A ctypes-based adapter to libarchive. The source-code is written to be clear and intuitive.
Even 7-Zip is supported for both reading and writing.
I could definitely use some help, if any is available. Completeness will require a bit more work (see libarchive's archive.h and archive_entry.h).
PyPI:
$ pip install libarchive
- The Ubuntu
libarchivepackage maintainer only provides a "libarchive.so" symlink in the dev package so you'll have to install thelibarchive-devpackage.
$ apt-get install libarchive-dev
-
Encryption is not currently supported since it's not supported in the underlying library (libarchive). Note this inquiry and the wishlist item.
-
OS X has a system version of
libarchivethat is very old. As a result, many users have encountered issues importing an alternate one. Specifically, often they install a different one via Brew but this will not be [sym]linked into the system like other packages. This is a precaution taken by Brew to prevent undefined behavior in the parts of OS X that depend on the factory version. In order to work around this, you should setLD_LIBRARY_PATH(or prepend ifLD_LIBRARY_PATHis already defined) with the path of the location of the library version you want to use. You'll want to set this from your user-profile script (unless your environment can not support this and you need to prepend something like "LD_LIBRARY_PATH=/some/path" to the front of the command-line or set it viaos.environabove where you import this package). A tool has been provided that will print the path of the first version oflibarchiveinstalled via Brew. Just copy-and-paste it. Thanks to @SkyLeach for discussing the issue and treatments.
To extract to the current directory from a physical file (and print each relative filepath):
import libarchive.public
for entry in libarchive.public.file_pour('/tmp/test.zip'):
print(entry)To extract to the current directory from memory:
import libarchive.public
with open('/tmp/test.zip', 'rb') as f:
for entry in libarchive.public.memory_pour(f.read()):
print(entry)To read files from a physical archive:
import libarchive.public
with libarchive.public.file_reader('test.7z') as e:
for entry in e:
with open('/tmp/' + str(entry), 'wb') as f:
for block in entry.get_blocks():
f.write(block)To read files from memory:
import libarchive.public
with open('test.7z', 'rb') as f:
buffer_ = f.read()
with libarchive.public.memory_reader(buffer_) as e:
for entry in e:
with open('/tmp/' + str(entry), 'wb') as f:
for block in entry.get_blocks():
f.write(block)To specify a format and/or filter for reads (rather than detecting it):
import libarchive.public
import libarchive.constants
with open('test.7z', 'rb') as f:
buffer_ = f.read()
with libarchive.public.memory_reader(
buffer_,
format_code=libarchive.constants.ARCHIVE_FORMAT_TAR_USTAR,
filter_code=libarchive.constants.ARCHIVE_FILTER_GZIP
) as e:
for entry in e:
with open('/tmp/' + str(entry), 'wb') as f:
for block in entry.get_blocks():
f.write(block)To read the "filetype" flag for each entry:
import libarchive.public
with open('test.7z', 'rb') as f:
buffer_ = f.read()
with libarchive.public.memory_reader(f.read()) as e:
for entry in e:
print(entry.filetype)The output of this is:
EntryFileType(IFREG=True, IFLNK=True, IFSOCK=True, IFCHR=False, IFBLK=False, IFDIR=False, IFIFO=False)
EntryFileType(IFREG=True, IFLNK=True, IFSOCK=True, IFCHR=False, IFBLK=False, IFDIR=False, IFIFO=False)
EntryFileType(IFREG=True, IFLNK=True, IFSOCK=True, IFCHR=False, IFBLK=False, IFDIR=False, IFIFO=False)To create a physical archive from physical files:
import libarchive.public
import libarchive.constants
libarchive.public.create_file(
'create.7z',
libarchive.constants.ARCHIVE_FORMAT_7ZIP,
['/etc/profile']):The path of the file to add will be recorded verbatim.
To create an archive in memory from physical files:
import libarchive.public
import libarchive.constants
with open('/tmp/new.7z', 'wb') as f:
def writer(buffer_, length):
f.write(buffer_)
return length
libarchive.public.create_generic(
writer,
format_name=libarchive.constants.ARCHIVE_FORMAT_7ZIP,
files=['/etc/profile']):libarchive uses nose/nose2 for testing:
tests$ ./run.py
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent