def test_anonymise_directory(tmp_path):
temp_filepath = tmp_path / "test.dcm"
temp_anon_filepath = label_dicom_filepath_as_anonymised(temp_filepath)
try:
copyfile(TEST_FILEPATH, temp_filepath)
assert not is_anonymised_directory(tmp_path)
# Test file deletion
anonymise_directory(
tmp_path, delete_original_files=False, anonymise_filenames=False
)
# # File should be anonymised but not dir, since original file
# # is still present.
assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
assert not is_anonymised_directory(tmp_path)
remove_file(temp_anon_filepath)
anonymise_directory(
tmp_path, delete_original_files=True, anonymise_filenames=False
)
# # File and dir should be anonymised since original file should
# # have been deleted.
assert is_anonymised_file(temp_anon_filepath)
assert not exists(temp_filepath)
assert is_anonymised_directory(tmp_path)
finally:
remove_file(temp_anon_filepath)
def _test_pseudonymise_cli_for_file(tmp_path, test_file_path):
temp_filepath = pjoin(tmp_path, "test.dcm")
try:
logging.info("CLI test on %s", test_file_path)
copyfile(test_file_path, temp_filepath)
# Basic file pseudonymisation
# Initially, just make sure it exits with zero and doesn't fail to generate output
assert not is_anonymised_file(temp_filepath)
# need the SOP Instance UID and SOP Class name to figure out the destination file name
# but will also be using the dataset to do some comparisons.
ds_input: pydicom.FileDataset = pydicom.dcmread(temp_filepath,
force=True)
pseudo_sop_instance_uid = pseudonymisation_api.pseudonymisation_dispatch[
"UI"](ds_input.SOPInstanceUID)
sop_class_uid: pydicom.dataelem.DataElement = ds_input.SOPClassUID
mode_prefix = DICOM_SOP_CLASS_NAMES_MODE_PREFIXES[
sop_class_uid.name # pylint: disable = no-member
]
temp_anon_filepath = pjoin(
tmp_path,
"{}.{}_Anonymised.dcm".format(mode_prefix,
pseudo_sop_instance_uid),
)
assert not exists(temp_anon_filepath)
anon_file_command = "pymedphys --verbose experimental dicom anonymise --pseudo".split(
) + [temp_filepath]
logging.info("Command line: %s", anon_file_command)
try:
subprocess.check_call(anon_file_command)
assert exists(temp_anon_filepath)
# assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
ds_pseudo = pydicom.dcmread(temp_anon_filepath, force=True)
assert ds_input["PatientID"].value != ds_pseudo["PatientID"].value
finally:
remove_file(temp_anon_filepath)
# Basic dir anonymisation
assert not is_anonymised_directory(tmp_path)
assert not exists(temp_anon_filepath)
anon_dir_command = "pymedphys --verbose experimental dicom anonymise --pseudo".split(
) + [str(tmp_path)]
try:
subprocess.check_call(anon_dir_command)
# assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
finally:
remove_file(temp_filepath)
def test_anonymise_file():
assert not is_anonymised_file(TEST_FILEPATH)
temp_basename = "{}_{}.dcm".format(".".join(TEST_FILEPATH.split(".")[:-1]),
uuid4())
temp_filepath = pjoin(dirname(TEST_FILEPATH), temp_basename)
anon_private_filepath = ""
anon_filepath_orig = ""
anon_filepath_pres = ""
try:
# Private tag handling
anon_private_filepath = anonymise_file(TEST_FILEPATH,
delete_private_tags=False)
assert not is_anonymised_file(anon_private_filepath,
ignore_private_tags=False)
assert is_anonymised_file(anon_private_filepath,
ignore_private_tags=True)
anon_private_filepath = anonymise_file(TEST_FILEPATH,
delete_private_tags=True)
assert is_anonymised_file(anon_private_filepath,
ignore_private_tags=False)
# Filename is anonymised?
assert basename(anon_private_filepath) == TEST_ANON_BASENAME
# Deletion of original file
copyfile(TEST_FILEPATH, temp_filepath)
anon_filepath_orig = anonymise_file(temp_filepath,
delete_original_file=True)
assert is_anonymised_file(anon_filepath_orig)
assert not exists(temp_filepath)
# Preservation of filename if desired
expected_filepath = "{}_Anonymised.dcm".format(".".join(
TEST_FILEPATH.split(".")[:-1]))
anon_filepath_pres = anonymise_file(TEST_FILEPATH,
anonymise_filename=False)
assert anon_filepath_pres == expected_filepath
finally:
remove_file(temp_filepath)
remove_file(anon_private_filepath)
remove_file(anon_filepath_orig)
remove_file(anon_filepath_pres)
def _check_is_anonymised_dataset_file_and_dir(
ds, tmp_path, anon_is_expected=True, ignore_private_tags=False
):
temp_filepath = str(tmp_path / "test.dcm")
try:
create.set_default_transfer_syntax(ds)
ds.file_meta = TEST_FILE_META
ds.save_as(temp_filepath, write_like_original=False)
if anon_is_expected:
assert is_anonymised_dataset(ds, ignore_private_tags)
assert is_anonymised_file(temp_filepath, ignore_private_tags)
assert is_anonymised_directory(tmp_path, ignore_private_tags)
else:
assert not is_anonymised_dataset(ds, ignore_private_tags)
assert not is_anonymised_file(temp_filepath, ignore_private_tags)
assert not is_anonymised_directory(tmp_path, ignore_private_tags)
finally:
remove_file(temp_filepath)
def test_alternative_identifying_keywords():
alternative_keyword_list = copy.copy(get_default_identifying_keywords())
alternative_keyword_list.append("SOPInstanceUID")
test_file_path = get_treatment_record_test_file_path()
ds_test = pydicom.dcmread(test_file_path, force=True)
with tempfile.TemporaryDirectory() as output_directory:
anon_private_filepath = anonymise_file(
test_file_path,
output_filepath=output_directory,
delete_private_tags=True,
identifying_keywords=alternative_keyword_list,
)
ds_anon = pydicom.dcmread(anon_private_filepath, force=True)
assert is_anonymised_file(anon_private_filepath, ignore_private_tags=False)
assert ds_test["SOPInstanceUID"].value != ds_anon["SOPInstanceUID"].value
def _test_pseudonymise_file_at_path(
test_file_path, test_identifying_keywords=None, test_replacement_strategy=None
):
assert not is_anonymised_file(test_file_path)
if test_identifying_keywords is None:
identifying_keywords_for_pseudo = (
pseudonymisation_api.get_default_pseudonymisation_keywords()
)
logging.info("Using pseudonymisation keywords")
else:
identifying_keywords_for_pseudo = test_identifying_keywords
if test_replacement_strategy is None:
replacement_strategy = pseudonymisation_api.pseudonymisation_dispatch
logging.info("Using pseudonymisation strategy")
else:
replacement_strategy = test_replacement_strategy
with tempfile.TemporaryDirectory() as output_directory:
pseudonymised_file_path = anonymise_file(
dicom_filepath=test_file_path,
output_filepath=output_directory,
delete_original_file=False,
anonymise_filename=True,
replace_values=True,
# keywords_to_leave_unchanged=None,
delete_private_tags=True,
delete_unknown_tags=True,
replacement_strategy=replacement_strategy,
identifying_keywords=identifying_keywords_for_pseudo,
)
# debug print + Assert to force the print
# print("Pseudonymised file at: ", pseudonymised_file_path)
# assert False
assert exists(pseudonymised_file_path)
ds_input = pydicom.dcmread(test_file_path, force=True)
ds_pseudo = pydicom.dcmread(pseudonymised_file_path, force=True)
# simplistic stand-in to make sure *something* is happening
assert ds_input["PatientID"].value != ds_pseudo["PatientID"].value
# make sure that we are not accidentally using the hardcode replacement approach
assert ds_pseudo["PatientID"].value not in ["", "Anonymous"]
def test_anonymise_cli(tmp_path):
temp_filepath = str(tmp_path / "test.dcm")
try:
copyfile(TEST_FILEPATH, temp_filepath)
temp_anon_filepath = str(tmp_path / TEST_ANON_BASENAME)
# Basic file anonymisation
assert not is_anonymised_file(temp_filepath)
assert not exists(temp_anon_filepath)
anon_file_command = "pymedphys dicom anonymise".split() + [temp_filepath]
try:
subprocess.check_call(anon_file_command)
assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
# File anonymisation - preserve filenames
assert not is_anonymised_file(temp_filepath)
expected_anon_filepath = label_dicom_filepath_as_anonymised(temp_filepath)
assert not exists(expected_anon_filepath)
anon_file_pres_command = "pymedphys dicom anonymise -f".split() + [
temp_filepath
]
try:
subprocess.check_call(anon_file_pres_command)
assert is_anonymised_file(expected_anon_filepath)
assert exists(temp_filepath)
finally:
remove_file(expected_anon_filepath)
# File anonymisation - clear values
assert not is_anonymised_file(temp_filepath)
assert not exists(temp_anon_filepath)
temp_cleared_anon_filepath = str(tmp_path / TEST_ANON_BASENAME)
anon_file_clear_command = "pymedphys dicom anonymise -c".split() + [
temp_filepath
]
try:
subprocess.check_call(anon_file_clear_command)
assert is_anonymised_file(temp_cleared_anon_filepath)
assert pydicom.dcmread(temp_cleared_anon_filepath).PatientName == ""
assert exists(temp_filepath)
finally:
remove_file(temp_cleared_anon_filepath)
# File anonymisation - leave keywords unchanged
assert not is_anonymised_file(temp_filepath)
assert not exists(temp_anon_filepath)
anon_file_keep_command = (
"pymedphys dicom anonymise".split()
+ [temp_filepath]
+ "-k PatientName".split()
)
try:
subprocess.check_call(anon_file_keep_command)
assert not is_anonymised_file(temp_anon_filepath)
ds = pydicom.dcmread(temp_anon_filepath)
ds.PatientName = "Anonymous"
assert is_anonymised_dataset(ds)
assert exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
# File anonymisation - private tag handling
assert not is_anonymised_file(temp_filepath)
assert not exists(temp_anon_filepath)
anon_file_private_command = "pymedphys dicom anonymise -p".split() + [
temp_filepath
]
try:
subprocess.check_call(anon_file_private_command)
assert not is_anonymised_file(temp_anon_filepath)
assert is_anonymised_file(temp_anon_filepath, ignore_private_tags=True)
assert exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
# TODO: File anonymisation - unknown tag handling
# # Calling a subprocess reloads BASELINE_DICOM_DICT...
# Basic dir anonymisation
assert not is_anonymised_directory(tmp_path)
assert not exists(temp_anon_filepath)
anon_dir_command = "pymedphys dicom anonymise".split() + [str(tmp_path)]
try:
subprocess.check_call(anon_dir_command)
assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
finally:
remove_file(temp_filepath)
def test_anonymise_directory(tmp_path):
temp_filepath = tmp_path / "test.dcm"
temp_anon_filepath = label_dicom_filepath_as_anonymised(temp_filepath)
temp_record_filepath = tmp_path / "test_record.dcm"
temp_anon_record_filepath = label_dicom_filepath_as_anonymised(temp_record_filepath)
try:
copyfile(get_rtplan_test_file_path(), temp_filepath)
assert not is_anonymised_directory(tmp_path)
# Test file deletion
anon_path_list = anonymise_directory(
tmp_path, delete_original_files=False, anonymise_filenames=False
)
# # File should be anonymised but not dir, since original file
# # is still present.
assert is_anonymised_file(temp_anon_filepath)
assert exists(temp_filepath)
assert not is_anonymised_directory(tmp_path)
assert anon_path_list is not None
assert anon_path_list[0] == temp_anon_filepath
remove_file(temp_anon_filepath)
anon_path_list = anonymise_directory(
tmp_path, delete_original_files=True, anonymise_filenames=False
)
# # File and dir should be anonymised since original file should
# # have been deleted.
assert is_anonymised_file(temp_anon_filepath)
assert not exists(temp_filepath)
assert is_anonymised_directory(tmp_path)
assert anon_path_list[0] == temp_anon_filepath
# Test fail fast vs. fail at last
# if the function fails fast, the specified removal
# will not take place
# if the function does not fail fail, the specified
# removal will take place
logging.warning("Testing fail fast")
remove_file(temp_anon_filepath)
copyfile(get_rtplan_test_file_path(), temp_filepath)
copyfile(get_treatment_record_test_file_path(), temp_record_filepath)
ds_record = pydicom.dcmread(temp_record_filepath, force=True)
# deliberately add a DICOM element that is not in the current
# dictionary
ds_record.add_new([0x300A, 0x9999], "FL", [1.0, 1.0])
pydicom.dcmwrite(temp_record_filepath, ds_record)
with pytest.raises((KeyError, ValueError)):
anon_path_list = anonymise_directory(
tmp_path,
delete_original_files=True,
anonymise_filenames=False,
fail_fast=True,
)
logging.warning(anon_path_list)
assert exists(temp_filepath)
with pytest.raises((KeyError, ValueError)):
anon_path_list = anonymise_directory(
tmp_path,
delete_original_files=True,
anonymise_filenames=False,
fail_fast=False,
)
logging.warning(anon_path_list)
assert not exists(temp_filepath)
finally:
remove_file(temp_anon_filepath)
remove_file(temp_anon_record_filepath)
