def _ReadFileHeader(self, file_object):
"""Reads the file header.
Args:
file_object: the file-like object to read from.
Raises:
FileFormatError: if file format related errors are detected.
"""
file_object.seek(0, os.SEEK_SET)
file_header = self._FILE_HEADER_STRUCT.parse_stream(file_object)
self._compressed_data_offset = file_object.get_offset()
if file_header.signature != self._FILE_SIGNATURE:
raise errors.FileFormatError(
u'Unsuppored file signature: 0x{0:04x}.'.format(
file_header.signature))
if file_header.compression_method != self._COMPRESSION_METHOD_DEFLATE:
raise errors.FileFormatError(
u'Unsuppored compression method: {0:d}.'.format(
file_header.compression_method))
self.modification_time = file_header.modification_time
self.operating_system = file_header.operating_system
if file_header.flags & self._FLAG_FEXTRA:
extra_field_data_size = construct.ULInt16(
u'extra_field_data_size').parse_stream(file_object)
file_object.seek(extra_field_data_size, os.SEEK_CUR)
self._compressed_data_offset += 2 + extra_field_data_size
if file_header.flags & self._FLAG_FNAME:
# Since encoding is set construct will convert the C string to Unicode.
# Note that construct 2 does not support the encoding to be a Unicode
# string.
self.original_filename = construct.CString(
u'original_filename', encoding='iso-8859-1').parse_stream(
file_object)
self._compressed_data_offset = file_object.get_offset()
if file_header.flags & self._FLAG_FCOMMENT:
# Since encoding is set construct will convert the C string to Unicode.
# Note that construct 2 does not support the encoding to be a Unicode
# string.
self.comment = construct.CString(
u'comment', encoding='iso-8859-1').parse_stream(file_object)
self._compressed_data_offset = file_object.get_offset()
if file_header.flags & self._FLAG_FHCRC:
self._compressed_data_offset += 2
self._compressed_data_size = (
file_object.get_size() - (self._compressed_data_offset + 8))
def _ReadMemberHeader(self, file_object):
"""Reads a member header.
Args:
file_object (FileIO): file-like object to read from.
Raises:
FileFormatError: if the member header cannot be read.
"""
file_offset = file_object.get_offset()
member_header = self._ReadStructure(file_object, file_offset,
self._MEMBER_HEADER_SIZE,
self._MEMBER_HEADER,
'member header')
if member_header.signature != self._GZIP_SIGNATURE:
raise errors.FileFormatError(
'Unsupported signature: 0x{0:04x}.'.format(
member_header.signature))
if member_header.compression_method != self._COMPRESSION_METHOD_DEFLATE:
raise errors.FileFormatError(
'Unsupported compression method: {0:d}.'.format(
member_header.compression_method))
self.modification_time = member_header.modification_time
self.operating_system = member_header.operating_system
if member_header.flags & self._FLAG_FEXTRA:
file_offset = file_object.get_offset()
extra_field_data_size = self._ReadStructure(
file_object, file_offset, self._UINT16LE_SIZE, self._UINT16LE,
'extra field data size')
file_object.seek(extra_field_data_size, os.SEEK_CUR)
if member_header.flags & self._FLAG_FNAME:
file_offset = file_object.get_offset()
string_value = self._ReadString(file_object, file_offset,
self._CSTRING, 'original filename')
self.original_filename = string_value.rstrip('\x00')
if member_header.flags & self._FLAG_FCOMMENT:
file_offset = file_object.get_offset()
string_value = self._ReadString(file_object, file_offset,
self._CSTRING, 'comment')
self.comment = string_value.rstrip('\x00')
if member_header.flags & self._FLAG_FHCRC:
file_object.read(2)
def _ReadStructureFromByteStream(
self, byte_stream, file_offset, data_type_map, description, context=None):
"""Reads a structure from a byte stream.
Args:
byte_stream (bytes): byte stream.
file_offset (int): offset of the data relative from the start of
the file-like object.
data_type_map (dtfabric.DataTypeMap): data type map of the structure.
description (str): description of the structure.
context (Optional[dtfabric.DataTypeMapContext]): data type map context.
Returns:
object: structure values object.
Raises:
FileFormatError: if the structure cannot be read.
ValueError: if file-like object or date type map are invalid.
"""
if not byte_stream:
raise ValueError('Invalid byte stream.')
if not data_type_map:
raise ValueError('Invalid data type map.')
try:
return data_type_map.MapByteStream(byte_stream, context=context)
except dtfabric_errors.MappingError as exception:
raise errors.FileFormatError((
'Unable to map {0:s} data at offset: 0x{1:08x} with error: '
'{2!s}').format(description, file_offset, exception))
def _ReadAndParseHeader(self, file_object):
"""Reads the member header and sets relevant member values.
Args:
file_object (FileIO): file-like object to read from.
Raises:
FileFormatError: if file format related errors are detected.
"""
member_header = self._MEMBER_HEADER_STRUCT.parse_stream(file_object)
if member_header.signature != self._GZIP_SIGNATURE:
raise errors.FileFormatError(
'Unsupported file signature: 0x{0:04x}.'.format(
member_header.signature))
if member_header.compression_method != self._COMPRESSION_METHOD_DEFLATE:
raise errors.FileFormatError(
'Unsupported compression method: {0:d}.'.format(
member_header.compression_method))
self.modification_time = member_header.modification_time
self.operating_system = member_header.operating_system
if member_header.flags & self._FLAG_FEXTRA:
extra_field_data_size = construct.ULInt16(
'extra_field_data_size').parse_stream(file_object)
file_object.seek(extra_field_data_size, os.SEEK_CUR)
if member_header.flags & self._FLAG_FNAME:
# Since encoding is set construct will convert the C string to Unicode.
# Note that construct 2 does not support the encoding to be a Unicode
# string.
self.original_filename = construct.CString(
'original_filename',
encoding=b'iso-8859-1').parse_stream(file_object)
if member_header.flags & self._FLAG_FCOMMENT:
# Since encoding is set construct will convert the C string to Unicode.
# Note that construct 2 does not support the encoding to be a Unicode
# string.
self.comment = construct.CString(
'comment', encoding=b'iso-8859-1').parse_stream(file_object)
if member_header.flags & self._FLAG_FHCRC:
file_object.read(2)
def _ReadData(self, file_object, file_offset, data_size, description):
"""Reads data.
Args:
file_object (FileIO): file-like object.
file_offset (int): offset of the data relative from the start of
the file-like object.
data_size (int): size of the data.
description (str): description of the data.
Returns:
bytes: byte stream containing the data.
Raises:
FileFormatError: if the structure cannot be read.
ValueError: if file-like object or date type map are invalid.
"""
if not file_object:
raise ValueError('Invalid file-like object.')
file_object.seek(file_offset, os.SEEK_SET)
read_error = ''
try:
data = file_object.read(data_size)
if len(data) != data_size:
read_error = 'missing data'
except IOError as exception:
read_error = '{0!s}'.format(exception)
if read_error:
raise errors.FileFormatError((
'Unable to read {0:s} data at offset: 0x{1:08x} with error: '
'{2:s}').format(description, file_offset, read_error))
return data
def _ReadFileEntry(self, file_object, file_offset):
"""Reads a file entry.
Args:
file_object (FileIO): file-like object.
file_offset (int): offset of the data relative from the start of
the file-like object.
Returns:
CPIOArchiveFileEntry: a file entry.
Raises:
FileFormatError: if the file entry cannot be read.
"""
if self.file_format == 'bin-big-endian':
data_type_map = self._CPIO_BINARY_BIG_ENDIAN_FILE_ENTRY
file_entry_data_size = self._CPIO_BINARY_BIG_ENDIAN_FILE_ENTRY_SIZE
elif self.file_format == 'bin-little-endian':
data_type_map = self._CPIO_BINARY_LITTLE_ENDIAN_FILE_ENTRY
file_entry_data_size = self._CPIO_BINARY_LITTLE_ENDIAN_FILE_ENTRY_SIZE
elif self.file_format == 'odc':
data_type_map = self._CPIO_PORTABLE_ASCII_FILE_ENTRY
file_entry_data_size = self._CPIO_PORTABLE_ASCII_FILE_ENTRY_SIZE
elif self.file_format in ('crc', 'newc'):
data_type_map = self._CPIO_NEW_ASCII_FILE_ENTRY
file_entry_data_size = self._CPIO_NEW_ASCII_FILE_ENTRY_SIZE
file_entry = self._ReadStructure(file_object, file_offset,
file_entry_data_size, data_type_map,
'file entry')
file_offset += file_entry_data_size
if self.file_format in ('bin-big-endian', 'bin-little-endian'):
file_entry.modification_time = (
(file_entry.modification_time.upper << 16)
| file_entry.modification_time.lower)
file_entry.file_size = ((file_entry.file_size.upper << 16)
| file_entry.file_size.lower)
if self.file_format == 'odc':
for attribute_name in self._CPIO_ATTRIBUTE_NAMES_ODC:
value = getattr(file_entry, attribute_name, None)
try:
value = int(value, 8)
except ValueError:
raise errors.FileFormatError(
'Unable to convert attribute: {0:s} into an integer'.
format(attribute_name))
value = setattr(file_entry, attribute_name, value)
elif self.file_format in ('crc', 'newc'):
for attribute_name in self._CPIO_ATTRIBUTE_NAMES_CRC:
value = getattr(file_entry, attribute_name, None)
try:
value = int(value, 16)
except ValueError:
raise errors.FileFormatError(
'Unable to convert attribute: {0:s} into an integer'.
format(attribute_name))
value = setattr(file_entry, attribute_name, value)
path_data = file_object.read(file_entry.path_size)
file_offset += file_entry.path_size
path = path_data.decode(self._encoding)
path, _, _ = path.partition('\x00')
if self.file_format in ('bin-big-endian', 'bin-little-endian'):
padding_size = file_offset % 2
if padding_size > 0:
padding_size = 2 - padding_size
elif self.file_format == 'odc':
padding_size = 0
elif self.file_format in ('crc', 'newc'):
padding_size = file_offset % 4
if padding_size > 0:
padding_size = 4 - padding_size
file_offset += padding_size
archive_file_entry = CPIOArchiveFileEntry()
archive_file_entry.data_offset = file_offset
archive_file_entry.data_size = file_entry.file_size
archive_file_entry.group_identifier = file_entry.group_identifier
archive_file_entry.inode_number = file_entry.inode_number
archive_file_entry.modification_time = file_entry.modification_time
archive_file_entry.path = path
archive_file_entry.mode = file_entry.mode
archive_file_entry.size = (file_entry_data_size +
file_entry.path_size + padding_size +
file_entry.file_size)
archive_file_entry.user_identifier = file_entry.user_identifier
file_offset += file_entry.file_size
if self.file_format in ('bin-big-endian', 'bin-little-endian'):
padding_size = file_offset % 2
if padding_size > 0:
padding_size = 2 - padding_size
elif self.file_format == 'odc':
padding_size = 0
elif self.file_format in ('crc', 'newc'):
padding_size = file_offset % 4
if padding_size > 0:
padding_size = 4 - padding_size
if padding_size > 0:
archive_file_entry.size += padding_size
return archive_file_entry