Problem with `cur_column()`

[olivroy]

The cur_column() function does not detect the values, using !!!. I read that in across, some parts may be evaluated too early.
Is there a way to replace the !!! call in that case?

The code works outside across.

---

Brief description of the problem

library(dplyr, warn.conflicts = FALSE)

d <- tibble(x = 1:2, y = 7:8)
values <- list(x = c("1" = "yes",  "2" = "no"),
                      y = c("7" = "YEs",  "8" = "No")
      )

d %>% 
    mutate(across(any_of(names(values)),
                  .fn = ~ recode(., !!!(values[[(cur_column())]])))
    )
#> Error: `cur_column()` must only be used inside `across()`.
d %>% mutate(x = recode(x, !!!(values[["x"]])))
#> # A tibble: 2 x 2
#>   x         y
#>   <chr> <int>
#> 1 yes       7
#> 2 no        8

^{Created on 2021-05-21 by the reprex package (v2.0.0)}

[romainfrancois]

!!! is spliced too early. You can work around this by defining the function upfront:

library(dplyr, warn.conflicts = FALSE)

d <- tibble(x = 1:2, y = 7:8)
values <- list(x = c("1" = "yes",  "2" = "no"),
               y = c("7" = "YEs",  "8" = "No")
)

f <- function(x) {
  recode(x, !!!(values[[(cur_column())]]))
}

d %>% 
  mutate(across(any_of(names(values)), f))
#> # A tibble: 2 x 2
#>   x     y    
#>   <chr> <chr>
#> 1 yes   YEs  
#> 2 no    No

^{Created on 2021-05-26 by the reprex package (v2.0.0)}

[netique]

I've just stumbled upon a similar issue, but with a different error message that is not helpful at all. Is it possible to somehow inform the user where the culprit is? In this case, there is nocur_collumn(), though:

library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(forcats)

# fails
tibble(a = factor(c("A1", "A2", "A3"))) |>
  mutate(
    across(
      a,
      \(x) {
        rename_vec <- c("a" = "A1", "b" = "A2", "c" = "A3")
        fct_recode(x, !!!rename_vec)
      }
    )
  )
#> Error: object 'rename_vec' not found

# works
rename_fun <- function(x) {
  rename_vec <- c("a" = "A1", "b" = "A2", "c" = "A3")
  fct_recode(x, !!!rename_vec)
}

tibble(a = factor(c("A1", "A2", "A3"))) |>
  mutate(
    across(
      a,
      rename_fun
    )
  )
#> # A tibble: 3 × 1
#>   a    
#>   <fct>
#> 1 a    
#> 2 b    
#> 3 c

^{Created on 2025-02-22 with reprex v2.1.1}