[Fix #7737] Add new Style/UnnecessaryArguments cop

[tejasbubane]

Closes #7737

---

Before submitting the PR make sure the following are checked:

• Wrote good commit messages.
• Commit message starts with [Fix #issue-number] (if the related issue exists).
• Feature branch is up-to-date with master (if not - rebase it).
• Squashed related commits together.
• Added tests.
• Added an entry to the Changelog if the new code introduces user-observable changes. See changelog entry format.
• The PR relates to only one subject with a clear title and description in grammatically correct, complete sentences.
• Run bundle exec rake default. It executes all tests and RuboCop for itself, and generates the documentation.

[bbatsov]

I think that's not the best approach to building such a cop. A much better one would be to avoid hardcoding anything and build the rules for the methods and their redundant params via the configuration of the cop.

[tejasbubane]

This is a bad workaround & will fail for string with multiple spaces. Can anyone help me fix this?

[bbatsov]

I guess first you can squeeze the extra spaces before passing them to the cases. Maybe someone else from @rubocop-hq/rubocop-core has better ideas.

[bquorning]

You could convert the configured values into AST nodes and compare directly against node.arguments.first.

I played a bit with it and got as far as this:

require 'parser/current'

def on_send(node)
  return unless redundant_argument?(node)
  # ...
end

def redundant_argument?(node)
  redundant_argument = redundant_arg_for_method(node.method_name.to_s)
  return false if redundant_argument.nil?

  node.arguments.first == redundant_argument
end

def redundant_arg_for_method(method_name)
  return nil unless cop_config['RedundantArguments'].key?(method_name)

  @mem ||= {}
  @mem[method_name] ||= begin
    arg = cop_config['RedundantArguments'].fetch(method_name)

    buffer = Parser::Source::Buffer.new('(string)', 1)
    buffer.source = arg.inspect
    builder = RuboCop::AST::Builder.new
    Parser::CurrentRuby.new(builder).parse(buffer)
  end
end

I am not completely sure that a new buffer and builder instance is needed for each configuration, so the memoization may need to be moved around a bit.

[tejasbubane]

@bbatsov I have changed cop logic to be configurable for any method & argument.

[bbatsov]

@tejasbubane Looks good to me overall. I'd only expand the docs to mention that people should be aware that the cop can't tell apart methods with the same name in different classes, so they should be careful with the cop.

[bbatsov]

And that this is limited to methods that take a single param. To make it more generic we should be passing both a param position and a param in the config (e.g. if the redundant param is the second one), but I guess such cases are quite uncommon.

[bbatsov]

Might be a good idea to add some validation logic for the config, so we can present the users with more meaningful error messages if they mess something up.

[bbatsov]

I guess it'd be clearer to say that the arg is redundant.

[bquorning]

I was going to write something about us not knowing the type of the receiver, but I noticed just now that it’s been covered already in the issue #7737 (comment)

[bquorning]

You could convert the configured values into AST nodes and compare directly against node.arguments.first.

I played a bit with it and got as far as this:

require 'parser/current'

def on_send(node)
  return unless redundant_argument?(node)
  # ...
end

def redundant_argument?(node)
  redundant_argument = redundant_arg_for_method(node.method_name.to_s)
  return false if redundant_argument.nil?

  node.arguments.first == redundant_argument
end

def redundant_arg_for_method(method_name)
  return nil unless cop_config['RedundantArguments'].key?(method_name)

  @mem ||= {}
  @mem[method_name] ||= begin
    arg = cop_config['RedundantArguments'].fetch(method_name)

    buffer = Parser::Source::Buffer.new('(string)', 1)
    buffer.source = arg.inspect
    builder = RuboCop::AST::Builder.new
    Parser::CurrentRuby.new(builder).parse(buffer)
  end
end

I am not completely sure that a new buffer and builder instance is needed for each configuration, so the memoization may need to be moved around a bit.

[bbatsov]

@tejasbubane ping :-)

[tejasbubane]

@bbatsov Changes done.

[marcandre]

Technically this cop is unsafe, as the actual default argument to join is $OUTPUT_FIELD_SEPARATOR || '' and that of split is $FIELD_SEPARATOR || ' '; if these globals are set, then join() and split() have different meaning than join('') and split(' ')

[bbatsov]

I'm closing this PR due to no recent activity. Feel free to re-open it if you ever come back to it.

[tejasbubane]

@bbatsov Looks like I don't have the rights to reopen this PR. Do you mind if I open a new one? Thought I should ask first because opening a new one means losing context from the discussions here.

[tejasbubane]

Technically this cop is unsafe, as the actual default argument to join is $OUTPUT_FIELD_SEPARATOR || '' and that of split is $FIELD_SEPARATOR || ' '; if these globals are set, then join() and split() have different meaning than join('') and split(' ')

@marcandre Thanks. The cop is marked unsafe. Also, these kernel variables are deprecated since 2.7

[marcandre]

I don't have the rights to reopen this PR. Do you mind if I open a new one?

Can't either. Says the branch was force-pushed or recreated. Not sure why github prevents it. Anyways, you can open a new one.

The cop is marked unsafe.

Good, I must have forgotten to check, sorry.