`pivot_wider()` has an issue with 0-row data frames when a manual spec specifies a nonexistent key

[DavisVaughan]

library(tidyr)

spec <- tibble(.name = "Release", .value = "seen", station = fish_encounters$station[1])
spec
#> # A tibble: 1 × 3
#>   .name   .value station
#>   <chr>   <chr>  <fct>  
#> 1 Release seen   Release

pivot_wider_spec(
  fish_encounters[0,], 
  spec
)
#> Error: Can't subset elements that don't exist.
#> x Location 1 doesn't exist.
#> ℹ There are only 0 elements.

# This kind of thing generally works and gives NA
pivot_wider_spec(
  fish_encounters[2,], 
  spec
)
#> # A tibble: 1 × 2
#>   fish  Release
#>   <fct>   <int>
#> 1 4842       NA

I think this is just due to an error in the way wrap_vec() computes the slice indices

[DavisVaughan]

Slightly more minimal reprex

library(tidyr)

spec <- tibble(.name = "name", .value = "value", x = 1L)

df <- tibble(key = "a", value = 1L, x = 2L)
pivot_wider_spec(df, spec)
#> # A tibble: 1 × 2
#>   key    name
#>   <chr> <int>
#> 1 a        NA

df <- tibble(key = character(), value = integer(), x = integer())
pivot_wider_spec(df, spec, id_cols = key)
#> Error: Can't subset elements that don't exist.
#> x Location 1 doesn't exist.
#> ℹ There are only 0 elements.

^{Created on 2021-12-03 by the reprex package (v2.0.1)}