dsoprea
PyEasyArchive
Python

A very intuitive and useful adapter to libarchive for universal archive access.

Last updated Jan 10, 2026
98
Stars
34
Forks
23
Issues
0
Stars/day
Attention Score
17
Language breakdown
Python 99.9%
Shell 0.1%
โ–ธ Files click to expand
README

Build Status Coverage Status

Introduction

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).

Installation

PyPI:

$ pip install libarchive

Notes

  • The Ubuntu libarchive package maintainer only provides a "libarchive.so" symlink in the dev package so you'll have to install the libarchive-dev package.
$ 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 libarchive that 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 set LDLIBRARYPATH (or prepend if LDLIBRARYPATH is 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 "LDLIBRARYPATH=/some/path" to the front of the command-line or set it via os.environ above where you import this package). A tool has been provided that will print the path of the first version of libarchive installed via Brew. Just copy-and-paste it. Thanks to @SkyLeach for discussing the issue and treatments.

Examples

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.memoryreader(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_, formatcode=libarchive.constants.ARCHIVEFORMATTARUSTAR, filtercode=libarchive.constants.ARCHIVEFILTER_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.ARCHIVEFORMAT7ZIP, ['/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, formatname=libarchive.constants.ARCHIVEFORMAT_7ZIP, files=['/etc/profile']):

Testing

libarchive uses nose/nose2 for testing:

tests$ ./run.py

ยฉ 2026 GitRepoTrend ยท dsoprea/PyEasyArchive ยท Updated daily from GitHub