Coding Style and Rubocop

In order to have similar style guide accross YaST modules rubocop is used for ruby sources. Not all modules using it, but purpose of this documentaiton is to provide guidenance how to setup rubocop for new or existing project to increase its usage.

How to Enable Rubocop

Initial Configuration

Create the initial .rubocop.yml file which inherits the common YaST style:

# use the shared Yast defaults
inherit_from:
  /usr/share/YaST2/data/devtools/data/rubocop_yast_style.yml

First Rubocop Run

Run rubocop and see how many issues Rubocop finds. If there are few issues (few dozens ;-)) you can run rubocop -a to autofix the issues, then run rubocop again and fix the remaining issues manually.

For easier code review please separate the automatic changes done by Rubocop and the manual changes. The automatic changes are usually just white space or indentation fixes where Rubocop is good at fixing them so this commit can be skipped in review.

Manual changes are usually non-trivial and there is a higher risk of regression or potentially changing the behavior so these should be reviewed more carefully. The reviewer could easily overlook them in the huge amount of spacing changes so commit them separately.

Disabling Checks

For the old code you will very likely see several hundreds, maybe even thousands issues. In this case it’s better to split the work into smaller parts and maybe even disable some checks which would need too much work (long methods/classes) or break the API (variable or method names).

Run rubocop --auto-gen-config, this will generate .rubocop_todo.yml file which disables all checks which failed. Now you can append that file to the default config:

  cat .rubocop_todo.yml >> .rubocop.yml

If you run rubocop now it should pass.

Now you should remove the disabled checks one by one, but it’s better first to just comment it out and see the result, if you decide to keep it disabled then you can just uncomment it back.

You should start with the checks which Rubocop is able to fix automatically, this will avoid updating your manual changes later. Such checks are marked with Cop supports --auto-correct text. Just enable the check and run rubocop -a.

Then manually fix the remaning issues.

It is recommended to commit each fix separately, see e.g. https://github.com/yast/yast-yast2/pull/478/commits, mark the the manual fixes in commit message (“Manually fixed” or something like that).

Commit Your Hard Work

Do not forget to commit also the new .rubocop.yml ;-) At Travis the script runs Rubocop automatically if .rubocop.yml is found, you do not need to do anything to enable it there.