Application of the mutate method to a listframe

mutate.listframe

listframe <- function(...){
  structure(
    tibble(...),
    class = c("listframe","tbl_df","tbl","data.frame")
  )
}

lf <- listframe(
  a = list(1,c("a","b","c"),matrix(1:4,2,2)),
  b = list(as.raw(1:3),"hello",c(1,3,4)),
  c = list("one","two","three")
)

lf



extract_dots <- function () 
{
  .pcall <- match.call(sys.function(sys.parent()), sys.call(sys.parent()), 
                       expand.dots = F, envir = parent.frame(2L))
  dots <- .pcall[["..."]]
  dots <- dots[names(dots) != ""]
  list2env(lapply(dots, eval, parent.frame(2)), parent.frame())
}

named_pmap <- function(.l,.f,...){
  .f <- as_mapper(.f,...)
  body(.f) <- call("{", quote(extract_dots()), body(.f))
  pmap(.l,.f)
}

mutate.listframe <- function(.data,...){
  .call <- match.call(expand.dots = FALSE)
  if(!...length()){
    .data
  } else {
    .call_out <- match.call()
    .call_out[[1]] <- quote(dplyr:::mutate.data.frame)
    nms <- names(.call[["..."]])
    for(c_nm in nms){
      if(is_formula(.call_out[[c_nm]])){
        .call_out[[c_nm]] <- call("named_pmap",
                                  substitute(.data),
                                  .call_out[[c_nm]])
      }
      #.call_out[[c_nm]] <- call("named_pmap",
      #                          substitute(.data),
      #                          call("~",.call_out[[c_nm]]))
      
    }
    
    eval.parent(.call_out)
  }
}

lf %>%
  mutate(amax = ~max(a),
         bclass = ~deparse(class(b)),
  )

mutate(lf,amax = ~max(a))

The commented out bit on lines 48-50 would create a call to named_pmap() regardless of whether a ~ is used and therefore means you wouldn't have to use e.g. amax = ~max(a) and instead would use amax = max(a), which is much more condensed, but less transparent about what it's doing. This wouldn't allow direct calls, e.g. mutate(lf, new_col = list(1,2,3)) would be converted to something like new_col = named_pmap(lf,~list(1,2,3)) which would re-create the list on each row, rather than the first row getting element 1, etc...