Stacking data masks?

[mikmart]

This is a re-post of this StackOverflow question: https://stackoverflow.com/q/79445476/4550695.

Given a function that defuses its arguments and evaluates them in a data mask:

foo <- function(...) {
  dots <- rlang::quos(...)
  expr <- rlang::quo(list(!!!dots))
  rlang::eval_tidy(expr, list(a = 1))
}

foo(a)
#> [[1]]
#> [1] 1

Is there a way to make a function that wraps the other function and adds another data mask?

A couple of attempts tried and failed:

bar <- function(...) {
  dots <- rlang::quos(...)
  expr <- rlang::quo(foo(!!!dots))
  rlang::eval_tidy(expr, list(b = 2))
}

try(bar(a, b))
#> Error : object 'b' not found

# Looks promising...
bar <- function(...) {
  eval(substitute(foo(...)), list(b = 2), parent.frame())
}

try(bar(a, b))
#> [[1]]
#> [1] 1
#> 
#> [[2]]
#> [1] 2

# ... but fails with injection.
baz <- function(x) {
  bar(a, {{ x }})
}

try(baz(b))
#> Error : object 'b' not found

[lionel-]

I don't remember exactly but I don't think so.