def test_invalid_sop_file_meta(self):
"""Test exception raised if SOP Class is not Media Storage Directory"""
ds = dcmread(get_testdata_file('CT_small.dcm'))
with pytest.raises(InvalidDicomError,
match=r"SOP Class is not Media Storage "
r"Directory \(DICOMDIR\)"):
DicomDir("some_name", ds, b'\x00' * 128, ds.file_meta, True, True)
def test_invalid_sop_no_file_meta(self):
"""Test exception raised if invalid sop class but no file_meta"""
ds = dcmread(get_testdata_file('CT_small.dcm'))
with pytest.raises(AttributeError,
match="'DicomDir' object has no attribute "
"'DirectoryRecordSequence'"):
DicomDir("some_name", ds, b'\x00' * 128, None, True, True)
def test_invalid_sop_no_file_meta(self, allow_reading_invalid_values):
"""Test exception raised if invalid sop class but no file_meta"""
ds = dcmread(get_testdata_file('CT_small.dcm'))
with pytest.raises(AttributeError,
match="'DicomDir' object has no attribute "
"'DirectoryRecordSequence'"):
with pytest.warns(UserWarning, match=r"Invalid transfer syntax"):
DicomDir("some_name", ds, b'\x00' * 128, None, True, True)
def read_partial(fileobj, stop_when=None, defer_size=None, force=False):
"""Parse a DICOM file until a condition is met.
Parameters
----------
fileobj : a file-like object
Note that the file will not close when the function returns.
stop_when :
Stop condition. See ``read_dataset`` for more info.
defer_size : int, str, None, optional
See ``read_file`` for parameter info.
force : boolean
See ``read_file`` for parameter info.
Notes
-----
Use ``read_file`` unless you need to stop on some condition other than
reaching pixel data.
Returns
-------
FileDataset instance or DicomDir instance.
See Also
--------
read_file
More generic file reading function.
"""
# Read File Meta Information
# Read preamble (if present)
preamble = read_preamble(fileobj, force)
# Read any File Meta Information group (0002,eeee) elements (if present)
file_meta_dataset = _read_file_meta_info(fileobj)
# Read Dataset
# Read any Command Set group (0000,eeee) elements (if present)
command_set = _read_command_set_elements(fileobj)
# Check to see if there's anything left to read
peek = fileobj.read(1)
fileobj.seek(-1, 1)
# `filobj` should be positioned at the start of the dataset by this point.
# Ensure we have appropriate values for `is_implicit_VR` and
# `is_little_endian` before we try decoding. We assume an initial
# transfer syntax of implicit VR little endian and correct it as necessary
is_implicit_VR = True
is_little_endian = True
transfer_syntax = file_meta_dataset.get("TransferSyntaxUID")
if peek == b'': # EOF
pass
elif transfer_syntax is None: # issue 258
# If no TransferSyntaxUID element then we have to try and figure out
# the correct values for `is_little_endian` and `is_implicit_VR`.
# Peek at the first 6 bytes to get the first element's tag group and
# (possibly) VR
group, _, VR = unpack("<HH2s", fileobj.read(6))
fileobj.seek(-6, 1)
# Test the VR to see if it's valid, and if so then assume explicit VR
from pydicom.values import converters
if not in_py2:
VR = VR.decode(default_encoding)
if VR in converters.keys():
is_implicit_VR = False
# Big endian encoding can only be explicit VR
# Big endian 0x0004 decoded as little endian will be 1024
# Big endian 0x0100 decoded as little endian will be 1
# Therefore works for big endian tag groups up to 0x00FF after
# which it will fail, in which case we leave it as little endian
# and hope for the best (big endian is retired anyway)
if group >= 1024:
is_little_endian = False
elif transfer_syntax == pydicom.uid.ImplicitVRLittleEndian:
pass
elif transfer_syntax == pydicom.uid.ExplicitVRLittleEndian:
is_implicit_VR = False
elif transfer_syntax == pydicom.uid.ExplicitVRBigEndian:
is_implicit_VR = False
is_little_endian = False
elif transfer_syntax == pydicom.uid.DeflatedExplicitVRLittleEndian:
# See PS3.6-2008 A.5 (p 71)
# when written, the entire dataset following
# the file metadata was prepared the normal way,
# then "deflate" compression applied.
# All that is needed here is to decompress and then
# use as normal in a file-like object
zipped = fileobj.read()
# -MAX_WBITS part is from comp.lang.python answer:
# groups.google.com/group/comp.lang.python/msg/e95b3b38a71e6799
unzipped = zlib.decompress(zipped, -zlib.MAX_WBITS)
fileobj = BytesIO(unzipped) # a file-like object
is_implicit_VR = False
else:
# Any other syntax should be Explicit VR Little Endian,
# e.g. all Encapsulated (JPEG etc) are ExplVR-LE
# by Standard PS 3.5-2008 A.4 (p63)
is_implicit_VR = False
# Try and decode the dataset
# By this point we should be at the start of the dataset and have
# the transfer syntax (whether read from the file meta or guessed at)
try:
dataset = read_dataset(fileobj,
is_implicit_VR,
is_little_endian,
stop_when=stop_when,
defer_size=defer_size)
except EOFError:
pass # error already logged in read_dataset
# Add the command set elements to the dataset (if any)
dataset.update(command_set)
class_uid = file_meta_dataset.get("MediaStorageSOPClassUID", None)
if class_uid and class_uid == "Media Storage Directory Storage":
return DicomDir(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)
else:
return FileDataset(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)
def read_partial(
fileobj: BinaryIO,
stop_when: Optional[Callable[[BaseTag, Optional[str], int], bool]] = None,
defer_size: Optional[Union[int, str, float]] = None,
force: bool = False,
specific_tags: Optional[List[BaseTag]] = None
) -> Union[FileDataset, DicomDir]:
"""Parse a DICOM file until a condition is met.
Parameters
----------
fileobj : a file-like object
Note that the file will not close when the function returns.
stop_when :
Stop condition. See :func:`read_dataset` for more info.
defer_size : int, str or float, optional
See :func:`dcmread` for parameter info.
force : bool
See :func:`dcmread` for parameter info.
specific_tags : list or None
See :func:`dcmread` for parameter info.
Notes
-----
Use :func:`dcmread` unless you need to stop on some condition other than
reaching pixel data.
Returns
-------
dataset.FileDataset or dicomdir.DicomDir
The read dataset.
See Also
--------
dcmread
More generic file reading function.
"""
# Read File Meta Information
# Read preamble (if present)
preamble = read_preamble(fileobj, force)
# Read any File Meta Information group (0002,eeee) elements (if present)
file_meta_dataset = _read_file_meta_info(fileobj)
# Read Dataset
# Read any Command Set group (0000,eeee) elements (if present)
command_set = _read_command_set_elements(fileobj)
# Check to see if there's anything left to read
peek = fileobj.read(1)
if peek != b'':
fileobj.seek(-1, 1)
# `filobj` should be positioned at the start of the dataset by this point.
# Ensure we have appropriate values for `is_implicit_VR` and
# `is_little_endian` before we try decoding. We assume an initial
# transfer syntax of implicit VR little endian and correct it as necessary
is_implicit_VR = True
is_little_endian = True
transfer_syntax = file_meta_dataset.get("TransferSyntaxUID")
if peek == b'': # EOF
pass
elif transfer_syntax is None: # issue 258
# If no TransferSyntaxUID element then we have to try and figure out
# the correct values for `is_little_endian` and `is_implicit_VR`.
# Peek at the first 6 bytes to get the first element's tag group and
# (possibly) VR
group, _, VR = unpack("<HH2s", fileobj.read(6))
fileobj.seek(-6, 1)
# Test the VR to see if it's valid, and if so then assume explicit VR
from pydicom.values import converters
VR = VR.decode(default_encoding)
if VR in converters.keys():
is_implicit_VR = False
# Big endian encoding can only be explicit VR
# Big endian 0x0004 decoded as little endian will be 1024
# Big endian 0x0100 decoded as little endian will be 1
# Therefore works for big endian tag groups up to 0x00FF after
# which it will fail, in which case we leave it as little endian
# and hope for the best (big endian is retired anyway)
if group >= 1024:
is_little_endian = False
elif transfer_syntax == pydicom.uid.ImplicitVRLittleEndian:
pass
elif transfer_syntax == pydicom.uid.ExplicitVRLittleEndian:
is_implicit_VR = False
elif transfer_syntax == pydicom.uid.ExplicitVRBigEndian:
is_implicit_VR = False
is_little_endian = False
elif transfer_syntax == pydicom.uid.DeflatedExplicitVRLittleEndian:
# See PS3.5 section A.5
# when written, the entire dataset following
# the file metadata was prepared the normal way,
# then "deflate" compression applied.
# All that is needed here is to decompress and then
# use as normal in a file-like object
zipped = fileobj.read()
# -MAX_WBITS part is from comp.lang.python answer:
# groups.google.com/group/comp.lang.python/msg/e95b3b38a71e6799
unzipped = zlib.decompress(zipped, -zlib.MAX_WBITS)
fileobj = BytesIO(unzipped) # a file-like object
is_implicit_VR = False
else:
# Any other syntax should be Explicit VR Little Endian,
# e.g. all Encapsulated (JPEG etc) are ExplVR-LE
# by Standard PS 3.5-2008 A.4 (p63)
is_implicit_VR = False
# Try and decode the dataset
# By this point we should be at the start of the dataset and have
# the transfer syntax (whether read from the file meta or guessed at)
try:
dataset = read_dataset(
fileobj,
is_implicit_VR,
is_little_endian,
stop_when=stop_when,
defer_size=defer_size,
specific_tags=specific_tags,
)
except EOFError:
if config.enforce_valid_values:
raise
# warning already logged in read_dataset
# Add the command set elements to the dataset (if any)
dataset.update(command_set)
class_uid = cast(
pydicom.uid.UID, file_meta_dataset.get("MediaStorageSOPClassUID", None)
)
ds: Union[DicomDir, FileDataset]
if class_uid and class_uid.name == "Media Storage Directory Storage":
warnings.warn(
"The 'DicomDir' class is deprecated and will be removed in v3.0, "
"after which 'dcmread()' will return a normal 'FileDataset' "
"instance for 'Media Storage Directory' SOP Instances.",
DeprecationWarning
)
ds = DicomDir(
fileobj,
dataset,
preamble,
file_meta_dataset,
is_implicit_VR,
is_little_endian,
)
else:
ds = FileDataset(
fileobj,
dataset,
preamble,
file_meta_dataset,
is_implicit_VR,
is_little_endian,
)
# save the originally read transfer syntax properties in the dataset
ds.set_original_encoding(
is_implicit_VR, is_little_endian, dataset._character_set
)
return ds
def read_partial(fileobj, stop_when=None, defer_size=None, force=False):
"""Parse a DICOM file until a condition is met.
Parameters
----------
fileobj : a file-like object
Note that the file will not close when the function returns.
stop_when :
Stop condition. See ``read_dataset`` for more info.
defer_size : int, str, None, optional
See ``read_file`` for parameter info.
force : boolean
See ``read_file`` for parameter info.
Notes
-----
Use ``read_file`` unless you need to stop on some condition
other than reaching pixel data.
Returns
-------
FileDataset instance or DicomDir instance.
See Also
--------
read_file
More generic file reading function.
"""
# Read preamble -- raise an exception if missing and force=False
preamble = read_preamble(fileobj, force)
file_meta_dataset = Dataset()
# Assume a transfer syntax, correct it as necessary
is_implicit_VR = True
is_little_endian = True
if preamble:
file_meta_dataset = _read_file_meta_info(fileobj)
transfer_syntax = file_meta_dataset.TransferSyntaxUID
if transfer_syntax == pydicom.uid.ImplicitVRLittleEndian:
pass
elif transfer_syntax == pydicom.uid.ExplicitVRLittleEndian:
is_implicit_VR = False
elif transfer_syntax == pydicom.uid.ExplicitVRBigEndian:
is_implicit_VR = False
is_little_endian = False
elif transfer_syntax == pydicom.uid.DeflatedExplicitVRLittleEndian:
# See PS3.6-2008 A.5 (p 71)
# when written, the entire dataset following
# the file metadata was prepared the normal way,
# then "deflate" compression applied.
# All that is needed here is to decompress and then
# use as normal in a file-like object
zipped = fileobj.read()
# -MAX_WBITS part is from comp.lang.python answer:
# groups.google.com/group/comp.lang.python/msg/e95b3b38a71e6799
unzipped = zlib.decompress(zipped, -zlib.MAX_WBITS)
fileobj = BytesIO(unzipped) # a file-like object
is_implicit_VR = False
else:
# Any other syntax should be Explicit VR Little Endian,
# e.g. all Encapsulated (JPEG etc) are ExplVR-LE
# by Standard PS 3.5-2008 A.4 (p63)
is_implicit_VR = False
else: # no header -- use the is_little_endian, implicit assumptions
file_meta_dataset.TransferSyntaxUID = pydicom.uid.ImplicitVRLittleEndian
endian_chr = "<"
element_struct = Struct(endian_chr + "HH2sH")
# Try reading first 8 bytes
group, elem, VR, length = element_struct.unpack(fileobj.read(8))
# Rewind file object
fileobj.seek(0)
# If the VR is a valid VR, assume Explicit VR transfer systax
from pydicom.values import converters
if not in_py2:
VR = VR.decode(default_encoding)
if VR in converters.keys():
is_implicit_VR = False
# Determine if group in low numbered range (Little vs Big Endian)
if group == 0: # got (0,0) group length. Not helpful.
# XX could use similar to http://www.dclunie.com/medical-image-faq/html/part2.html code example
msg = ("Not able to guess transfer syntax when first item "
"is group length")
raise NotImplementedError(msg)
if group < 2000:
file_meta_dataset.TransferSyntaxUID = pydicom.uid.ExplicitVRLittleEndian
else:
file_meta_dataset.TransferSyntaxUID = pydicom.uid.ExplicitVRBigEndian
is_little_endian = False
try:
dataset = read_dataset(fileobj,
is_implicit_VR,
is_little_endian,
stop_when=stop_when,
defer_size=defer_size)
except EOFError:
pass # error already logged in read_dataset
class_uid = file_meta_dataset.get("MediaStorageSOPClassUID", None)
if class_uid and class_uid == "Media Storage Directory Storage":
return DicomDir(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)
else:
return FileDataset(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)
def read_partial(fileobj, stop_when=None, defer_size=None, force=False):
"""Parse a DICOM file until a condition is met.
Parameters
----------
fileobj : a file-like object
Note that the file will not close when the function returns.
stop_when :
Stop condition. See ``read_dataset`` for more info.
defer_size : int, str, None, optional
See ``read_file`` for parameter info.
force : boolean
See ``read_file`` for parameter info.
Notes
-----
Use ``read_file`` unless you need to stop on some condition
other than reaching pixel data.
Returns
-------
FileDataset instance or DicomDir instance.
See Also
--------
read_file
More generic file reading function.
"""
# Read preamble -- raise an exception if missing and force=False
preamble = read_preamble(fileobj, force)
file_meta_dataset = Dataset()
# Assume a transfer syntax, correct it as necessary
is_implicit_VR = True
is_little_endian = True
if preamble:
file_meta_dataset = _read_file_meta_info(fileobj)
transfer_syntax = file_meta_dataset.TransferSyntaxUID
if transfer_syntax == pydicom.uid.ImplicitVRLittleEndian:
pass
elif transfer_syntax == pydicom.uid.ExplicitVRLittleEndian:
is_implicit_VR = False
elif transfer_syntax == pydicom.uid.ExplicitVRBigEndian:
is_implicit_VR = False
is_little_endian = False
elif transfer_syntax == pydicom.uid.DeflatedExplicitVRLittleEndian:
# See PS3.6-2008 A.5 (p 71)
# when written, the entire dataset following
# the file metadata was prepared the normal way,
# then "deflate" compression applied.
# All that is needed here is to decompress and then
# use as normal in a file-like object
zipped = fileobj.read()
# -MAX_WBITS part is from comp.lang.python answer:
# groups.google.com/group/comp.lang.python/msg/e95b3b38a71e6799
unzipped = zlib.decompress(zipped, -zlib.MAX_WBITS)
fileobj = BytesIO(unzipped) # a file-like object
is_implicit_VR = False
else:
# Any other syntax should be Explicit VR Little Endian,
# e.g. all Encapsulated (JPEG etc) are ExplVR-LE
# by Standard PS 3.5-2008 A.4 (p63)
is_implicit_VR = False
else: # no header -- use the is_little_endian, implicit assumptions
file_meta_dataset.TransferSyntaxUID = pydicom.uid.ImplicitVRLittleEndian
try:
dataset = read_dataset(fileobj, is_implicit_VR, is_little_endian,
stop_when=stop_when, defer_size=defer_size)
except EOFError:
pass # error already logged in read_dataset
class_uid = file_meta_dataset.get("MediaStorageSOPClassUID", None)
if class_uid and class_uid == "Media Storage Directory Storage":
return DicomDir(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)
else:
return FileDataset(fileobj, dataset, preamble, file_meta_dataset,
is_implicit_VR, is_little_endian)