respond_to? should be instance method

[Yuki-Inoue]

No description provided.

[benlangfeld]

Could you explain why, and perhaps propose a test to demonstrate how this is broken?

[Yuki-Inoue]

Currently, it does not break anything, as far as I can tell, if the whenever is to be invoked from commandline.

However, in ruby it is always better to have consistent response between method call and the respond_to? method. Some blog posts explaining this are: http://blog.enriquez.me/2010/2/21/dont-forget-about-respond-to-when-implementing-method-missing/

Actually, I currently have one motivation to fix this; I want to query job_list whether the method identifier is available against this class.

Consider following snippet:

bundle exec ruby -I lib -r whenever -r whenever/job_list -e "job_list = Whenever::JobList.new('set :identifier, %(foo)'); puts job_list.respond_to? :identifier; puts job_list.identifier"

When executed at top level of this repo, before this fix, it prints:

false
foo

With this fix, it prints:

true
foo

Without this fix, I'd have to actually call JobList#identifier in order to determine whether identifier exist, and rescue NoMethodError (or, created a method called identifier_defined?) for those cases when :identifier is not set in the schedule.rb. Generally, it is not good to use exceptions unless they really are the exceptional cases.

With this fix, I can determine by just calling job_list.respond_to? :identifier.
(Which is actually done on here on the PR #758)

P.S. And while I was looking for blogs explaining how respond_to? work, I found we should use respond_to_missing? instead of respond_to?, as stated in https://stackoverflow.com/questions/13793060/respond-to-vs-respond-to-missing

So, I've fixed it and pushed it to this PR.

[benlangfeld]

Hi @Yuki-Inoue . I'd still like to see a test for this. With that and a changelog entry, I'll get this merged right away.

[KjellMorgenstern]

irb actually has a problem here.

Start the irb:

Running via Spring preloader in process 141262
Loading development environment (Rails 7.0.4)
irb(main):001:0* helper.

The moment I type the dot in helper. , irb exits with a backtrace:

./home/user/.rbenv/versions/3.1.2/lib/ruby/gems/3.1.0/gems/whenever-1.0.0/lib/whenever/job_list.rb:41:in `respond_to?': undefined method `has_key?' for nil:NilClass (NoMethodError)
                           
      @set_variables.has_key?(name) || super
                    ^^^^^^^^^
        from /home/user/.rbenv/versions/3.1.2/lib/ruby/gems/3.1.0/gems/irb-1.5.0/lib/irb/completion.rb:364:in `block in retrieve_completion_data'
	from /home/user/.rbenv/versions/3.1.2/lib/ruby/gems/3.1.0/gems/irb-1.5.0/lib/irb/completion.rb:362:in `each_object'
	from /home/user/.rbenv/versions/3.1.2/lib/ruby/gems/3.1.0/gems/irb-1.5.0/lib/irb/completion.rb:362:in `retrieve_completion_data'
	from /home/user/.rbenv/versions/3.1.2/lib/ruby/gems/3.1.0/gems/irb-1.5.0/lib/irb/completion.rb:161:in `block in <module:InputCompletor>'

[bragamat]

@Yuki-Inoue can you provide some update on this or @benlangfeld can we close this one out ? 😄

[Yuki-Inoue]

Sorry, recently, I don't have much spare time.
I guess this PR can be closed, until someone (or future me, maybe) comes up with test.

[bragamat]

Sorry, recently, I don't have much spare time. I guess this PR can be closed, until someone (or future me, maybe) comes up with test.

Thank you for the update @Yuki-Inoue!

@benlangfeld let's close this one as well 😄