def test_charset_patient_names(self, filename, patient_name):
"""Test patient names are correctly decoded and encoded."""
# check that patient names are correctly read
file_path = get_charset_files(filename + '.dcm')[0]
ds = dcmread(file_path)
ds.decode()
assert patient_name == ds.PatientName
# check that patient names are correctly written back
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds = dcmread(fp)
assert patient_name == ds.PatientName
# check that patient names are correctly written back
# without original byte string (PersonName3 only)
if hasattr(ds.PatientName, 'original_string'):
ds.PatientName.original_string = None
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds = dcmread(fp)
assert patient_name == ds.PatientName
def test_seq_item_looks_like_explicit_VR(self):
# For issue 999.
# Set up an implicit VR dataset with a "normal" group 8 tag,
# followed by a sequence with an item (dataset) having
# a data element length that looks like a potential valid VR
ds = Dataset()
ds.file_meta = Dataset()
ds.file_meta.MediaStorageSOPClassUID = "1.1.1"
ds.file_meta.MediaStorageSOPInstanceUID = "2.2.2"
ds.is_implicit_VR = True
ds.is_little_endian = True
ds.SOPClassUID = '9.9.9' # First item group 8 in top-level dataset
seq = Sequence()
seq_ds = Dataset()
seq_ds.BadPixelImage = b"\3" * 0x5244 # length looks like "DR"
seq.append(seq_ds)
ds.ReferencedImageSequence = seq
dbio = DicomBytesIO()
ds.save_as(dbio, write_like_original=False)
# Now read the constructed dataset back in
# In original issue, shows warning that has detected what appears
# to be Explicit VR, then throws NotImplemented for the unknown VR
dbio.seek(0)
ds = dcmread(dbio)
ds.remove_private_tags() # forces it to actually parse SQ
def test_japanese_multi_byte_personname(self):
"""Test japanese person name which has multi byte strings are
correctly encoded."""
file_path = get_charset_files('chrH32.dcm')[0]
ds = dcmread(file_path)
ds.decode()
if hasattr(ds.PatientName, 'original_string'):
original_string = ds.PatientName.original_string
ds.PatientName.original_string = None
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds_out = dcmread(fp)
assert original_string == ds_out.PatientName.original_string
japanese_pn = PersonName(u"Mori^Ogai=森^鷗外=もり^おうがい")
pyencs = pydicom.charset.convert_encodings(
["ISO 2022 IR 6", "ISO 2022 IR 87", "ISO 2022 IR 159"])
actual_encoded = bytes(japanese_pn.encode(pyencs))
expect_encoded = (
b"\x4d\x6f\x72\x69\x5e\x4f\x67\x61\x69\x3d\x1b\x24\x42\x3f"
b"\x39\x1b\x28\x42\x5e\x1b\x24\x28\x44\x6c\x3f\x1b\x24\x42"
b"\x33\x30\x1b\x28\x42\x3d\x1b\x24\x42\x24\x62\x24\x6a\x1b"
b"\x28\x42\x5e\x1b\x24\x42\x24\x2a\x24\x26\x24\x2c\x24\x24"
b"\x1b\x28\x42")
assert expect_encoded == actual_encoded
def test_charset_patient_names(self, filename, patient_name):
"""Test patient names are correctly decoded and encoded."""
# check that patient names are correctly read
file_path = get_charset_files(filename + '.dcm')[0]
ds = dcmread(file_path)
ds.decode()
assert patient_name == ds.PatientName
# check that patient names are correctly written back
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds = dcmread(fp)
assert patient_name == ds.PatientName
# check that patient names are correctly written back
# without original byte string (PersonName3 only)
if hasattr(ds.PatientName, 'original_string'):
ds.PatientName.original_string = None
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds = dcmread(fp)
assert patient_name == ds.PatientName
def test_changed_character_set(self):
# Regression test for #629
multiPN_name = get_charset_files("chrFrenMulti.dcm")[0]
ds = dcmread(multiPN_name) # is Latin-1
ds.SpecificCharacterSet = 'ISO_IR 192'
from pydicom.filebase import DicomBytesIO
fp = DicomBytesIO()
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds_out = dcmread(fp)
# we expect UTF-8 encoding here
assert b'Buc^J\xc3\xa9r\xc3\xb4me' == ds_out.get_item(0x00100010).value
def test_changed_character_set(self):
# Regression test for #629
multiPN_name = get_charset_files("chrFrenMulti.dcm")[0]
ds = dcmread(multiPN_name) # is Latin-1
ds.SpecificCharacterSet = 'ISO_IR 192'
from pydicom.filebase import DicomBytesIO
fp = DicomBytesIO()
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds_out = dcmread(fp)
# we expect UTF-8 encoding here
assert b'Buc^J\xc3\xa9r\xc3\xb4me' == ds_out.get_item(0x00100010).value
def test_read_file_does_not_raise(self):
"""Test that reading from DicomBytesIO does not raise on EOF.
Regression test for #358."""
ds = dcmread(mr_name)
fp = DicomBytesIO()
ds.save_as(fp)
fp.seek(0)
de_gen = data_element_generator(fp, False, True)
try:
while True:
next(de_gen)
except StopIteration:
pass
except EOFError:
self.fail('Unexpected EOFError raised')
def test_dcmread_does_not_raise(self):
"""Test that reading from DicomBytesIO does not raise on EOF.
Regression test for #358."""
ds = dcmread(mr_name)
fp = DicomBytesIO()
ds.save_as(fp, write_like_original=True)
fp.seek(0)
de_gen = data_element_generator(fp, False, True)
try:
while True:
next(de_gen)
except StopIteration:
pass
except EOFError:
self.fail('Unexpected EOFError raised')
def test_japanese_multi_byte_personname(self):
"""Test japanese person name which has multi byte strings are
correctly encoded."""
file_path = get_charset_files('chrH32.dcm')[0]
ds = dcmread(file_path)
ds.decode()
if hasattr(ds.PatientName, 'original_string'):
original_string = ds.PatientName.original_string
ds.PatientName.original_string = None
fp = DicomBytesIO()
fp.is_implicit_VR = False
fp.is_little_endian = True
ds.save_as(fp, write_like_original=False)
fp.seek(0)
ds_out = dcmread(fp)
assert original_string == ds_out.PatientName.original_string
def write_file_meta_info(fp, file_meta, enforce_standard=True):
"""Write the File Meta Information elements in `file_meta` to `fp`.
If `enforce_standard` is ``True`` then the file-like `fp` should be
positioned past the 128 byte preamble + 4 byte prefix (which should
already have been written).
**DICOM File Meta Information Group Elements**
From the DICOM standard, Part 10,
:dcm:`Section 7.1<part10/chapter_7.html#sect_7.1>`, any DICOM file shall
contain a 128-byte preamble, a 4-byte DICOM prefix 'DICM' and (at a
minimum) the following Type 1 DICOM Elements (from
:dcm:`Table 7.1-1<part10/chapter_7.html#table_7.1-1>`):
* (0002,0000) *File Meta Information Group Length*, UL, 4
* (0002,0001) *File Meta Information Version*, OB, 2
* (0002,0002) *Media Storage SOP Class UID*, UI, N
* (0002,0003) *Media Storage SOP Instance UID*, UI, N
* (0002,0010) *Transfer Syntax UID*, UI, N
* (0002,0012) *Implementation Class UID*, UI, N
If `enforce_standard` is ``True`` then (0002,0000) will be added/updated,
(0002,0001) and (0002,0012) will be added if not already present and the
other required elements will be checked to see if they exist. If
`enforce_standard` is ``False`` then `file_meta` will be written as is
after minimal validation checking.
The following Type 3/1C Elements may also be present:
* (0002,0013) *Implementation Version Name*, SH, N
* (0002,0016) *Source Application Entity Title*, AE, N
* (0002,0017) *Sending Application Entity Title*, AE, N
* (0002,0018) *Receiving Application Entity Title*, AE, N
* (0002,0102) *Private Information*, OB, N
* (0002,0100) *Private Information Creator UID*, UI, N
If `enforce_standard` is ``True`` then (0002,0013) will be added/updated.
*Encoding*
The encoding of the *File Meta Information* shall be *Explicit VR Little
Endian*.
Parameters
----------
fp : file-like
The file-like to write the File Meta Information to.
file_meta : pydicom.dataset.Dataset
The File Meta Information elements.
enforce_standard : bool
If ``False``, then only the *File Meta Information* elements already in
`file_meta` will be written to `fp`. If ``True`` (default) then a DICOM
Standards conformant File Meta will be written to `fp`.
Raises
------
ValueError
If `enforce_standard` is ``True`` and any of the required *File Meta
Information* elements are missing from `file_meta`, with the
exception of (0002,0000), (0002,0001) and (0002,0012).
ValueError
If any non-Group 2 Elements are present in `file_meta`.
"""
validate_file_meta(file_meta, enforce_standard)
if enforce_standard and 'FileMetaInformationGroupLength' not in file_meta:
# Will be updated with the actual length later
file_meta.FileMetaInformationGroupLength = 0
# Write the File Meta Information Group elements
# first write into a buffer to avoid seeking back, that can be
# expansive and is not allowed if writing into a zip file
buffer = DicomBytesIO()
buffer.is_little_endian = True
buffer.is_implicit_VR = False
write_dataset(buffer, file_meta)
# If FileMetaInformationGroupLength is present it will be the first written
# element and we must update its value to the correct length.
if 'FileMetaInformationGroupLength' in file_meta:
# Update the FileMetaInformationGroupLength value, which is the number
# of bytes from the end of the FileMetaInformationGroupLength element
# to the end of all the File Meta Information elements.
# FileMetaInformationGroupLength has a VR of 'UL' and so has a value
# that is 4 bytes fixed. The total length of when encoded as
# Explicit VR must therefore be 12 bytes.
file_meta.FileMetaInformationGroupLength = buffer.tell() - 12
buffer.seek(0)
write_data_element(buffer, file_meta[0x00020000])
fp.write(buffer.getvalue())
class TestWriteFileMetaInfo(unittest.TestCase):
"""Unit tests for _write_file_meta_info."""
def setUp(self):
"""Create an empty file-like for use in testing."""
self.fp = DicomBytesIO()
def test_bad_elements(self):
"""Test that non-group 2 elements aren't written to the file meta."""
meta = Dataset()
meta.PatientID = '12345678'
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
def test_missing_elements(self):
"""Test that missing required elements raises ValueError."""
meta = Dataset()
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.MediaStorageSOPClassUID = '1.1'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.MediaStorageSOPInstanceUID = '1.2'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.TransferSyntaxUID = '1.3'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
def test_prefix(self):
"""Test that the 'DICM' prefix is present."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(0)
prefix = self.fp.read(4)
self.assertEqual(prefix, b'DICM')
def test_group_length(self):
"""Test that the value for FileMetaInformationGroupLength is OK."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
# 78 in total, - 4 for prefix, - 12 for group length = 62
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x3E\x00\x00\x00')
def test_group_length_updated(self):
"""Test that FileMetaInformationGroupLength gets updated if present."""
meta = Dataset()
meta.FileMetaInformationGroupLength = 100 # Actual length
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x3E\x00\x00\x00')
# Check original file meta is unchanged/updated
self.assertEqual(meta.FileMetaInformationGroupLength, 62)
self.assertEqual(meta.FileMetaInformationVersion, b'\x00\x01')
self.assertEqual(meta.MediaStorageSOPClassUID, '1.1')
self.assertEqual(meta.MediaStorageSOPInstanceUID, '1.2')
self.assertEqual(meta.TransferSyntaxUID, '1.3')
self.assertEqual(meta.ImplementationClassUID, '1.4')
def test_version(self):
"""Test that the value for FileMetaInformationVersion is OK."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(16 + 12)
self.assertEqual(self.fp.read(2), b'\x00\x01')
def test_filelike_position(self):
"""Test that the file-like's ending position is OK."""
# 4 bytes prefix
# 8 + 4 bytes FileMetaInformationGroupLength
# 12 + 2 bytes FileMetaInformationVersion
# 8 + 4 bytes MediaStorageSOPClassUID
# 8 + 4 bytes MediaStorageSOPInstanceUID
# 8 + 4 bytes TransferSyntaxUID
# 8 + 4 bytes ImplementationClassUID
# 78 bytes total
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.assertEqual(self.fp.tell(), 78)
# 8 + 6 bytes ImplementationClassUID
# 80 bytes total, group length 64
self.fp.seek(0)
meta.ImplementationClassUID = '1.4.1'
_write_file_meta_info(self.fp, meta)
# Check File Meta length
self.assertEqual(self.fp.tell(), 80)
# Check Group Length
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x40\x00\x00\x00')
_write_file_meta_info(self.fp, meta)
def write_file_meta_info(fp, file_meta, enforce_standard=True):
"""Write the File Meta Information elements in `file_meta` to `fp`.
If `enforce_standard` is True then the file-like `fp` should be positioned
past the 128 byte preamble + 4 byte prefix (which should already have been
written).
DICOM File Meta Information Group Elements
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From the DICOM standard, Part 10 Section 7.1, any DICOM file shall contain
a 128-byte preamble, a 4-byte DICOM prefix 'DICM' and (at a minimum) the
following Type 1 DICOM Elements (from Table 7.1-1):
* (0002,0000) FileMetaInformationGroupLength, UL, 4
* (0002,0001) FileMetaInformationVersion, OB, 2
* (0002,0002) MediaStorageSOPClassUID, UI, N
* (0002,0003) MediaStorageSOPInstanceUID, UI, N
* (0002,0010) TransferSyntaxUID, UI, N
* (0002,0012) ImplementationClassUID, UI, N
If `enforce_standard` is True then (0002,0000) will be added/updated,
(0002,0001) and (0002,0012) will be added if not already present and the
other required elements will be checked to see if they exist. If
`enforce_standard` is False then `file_meta` will be written as is after
minimal validation checking.
The following Type 3/1C Elements may also be present:
* (0002,0013) ImplementationVersionName, SH, N
* (0002,0016) SourceApplicationEntityTitle, AE, N
* (0002,0017) SendingApplicationEntityTitle, AE, N
* (0002,0018) ReceivingApplicationEntityTitle, AE, N
* (0002,0100) PrivateInformationCreatorUID, UI, N
* (0002,0102) PrivateInformation, OB, N
If `enforce_standard` is True then (0002,0013) will be added/updated.
Encoding
~~~~~~~~
The encoding of the File Meta Information shall be Explicit VR Little
Endian
Parameters
----------
fp : file-like
The file-like to write the File Meta Information to.
file_meta : pydicom.dataset.Dataset
The File Meta Information DataElements.
enforce_standard : bool
If False, then only the File Meta Information elements already in
`file_meta` will be written to `fp`. If True (default) then a DICOM
Standards conformant File Meta will be written to `fp`.
Raises
------
ValueError
If `enforce_standard` is True and any of the required File Meta
Information elements are missing from `file_meta`, with the
exception of (0002,0000), (0002,0001) and (0002,0012).
ValueError
If any non-Group 2 Elements are present in `file_meta`.
"""
validate_file_meta(file_meta, enforce_standard)
if enforce_standard and 'FileMetaInformationGroupLength' not in file_meta:
# Will be updated with the actual length later
file_meta.FileMetaInformationGroupLength = 0
# Write the File Meta Information Group elements
# first write into a buffer to avoid seeking back, that can be
# expansive and is not allowed if writing into a zip file
buffer = DicomBytesIO()
buffer.is_little_endian = True
buffer.is_implicit_VR = False
write_dataset(buffer, file_meta)
# If FileMetaInformationGroupLength is present it will be the first written
# element and we must update its value to the correct length.
if 'FileMetaInformationGroupLength' in file_meta:
# Update the FileMetaInformationGroupLength value, which is the number
# of bytes from the end of the FileMetaInformationGroupLength element
# to the end of all the File Meta Information elements.
# FileMetaInformationGroupLength has a VR of 'UL' and so has a value
# that is 4 bytes fixed. The total length of when encoded as
# Explicit VR must therefore be 12 bytes.
file_meta.FileMetaInformationGroupLength = buffer.tell() - 12
buffer.seek(0)
write_data_element(buffer, file_meta[0x00020000])
fp.write(buffer.getvalue())
class TestWriteFileMetaInfo(unittest.TestCase):
"""Unit tests for _write_file_meta_info."""
def setUp(self):
"""Create an empty file-like for use in testing."""
self.fp = DicomBytesIO()
def test_bad_elements(self):
"""Test that non-group 2 elements aren't written to the file meta."""
meta = Dataset()
meta.PatientID = '12345678'
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
def test_missing_elements(self):
"""Test that missing required elements raises ValueError."""
meta = Dataset()
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.MediaStorageSOPClassUID = '1.1'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.MediaStorageSOPInstanceUID = '1.2'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.TransferSyntaxUID = '1.3'
self.assertRaises(ValueError, _write_file_meta_info, self.fp, meta)
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
def test_prefix(self):
"""Test that the 'DICM' prefix is present."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(0)
prefix = self.fp.read(4)
self.assertEqual(prefix, b'DICM')
def test_group_length(self):
"""Test that the value for FileMetaInformationGroupLength is OK."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
# 78 in total, - 4 for prefix, - 12 for group length = 62
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x3E\x00\x00\x00')
def test_group_length_updated(self):
"""Test that FileMetaInformationGroupLength gets updated if present."""
meta = Dataset()
meta.FileMetaInformationGroupLength = 100 # Actual length
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x3E\x00\x00\x00')
# Check original file meta is unchanged/updated
self.assertEqual(meta.FileMetaInformationGroupLength, 62)
self.assertEqual(meta.FileMetaInformationVersion, b'\x00\x01')
self.assertEqual(meta.MediaStorageSOPClassUID, '1.1')
self.assertEqual(meta.MediaStorageSOPInstanceUID, '1.2')
self.assertEqual(meta.TransferSyntaxUID, '1.3')
self.assertEqual(meta.ImplementationClassUID, '1.4')
def test_version(self):
"""Test that the value for FileMetaInformationVersion is OK."""
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.fp.seek(16 + 12)
self.assertEqual(self.fp.read(2), b'\x00\x01')
def test_filelike_position(self):
"""Test that the file-like's ending position is OK."""
# 4 bytes prefix
# 8 + 4 bytes FileMetaInformationGroupLength
# 12 + 2 bytes FileMetaInformationVersion
# 8 + 4 bytes MediaStorageSOPClassUID
# 8 + 4 bytes MediaStorageSOPInstanceUID
# 8 + 4 bytes TransferSyntaxUID
# 8 + 4 bytes ImplementationClassUID
# 78 bytes total
meta = Dataset()
meta.MediaStorageSOPClassUID = '1.1'
meta.MediaStorageSOPInstanceUID = '1.2'
meta.TransferSyntaxUID = '1.3'
meta.ImplementationClassUID = '1.4'
_write_file_meta_info(self.fp, meta)
self.assertEqual(self.fp.tell(), 78)
# 8 + 6 bytes ImplementationClassUID
# 80 bytes total, group length 64
self.fp.seek(0)
meta.ImplementationClassUID = '1.4.1'
_write_file_meta_info(self.fp, meta)
# Check File Meta length
self.assertEqual(self.fp.tell(), 80)
# Check Group Length
self.fp.seek(12)
self.assertEqual(self.fp.read(4), b'\x40\x00\x00\x00')
_write_file_meta_info(self.fp, meta)
