def transcode_file(
source_file: PathOrFile,
dest_file: PathOrFile,
source_compression: CompressionArg = True,
dest_compression: CompressionArg = True,
use_system: bool = True,
source_open_args: Optional[dict] = None,
dest_open_args: Optional[dict] = None,
) -> None:
"""Convert from one file format to another.
Args:
source_file: The path or file-like object to read from. If a file, it
must be opened in mode 'rb'.
dest_file: The path or file-like object to write to. If a file, it
must be opened in binary mode.
source_compression: The compression type of the source file. If True,
guess compression format from the file name, otherwise the name of
any supported compression format.
dest_compression: The compression type of the dest file. If True,
guess compression format from the file name, otherwise the name of
any supported compression format.
use_system: Whether to use system-level compression.
source_open_args: Additional arguments to pass to xopen for the source
file.
dest_open_args: Additional arguments to pass to xopen for the
destination file.
"""
src_args = copy.copy(source_open_args) if source_open_args else {}
if "mode" not in src_args:
src_args["mode"] = "rb"
dst_args = copy.copy(dest_open_args) if dest_open_args else {}
if "mode" not in dst_args:
dst_args["mode"] = "wb"
with open_(source_file,
compression=source_compression,
use_system=use_system,
**src_args) as src, open_(dest_file,
compression=dest_compression,
use_system=use_system,
**dst_args) as dst:
for chunk in iter_file_chunked(src):
dst.write(chunk)
def read_bytes(path_or_file: PathOrFile,
chunksize: int = 1024,
**kwargs) -> Generator[bytes, None, None]:
"""Iterate over a file in chunks. The mode will always be overridden
to 'rb'.
Args:
path_or_file: Path to the file, or a file-like object.
chunksize: Number of bytes to read at a time.
kwargs: Additional arguments to pass top :method:`xphyle.open_`.
Yields:
Chunks of the input file as bytes. Each chunk except the last should
be of size `chunksize`.
"""
kwargs["mode"] = "rb"
with open_(path_or_file, **kwargs) as fileobj:
if fileobj is None:
return
yield from iter_file_chunked(fileobj, chunksize)
def decompress_file(self,
source: PathOrFile,
dest: PathOrFile = None,
keep: bool = True,
use_system: bool = True,
**kwargs) -> str:
"""Decompress data from one file and write to another.
Args:
source: Source file, either a path or an open file-like object.
dest: Destination file, either a path or an open file-like object.
If None, the file name is determined from ``source``.
keep: Whether to keep the source file
use_system: Whether to try to use system-level compression
kwargs: Additional arguments to passs to the open method when
opening the compressed file
Returns:
Path to the destination file
"""
source_is_path = isinstance(source, str)
if source_is_path:
source_path = str(source)
check_readable_file(source_path)
else:
source_path = cast(FileLike, source).name
source_parts = split_path(source_path)
if dest is None:
if len(source_parts) > 2:
dest = (os.path.join(*source_parts[0:2]) +
''.join(source_parts[2:-1]))
dest_is_path = True
else:
raise Exception("Cannot determine path for decompressed file")
else:
dest_is_path = isinstance(dest, str)
if dest_is_path:
dest_name = str(check_writable_file(str(dest)))
dest_file = open(dest_name, 'wb')
else:
dest_file = cast(FileLike, dest)
dest_name = dest_file.name
# pragma: no-cover
try:
dest_file.fileno()
except OSError:
use_system = False
try:
if use_system and self.can_use_system_decompression:
src = str(source) if source_is_path else STDIN
cmd = self.get_command('d', src=src)
psrc = None if source_is_path else cast(FileLike, source)
proc = PROCESS_PROGRESS.wrap(cmd, stdin=psrc, stdout=dest_file)
proc.communicate()
else:
source_file = self.open_file_python(source, 'rb', **kwargs)
try:
# Perform sequential decompression as the source
# file might be quite large
for chunk in iter_file_chunked(source_file):
dest_file.write(chunk)
finally:
if source_is_path:
source_file.close()
if not keep:
if not source_is_path:
cast(FileLike, source).close()
os.remove(source_path)
finally:
if dest_is_path:
dest_file.close()
return dest_name
def compress_file(self,
source: PathOrFile,
dest: PathOrFile = None,
keep: bool = True,
compresslevel: int = None,
use_system: bool = True,
**kwargs) -> str:
"""Compress data from one file and write to another.
Args:
source: Source file, either a path or an open file-like object.
dest: Destination file, either a path or an open file-like object.
If None, the file name is determined from ``source``.
keep: Whether to keep the source file
compresslevel: Compression level
use_system: Whether to try to use system-level compression
kwargs: Additional arguments to pass to the open method when opening
the destination file
Returns:
Path to the destination file.
"""
source_is_path = isinstance(source, str)
if source_is_path:
source_path = str(source)
check_readable_file(source_path)
else:
source_io = cast(IO, source)
source_path = source_io.name
# pragma: no-cover
try:
source_io.fileno()
except OSError:
use_system = False
if dest is None:
dest = "{}.{}".format(source_path, self.default_ext)
dest_is_path = True
else:
dest_is_path = isinstance(dest, str)
if dest_is_path:
check_writable_file(str(dest))
try:
if use_system and self.can_use_system_compression:
if source_is_path:
cmd_src = str(source)
prc_src = None
else:
cmd_src = STDIN
prc_src = cast(FileLike, source)
if dest_is_path:
dest_name = str(dest)
dest_file = open(dest_name, 'wb')
else:
dest_file = cast(FileLike, dest)
dest_name = dest_file.name
cmd = self.get_command('c',
src=cmd_src,
compresslevel=compresslevel)
proc = PROCESS_PROGRESS.wrap(cmd,
stdin=prc_src,
stdout=dest_file)
proc.communicate()
else:
if source_is_path:
source_file = open(str(source), 'rb')
else:
source_file = cast(FileLike, source)
dest_name = str(dest)
dest_file = self.open_file_python(dest, 'wb', **kwargs)
try:
# Perform sequential compression as the source
# file might be quite large
for chunk in iter_file_chunked(source_file):
dest_file.write(chunk)
finally:
if source_is_path:
source_file.close()
if not keep:
if not source_is_path:
cast(FileLike, source).close()
os.remove(source_path)
finally:
if dest_is_path:
dest_file.close()
return dest_name