Closing Heredoc Indentation

[siegfault]

When autocorrecting heredoc indentation, the heredoc close is ignored. I think I would expect it to be realigned with the heredoc open.

The following examples use

class Foo
  <<-SQL.strip_indent
    foo
SQL
end

as starting code.

---

Expected behavior

When running rubocop, I'd expect to see a violation for SQL needing to be indented two spaces and for the autocorrect to result in

class Foo
  <<-SQL.strip_indent
    foo
  SQL
end

Actual behavior

Instead, there are no violations.

Steps to reproduce the problem

See above.

RuboCop version

$ rubocop -V
0.52.1 (using Parser 2.4.0.2, running on ruby 2.3.1 x86_64-linux)

Further Discussion

I'm not sure whether this would fall on the IndentHeredoc rule (which currently only handles the innards of the heredoc), the IndentationWidth rule (which handles the opening), a new ClosingHeredocIndentation rule, or somewhere else. I'd be happy to work on this if given some indication of a preferred location.

[siegfault]

Any thoughts here?

[Drenmi]

Either IndentationWidth or ClosingHeredocIndentation seem like a good place. 🙂

[siegfault]

Sounds good - thanks! I'll see if I can take a look at this this weekend.

[Drenmi]

Swell! Feel free to ping me here if you have any questions.

[siegfault]

@Drenmi I opened #5557 and would love comments! I started with IndentationWidth since I thought that might be preferable, but got a bit stumped trying to make it work there. Then I got ClosingHeredocIndenation working. Knowing what I know now, I might revisit an IndentationWidth implementation, but that will probably have to wait until the weekend.

[Drenmi]

I opened #5557 and would love comments!

Will have a look!

I started with IndentationWidth since I thought that might be preferable, but got a bit stumped trying to make it work there.

Expected so. It's already an insanely complex thing. 🙂