Recursively walk into directories and archives
LGPL-3.0 License
This module primarily provides the function unzipwalk()
, which recursively walks
into directories and compressed files and returns all files, directories, etc. found,
together with binary file handles (file objects) for reading the files.
Currently supported are ZIP, tar, tgz, and gz compressed files.
File types are detected based on their extensions.
>>> from unzipwalk import unzipwalk
>>> results = []
>>> for result in unzipwalk('.'):
... names = tuple( name.as_posix() for name in result.names )
... if result.hnd: # result is a file opened for reading (binary)
... # could use result.hnd.read() here, or for line-by-line:
... for line in result.hnd:
... pass # do something interesting with the data here
... results.append(names + (result.typ.name,))
>>> print(sorted(results))
[('bar.zip', 'ARCHIVE'),
('bar.zip', 'bar.txt', 'FILE'),
('bar.zip', 'test.tar.gz', 'ARCHIVE'),
('bar.zip', 'test.tar.gz', 'hello.csv', 'FILE'),
('bar.zip', 'test.tar.gz', 'test', 'DIR'),
('bar.zip', 'test.tar.gz', 'test/cool.txt.gz', 'ARCHIVE'),
('bar.zip', 'test.tar.gz', 'test/cool.txt.gz', 'test/cool.txt', 'FILE'),
('foo.txt', 'FILE')]
Note that unzipwalk()
automatically closes files as it goes from file to file.
This means that you must use the handles as soon as you get them from the generator -
something as seemingly simple as sorted(unzipwalk('.'))
would cause the code above to fail,
because all files will have been opened and closed during the call to sorted()
and the handles to read the data would no longer be available in the body of the loop.
This is why the above example first processes all the files before sorting the results.
You can also use recursive_open()
to open the files later.
The yielded file handles can be wrapped in io.TextIOWrapper
to read them as text files.
For example, to read all CSV files in the current directory and below, including within compressed files:
>>> from unzipwalk import unzipwalk, FileType
>>> from io import TextIOWrapper
>>> import csv
>>> for result in unzipwalk('.'):
... if result.typ==FileType.FILE and result.names[-1].suffix.lower()=='.csv':
... print([ name.as_posix() for name in result.names ])
... with TextIOWrapper(result.hnd, encoding='UTF-8', newline='') as handle:
... csv_rd = csv.reader(handle, strict=True)
... for row in csv_rd:
... print(repr(row))
['bar.zip', 'test.tar.gz', 'hello.csv']
['Id', 'Name', 'Address']
['42', 'Hello', 'World']
The original name of a gzip-compressed file is derived from the compressed file’s name
by simply removing the .gz
extension. Using the original filename from the gzip
file’s header is currently not possible due to
limitations in the underlying library.
This generator recursively walks into directories and compressed files and yields named tuples of type UnzipWalkResult
.
UnzipWalkResult
of type FileType.SKIP
is yielded.UnzipWalkResult
of type FileType.ERROR
is yielded for those files instead.gzip.BadGzipFile
errors are not raised until the file is actually read,Do not rely on the order of results! But see also the discussion in the main documentation about why
e.g. sorted(unzipwalk(...))
automatically closes files and so may not be what you want.
Return type for unzipwalk()
.
A tuple of the filename(s) as pathlib
objects. The first element is always the physical file in the file system.
If the tuple has more than one element, then the yielded file is contained in a compressed file, possibly nested in
other compressed file(s), and the last element of the tuple will contain the file’s actual name.
A FileType
value representing the type of the current file.
When typ
is FileType.FILE
, this is a ReadOnlyBinary
file handle (file object)
for reading the file contents in binary mode. Otherwise, this is None
.
If this object was produced by from_checksum_line()
, this handle will read the checksum of the data, not the data itself!
Validate whether the object’s fields are set properly and throw errors if not.
Intended for internal use, mainly when type checkers are not being used.
unzipwalk()
validates all the results it returns.
Encodes this object into a line of text suitable for use as a checksum line.
Intended mostly for internal use by the --checksum
CLI option.
See from_checksum_line()
for the inverse operation.
Requires that the file handle be open (for files), and will read from it to generate the checksum!
hashlib.new()
.Decodes a checksum line as produced by checksum_line()
.
Intended as a utility function for use when reading files produced by the --checksum
CLI option.
The hnd
of the returned object will not be a handle to
the data from the file, instead it will be a handle to read the checksum of the file!
(You could use recursive_open()
to open the files themselves.)
True
if the pathname in the line is in Windows format,UnzipWalkResult
object, or None
for empty or comment lines.Used in UnzipWalkResult
to indicate the type of the file.
Don’t rely on the numeric value of the enum elements, they are automatically generated and may change!
A regular file.
An archive file, will be descended into.
A directory.
A symbolic link.
Some other file type (e.g. FIFO).
A file was skipped due to the matcher
filter.
An error was encountered with this file, when the raise_errors
option is off.
This context manager allows opening files nested inside archives directly.
unzipwalk()
automatically closes files as it iterates through directories and archives;
this function exists to allow you to open the returned files after the iteration.
If any of encoding
, errors
, or newline
is specified, the returned
file is wrapped in io.TextIOWrapper
!
If the last file in the list of files is an archive file, then it won’t be decompressed, instead you’ll be able to read the archive’s raw compressed data from the handle.
In this example, we open a gzip-compressed file, stored inside a tgz archive, which in turn is stored in a Zip file:
>>> from unzipwalk import recursive_open
>>> with recursive_open(('bar.zip', 'test.tar.gz', 'test/cool.txt.gz', 'test/cool.txt'), encoding='UTF-8') as fh:
... print(fh.read())
Hi, I'm a compressed file!
Interface for the file handle (file object) used in UnzipWalkResult
.
The interface is the intersection of typing.BinaryIO
, gzip.GzipFile
, and zipfile.ZipExtFile
.
Because gzip.GzipFile
doesn’t implement .tell()
, that method isn’t available here.
Whether the handle supports seeking depends on the underlying library.
Note unzipwalk()
automatically closes files.
usage: unzipwalk [-h] [-a] [-d | -c ALGO] [-e EXCLUDE] [-r] [-o OUTFILE]
[PATH ...]
Recursively walk into directories and archives
positional arguments:
PATH paths to process (default is current directory)
optional arguments:
-h, --help show this help message and exit
-a, --all-files also list dirs, symlinks, etc.
-d, --dump also dump file contents
-c ALGO, --checksum ALGO
generate a checksum for each file**
-e EXCLUDE, --exclude EXCLUDE
filename globs to exclude*
-r, --raise-errors raise errors instead of reporting them in output
-o OUTFILE, --outfile OUTFILE
output filename
* Note --exclude currently only matches against the final name in the
sequence, excluding path names, but this interface may change in future
versions. For more control, use the library instead of this command-line tool.
** Possible values for ALGO: blake2b, blake2s, md4, md5, md5-sha1, ripemd160,
sha1, sha224, sha256, sha384, sha3_224, sha3_256, sha3_384, sha3_512, sha512,
sha512_224, sha512_256, shake_128, shake_256, sm3, whirlpool
The available checksum algorithms may vary depending on your system and Python version.
Run the command with --help
to see the list of currently available algorithms.
Copyright (c) 2022-2024 Hauke Dämpfling ([email protected]) at the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany, https://www.igb-berlin.de/
This library is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public License along with this program. If not, see https://www.gnu.org/licenses/