These scripts are automatically generated from YAML files which can be found under specs/.

The format supports nested subcommands, specifying enums for possible parameters/option values and dynamic completion calling an external command.

The generation is done with the appspec tool (see "Developing" below).

Currently it supports bash and zsh completions.

• cpan-upload - upload a distribution to the CPAN
• cpan - easily interact with CPAN from the command line
• cpanm - Install CPAN modules
• dancer2 - Dancer2 command line interface
• dzil - do stuff with your dist
• fallocate - Preallocate or deallocate space to a file
• fatpack - Pack perl script and modules into standalone script
• file - determine file type
• fzf - a command-line fuzzy finder
• h2xs - convert .h C header files to Perl extensions
• htpasswd - Manage user files for basic authentication
• hypnotoad - Hypnotoad HTTP and WebSocket server
• jq - Command-line JSON processor
• json_pp - JSON::PP command utility
• json_xs - JSON::XS commandline utility
• locate - Find files by name
• lwp-request - Simple command line user agent (Aliases GET, HEAD, POST)
• mojo - The Mojolicious command system
• morbo - Morbo HTTP and WebSocket development server
• mpath - display the full path to a perl module (installed locally)
• pip - A tool for installing and managing Python packages
• plackup - Run PSGI application with Plack handlers
• pod2html - convert .pod files to .html files
• prove - Run tests through a TAP harness
• starman - High-performance preforking PSGI/Plack web server
• tower-cli - Ansible Tower Commandline
• yamllint - A linter for YAML files
• ysh - The YAML Test Shell
• zild - Dist::Zilla Mixed Up

git clone https://github.com/perlpunk/shell-completions

Look into the specs/commandname.yaml for specific usage instructions.

For a specific script, just source it:

$ source /path/to/shell-completions/bash/commandname.bash

Put this in your .bashrc for using all completions:

for i in /path/to/shell-completions/bash/*.bash; do source $i; done

You can enable this by putting the following in your ~/.inputrc:

set show-all-if-ambiguous on

See GNU Readline Init File Syntax.

For activating the completions, you can do:

% fpath=(/path/to/shell-completions/zsh $fpath)
% compinit

To just try out completion for one specific script, you can also do:

% source /path/to/shell-completions/zsh/_commandname
% compdef _commandname commandname

Put this in your .zshrc for activating all completions:

# before the compinit call
fpath=('/path/to/shell-completions/zsh' $fpath)

This way zsh won't load all functions at startup, but the completions will get autoloaded dynamically when the command is first used.

When developing a new completion, after a change it's enough to source it, without calling compinit again:

% source /path/to/shell-completions/zsh/_commandname

The appspec tool is used to generate the completions. You don't need it if you just want to use the completions.

If you want to modify or add a new completion, install the App::AppSpec perl module, or use the standalone script:

% git clone https://github.com/perlpunk/App-AppSpec-p5 -b standalone appspec
% PATH=$PWD/appspec/bin:$PATH

For more details see https://github.com/perlpunk/App-AppSpec-p5.

Create a new specification specs/commandname.yaml and type:

make commandname

• https://github.com/zsh-users/zsh-completions - Over 130 completion scripts for zsh
• https://github.com/unixorn/awesome-zsh-plugins - A collection of ZSH frameworks, plugins & themes
• https://github.com/complete-shell - A different approach for generating and installing completions for bash