[Fix #8948] Fix autocorrection for Style/RedundantRegexpCharacterClass with %r

[ysakasin]

This commit uses RegexpParser::Map#body instead of Parser::Source::Range#adjust in Style/RedundantRegexpCharacterClass.

When regular expression are written by %r literal, RedundantRegexpCharacterClass made incorrect loc as following.

%r{abc[\d]+}
    ^^^^^^

This bug are caused by callingnode.loc.begin.adjust to get range of character class with %r regexp.
Parser::Source::Range#adjust adjust based on Parser::Source::Range#begin_pos but it with %r literal doesn't indicate start of regular expression source, point to location of '%'.
So, RedundantRegexpCharacterClass makes incorrect loc if %r literal are used in code.

r = %r{abc}
    ^      # `node.loc.begin.begin_pos`

r = %r{abs}
       ^   # `node.loc.begin.end_pos`

Ref. #8948

---

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.

[owst]

Thanks for fixing this @ysakasin!

I recently added RegexpNode#parsed_tree_expr_loc: https://github.com/rubocop-hq/rubocop/blob/0259bed3df30e7d0e18645e651a683b890d03db4/lib/rubocop/ext/regexp_node.rb#L42-L44 which correctly uses loc.begin.end. I wonder if we could alter that method so that it accepts an optional explicit length argument that you could use in this PR?

What about something like:

def parsed_tree_expr_loc(expr, loc_size: expr.full_length)
  loc.begin.end.adjust(begin_pos: expr.ts, end_pos: expr.ts).resize(loc_size)
end

[ysakasin]

@owst Thanks reviewing!

I modify to use RegexpNode#parsed_tree_expr_loc and add keyword argument loc_size to it.

[marcandre]

It's very clucky to have to provide parsed_tree_expr_loc with the right answer.
At minimum, that method should check the type of the node and adjust if need be (like here).
Probably better would be to monkey patch Regexp's parser Nodes and add a loc method that would provide us with the right information.
For a char class like here, you could use loc.expression and that's it.

[ysakasin]

Like following?

        def each_redundant_character_class(node)
          each_single_element_character_class(node) do |char_class|
            next unless redundant_single_element_character_class?(node, char_class)

            # This means that: '['.length + len_char_class_source + ']'.length
            len_char_class = 1 + char_class.expressions.first.text.length + 1

            yield node.loc.expression.adjust(begin_pos: char_class.ts).resize(len_char_class)
          end
        end

[marcandre]

Not what I meant. Let me prepare a PR

[marcandre]

I created #8960. I believe that the correction would then simply be to remove the loc.begin and loc.end.

[ysakasin]

@marcandre Thanks adding awesome method!
I replaced RegexpNode#parsed_tree_expr_loc with RegexpParser::Map#body.

[bbatsov]

Looks great! Thanks!