Don't add braces to function bodies specifically for S4 generics #1141

Activity

1. lorenzwalthert commented on Nov 29, 2023

   

   lorenzwalthert

   on Nov 29, 2023

   Collaborator

   More actions

   This is a bit more complex with current code since the rule is implemented to check if there is a function token and If yes, add the braces, without knowing about the fact that we are inside setGeneric. Can you identify we are inside setGeneric if standardGeneric is in the function body or is that not true? If it is mostly true, we could avoid adding the curly braces based on a condition where stadnardGeneric() is called inside the function. That would still add braces to a situation like this:

    setGeneric("myGeneric", function(x) {
     othergeneric("myGeneric")
   })
   
2. MichaelChirico commented on Feb 24, 2025

   

   MichaelChirico

   on Feb 24, 2025

   ContributorAuthor

   More actions

   I don't see any CRAN packages using standardGeneric() without setGeneric() (at least, setGeneric() is in the same file):

   https://github.com/search?q=org%3Acran+lang%3AR+%2FstandardGeneric%5B%28%5D%2F+-%22setGeneric%22&type=code

   I don't think there's anything stopping authors from doing:

   myGenericImpl <- function(x)
     standardGeneric("myGeneric")
   
   setGeneric("myGeneric", myGenericImpl)

   It seems safe enough to me to go forward with the rule you proposed.
3. MichaelChirico commented on Jun 5, 2026

   

   MichaelChirico

   on Jun 5, 2026

   ContributorAuthor

   More actions

   I filed https://bugs.r-project.org/show_bug.cgi?id=19083