processing URL as RDOC bug

[adrianomitre]

For some reason, YARD is processing the URL of references as RDOC in such a way that {foo}[http://baz.com/+bar+] result in foo, which simply breaks the link.

For an example, look at the end of the description section of http://rubydoc.info/gems/retroactive_module_inclusion/1.1.0/frames for a link named Dynamic Module Include Problem, which should point to http://eigenclass.org/hiki/The+double+inclusion+problem but rather point to http://eigenclass.org/hiki/The%3Ctt%3Edouble%3C/tt%3Einclusion+problem.

[lsegal]

Don't process ++ typewriter text inside of HTML attributes

Closed by 4ac099c