test: refactor and update test cases for dataset and get_random_selection() function

tests/test_dataset.py

@@ -1,49 +0,0 @@
-import json
-import unittest
-
-from aip_trainer.lambdas import lambdaGetSample
-from tests import test_logger, TEST_ROOT_FOLDER
-
-
-def helper_category(category: int, threshold_min: int, threshold_max: int, n: int = 1000):
-    for _ in range(n):
-        event = {'body': json.dumps({'category': category, 'language': 'de'})}
-        response = lambdaGetSample.lambda_handler(event, [])
-        response_dict = json.loads(response)
-        number_of_words = len(response_dict['real_transcript'].split())
-        try:
-            assert threshold_min < number_of_words <= threshold_max
-        except AssertionError:
-            test_logger.error(
-                f"Category: {category} had a sentence with length {number_of_words}.")
-            raise AssertionError
-
-
-class TestDataset(unittest.TestCase):
-    def test_random_sentences(self):
-        helper_category(0, 0, 40)
-
-    def test_easy_sentences(self):
-        helper_category(1, 0, 8)
-
-    def test_normal_sentences(self):
-        helper_category(2, 8, 20)
-
-    def test_hard_sentences(self):
-        helper_category(3, 20, 10000)
-
-    def test_get_pickle2json_dataframe(self):
-        import os
-
-        custom_filename = 'test_data_de_en_2'
-        lambdaGetSample.get_pickle2json_dataframe(custom_filename, TEST_ROOT_FOLDER)
-        with open(TEST_ROOT_FOLDER / f'{custom_filename}.json', 'r') as src1:
-            with open(TEST_ROOT_FOLDER / f'{custom_filename}_expected.json', 'r') as src2:
-                json1 = json.load(src1)
-                json2 = json.load(src2)
-                assert json1 == json2
-        os.remove(TEST_ROOT_FOLDER / f'{custom_filename}.json')
-
-
-if __name__ == '__main__':
-    unittest.main()

tests/test_lambdaGetSample.py

@@ -0,0 +1,103 @@
+import json
+import unittest
+
+from aip_trainer.lambdas import lambdaGetSample
+from tests import TEST_ROOT_FOLDER
+from numpy.random import seed
+
+expected_output__get_random_selection = [
+    "Er ist ein begeisterter Theaterliebhaber.",
+    ["Du hast schon den ganzen Morgen gute Laune."],
+    "Tom hat mir gerade das Leben gerettet.",
+    ["Kannst du mich jetzt hören?"],
+    "Falls es eine bessere Lösung geben sollte, werde ich Sie umgehend informieren.",
+    ["Die Kette, die du mir vor zehn Jahren geschenkt hast, liegt mir bis heute sehr am Herzen."],
+    "Es lässt sich nicht in Worte fassen, wie glücklich ich gerade bin!“ – So ist es aber ganz gut in Worte gefasst.“",
+    ["Als nur ein Beispiel dafür, wie Automatisierung der Menschheit zugute kommen kann, nehme man eine Roboterhand, mit deren Hilfe man unreine oder gefährliche Arbeiten verrichten kann."],
+    "Since I was really tired I went to sleep early.",
+    ["She accused me of being a liar."],
+    "They lived in the land of Cockaigne.",
+    ["This is a complicated question to answer."],
+    "We are still hoping that Tom will get better.",
+    ["Tom said something, but I couldn't hear what he said."],
+    "People who starve themselves and mask themselves so that they can conform to what society deems beautiful are only hurting themselves.",
+    ["The queen has the ability to move as much as if she were a rook, that is, over the rows and columns, as if she were a bishop, that is, over the diagonals."],
+]
+
+
+def helper_category(cls, category: int, language: str, expected_output: dict):
+    seed(0)
+    event = {"body": json.dumps({"category": category, "language": language})}
+    response = lambdaGetSample.lambda_handler(event, [])
+    response_dict = json.loads(response)
+    try:
+        cls.assertDictEqual(response_dict, expected_output)
+    except AssertionError as ae:
+        print(f"category: {category}, language: {language}.")
+        print(f"response_dict: {response_dict}")
+        print(f"expected_output: {expected_output}")
+        raise ae
+
+
+class TestLambdaGetSample(unittest.TestCase):
+    def test_get_random_selection(self):
+        seed(0)
+        count = 0
+        for lang in ["de", "en"]:
+            for cat in range(4):
+                for is_gradio_output in [True, False]:
+                    output = lambdaGetSample.get_random_selection(
+                        lang, cat, is_gradio_output=is_gradio_output
+                    )
+                    self.assertEqual(output, expected_output__get_random_selection[count])
+                    count += 1
+
+
+class TestDataset(unittest.TestCase):
+    def test_random_sentences_de(self):
+        expected = {'real_transcript': 'Er ist ein begeisterter Theaterliebhaber.', 'ipa_transcript': 'ɛɐ̯ ɪst aɪ̯n bɛːɡaɪ̯stɛːrtɛːr tɛːaːtɛːrliːbhaːbɛːr.', 'transcript_translation': ''}
+        helper_category(self, 0, "de", expected_output=expected)
+
+    def test_easy_sentences_easy_de(self):
+        expected = {'real_transcript': 'Sie will niemanden heiraten.', 'ipa_transcript': 'ziː vɪl niːmandɛːn haɪ̯raːtɛːn.', 'transcript_translation': ''}
+        helper_category(self, 1, "de", expected_output=expected)
+
+    def test_normal_sentences_medium_de(self):
+        expected = {'real_transcript': 'Leg das Buch dorthin, wo du es gefunden hast.', 'ipa_transcript': 'lɛːɡ daːs bʊx doːrtiːn, voː duː ɛːs ɡɛːfʊndɛːn hast.', 'transcript_translation': ''}
+        helper_category(self, 2, "de", expected_output=expected)
+
+    def test_hard_sentences_hard_de(self):
+        expected = {'real_transcript': 'Eine Frau braucht neun Monate, um ein Kind zur Welt zu bringen, aber das heißt nicht, dass es neun zusammen in einem Monat schaffen könnten.', 'ipa_transcript': 'aɪ̯nɛː fraʊ̯ braʊ̯xt nɔɪ̯n moːnaːtɛː, uːm aɪ̯n kɪnd t͡suːr vɛlt t͡suː brɪŋɛːn, aːbɛːr daːs haɪ̯st nɪçt, das ɛːs nɔɪ̯n t͡suːzamɛːn iːn aɪ̯nɛːm moːnaːt ʃafɛːn kœntɛːn.', 'transcript_translation': ''}
+        helper_category(self, 3, "de", expected_output=expected)
+    
+    def test_random_sentences_en(self):
+        expected = {'real_transcript': 'He is a passionate theatregoer.', 'ipa_transcript': 'hi ɪz ə ˈpæʃənət theatregoer.', 'transcript_translation': ''}
+        helper_category(self, 0, "en", expected_output=expected)
+
+    def test_easy_sentences_easy_en(self):
+        expected = {'real_transcript': 'Lemons are usually sour.', 'ipa_transcript': 'ˈlɛmənz ər ˈjuʒəwəli saʊər.', 'transcript_translation': ''}
+        helper_category(self, 1, "en", expected_output=expected)
+
+    def test_normal_sentences_medium_en(self):
+        expected = {'real_transcript': 'Tom read the Bible in its entirety, from the beginning to the end.', 'ipa_transcript': 'tɑm rɛd ðə ˈbaɪbəl ɪn ɪts ɪnˈtaɪərti, frəm ðə bɪˈgɪnɪŋ tɪ ðə ɛnd.', 'transcript_translation': ''}
+        helper_category(self, 2, "en", expected_output=expected)
+
+    def test_hard_sentences_hard_en(self):
+        expected = {'real_transcript': 'That was the first time, in the history of chess, that a machine (Deep Blue) defeated a Grand Master (Garry Kasparov).', 'ipa_transcript': 'ðət wɑz ðə fərst taɪm, ɪn ðə ˈhɪstəri əv ʧɛs, ðət ə məˈʃin (dip blu) dɪˈfitɪd ə grænd ˈmæstər (ˈgɛri ˈkæspərɑv).', 'transcript_translation': ''}
+        helper_category(self, 3, "en", expected_output=expected)
+
+    def test_get_pickle2json_dataframe(self):
+        import os
+
+        custom_filename = 'test_data_de_en_2'
+        lambdaGetSample.get_pickle2json_dataframe(custom_filename, TEST_ROOT_FOLDER)
+        with open(TEST_ROOT_FOLDER / f'{custom_filename}.json', 'r') as src1:
+            with open(TEST_ROOT_FOLDER / f'{custom_filename}_expected.json', 'r') as src2:
+                json1 = json.load(src1)
+                json2 = json.load(src2)
+                assert json1 == json2
+        os.remove(TEST_ROOT_FOLDER / f'{custom_filename}.json')
+
+
+if __name__ == "__main__":
+    unittest.main()