Include base type (e.g. list) as a class attribute?

Question

I am writing an R package which revolves around analyzing some around data stored in a list. For example,

myData <- list(x = "x vector", y = "another vector", 
    z = "function(x,y)", meta = "info about this dataset")

The character values are just descriptions here since the values are not relevant. I have a few different kinds of data stored in a list format, but there is some overlap so I'm giving each kind a class attribute, like this

class(myData) <- "datatype1"

Now I've tested this, and seen that things like

length(myData) ; myData$x

still work, and I am under the impression that this is because the storage mode is still a list, and the operations I've tried are .Primitve or .Internal, but I don't know when/why this works.

My question is the following: Is there any circumstance in which I should worry about getting rid of 'list' as a class name? Should I instead use:

class(myData) <- c("datatype1", "list")

?

Basically I don't know if/when function dispatch fails if the class 'list' is not found but the storage mode is still a list. I've tried to read documentation and haven't seen this particular issue. If someone can point me to the relevant info, I'm happy to read the manual. Thanks!

Answer 1

In general some functions in some packages may be dispatched based on "list" class. For example as.data.frame or within in base R.

list1 <- list(a = 1)
class(list1) <- "nolist" 
within(list1, a <- 4)

Error in UseMethod("within") : no applicable method for 'within' applied to an object of class "nolist"

For a "list" dependent methods check

methods(class = "list")