Autojump is a faster way to navigate your filesystem. It extends the idea of cd command but enables you to jump to any directory from anywhere.

Basic Usage

Jump To A Directory That Contains "foo". For example: j foo.

Jump To A Child Directory (sub-directory of the current directory). For example, j bar.

Open a file explorer window instead of jumping to the directory. For example, jo music.

Jump to A Directory using multiple arguments. For example, j gh ez jumps to /Users/soasme/github/enqueuezero, while j n ez jumps to /Users/soasme/notes/enqueuezero.


It's very easy to install autojump depending on your platform.

# Debian, Ubuntu
$ apt-get install autojump

# CentOS, RedHat, Fedora
$ yum install autojump

# macOS
$ brew install autojump

Read the source code

The entry point of the app is a shell script that executes a Python script, autojump and then cd to the printed path.

Autojump is as a mixed shell and Python app without library dependencies, instead of a standard Python module. As a result, all utility modules are in an absolute-importing style.

import os, sys # and other libraries from Python standard libraries.

from autojump_argparse import ...
from autojump_data import ...
from autojump_match import ...
from autojump_utils import ...

The main function reads from arguments parsing from the command-line. Below is a simplified Python code of how autojump works.

def parse_arguments():
    parser = ArgumentParser()
    parser.add_argument('directory', nargs='*', default='', help='directory to jump to')
    # ...
    return parser.parse_args()

def main(args):
    haystack = load(config)
    needles =
    print(find_matches(needles, haystack))

if __name__ == '__main__':

The matching algorithms are implemented in module. All matching functions accept parameters needles and haystack, former of which is the keyword to match and latter of which includes the weighted paths to be matched. For example, Below variables show how the fuzzy searching generates the result for needles in the haystack.

needles = ['foo', 'bar']
haystack = [
    (path='/foo/bar/baz', weight=11),
    (path='/foo/baz/moo', weight=10),
    (path='/moo/foo/baz', weight=10),
    (path='/foo/baz', weight=10),
    (path='/foo/bar', weight=10),
result = [
    (path='/foo/bar/baz', weight=11),
    (path='/moo/foo/baz', weight=10),
    (path='/foo/baz', weight=10),
    (path='/foo/bar', weight=10),


The autojump is a useful utility for switching directory faster. It's based on a traditional finding-needles-in-haystack problem but is far beyond that, which makes it a gorgeous software.