Defining your command driver¶
It’s possible to run your OptparseInterfaces using the pyqi command, as illustrated in Running our Command via its OptparseInterface, but that mechanism is clunky and not how you’d want your users to interact with your software. To handle this more gracefully, you can create a shell script that can be distributed with your package and used as the primary driver for all OptparseInterfaces.
Creating the driver shell script¶
To define a driver command for your project, create a new file named as you’d like your users to access your code. For example, the driver for the biom-format package is called biom, and the driver for the pyqi package is called pyqi. In this example our driver name will be my-project. Add the following two lines to that file, replacing my-project with your driver name:
#!/bin/sh exec pyqi --driver-name my-project --command-config-module my_project.interfaces.optparse.config -- "$@"
The value passed with --command-config-module must be the directory where the OptparseInterface configuration files can be found. If you followed the suggestions in Organizing your repository the above should work.
The driver script should then be made executable with:
chmod +x my-project
You’ll next need to ensure that the directory containing this driver file is in your PATH environment variable. Again, if you followed the recommendations in Organizing your repository and if your project directory is under $HOME/code, you can do this by running:
You should now be able to run:
This will print a list of the commands that are available via the driver script, which will be all of the Commands for which you’ve defined OptparseInterfaces. If one of these commands is called my-command, you can now run it as follows to get the help text associated with that command:
my-project my-command -h
The command names that you pass to the driver (my-command, in this example) match the name of the OptparseInterface config file, minus the .py. The driver also matches the dashed version of a command name, so my-command and my_command both map to the same command.
Configuring bash completion¶
One very useful feature for your driver script is to enable tab-completion of commands and command line options (meaning that when a user starts typing the name of a command or an option, they can hit the tab key to complete it without typing the full name, if the name is unique). pyqi facilitates this with the pyqi make-bash-completion command. There are two steps in enabling tab completion. First, you’ll need to generate the tab completion file, and then you’ll need to edit your $HOME/.bash_profile file.
To create the tab completion file for my-project, run the following commands (again, this is assuming that your OptparseInterface config files are located as described in Organizing your repository):
mkdir ~/.bash_completion.d pyqi make-bash-completion --command-config-module my_project.interfaces.optparse.config --driver-name my-project -o ~/.bash_completion.d/my-project
Then, add the following lines to your $HOME/.bash_profile file:
# enable bash completion for pyqi-based scripts for f in ~/.bash_completion.d/*; do source $f; done
When you open a new terminal, tab completion should work for the my-project commands and their options.