def test_should_return_same_df_when_column_not_exists(self):
transformation = Transformation(self.test_data)
transformation.split_content("time", ["day", "month", "year"])
current_result = transformation.dataframe.columns
expected_result = self.test_data.columns
self.assertEqual(current_result, expected_result)
def test_should_return_same_df_when_columns_param_is_empty(self):
transformation = Transformation(self.test_data)
transformation.split_content("", [])
current_result = transformation.dataframe.columns
expected_result = self.test_data.columns
self.assertEqual(current_result, expected_result)
def test_should_split_column_content_into_three_new_columns(self):
transformation = Transformation(self.test_data)
transformation.split_content("coordinates",
["longitude", "latitude", "depth"])
current_result = transformation.dataframe.columns
expected_result = [
"date", "place", "mag", "status", "longitude", "latitude", "depth",
"alert"
]
self.assertCountEqual(current_result, expected_result)
def test_should_return_transformed_data_using_all_pipeline_components(
self, mock_get_data):
self.create_tmp_folder()
fake_api_input = ApiInput(self.FAKE_URL)
mock_get_data.return_value = self.FAKE_INPUT_DATA
extraction_process = Extraction(fake_api_input)
extraction_process.extract()
raw_data = extraction_process.data
raw_df = self.spark.createDataFrame(
raw_data,
["date", "place", "mag", "status", "coordinates", "alert"])
transformation_process = Transformation(raw_df)
transformation_process.drop(["alert"])
transformation_process.rename({"mag": "magnitude", "place": "city"})
transformation_process.replace_null_values({"status": "Automatic"})
transformation_process.lowercase(["status"])
transformation_process.convert_data_type({"date": IntegerType()})
transformation_process.split_content(
"coordinates", ["longitude", "latitude", "depth"])
transformed_df = transformation_process.dataframe
csv_storage = CsvStorage(self.OUTPUT_FILEPATH)
loading_process = Loading(csv_storage)
loading_process.load(transformed_df)
current_result = self.spark \
.read \
.csv(self.OUTPUT_FILEPATH, header=True, inferSchema=True) \
.collect()
expected_result = self.spark \
.createDataFrame(self.FAKE_EXPECTED_DATA,
["date", "city", "magnitude", "status", "longitude", "latitude", "depth"]) \
.collect()
self.assertEqual(current_result, expected_result)
self.delete_test_file()