One of the fantastic new features in the upcoming RStudio 1.3 release is real-time spellchecking. According to the announcement blog post:
The application will only check comments in code files and raw text in markdown files.
I agree this is the right decision. But since using RStudio 1.3 Preview, I've noticed that I frequently have code blocks within comments (which of course are highlighted as misspelled). I'm specifically thinking of the @example and @examples sections of Roxygen comments.
I don't think it would be feasible to automatically recognize all code blocks in comments. But perhaps we could disable the spellchecker from these common Roxygen sections? These sections themselves can contain comments, so perhaps you'd prefer to apply the logic recursively?