Clitic aims to be a collection of templates for Unix command line tools written in different programming languages. A template should be a starting point to write program that behaves like one would expect a Unix command line tool to behave.

It also provides a test suite to test just that behaviour.

Currently, templates for the following languages (and argument parsing suites) are available:

• Python (using argparse)
• Python (using docopt)
• Rust (using docopt)

Roughly spoken, the tests check if the templates behave like this:

• run without arguments and output Hello World
• run with one or several positional arguments and output them stdout each on a new line
• run with one or several positional arguments taken from stdin (using the traditional "-" switch or without using it) and output each on a new line
• run with the --version or -V argument and output foo 0.1.0
• run with the --help or -h argument and begin output with usage or Usage
• send output to stdout without causing a broken pipe

See clitic.bats for details about the tests.

To run all the tests: make test-all

To test a particular template, e.g. the Rust template: make rust

Refer to the makefile to see the valid targets for the makefile.

I'd be happy about useful templates that pass the tests, as well as refinements to the tests themselves. I'm no Unix wizard, there are probably a lot of sensible tests I haven't thought of.

• bash >= 0.4.0
• bats = 0.4.0
• make

• argparse
• docopt

• cargo

In linguistics, a Clitic is something that can't stand on its own, but takes it's meaning from the rest. Like the "'em" in "Beat'em up!". The same is true for these templates. They'll only mean something if someone uses them to actually writes cli tools.