False negative on Style/FormatString?

[kakutani]

Rubocop(master) looks make false nagative against typical sprintf/format cases.

Some example cases(picked from ruby-doc.org) will fail tests.

This patch will make the test suite is passing:

diff --git a/lib/rubocop/cop/style/format_string.rb b/lib/rubocop/cop/style/format_string.rb
index e75095a..37c12df 100644
--- a/lib/rubocop/cop/style/format_string.rb
+++ b/lib/rubocop/cop/style/format_string.rb
@@ -31,13 +31,13 @@ def offending_node?(node)
         end

         def format_method?(name, node)
-          receiver, method_name, args = *node
+          receiver, method_name, *args = *node

           # commands have no explicit receiver
           return false unless !receiver && method_name == name

           # we do an argument count check to reduce false positives
-          args && args.children.size >= 2
+          args.size >= 2
         end

         def format?(node)

I'm not sure why format(something, a, b) registers an offense in spec but format("%d %04x", 123, 123) does not.

It would be great if you guys look into it.

[eitoball]

👍