autojump

Installs and activates autojump, which learns the directories you visit and lets you return to one by a fragment of its name instead of its path.

Enable it

PLUGINS=(
  https://github.com/jakubro/profile.d-autojump
)

profile.d-install
\. ~/.bashrc

What it does

At install time it clones autojump into a temporary directory and runs its installer with SHELL=bash ./install.py -d ~/.autojump, then deletes the temporary clone. The installer prints setup instructions for a manual install; profile.d has already done that part, and says so.

In every interactive shell it sources ~/.autojump/share/autojump/autojump.bash, which defines j and hooks the directory database into cd. It then runs j --purge, dropping entries for directories that no longer exist.

Both steps are skipped in a non-interactive shell.

Using it

There is nothing to configure. Use cd as you always have, and the database builds itself:

$ cd /var/log/nginx

Later, from anywhere:

$ j nginx
/var/log/nginx

Matching is on a fragment of the path, weighted by how often and how recently you have been there, so the more you use cd the better the guesses get. Other subcommands:

$ j -s            # the database, with each directory's weight
$ j --purge       # drop entries for directories that no longer exist
$ jo nginx        # open the match in a file manager

A fresh install knows nothing and j will not find anything until you have visited a few directories normally.

Requirements

Python, for autojump's installer and for autojump itself. The install script does not check for it and does not install it - if Python is missing, the run fails with the installer's own error.

Known issues

Every update is a fresh install. The install script clones autojump and runs its installer on every profile.d-install, with no check for whether it is already present or unchanged. Nothing breaks, but a run that should have been a no-op clones a repository and re-runs a Python installer.