected from separate calculation on same values?"""
self.assertIsNotNone(self.chat_obj.sentiment)
expected_mean_sentiment = round((self.first_message_sentiment + self.second_message_sentiment) / 2, SENTIMENT_DECIMAL_PLACES)
self.assertEqual(expected_mean_sentiment, self.chat_obj.sentiment)
class TestHelloWorldThings(unittest.TestCase):
def setUp(self):
# Boilerplate mock data environment stuff
data_map = { # todo DRY, this is repeated in every model interface's tests module
"user_data": "test/testing_mockUserDB.json",
"datespot_data": "test/testing_mockDatespotDB.json",
"match_data": "test/testing_mockMatchData.json",
"review_data": "test/testing_mockReviewData.json",
"message_data": "test/testing_mockMessageData.json",
"chat_data": "test/testing_mockChatData.json"
}
with open(TEST_JSON_DB_NAME, 'w') as fobj:
json.dump(data_map, fobj)
fobj.seek(0)
# make sure all the test-mock JSONs exist:
for filename in data_map:
with open(data_map[filename], 'w') as fobj:
json.dump({}, fobj)
fobj.seek(0)
# Mock DB
self.db = DatabaseAPI(json_map_filename=TEST_JSON_DB_NAME)
self.user_data = model_interfaces.UserModelInterface(
json_map_filename=TEST_JSON_DB_NAME)
# Make three mock users
self.akatosh_name = "Akatosh"
self.akatosh_location = (40.73517750328247, -74.00683227856715)
self.akatosh_id = "1"
akatosh_json = {
"name": self.akatosh_name,
"current_location": self.akatosh_location,
"force_key": self.akatosh_id
}
self.db.post_object({
"object_model_name": "user",
"object_data": akatosh_json
}) # Don't need to store the key returned by this, forced it to "1"
self.stendarr_name = "Stendarr"
self.stendarr_location = (40.74769591216627, -73.99447266003756)
self.stendarr_id = "2"
stendarr_json = {
"name": self.stendarr_name,
"current_location": self.stendarr_location,
"force_key": self.stendarr_id
}
self.db.post_object({
"object_model_name": "user",
"object_data": stendarr_json
})
self.talos_name = "Talos"
self.talos_location = (40.76346250260515, -73.98013893542904)
self.talos_id = "3"
talos_json = {
"name": self.talos_name,
"current_location": self.talos_location,
"force_key": self.talos_id
}
self.db.post_object({
"object_model_name": "user",
"object_data": talos_json
})
# Instantiate mock simple bilateral message (one recipient)
self.mock_bilateral_timestamp = time.time()
self.single_sentence_text = "Worship the Nine, do your duty, and heed the commands of the saints and priests."
self.expected_sentiment_single_sentence = 0.296 # todo hardcoded
self.mock_chat_id_1 = "1a"
self.message_obj = models.Message(
time_sent=self.mock_bilateral_timestamp,
sender=self.user_data.lookup_obj(self.akatosh_id),
chat_id=self.mock_chat_id_1,
text=self.single_sentence_text)
# Instantiate mock multi-sentence message.
self.multisentence_text = "I'm Akatosh blah blah blah. Lord Akatosh lends you his might. When your own strength fails you, trust in the Nine."
self.expected_sentiment_multisentence = 0.092 # todo hardcoded
self.mock_chat_id_2 = "2a"
self.multisentence_message_obj = models.Message(
time_sent=time.time(),
sender=self.user_data.lookup_obj(self.akatosh_id),
chat_id=self.mock_chat_id_2,
text=self.multisentence_text)
def test_init(self):
self.assertIsInstance(self.message_obj, models.Message)
def test_eq(self):
"""Does the custom __eq__() behave as expected?"""
self.assertTrue(self.message_obj == self.message_obj)
self.assertFalse(self.message_obj == self.multisentence_message_obj)
def test_hash(self):
"""Does the integer returned by __hash__ match the results of mimicing the same hashing steps
manually?"""
expected_hash = hash(
str(self.mock_bilateral_timestamp) + self.akatosh_id)
self.assertEqual(expected_hash, hash(self.message_obj))
def test_str(self):
"""Does the __str__() method return the expected string for a known message?"""
expected_string = f"{self.message_obj.time_sent}:\t{self.message_obj.sender.id}:\t{self.message_obj.text}"
actual_string = str(self.message_obj)
self.assertEqual(expected_string, str(self.message_obj))
self.assertEqual(
expected_string, __str__(self.message_obj)
) # Todo for some reason, the single line of the __str__ method doesn't register as covered
def test_tokenize(self):
"""Does _tokenize split a multisentence text into the expected sentences?"""
expected_sentences = [
"I'm Akatosh blah blah blah.", "Lord Akatosh lends you his might.",
"When your own strength fails you, trust in the Nine."
]
self.multisentence_message_obj._tokenize()
for i in range(len(expected_sentences)):
self.assertEqual(expected_sentences[i],
self.multisentence_message_obj._sentences[i])
# Todo test other sentence-ending punctuation.
def test_analyze_sentiment(self):
"""Does the sentiment match the hardcoded expected sentiment?"""
self.message_obj._analyze_sentiment()
self.assertAlmostEqual(self.expected_sentiment_single_sentence,
self.message_obj._sentiment_avg)
self.multisentence_message_obj._analyze_sentiment()
self.assertAlmostEqual(self.expected_sentiment_multisentence,
self.multisentence_message_obj._sentiment_avg)
def test_str(self):
"""Does the __str__ method return the expected string?"""
expected_string = f"{self.mock_bilateral_timestamp}:\t{self.akatosh_id}:\t{self.single_sentence_text}"
class TestHelloWorldThings(unittest.TestCase):
"""Basic non-brokenness tests."""
@freeze_time(datetime.datetime.now())
def setUp(self):
# Blank out the test JSON files:
data_map = { # todo DRY, this is repeated in every model interface's tests module
"user_data": "test/testing_mockUserDB.json",
"datespot_data": "test/testing_mockDatespotDB.json",
"match_data": "test/testing_mockMatchData.json",
"review_data": "test/testing_mockReviewData.json",
"message_data": "test/testing_mockMessageData.json",
"chat_data": "test/testing_mockChatData.json"
}
with open(TEST_JSON_DB_NAME, 'w') as fobj:
json.dump(data_map, fobj)
fobj.seek(0)
# make sure all the test-mock JSONs exist
for filename in data_map:
with open(data_map[filename], 'w') as fobj:
json.dump({}, fobj)
fobj.seek(0)
# Instantiate DatabaseAPI object and model interfaces
self.db = DatabaseAPI(json_map_filename=TEST_JSON_DB_NAME)
self.user_data = model_interfaces.UserModelInterface(
json_map_filename=TEST_JSON_DB_NAME
) # These are cumbersome, but no implementation code actually needs to ask the DB API for
self.datespot_data = model_interfaces.DatespotModelInterface(
json_map_filename=TEST_JSON_DB_NAME
) # ...a model object. DB API having a "get_object()" method would be convenient
self.match_data = model_interfaces.MatchModelInterface(
json_map_filename=TEST_JSON_DB_NAME
) # ...for the tests code but have no other use.
self.review_data = model_interfaces.ReviewModelInterface(
json_map_filename=TEST_JSON_DB_NAME)
self.chat_data = model_interfaces.ChatModelInterface(
json_map_filename=TEST_JSON_DB_NAME)
self.message_data = model_interfaces.MessageModelInterface(
json_map_filename=TEST_JSON_DB_NAME)
# Data for mock users
self.azura_name = "Azura"
self.azura_location = (40.73517750328247, -74.00683227856715)
self.azura_id = "1"
self.azura_existing_taste_name = "dawn"
self.azura_existing_taste_strength = 0.9
self.azura_existing_taste_datapoints = 3
self.azura_existing_tastes = { # This is the data in the object's internal format, for testing convenience not for a call to the external create_user()
self.azura_existing_taste_name: [
self.azura_existing_taste_strength,
self.azura_existing_taste_datapoints
]
}
self.azura_data = {
"name": self.azura_name,
"current_location": self.azura_location,
"force_key": self.azura_id
}
self.boethiah_name = "Boethiah"
self.boethiah_location = (40.76346250260515, -73.98013893542904)
self.boethiah_id = "2"
self.boethiah_data = {
"name": self.boethiah_name,
"current_location": self.boethiah_location,
"force_key": self.boethiah_id
}
self.hircine_name = "Hircine"
self.hircine_location = (40.76525023033338, -73.96722141608099)
self.hircine_id = "3"
self.hircine_data = {
"name": self.hircine_name,
"current_location": self.hircine_location,
"force_key": self.hircine_id
}
# Data for mock Datespot
self.terrezanos_location = (40.737291166191476, -74.00704685527774)
self.terrezanos_name = "Terrezano's"
self.terrezanos_traits = {
"italian": [1.0, "discrete"],
"wine": [0.5, 1],
"pasta": [0.6, 2],
"NOT FROM PIZZA HUT": [0.01, 2],
"authentic": [-0.05, 3],
"warehouse": [1.0, "discrete"]
}
self.terrezanos_price_range = 2
self.terrezanos_hours = [
[14, 22], [14, 21], [14, 21], [14, 21], [14, 23], [14,
23], [14, 20]
] # ints in [0..23] representing hours, for now
self.terrezanos_data = {
"location": self.terrezanos_location,
"name": self.terrezanos_name,
"traits": self.terrezanos_traits,
"price_range": self.terrezanos_price_range,
"hours": self.terrezanos_hours,
}
self.terrezanos_id = self.db.post_object({
"object_model_name":
"datespot",
"object_data":
self.terrezanos_data
})
# Data for mock Review of Terrezano's
self.mock_text_positive_relevant = "This was a wonderful place to go on a date. I had the pasta. It was authentic and not from Pizza Hut."
self.expected_sentiment = 0.1906 # todo hardcoded
self.expected_relevance = round(1 /
len(self.mock_text_positive_relevant),
4) # i.e. "date" appears once.
self.terrezanos_review_data = {
"datespot_id": self.terrezanos_id,
"text": self.mock_text_positive_relevant
}
# Add three users for use in testing compound objects
self.db.post_object({
"object_model_name": "user",
"object_data": self.azura_data
})
self.db.post_object({
"object_model_name": "user",
"object_data": self.boethiah_data
})
self.db.post_object({
"object_model_name": "user",
"object_data": self.hircine_data
})
# Data for mock Message and Chat
self.mock_bilateral_timestamp = time.time()
self.quick_mock_chat_data = {
"start_time": time.time(),
"participant_ids": [self.azura_id, self.boethiah_id]
}
self.mock_chat_id_1 = self.db.post_object({
"object_model_name":
"chat",
"object_data":
self.quick_mock_chat_data
}) # Need a Chat to create a Message
self.single_sentence_text = "Worship the Nine, do your duty, and heed the commands of the saints and priests."
self.expected_sentiment_single_sentence = 0.296 # todo hardcoded
self.mock_bilateral_message_data = {
"time_sent": self.mock_bilateral_timestamp,
"sender_id": self.azura_id,
"chat_id": self.mock_chat_id_1,
"text": self.single_sentence_text
}
# Add two matches for Azura user
@freeze_time("2021-05-01 12:00:01")
def freezetime_match_1(match_data: dict) -> str:
"""Creates the match at a specified timestamp to avoid unittest forcing the timestamps to be identical;
returns the id string."""
return self.db.post_object(match_data)
@freeze_time("2021-05-01 12:00:02") # One second later
def freezetime_match_2(match_data: dict) -> str:
"""Creates the second match at a later timestamp."""
return self.db.post_object(match_data)
match_data_azura_boethiah = {
"object_model_name": "match",
"object_data": {
"user1_id": self.azura_id,
"user2_id": self.boethiah_id
}
}
match_data_hircine_azura = {
"object_model_name": "match",
"object_data": {
"user1_id": self.hircine_id,
"user2_id": self.azura_id
}
}
self.match_id_azura_boethiah = freezetime_match_1(
match_data_azura_boethiah)
self.match_id_hircine_azura = freezetime_match_2(
match_data_hircine_azura)
self.match_obj_azura_boethiah = self.match_data.lookup_obj(
self.match_id_azura_boethiah)
self.match_obj_hircine_azura = self.match_data.lookup_obj(
self.match_id_hircine_azura)
def test_init(self):
"""Was an object of the expected type instantiated?"""
self.assertIsInstance(self.db, DatabaseAPI)
def test_model_interface_constructor_calls(self):
"""Does the DB interface call the expected model interface for each model name?"""
expected_interfaces = {
"user": model_interfaces.UserModelInterface,
"datespot": model_interfaces.DatespotModelInterface,
"match": model_interfaces.MatchModelInterface,
"review": model_interfaces.ReviewModelInterface,
"message": model_interfaces.MessageModelInterface,
"chat": model_interfaces.ChatModelInterface
}
for model_name in expected_interfaces:
actual_interface = self.db._model_interface(model_name)
self.assertIsInstance(actual_interface,
expected_interfaces[model_name])
def test_validate_model_name(self):
"""Does the validator raise the expected error for a bad model name?"""
with self.assertRaises(ValueError):
self.db._validate_model_name("foo")
### Tests for post_object() and get_object() ###
def test_post_obj_user(self):
talos_name = "Talos"
talos_location = (40.76346250260515, -73.98013893542904)
expected_talos_id = "4"
talos_data = {
"name": talos_name,
"current_location": talos_location,
"force_key": expected_talos_id
}
actual_talos_id = self.db.post_object({
"object_model_name": "user",
"object_data": talos_data
})
self.assertIsInstance(actual_talos_id, str)
talos_obj = self.user_data.lookup_obj(actual_talos_id)
self.assertIsInstance(talos_obj, models.User)
self.assertEqual(expected_talos_id, actual_talos_id)
with self.assertRaises(
ValueError
): # Trying with key already in DB should raise error
talos_data["force_key"] = self.hircine_id # Used in setUp
actual_talos_id = self.db.post_object({
"object_model_name": "user",
"object_data": talos_data
})
def test_post_obj_datespot(self):
domenicos_location = (40.723889184134926, -73.97613846772394)
domenicos_name = "Domenico's"
domenicos_traits = {
"coffee": [
1.0, 1
], # todo...So we're imagining this as ~ how good the coffee is, rather than the discrete fact that they do serve coffee?
"coffee shop": [1.0, "discrete"],
"gourmet": [0.25, 1],
"americano": [0.15, 1],
"knows coffee": [0.3, 1],
"bricks": [0.6, 1],
"burger juice": [0.9, 1]
}
domenicos_price_range = 1
domenicos_hours = [[8, 19], [8, 19], [8, 19], [8, 19], [8, 19],
[8, 19], [10, 17]]
domenicos_data = {
"location": domenicos_location,
"name": domenicos_name,
"traits": domenicos_traits,
"price_range": domenicos_price_range,
"hours": domenicos_hours
}
domenicos_id = self.db.post_object({
"object_model_name": "datespot",
"object_data": domenicos_data
})
domenicos_obj = self.datespot_data.lookup_obj(domenicos_id)
self.assertIsInstance(domenicos_obj, models.Datespot)
def test_post_obj_match(self):
match_data = {"user1_id": self.azura_id, "user2_id": self.boethiah_id}
match_id = self.db.post_object({
"object_model_name": "match",
"object_data": match_data
})
match_obj = self.match_data.lookup_obj(match_id)
self.assertIsInstance(match_obj, models.Match)
def test_post_obj_review(self):
review_id = self.db.post_object({
"object_model_name":
"review",
"object_data":
self.terrezanos_review_data
})
review_obj = self.review_data.lookup_obj(review_id)
self.assertIsInstance(review_obj, models.Review)
def test_post_obj_message(self):
message_id = self.db.post_object({
"object_model_name":
"message",
"object_data":
self.mock_bilateral_message_data
})
message_obj = self.message_data.lookup_obj(message_id)
self.assertIsInstance(message_obj, models.Message)
def test_post_obj_chat(self):
chat_id = self.db.post_object({
"object_model_name": "chat",
"object_data": self.quick_mock_chat_data
})
chat_obj = self.chat_data.lookup_obj(chat_id)
self.assertIsInstance(chat_obj, models.Chat)
### Tests for put_json() ###
def test_put_json_update_user(self):
# Test updating a User's location:
new_data = {
"current_location": (40.737291166191476, -74.00704685527774),
}
args_data = {
"object_model_name": "user",
"object_id": self.azura_id,
"update_data": new_data
}
self.db.put_data(args_data)
azura_obj = self.user_data.lookup_obj(self.azura_id)
self.assertAlmostEqual(new_data["current_location"],
azura_obj.current_location)
def test_updating_unsupported_model_raises_error(self):
"""Does attempting to update a model for which updates aren't supported raise the
expected error?"""
unsupported_models = ["review", "message"]
arbitrary_object_id = "a"
arbitrary_data = {"foo": "bar"}
for model in unsupported_models:
with self.assertRaises(ValueError):
self.db.put_data({
"object_model_name": model,
"object_id": arbitrary_object_id,
"update_data": arbitrary_data
})
# TODO complete for other models and their main anticipated update cases
### Tests for post_decision() ###
def test_post_yes_decision_no_match(self):
"""Does posting a "yes" decision that doesn't create a match return the expected JSON?"""
decision_yes_data = {
"user_id": self.azura_id,
"candidate_id": self.boethiah_id,
"outcome": True
}
expected_response_data = {"match_created": False}
actual_response_data = self.db.post_decision(decision_yes_data)
self.assertEqual(expected_response_data, actual_response_data)
def test_post_yes_decision_yes_match(self):
"""Does posting a "yes" decision that creates a match return the expected JSON?"""
# Post a decision of Azura liking Boethiah:
azura_decision_yes_data = {
"user_id": self.azura_id,
"candidate_id": self.boethiah_id,
"outcome": True
}
self.db.post_decision(azura_decision_yes_data)
# Post a second decision of Boethiah liking Azura
boethiah_decision_yes_data = {
"user_id": self.boethiah_id,
"candidate_id": self.azura_id,
"outcome": True
}
expected_response_data = {"match_created": True}
actual_response_data = self.db.post_decision(
boethiah_decision_yes_data)
self.assertEqual(expected_response_data, actual_response_data)
def test_post_no_decision(self):
"""Does posting a "no" decision return the expected JSON?"""
decision_no_data = {
"user_id": self.azura_id,
"candidate_id": self.boethiah_id,
"outcome": False
}
expected_response_data = {"match_created": False}
actual_response_data = self.db.post_decision(decision_no_data)
self.assertEqual(expected_response_data, actual_response_data)
def test_non_boolean_outcome_raises_error(self):
"""Does posting JSON without a boolean outcome raise the expected error?"""
bad_data = {
"user_id": self.azura_id,
"candidate_id": self.boethiah_id,
"outcome": 2
}
with self.assertRaises(TypeError):
self.db.post_decision(bad_data)
### Tests for get_datespots_near() ###
# Code that makes live API calls isn't covered by the main tests suite
def test_get_datespots_near_cache_only_default_radius(self):
"""Does the method return a string matching the expected shape of a JSON-ified list of Datespots
in response to a query that provides valid location but no radius?"""
location_query_data = {
"location": (40.737291166191476, -74.00704685527774)
}
results = self.db.get_datespots_near(location_query_data)
self.assertIsInstance(results, list)
self.assertGreater(len(results), 0)
first_result = results[0]
distance, datespot = first_result # it's a two-element tuple
self.assertIsInstance(distance, float)
self.assertIsInstance(datespot, models.Datespot)
def test_get_datespots_near_cache_only_nondefault_radius(self):
"""Does the method return the expected JSON in response to a query that provides valid location
and specifies a non-default_radius?"""
location_query_data = {
"location": (40.737291166191476, -74.00704685527774),
"radius": 4000
}
results = self.db.get_datespots_near(location_query_data)
self.assertIsInstance(results, list)
self.assertGreater(len(results), 0)
first_result = results[0]
distance, datespot = first_result # it's a two-element tuple
self.assertIsInstance(distance, float)
self.assertIsInstance(datespot, models.Datespot)
### Tests for get_datespot_suggestions() ###
def test_get_candidate_datespots(self):
"""Does the method return the expected JSON in response to JSON matching with a valid Match?"""
# Put a Match in the mock DB
match_data = {"user1_id": self.azura_id, "user2_id": self.boethiah_id}
match_id = self.db.post_object({
"object_model_name": "match",
"object_data": match_data
})
match_obj = self.match_data.lookup_obj(match_id)
query_data = {"match_id": match_id}
results = self.db.get_candidate_datespots(query_data)
self.assertIsInstance(results, list)
self.assertGreater(len(results), 0)
# Terrezanos should be the only Datespot known to the DB here:
self.assertEqual(results[0][1].id, self.terrezanos_id)
### Tests for other public methods ###
def test_get_next_candidate(
self
): # We have two Users in the DB, so one will be the other's candidate
query_data = {"user_id": self.azura_id}
result = self.db.get_next_candidate(query_data)
candidate_name = result["name"]
self.assertEqual(self.boethiah_name, candidate_name)
# TODO Post / get obj / get json for all of:
# user
# datespot
# match
# review
# message
# chat
def test_get_matches_list(self):
expected_result_data = [
{ # Expecte the Match created second to appear first in the list
"match_id":
self.match_id_hircine_azura,
"match_timestamp":
self.match_obj_hircine_azura.timestamp,
"match_partner_info":
self.user_data.render_candidate(self.hircine_id)
},
{
"match_id":
self.match_id_azura_boethiah,
"match_timestamp":
self.match_obj_hircine_azura.timestamp,
"match_partner_info":
self.user_data.render_candidate(self.boethiah_id)
}
]
actual_result_data = self.db.get_matches_list(
query_data={"user_id": self.azura_id})
#self.assertEqual(actual_result_data, expected_result_data)
# TODO Had to give up for now on asserting about the timestamps due to weird behavior--keep getting them created with identical timestamps
# when created in setUp, even though the timestamps increment in match.py. Unittests for match.py confirmed that the underlying sort works.
for result in actual_result_data:
self.assertIsInstance(result, dict)
self.assertEqual(len(result), len(expected_result_data[0]))
def test_get_suggestions(self):
expected_result_data = [ # Terrezanos is the only Datespot in the DB here
self.datespot_data.render_obj(self.terrezanos_id)
]
actual_result_data = self.db.get_suggestions_list(
{"match_id": self.match_id_azura_boethiah})
self.assertEqual(actual_result_data, expected_result_data)
class TestHelloWorldThings(unittest.TestCase):
def setUp(self):
self.db = DatabaseAPI() # Testing on the real DB, to have restaurants
# TODO: script that populates the test DBs with realistic restaurants en masse. And/or separate JSON map for this test
# (pointing to same test DB filenames for some like users, different one for datespots)
self.user_data = model_interfaces.UserModelInterface()
# Need user objects to instantiate a Match
grortName = "Grort"
self.grortCurrentLocation = (40.746667, -74.001111)
grort_data = {
"name": grortName,
"current_location": self.grortCurrentLocation
}
self.grort_user_id = self.db.post_object({
"object_model_name": "user",
"object_data": grort_data
})
self.userGrort = self.user_data.lookup_obj(self.grort_user_id)
drobbName = "Drobb"
self.drobbCurrentLocation = (40.767376158866554, -73.98615327558278)
drobb_data = {
"name": drobbName,
"current_location": self.drobbCurrentLocation
}
self.drobb_user_id = self.db.post_object({
"object_model_name": "user",
"object_data": drobb_data
})
self.userDrobb = self.user_data.lookup_obj(self.drobb_user_id)
# distance should be approx 2610m
# midpoint should be circa (40.75827478958617, -73.99310556132602)
self.matchGrortDrobb = models.Match(self.userGrort, self.userDrobb)
assert self.matchGrortDrobb.midpoint is not None
# Get the candidates list that the DatabaseAPI would be giving to Match:
self.candidate_datespots_list = self.db.get_datespots_near(
{"location": self.matchGrortDrobb.midpoint})
def test_hash(self):
"""Does the __hash__() method's return value match the value obtained by mimicking its logic in the test code?"""
expected_hash = hash((self.grort_user_id, self.drobb_user_id))
actual_hash = hash(self.matchGrortDrobb)
self.assertEqual(actual_hash, expected_hash)
def test_hash_output_consistent_regardless_of_user_order(self):
"""Does Match(Alice, Bob) hash to same value as Match(Bob, Alice)?"""
expected_hash = hash(self.matchGrortDrobb)
match_obj_flipped_members = models.Match(
self.userDrobb, self.userGrort
) # Reverse the user1 and user2 roles from those in setUp
assert match_obj_flipped_members.user1 == self.matchGrortDrobb.user2 and match_obj_flipped_members.user2 == self.matchGrortDrobb.user1
actual_hash = hash(match_obj_flipped_members)
self.assertEqual(actual_hash, expected_hash)
# Test same for the public id property attribute
expected_id = self.matchGrortDrobb.id
actual_id = match_obj_flipped_members.id
self.assertEqual(actual_id, expected_id)
# TODO The implementation code isn't correct as of 6/10. Observed same user ID strings producing differnent Match id hashes
# during Postman endpoint testing.
def test_public_id_attribute_matches_hash(self):
"""Does the public Match.id attribute-property bear the expected relationship to the return value Match.__hash__()?"""
expected_id = str(hex(hash(self.matchGrortDrobb)))[
2:] # Mimic logic in Match._id() private method
actual_id = self.matchGrortDrobb.id
self.assertEqual(actual_id, expected_id)
def test_compute_midpoint(self):
maxDelta = 0.01
approxExpectedMidpoint = (40.75827478958617, -73.99310556132602)
expectedLat, expectedLon = approxExpectedMidpoint
actualLat, actualLon = self.matchGrortDrobb.midpoint
self.assertAlmostEqual(actualLat,
expectedLat,
delta=expectedLat * maxDelta)
self.assertAlmostEqual(actualLon,
expectedLon,
delta=expectedLat * maxDelta)
def test_public_distance_attribute(self):
"""Does the public distance attribute-property return the expected distance?"""
expected_distance = geo_utils.haversine(self.grortCurrentLocation,
self.drobbCurrentLocation)
actual_distance = self.matchGrortDrobb.distance
self.assertAlmostEqual(actual_distance, expected_distance)
def test_get_suggestions_return_type(self):
"""Does Match.get_suggestions() external method return the expected type?"""
expected_return_type = list
returned_obj = self.matchGrortDrobb.suggestions(
self.candidate_datespots_list)
self.assertIsInstance(returned_obj, expected_return_type)
def test_get_suggestions_return_not_null(self):
returned_obj = self.matchGrortDrobb.suggestions(
self.candidate_datespots_list)
self.assertGreater(len(returned_obj), 0)
def test_get_suggestions_return_shape(self):
"""Does the returned object's shape (nested lists/tuples) match the expected structure?"""
returned_obj = self.matchGrortDrobb.suggestions(
self.candidate_datespots_list)
# each "suggestion" should be a Datespot object literal:
for element in returned_obj:
self.assertIsInstance(element, models.Datespot)
# def test_db_user_method_returns_expected_query_results(self):
# """Does the method that calls the main database API return the expected query results, i.e.
# a list of serialized restaurant object dicts sorted by distance?"""
# returned_obj = self.matchGrortDrobb._get_datespots_by_geography()
# #print(returned_obj)
# self.assertIsInstance(returned_obj, list)
# self.assertGreater(len(returned_obj), 0)
def test_internal_scorer_method_returns_expected_data(self):
"""Does the internal "private" method responsible for scoring each nearby datespot return a non-empty
list?"""
returned_obj = self.matchGrortDrobb._score_nearby_datespots(
self.candidate_datespots_list)
self.assertIsInstance(returned_obj, list)
self.assertGreater(len(returned_obj), 0)