Skip to content

Commit

Permalink
add unit tests for tag coherence fix
Browse files Browse the repository at this point in the history
  • Loading branch information
@mvesin
mvesin committed Jan 31, 2023
1 parent 0463fb1 commit 3c4aaf5
Showing 1 changed file with 96 additions and 37 deletions.
133 changes: 96 additions & 37 deletions tests/test_dataset_manager.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -596,10 +596,6 @@ def test_dataset_manager_20_modify_database_info(self,
Tests modify_database_info (normal case scenario),
where one replaces an existing dataset by another one
"""
# defining a fake database
fake_database = copy.deepcopy(self.fake_database)


def tinydb_update_side_effect(new_dataset: dict, existing_dataset):
"""
side effect function that mimics the update of the database
Expand All @@ -611,36 +607,99 @@ def tinydb_update_side_effect(new_dataset: dict, existing_dataset):
"""
fake_database['2'] = new_dataset

new_dataset = {"name": "anther_test",
"data_type": "csv",
"tags": [ "other_tags"],
"description": "another_description",
"shape": [2000, 2],
"path": "/path/to/my/other/data",
"dataset_id": "dataset_9876",
"dtypes": ["int64", "float64"]}
def conflicting_tags_side_effect(tags):
if all(t in fake_database.get('2')['tags'] for t in tags):
return [{'dataset_id': dataset_id}]
else:
return []

dataset_id = fake_database.get('2')['dataset_id']
tinydb_update_patch.side_effect = tinydb_update_side_effect
conflicting_tags_patch.return_value = []
conflicting_tags_patch.side_effect = conflicting_tags_side_effect

new_dataset_list = [
{
"name": "anther_test",
"data_type": "csv",
"tags": [ "other_tags"],
"description": "another_description",
"shape": [2000, 2],
"path": "/path/to/my/other/data",
"dataset_id": "dataset_9876",
"dtypes": ["int64", "float64"]
},
{
"tags": self.fake_database['2']['tags']
},
{
"tags": self.fake_database['1']['tags'][:-1] + ['yet another tag']
},
]

for new_dataset in new_dataset_list:
# defining a fake database
fake_database = copy.deepcopy(self.fake_database)
dataset_id = fake_database.get('2')['dataset_id']

# action
self.dataset_manager.modify_database_info(dataset_id, new_dataset)
# action
self.dataset_manager.modify_database_info(dataset_id, new_dataset)

# checks
# check that correct calls are made
tinydb_update_patch.assert_called_once_with(new_dataset,
Query().dataset_id == dataset_id)
# check database status after updating
# first entry in database should be left unchanged ...
self.assertEqual(fake_database.get('1'), self.fake_database.get('1'))
# ... whereas second entry should be updated with variable `new_dataset`
self.assertNotEqual(fake_database.get('2'), self.fake_database.get('2'))
self.assertEqual(fake_database.get('2'), new_dataset)
# checks
# check that correct calls are made
tinydb_update_patch.assert_called_once_with(new_dataset,
Query().dataset_id == dataset_id)
# check database status after updating
# first entry in database should be left unchanged ...
self.assertEqual(fake_database.get('1'), self.fake_database.get('1'))
# ... whereas second entry should be updated with variable `new_dataset`
self.assertNotEqual(fake_database.get('2'), self.fake_database.get('2'))
self.assertEqual(fake_database.get('2'), new_dataset)

tinydb_update_patch.reset_mock()


@patch('fedbiomed.node.dataset_manager.DatasetManager.search_conflicting_tags')
def test_dataset_manager_20_modify_database_info_errpr(self,
conflicting_tags_patch):
"""
Tests modify_database_info (error case scenario),
where one replaces an existing dataset by another one
"""
def conflicting_tags_side_effect(tags):
return [{'dataset_id': 'one dataset is conflicting', 'name': 'fake dataset with conflicting tags'}]

conflicting_tags_patch.side_effect = conflicting_tags_side_effect

new_dataset_list = [
{
"name": "anther_test",
"data_type": "csv",
"tags": [ "other_tags"],
"description": "another_description",
"shape": [2000, 2],
"path": "/path/to/my/other/data",
"dataset_id": "dataset_9876",
"dtypes": ["int64", "float64"]
},
{
"tags": self.fake_database['2']['tags']
},
{
"tags": self.fake_database['1']['tags'][:-1] + ['yet another tag']
},
]

for new_dataset in new_dataset_list:
# defining a fake database
fake_database = copy.deepcopy(self.fake_database)
dataset_id = fake_database.get('2')['dataset_id']

# action + check
with self.assertRaises(FedbiomedDatasetManagerError):
self.dataset_manager.modify_database_info(dataset_id, new_dataset)


@patch('tinydb.table.Table.all')
def test_dataset_manager_21_list_my_data(self,
def test_dataset_manager_22_list_my_data(self,
query_all_patch):
"""
Checks `list_my_data` method in the normal case scenario
Expand Down Expand Up @@ -681,7 +740,7 @@ def test_dataset_manager_21_list_my_data(self,


@patch('fedbiomed.node.dataset_manager.DatasetManager.load_default_database')
def test_dataset_manager_22_load_as_dataloader_default(self,
def test_dataset_manager_23_load_as_dataloader_default(self,
load_default_database_patch):
"""
Tests `load_as_dataloader` method where the input
Expand Down Expand Up @@ -711,7 +770,7 @@ def test_dataset_manager_22_load_as_dataloader_default(self,


@patch('fedbiomed.node.dataset_manager.DatasetManager.load_images_dataset')
def test_dataset_manager_23_load_as_dataloader_images(self, load_images_dataset_patch):
def test_dataset_manager_24_load_as_dataloader_images(self, load_images_dataset_patch):
"""
Tests `load_as_dataloader` method where the input
dataset is a images dataset
Expand Down Expand Up @@ -742,7 +801,7 @@ def test_dataset_manager_23_load_as_dataloader_images(self, load_images_dataset_
@patch('fedbiomed.node.dataset_manager.DatasetManager.read_csv')
@patch('os.path.isfile')
@patch('fedbiomed.node.dataset_manager.DatasetManager.search_by_tags')
def test_dataset_manager_24_load_data_file(self,
def test_dataset_manager_25_load_data_file(self,
search_by_tags_patch,
os_isfile_patch,
read_csv_patch
Expand Down Expand Up @@ -809,7 +868,7 @@ def test_dataset_manager_24_load_data_file(self,
@patch('os.path.isdir')
@patch('os.path.isfile')
@patch('fedbiomed.node.dataset_manager.DatasetManager.search_by_tags')
def test_dataset_manager_25_load_data_folder(self,
def test_dataset_manager_26_load_data_folder(self,
search_by_tags_patch,
os_isfile_patch,
os_isdir_patch,
Expand Down Expand Up @@ -848,7 +907,7 @@ def test_dataset_manager_25_load_data_folder(self,
with self.assertRaises(NotImplementedError):
self.dataset_manager.load_data(tags, mode='pandas')

def test_dataset_manager_26_load_data_exception(self):
def test_dataset_manager_27_load_data_exception(self):
"""
Tests if an exception is raised when passing an
unknown mode to `load_data` method
Expand All @@ -860,7 +919,7 @@ def test_dataset_manager_26_load_data_exception(self):
with self.assertRaises(NotImplementedError):
self.dataset_manager.load_data(tags, mode='unknown_mode')

def test_dataset_manager_27_load_existing_mednist_dataset(self):
def test_dataset_manager_28_load_existing_mednist_dataset(self):
"""
Tests case where one is loading mednist dataset without downloading it
"""
Expand Down Expand Up @@ -921,7 +980,7 @@ def _create_fake_mednist_dataset(self):


@patch('fedbiomed.node.dataset_manager.urlretrieve')
def test_dataset_manager_28_download_extrat_mednist(self,
def test_dataset_manager_29_download_extrat_mednist(self,
urlretrieve_patch):
"""
Tests the correct process of data download and extraction
Expand Down Expand Up @@ -962,7 +1021,7 @@ def urlretieve_side_effect(url, path):
if i % 2 != 0:
label += 1

def test_dataset_manager_29_load_mednist_database_exception(self):
def test_dataset_manager_30_load_mednist_database_exception(self):
"""
Tests if exception `FedbiomedDatasetManagerError` is triggered
when mednist dataset folder is empty
Expand All @@ -977,7 +1036,7 @@ def test_dataset_manager_29_load_mednist_database_exception(self):
# and checking if method raises FedbiomedDatasetManagerError
self.dataset_manager.load_mednist_database(self.tempdir)

def test_dataset_manager_30_obfuscate_private_information(self):
def test_dataset_manager_31_obfuscate_private_information(self):
"""Tests if error is raised if dataset is not parsable when calling `obfuscate_privte_information"""
metadata_with_private_info = [{
'path': 'private/info',
Expand All @@ -1001,7 +1060,7 @@ def test_dataset_manager_30_obfuscate_private_information(self):
_ = DatasetManager.obfuscate_private_information([*metadata_with_private_info, 'non-dict-like'])

@patch('os.path.isdir')
def test_dataset_manager_31_data_loading_plan_save(self, patch_isdir):
def test_dataset_manager_32_data_loading_plan_save(self, patch_isdir):
"""Tests that DatasetManager correctly saves a DataLoadingPlan"""
patch_isdir.return_value = True
from test_data_loading_plan import LoadingBlockForTesting
Expand Down

0 comments on commit 3c4aaf5

Please sign in to comment.