Fix false negatives for Style/EmptyLiteral

[vlad-pisanov]

Improve Style/EmptyLiteral to detect other common ways of creating empty literals.

Array.new      # ✔️ existing offense
Hash.new       # ✔️ existing offense

Array[]        # ⚠️ new offense
Array([])      # ⚠️ new offense
Array.new([])  # ⚠️ new offense
Hash[]         # ⚠️ new offense
Hash([])       # ⚠️ new offense
Hash.new([])   # ⚠️ new offense

Also: improve offense message by including the actual offense source

---

Before submitting the PR make sure the following are checked:

• The PR relates to only one subject with a clear title and description in grammatically correct, complete sentences.
• 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.
• Ran bundle exec rake default. It executes all tests and runs RuboCop on its own code.
• Added an entry (file) to the changelog folder named {change_type}_{change_description}.md if the new code introduces user-observable changes. See changelog entry format for details.

[koic]

Can you update as below and adjust the specs to have more appropriate message?

Suggested change

	current = array_with_index(node) ? 'Array[]' : 'Array.new'
	format(ARR_MSG, current: current)
	format(ARR_MSG, current: node.source)

[koic]

Ditto.

Suggested change

	current = hash_with_index(node) ? 'Hash[]' : 'Hash.new'
	format(HASH_MSG, current: current)
	format(HASH_MSG, current: node.source)

[vlad-pisanov]

combined existing and new cases into shared examples for readability

[koic]

Thank you for being considerate. However, as I mentioned in the comment below, shared_examples do not always improve readability. Can you respect the existing spec descriptions?

[bbatsov]

I think in that particular case the shared examples are OK, given that we're essentially testing 6 times for the same thing.

[koic]

shared_examples don't always improve readability and reproducibility. Could you expand them into individual tests, as in the original specs?

[franzliedke]

@vlad-pisanov @bbatsov Why is Hash.new([]) an offense? That parameter is the default value for unknown keys. (It's better to do this with the block form, but that's a different discussion.) That's not the same as a literal {}, or what am I missing?

[bbatsov]

@franzliedke That's an oversight that's fixed on master.

[bbatsov]

See #13178