Class TestDataModule

data_test.py:35–295

    df.to_csv(file_path, index=False)

class TestDataModule(unittest.TestCase):
    @classmethod
    def setUpClass(cls):
        cls.tokenizer = MockTokenizer()
        
        # Create temporary directory for test files
        cls.temp_dir = tempfile.TemporaryDirectory()
        cls.temp_path = cls.temp_dir.name
        
        # Create sample CSV data
        cls.csv_data = {
            'history_item_title': ["['Item A', 'Item B']", "['Item C', 'Item D']"],
            'item_title': ['Item E', 'Item F'],
            'history_item_id': ["['1', '2']", "['3', '4']"],
            'item_id': ['5', '6'],
            'history_item_sid': ["['SID1', 'SID2']", "['SID3', 'SID4']"],
            'item_sid': ['SID5', 'SID6'],
            'user_id_original_str': ['user1', 'user2'],
            'e_token': ['[CTX_HOMEPAGE]', '[CTX_SEARCH]']
        }
        cls.csv_file = os.path.join(cls.temp_path, 'test_data.csv')
        create_minimal_csv(cls.csv_file, cls.csv_data)
        
        # Create sample item features JSON
        cls.item_features = {
            '5': {'title': 'Item E', 'description': 'Description of Item E', 'item_type': 'O'},
            '6': {'title': 'Item F', 'description': 'Description of Item F', 'item_type': 'I'}
        }
        cls.item_file = os.path.join(cls.temp_path, 'test.item.json')
        with open(cls.item_file, 'w') as f:
            json.dump(cls.item_features, f)
        
        # Create sample indices JSON
        cls.indices = {
            '5': ['SID5_1', 'SID5_2', 'SID5_3'],
            '6': ['SID6_1', 'SID6_2', 'SID6_3']
        }
        cls.index_file = os.path.join(cls.temp_path, 'test.index.json')
        with open(cls.index_file, 'w') as f:
            json.dump(cls.indices, f)
        
        # Create sample user preference JSON
        cls.user_preference_data = [
            {
                'user': 'user1',
                'user_preference': 'Likes action games',
                'context': {
                    'history_items': ['1', '2'],
                    'target_item': '5'
                },
                'split': 'train'
            },
            {
                'user': 'user2',
                'user_preference': 'Prefers strategy games',
                'context': {
                    'history_items': ['3', '4'],
                    'target_item': '6'