def test_column_name_change(self):
"""Tests correct renaming of snowflake columns after source change"""
tap_lines_before_column_name_change = test_utils.get_test_tap_lines('messages-with-three-streams.json')
tap_lines_after_column_name_change = test_utils.get_test_tap_lines(
'messages-with-three-streams-modified-column.json')
# Load with default settings
self.persist_lines_with_cache(tap_lines_before_column_name_change)
self.persist_lines_with_cache(tap_lines_after_column_name_change)
# Get loaded rows from tables
snowflake = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
table_one = snowflake.query("SELECT * FROM {}.test_table_one ORDER BY c_pk".format(target_schema))
table_two = snowflake.query("SELECT * FROM {}.test_table_two ORDER BY c_pk".format(target_schema))
table_three = snowflake.query("SELECT * FROM {}.test_table_three ORDER BY c_pk".format(target_schema))
# Get the previous column name from information schema in test_table_two
previous_column_name = snowflake.query("""
SELECT column_name
FROM information_schema.columns
WHERE table_catalog = '{}'
AND table_schema = '{}'
AND table_name = 'TEST_TABLE_TWO'
AND ordinal_position = 1
""".format(
self.config.get('dbname', '').upper(),
target_schema.upper()))[0]["COLUMN_NAME"]
# Table one should have no changes
self.assertEqual(
table_one,
[{'C_INT': 1, 'C_PK': 1, 'C_VARCHAR': '1'}])
# Table two should have versioned column
self.assertEquals(
table_two,
[
{previous_column_name: datetime.datetime(2019, 2, 1, 15, 12, 45), 'C_INT': 1, 'C_PK': 1,
'C_VARCHAR': '1', 'C_DATE': None},
{previous_column_name: datetime.datetime(2019, 2, 10, 2), 'C_INT': 2, 'C_PK': 2, 'C_VARCHAR': '2',
'C_DATE': '2019-02-12 02:00:00'},
{previous_column_name: None, 'C_INT': 3, 'C_PK': 3, 'C_VARCHAR': '2', 'C_DATE': '2019-02-15 02:00:00'}
]
)
# Table three should have renamed columns
self.assertEqual(
table_three,
[
{'C_INT': 1, 'C_PK': 1, 'C_TIME': datetime.time(4, 0), 'C_VARCHAR': '1', 'C_TIME_RENAMED': None},
{'C_INT': 2, 'C_PK': 2, 'C_TIME': datetime.time(7, 15), 'C_VARCHAR': '2', 'C_TIME_RENAMED': None},
{'C_INT': 3, 'C_PK': 3, 'C_TIME': datetime.time(23, 0, 3), 'C_VARCHAR': '3',
'C_TIME_RENAMED': datetime.time(8, 15)},
{'C_INT': 4, 'C_PK': 4, 'C_TIME': None, 'C_VARCHAR': '4', 'C_TIME_RENAMED': datetime.time(23, 0, 3)}
])
def test_nested_schema_unflattening(self):
"""Loading nested JSON objects with no props without flattening"""
tap_lines = test_utils.get_test_tap_lines('messages-with-nested-schema.json')
# Load with default settings - Flattening disabled
self.persist_lines(tap_lines)
# Get loaded rows from tables - Transform JSON to string at query time
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
unflattened_table = query(bigquery, """
SELECT c_pk
, c_array c_array
, c_object c_object
, c_object c_object_with_props
, c_nested_object c_nested_object
FROM {}.test_table_nested_schema
ORDER BY c_pk""".format(target_schema))
# Should be valid nested JSON strings
self.assertEqual(
unflattened_table,
[{
'c_pk': 1,
'c_array': '[1, 2, 3]',
'c_object': '{"key_1": "value_1"}',
'c_object_with_props': '{"key_1": "value_1"}',
'c_nested_object': {'nested_prop_1': 'nested_value_1',
'nested_prop_2': 'nested_value_2',
'nested_prop_3': {'multi_nested_prop_1': 'multi_value_1',
'multi_nested_prop_2': 'multi_value_2'}},
}])
def test_flush_streams_with_no_intermediate_flushes(self, mock_emit_state):
"""Test emitting states when no intermediate flush required"""
mock_emit_state.get.return_value = None
tap_lines = test_utils.get_test_tap_lines('messages-pg-logical-streams.json')
# Set batch size big enough to never has to flush in the middle
self.config['hard_delete'] = True
self.config['batch_size_rows'] = 1000
self.persist_lines(tap_lines)
# State should be emitted only once with the latest received STATE message
self.assertEquals(
mock_emit_state.mock_calls,
[
mock.call({"currently_syncing": None, "bookmarks": {
"logical1-logical1_edgydata": {"last_replication_method": "LOG_BASED", "lsn": 108240872, "version": 1570922723596, "xmin": None},
"logical1-logical1_table1": {"last_replication_method": "LOG_BASED", "lsn": 108240872, "version": 1570922723618, "xmin": None},
"logical1-logical1_table2": {"last_replication_method": "LOG_BASED", "lsn": 108240872, "version": 1570922723635, "xmin": None},
"logical2-logical2_table1": {"last_replication_method": "LOG_BASED", "lsn": 108240872, "version": 1570922723651, "xmin": None},
"public-city": {"last_replication_method": "INCREMENTAL", "replication_key": "id", "version": 1570922723667, "replication_key_value": 4079},
"public-country": {"last_replication_method": "FULL_TABLE", "version": 1570922730456, "xmin": None},
"public2-wearehere": {}}})
])
# Every table should be loaded correctly
self.assert_logical_streams_are_in_bigquery(True)
def test_nested_schema_unflattening(self):
"""Loading nested JSON objects into VARIANT columns without flattening"""
tap_lines = test_utils.get_test_tap_lines(
'messages-with-nested-schema.json')
# Load with default settings - Flattening disabled
self.persist_lines_with_cache(tap_lines)
# Get loaded rows from tables - Transform JSON to string at query time
snowflake = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
unflattened_table = snowflake.query("""
SELECT c_pk
,TO_CHAR(c_array) c_array
,TO_CHAR(c_object) c_object
,TO_CHAR(c_object) c_object_with_props
,TO_CHAR(c_nested_object) c_nested_object
FROM {}.test_table_nested_schema
ORDER BY c_pk""".format(target_schema))
# Should be valid nested JSON strings
self.assertEqual(unflattened_table, [{
'C_PK':
1,
'C_ARRAY':
'[1,2,3]',
'C_OBJECT':
'{"key_1":"value_1"}',
'C_OBJECT_WITH_PROPS':
'{"key_1":"value_1"}',
'C_NESTED_OBJECT':
'{"nested_prop_1":"nested_value_1","nested_prop_2":"nested_value_2","nested_prop_3":{"multi_nested_prop_1":"multi_value_1","multi_nested_prop_2":"multi_value_2"}}'
}])
def test_nested_schema_flattening(self):
"""Loading nested JSON objects with flattening and not not flattening"""
tap_lines = test_utils.get_test_tap_lines('messages-with-nested-schema.json')
# Turning on data flattening
self.config['data_flattening_max_level'] = 10
# Load with default settings - Flattening disabled
self.persist_lines_with_cache(tap_lines)
# Get loaded rows from tables
snowflake = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
flattened_table = snowflake.query(
"SELECT * FROM {}.test_table_nested_schema ORDER BY c_pk".format(target_schema))
# Should be flattened columns
self.assertEqual(
flattened_table,
[{
'C_PK': 1,
'C_ARRAY': '[\n 1,\n 2,\n 3\n]',
'C_OBJECT': None,
# Cannot map RECORD to SCHEMA. SCHEMA doesn't have properties that requires for flattening
'C_OBJECT_WITH_PROPS__KEY_1': 'value_1',
'C_NESTED_OBJECT__NESTED_PROP_1': 'nested_value_1',
'C_NESTED_OBJECT__NESTED_PROP_2': 'nested_value_2',
'C_NESTED_OBJECT__NESTED_PROP_3__MULTI_NESTED_PROP_1': 'multi_value_1',
'C_NESTED_OBJECT__NESTED_PROP_3__MULTI_NESTED_PROP_2': 'multi_value_2',
}])
def test_loading_csv_files(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines(
'messages-with-three-streams.json')
self.persist_messages(tap_lines)
self.assert_three_streams_are_in_s3_bucket()
def test_information_schema_cache_outdated(self):
"""If informations schema cache is not up to date then it should fail"""
tap_lines_with_multi_streams = test_utils.get_test_tap_lines(
"messages-with-three-streams.json")
# 1) Simulate an out of data cache:
# Table is in cache but not exists in database
snowflake = DbSync(self.config)
target_schema = self.config.get("default_target_schema", "").upper()
snowflake.query("""
CREATE TABLE IF NOT EXISTS {}.columns (table_schema VARCHAR, table_name VARCHAR, column_name VARCHAR, data_type VARCHAR)
""".format(snowflake.pipelinewise_schema))
snowflake.query("""
INSERT INTO {0}.columns (table_schema, table_name, column_name, data_type)
SELECT '{1}', 'TEST_TABLE_ONE', 'DUMMY_COLUMN_1', 'TEXT' UNION
SELECT '{1}', 'TEST_TABLE_ONE', 'DUMMY_COLUMN_2', 'TEXT' UNION
SELECT '{1}', 'TEST_TABLE_TWO', 'DUMMY_COLUMN_3', 'TEXT'
""".format(snowflake.pipelinewise_schema, target_schema))
# Loading into an outdated information_schema cache should fail with table not exists
with self.assertRaises(Exception):
self.persist_lines_with_cache(tap_lines_with_multi_streams)
# 2) Simulate an out of data cache:
# Table is in cache structure is not in sync with the actual table in the database
snowflake.query("CREATE SCHEMA IF NOT EXISTS {}".format(target_schema))
snowflake.query(
"CREATE OR REPLACE TABLE {}.test_table_one (C_PK NUMBER, C_INT NUMBER, C_VARCHAR TEXT)"
.format(target_schema))
# Loading into an outdated information_schema cache should fail with columns exists
# It should try adding the new column based on the values in cache but the column already exists
with self.assertRaises(Exception):
self.persist_lines_with_cache(tap_lines_with_multi_streams)
def test_nested_schema_flattening(self):
"""Loading nested JSON objects with flattening and not not flattening"""
tap_lines = test_utils.get_test_tap_lines('messages-with-nested-schema.json')
# Turning on data flattening
self.config['data_flattening_max_level'] = 10
# Load with default settings - Flattening disabled
self.persist_lines(tap_lines)
# Get loaded rows from tables
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
flattened_table = query(bigquery,
"SELECT * FROM {}.test_table_nested_schema ORDER BY c_pk".format(target_schema))
# Should be flattened columns
self.assertEqual(
flattened_table,
[{
'c_pk': 1,
'c_array': '[1, 2, 3]',
'c_object': None,
'c_object_with_props__key_1': 'value_1',
'c_nested_object__nested_prop_1': 'nested_value_1',
'c_nested_object__nested_prop_2': 'nested_value_2',
'c_nested_object__nested_prop_3__multi_nested_prop_1': 'multi_value_1',
'c_nested_object__nested_prop_3__multi_nested_prop_2': 'multi_value_2',
}])
def test_naming_convention(self):
tap_lines = test_utils.get_test_tap_lines(
'messages-with-three-streams.json')
self.config['naming_convention'] = "tester/{stream}/{timestamp}.csv"
self.persist_messages(tap_lines)
self.assert_three_streams_are_in_s3_bucket()
def test_table_with_pk_multi_column_removed(self):
"""Test table with a pk with multiple columns gets clustered by those and removing the pk doesnt cause errors"""
tap_lines = test_utils.get_test_tap_lines('table_with_multi_pk_cluster.json')
self.persist_lines(tap_lines)
# Get loaded rows from tables
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
table = query(bigquery, "SELECT * FROM {}.test_table_cluster_multi ORDER BY c_pk".format(target_schema))
cluster_columns = query(bigquery, "SELECT clustering_ordinal_position, column_name FROM {}.INFORMATION_SCHEMA.COLUMNS WHERE table_name = 'test_table_cluster_multi' AND clustering_ordinal_position > 0 ORDER BY 1".format(target_schema))
# ----------------------------------------------------------------------
# Check rows in table
# ----------------------------------------------------------------------
expected_table = [
{'c_pk': 2, 'c_int': 2, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 12, 2, 0, 0, tzinfo=timezone.utc)},
{'c_pk': 3, 'c_int': 3, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 15, 2, 0, 0, tzinfo=timezone.utc)}
]
expected_cluster_columns = [
{'clustering_ordinal_position': 1, 'column_name': 'c_pk'},
{'clustering_ordinal_position': 2, 'column_name': 'c_varchar'}
]
self.assertEqual(self.remove_metadata_columns_from_rows(table), expected_table)
self.assertEqual(cluster_columns, expected_cluster_columns)
# ----------------------------------------------------------------------
# Remove the primary key and check if clustering stayed unchanged
# ----------------------------------------------------------------------
self.config['primary_key_required'] = False
tap_lines = test_utils.get_test_tap_lines('table_with_multi_pk_cluster_changed.json')
self.persist_lines(tap_lines)
table_changed = query(bigquery, "SELECT * FROM {}.test_table_cluster_multi ORDER BY c_pk".format(target_schema))
cluster_columns_changed = query(bigquery, "SELECT clustering_ordinal_position, column_name FROM {}.INFORMATION_SCHEMA.COLUMNS WHERE table_name = 'test_table_cluster_multi' AND clustering_ordinal_position > 0 ORDER BY 1".format(target_schema))
expected_table_changed = [
{'c_pk': 2, 'c_int': 2, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 12, 2, 0, 0, tzinfo=timezone.utc)},
{'c_pk': 2, 'c_int': 2, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 12, 2, 0, 0, tzinfo=timezone.utc)},
{'c_pk': 3, 'c_int': 3, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 15, 2, 0, 0, tzinfo=timezone.utc)}
]
expected_cluster_columns_changed = []
self.assertEqual(self.remove_metadata_columns_from_rows(table_changed), expected_table_changed)
self.assertEqual(cluster_columns_changed, expected_cluster_columns)
def test_loading_csv_files_with_gzip_compression(self):
"""Loading multiple tables from the same input tap with gzip compression"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning on gzip compression
self.config['compression'] = 'gzip'
self.persist_messages(tap_lines)
self.assert_three_streams_are_in_s3_bucket(compression='gzip')
def test_loading_tables_with_client_side_encryption_and_wrong_master_key(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning on client-side encryption and load but using a well formatted but wrong master key
self.config['client_side_encryption_master_key'] = "Wr0n6m45t3rKeY0123456789a0123456789a0123456="
with assert_raises(ProgrammingError):
self.persist_lines_with_cache(tap_lines)
def test_loading_tables_with_no_encryption(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning off client-side encryption and load
self.config['client_side_encryption_master_key'] = ''
self.persist_lines(tap_lines)
self.assert_three_streams_are_into_bigquery()
def test_logical_streams_from_pg_with_hard_delete_and_default_batch_size_should_pass(self):
"""Tests logical streams from pg with inserts, updates and deletes"""
tap_lines = test_utils.get_test_tap_lines('messages-pg-logical-streams.json')
# Turning on hard delete mode
self.config['hard_delete'] = True
self.persist_lines(tap_lines)
self.assert_logical_streams_are_in_bigquery(True)
def test_loading_tables_with_custom_temp_dir(self):
"""Loading multiple tables from the same input tap using custom temp directory"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Use custom temp_dir
self.config['temp_dir'] = ('~/.pipelinewise/tmp')
self.persist_messages(tap_lines)
self.assert_three_streams_are_in_s3_bucket()
def test_loading_unicode_characters(self):
"""Loading unicode encoded characters"""
tap_lines = test_utils.get_test_tap_lines(
'messages-with-unicode-characters.json')
# Load with default settings
target_snowflake.persist_lines(self.config, tap_lines)
# Get loaded rows from tables
snowflake = DbSync(self.config)
target_schema = self.config.get('schema', '')
table_unicode = snowflake.query(
"SELECT * FROM {}.test_table_unicode".format(target_schema))
self.assertEqual(table_unicode, [{
'C_INT':
1,
'C_PK':
1,
'C_VARCHAR':
'Hello world, Καλημέρα κόσμε, コンニチハ'
}, {
'C_INT':
2,
'C_PK':
2,
'C_VARCHAR':
'Chinese: 和毛泽东 <<重上井冈山>>. 严永欣, 一九八八年.'
}, {
'C_INT':
3,
'C_PK':
3,
'C_VARCHAR':
'Russian: Зарегистрируйтесь сейчас на Десятую Международную Конференцию по'
}, {
'C_INT':
4,
'C_PK':
4,
'C_VARCHAR':
'Thai: แผ่นดินฮั่นเสื่อมโทรมแสนสังเวช'
}, {
'C_INT':
5,
'C_PK':
5,
'C_VARCHAR':
'Arabic: لقد لعبت أنت وأصدقاؤك لمدة وحصلتم علي من إجمالي النقاط'
}, {
'C_INT':
6,
'C_PK':
6,
'C_VARCHAR':
'Special Characters: [",\'!@£$%^&*()]'
}])
def test_loading_tables_with_client_side_encryption(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning on client-side encryption and load
self.config['client_side_encryption_master_key'] = os.environ.get('CLIENT_SIDE_ENCRYPTION_MASTER_KEY')
self.persist_lines(tap_lines)
self.assert_three_streams_are_into_bigquery()
def test_column_name_change(self):
"""Tests correct renaming of bigquery columns after source change"""
tap_lines_before_column_name_change = test_utils.get_test_tap_lines('messages-with-three-streams.json')
tap_lines_after_column_name_change = test_utils.get_test_tap_lines(
'messages-with-three-streams-modified-column.json')
# Load with default settings
self.persist_lines(tap_lines_before_column_name_change)
self.persist_lines(tap_lines_after_column_name_change)
# Get loaded rows from tables
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
table_one = query(bigquery, "SELECT * FROM {}.test_table_one ORDER BY c_pk".format(target_schema))
table_two = query(bigquery, "SELECT * FROM {}.test_table_two ORDER BY c_pk".format(target_schema))
table_three = query(bigquery, "SELECT * FROM {}.test_table_three ORDER BY c_pk".format(target_schema))
# Table one should have no changes
self.assertEqual(
table_one,
[{'c_int': 1, 'c_pk': 1, 'c_varchar': '1'}])
# Table two should have versioned column
self.assertEquals(
table_two,
[
{'c_int': 1, 'c_pk': 1,
'c_varchar': '1', 'c_date': datetime.datetime(2019, 2, 1, 15, 12, 45, tzinfo=timezone.utc), 'c_date__st': None},
{'c_int': 2, 'c_pk': 2, 'c_varchar': '2',
'c_date': datetime.datetime(2019, 2, 10, 2, tzinfo=timezone.utc), 'c_date__st': '2019-02-12 02:00:00'},
{'c_int': 3, 'c_pk': 3, 'c_varchar': '2', 'c_date': None, 'c_date__st': '2019-02-15 02:00:00'}
]
)
# Table three should have renamed columns
self.assertEqual(
table_three,
[
{'c_int': 1, 'c_pk': 1, 'c_time': datetime.time(4, 0), 'c_varchar': '1', 'c_time_renamed': None},
{'c_int': 2, 'c_pk': 2, 'c_time': datetime.time(7, 15), 'c_varchar': '2', 'c_time_renamed': None},
{'c_int': 3, 'c_pk': 3, 'c_time': datetime.time(23, 0, 3), 'c_varchar': '3',
'c_time_renamed': datetime.time(8, 15)},
{'c_int': 4, 'c_pk': 4, 'c_time': None, 'c_varchar': '4', 'c_time_renamed': datetime.time(23, 0, 3)}
])
def test_loading_unicode_characters(self):
"""Loading unicode encoded characters"""
tap_lines = test_utils.get_test_tap_lines(
"messages-with-unicode-characters.json")
# Load with default settings
self.persist_lines_with_cache(tap_lines)
# Get loaded rows from tables
snowflake = DbSync(self.config)
target_schema = self.config.get("default_target_schema", "")
table_unicode = snowflake.query(
"SELECT * FROM {}.test_table_unicode ORDER BY C_INT".format(
target_schema))
self.assertEqual(
table_unicode,
[
{
"C_INT": 1,
"C_PK": 1,
"C_VARCHAR": "Hello world, Καλημέρα κόσμε, コンニチハ"
},
{
"C_INT": 2,
"C_PK": 2,
"C_VARCHAR": "Chinese: 和毛泽东 <<重上井冈山>>. 严永欣, 一九八八年."
},
{
"C_INT":
3,
"C_PK":
3,
"C_VARCHAR":
"Russian: Зарегистрируйтесь сейчас на Десятую Международную Конференцию по",
},
{
"C_INT": 4,
"C_PK": 4,
"C_VARCHAR": "Thai: แผ่นดินฮั่นเสื่อมโทรมแสนสังเวช"
},
{
"C_INT":
5,
"C_PK":
5,
"C_VARCHAR":
"Arabic: لقد لعبت أنت وأصدقاؤك لمدة وحصلتم علي من إجمالي النقاط",
},
{
"C_INT": 6,
"C_PK": 6,
"C_VARCHAR": "Special Characters: [\",'!@£$%^&*()]"
},
],
)
def test_loading_tables_with_metadata_columns(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning on adding metadata columns
self.config['add_metadata_columns'] = True
self.persist_lines(tap_lines)
# Check if data loaded correctly and metadata columns exist
self.assert_three_streams_are_into_bigquery(should_metadata_columns_exist=True)
def test_loading_csv_files_with_invalid_compression(self):
"""Loading multiple tables from the same input tap with invalid compression"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Turning on a not supported compression method
self.config['compression'] = 'INVALID_COMPRESSION_METHOD'
# Invalid compression method should raise exception
with assert_raises(NotImplementedError):
self.persist_messages(tap_lines)
def test_logical_streams_from_pg_with_hard_delete_and_batch_size_of_5_and_no_records_should_pass(self):
"""Tests logical streams from pg with inserts, updates and deletes"""
tap_lines = test_utils.get_test_tap_lines('messages-pg-logical-streams-no-records.json')
# Turning on hard delete mode
self.config['hard_delete'] = True
self.config['batch_size_rows'] = 5
self.persist_lines(tap_lines)
self.assert_logical_streams_are_in_bigquery_and_are_empty()
def test_loading_tables_with_no_encryption(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines(
"messages-with-three-streams.json")
# Turning off client-side encryption and load
self.config["client_side_encryption_master_key"] = ""
self.persist_lines_with_cache(tap_lines)
self.assert_three_streams_are_into_snowflake()
def test_logical_streams_from_pg_with_hard_delete_and_batch_size_of_5(self):
"""Tests logical streams from pg with inserts, updates and deletes"""
tap_lines = test_utils.get_test_tap_lines('messages-logical-streams.json')
# Turning on hard delete mode
self.config['hard_delete'] = True
self.config['batch_size_rows'] = 5
self.persist_lines_with_cache(tap_lines)
self.assert_logical_streams_are_in_snowflake(True)
def test_loading_tables_with_defined_parallelism(self):
"""Loading multiple tables from the same input tap with various columns types"""
tap_lines = test_utils.get_test_tap_lines('messages-with-three-streams.json')
# Using fixed 1 thread parallelism
self.config['parallelism'] = 1
self.persist_lines(tap_lines)
# Check if data loaded correctly and metadata columns exist
self.assert_three_streams_are_into_bigquery()
def test_table_with_pk_adds_clustering(self):
"""Tests table with a primary key gets clustered on those fields"""
tap_lines = test_utils.get_test_tap_lines('table_with_pk_cluster.json')
self.persist_lines(tap_lines)
# Get loaded rows from tables
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
table = query(bigquery, "SELECT * FROM {}.test_table_cluster ORDER BY c_pk".format(target_schema))
cluster_columns = query(bigquery, "SELECT clustering_ordinal_position, column_name FROM {}.INFORMATION_SCHEMA.COLUMNS WHERE table_name = 'test_table_cluster' AND clustering_ordinal_position > 0 ORDER BY 1".format(target_schema))
# ----------------------------------------------------------------------
# Check rows in table
# ----------------------------------------------------------------------
expected_table = [
{'c_pk': 2, 'c_int': 2, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 12, 2, 0, 0, tzinfo=timezone.utc)},
{'c_pk': 3, 'c_int': 3, 'c_varchar': '2', 'c_date': datetime.datetime(2019, 2, 15, 2, 0, 0, tzinfo=timezone.utc)}
]
expected_cluster_columns = [
{'clustering_ordinal_position': 1, 'column_name': 'c_pk'},
]
self.assertEqual(self.remove_metadata_columns_from_rows(table), expected_table)
self.assertEqual(cluster_columns, expected_cluster_columns)
# ----------------------------------------------------------------------
# Change the primary key and check if clustering stayed unchanged
# ----------------------------------------------------------------------
tap_lines = test_utils.get_test_tap_lines('table_with_pk_cluster_changed.json')
self.persist_lines(tap_lines)
table_changed = query(bigquery, "SELECT * FROM {}.test_table_cluster ORDER BY c_pk".format(target_schema))
cluster_columns_changed = query(bigquery, "SELECT clustering_ordinal_position, column_name FROM {}.INFORMATION_SCHEMA.COLUMNS WHERE table_name = 'test_table_cluster' AND clustering_ordinal_position > 0 ORDER BY 1".format(target_schema))
expected_table_changed = [
{'c_pk': 2, 'c_int': 2, 'c_varchar': 'c', 'c_date': datetime.datetime(2019, 2, 12, 2, 0, 0, tzinfo=timezone.utc)},
{'c_pk': 3, 'c_int': 3, 'c_varchar': 'c', 'c_date': datetime.datetime(2022, 5, 15, 5, 0, 0, tzinfo=timezone.utc)}
]
self.assertEqual(self.remove_metadata_columns_from_rows(table_changed), expected_table_changed)
self.assertEqual(cluster_columns_changed, expected_cluster_columns)
def test_table_with_no_pk(self):
"""Tests table with a primary key gets clustered on those fields"""
tap_lines = test_utils.get_test_tap_lines('table_with_no_pk.json')
self.config['primary_key_required'] = False
self.persist_lines(tap_lines)
# Get loaded rows from tables
bigquery = DbSync(self.config)
target_schema = self.config.get('default_target_schema', '')
table = query(bigquery, "SELECT * FROM {}.test_table_no_pk ORDER BY c_id".format(target_schema))
self.assertEqual(len(table), 2)
def test_loading_tables_with_hard_delete(self):
"""Loading multiple tables from the same input tap with deleted rows"""
tap_lines = test_utils.get_test_tap_lines(
"messages-with-three-streams.json")
# Turning on hard delete mode
self.config["hard_delete"] = True
self.persist_lines_with_cache(tap_lines)
# Check if data loaded correctly and metadata columns exist
self.assert_three_streams_are_into_snowflake(
should_metadata_columns_exist=True, should_hard_deleted_rows=True)
def test_loading_with_multiple_schema(self):
"""Loading table with multiple SCHEMA messages"""
tap_lines = test_utils.get_test_tap_lines(
"messages-with-multi-schemas.json")
# Load with default settings
self.persist_lines_with_cache(tap_lines)
# Check if data loaded correctly
self.assert_three_streams_are_into_snowflake(
should_metadata_columns_exist=False,
should_hard_deleted_rows=False)
def test_non_db_friendly_columns(self):
"""Loading non-db friendly columns like, camelcase, minus signs, etc."""
tap_lines = test_utils.get_test_tap_lines(
"messages-with-non-db-friendly-columns.json")
# Load with default settings
self.persist_lines_with_cache(tap_lines)
# Get loaded rows from tables
snowflake = DbSync(self.config)
target_schema = self.config.get("default_target_schema", "")
table_non_db_friendly_columns = snowflake.query(
"SELECT * FROM {}.test_table_non_db_friendly_columns ORDER BY c_pk"
.format(target_schema))
self.assertEqual(
table_non_db_friendly_columns,
[
{
"C_PK": 1,
"CAMELCASECOLUMN": "Dummy row 1",
"MINUS-COLUMN": "Dummy row 1"
},
{
"C_PK": 2,
"CAMELCASECOLUMN": "Dummy row 2",
"MINUS-COLUMN": "Dummy row 2"
},
{
"C_PK": 3,
"CAMELCASECOLUMN": "Dummy row 3",
"MINUS-COLUMN": "Dummy row 3"
},
{
"C_PK": 4,
"CAMELCASECOLUMN": "Dummy row 4",
"MINUS-COLUMN": "Dummy row 4"
},
{
"C_PK": 5,
"CAMELCASECOLUMN": "Dummy row 5",
"MINUS-COLUMN": "Dummy row 5"
},
],
)
