Using related columns in `cur_column()` with `across()`

[olivroy]

Hello!
I would like to write code with dplyr + across. for this:

Where x = x_1 + x_2, y = y_1 + y+2, etc.

tibble(
  x = c(100, 50, 20),
  x_1 = c(20, 30, 10),
  x_2 = c(80, 20, 10),
  y = c(20, 12, 12),
  y_1 = c(15, 10, 2),
  y_2 = c(5, 2, 10)
) |>
   mutate(
    across(matches("\\d$"), \(column) column / .data[[stringr::str_remove(col_name, "_\\d")
  )

I wonder it would be possible to add a helper like cur_col_name() for example that returns a string that I could manipulate to extract another column name

Thanks