pivot_wider: rethrow_id_cols_oob() - new error

[gevro]

Hi, I found this error for a different situation than #1609 that I'm not sure are addressed by the fixes for that issue.

df <- tibble(col1 = c("x", "y"), col2 = c(1, 2), col3 = c("a", "b"), col4 = c("c", "d"))

df %>% pivot_wider(id_cols = all_of(c("col1","col2")), names_from = col3, values_from = -all_of(c("col1", "col4")))

Error:

Error in `rethrow_id_cols_oob()`:
! `i` must be a single string, not a character vector.
ℹ This is an internal error that was detected in the tidyr package.
  Please report it at <https://github.com/tidyverse/tidyr/issues> with a reprex (<https://tidyverse.org/help/>) and the full backtrace.
Run `rlang::last_trace()` to see where the error occurred.

No error:

df %>% pivot_wider(id_cols = all_of(c("col1","col2")), names_from = col3, values_from = -all_of(c("col1", "col2", "col4")))

[moxboxwa]

I also came across this error -- but found the problem on line 8: failed to capitalize the varname "year". So this was a simple coding error on my part (in case that helps dx other reports)

rlang::last_trace(drop=F)
<error/rlang_error>
Error in rethrow_id_cols_oob():
! i must be a single string, not an integer vector.
ℹ This is an internal error that was detected in the tidyr package.
Please report it at https://github.com/tidyverse/tidyr/issues with a reprex and the full backtrace.

---

Backtrace:
     ▆
  1. ├─... %>% ...
  2. ├─dplyr::bind_rows(...)
  3. │ └─rlang::list2(...)
  4. ├─dplyr::select(., Year, Number = Suicide, Percent)
  5. ├─dplyr::mutate(...)
  6. ├─dplyr::mutate(., Year = as.character(Year))
  7. ├─tidyr::replace_na(., list(Suicide = 0))
  8. ├─tidyr::pivot_wider(., id_cols = year, names_from = suicide, values_from = Number)
  9. ├─tidyr:::pivot_wider.data.frame(...)
 10. │ └─tidyr:::build_wider_id_cols_expr(...)
 11. │   └─tidyr:::select_wider_id_cols(...)
 12. │     ├─rlang::try_fetch(...)
 13. │     │ └─base::withCallingHandlers(...)
 14. │     └─tidyselect::eval_select(...)
 15. │       └─tidyselect:::eval_select_impl(...)
 16. │         ├─tidyselect:::with_subscript_errors(...)
 17. │         │ └─base::withCallingHandlers(...)
 18. │         └─tidyselect:::vars_select_eval(...)
 19. │           └─tidyselect:::loc_validate(pos, vars, call = error_call)
 20. │             └─vctrs::vec_as_location(pos, n = length(vars), call = call)
 21. ├─vctrs (local) `<fn>`()
 22. │ └─vctrs:::stop_subscript_oob(...)
 23. │   └─vctrs:::stop_subscript(...)
 24. │     └─rlang::abort(...)
 25. │       └─rlang:::signal_abort(cnd, .file)
 26. │         └─base::signalCondition(cnd)
 27. ├─tidyselect (local) `<fn>`(`<vctrs___>`)
 28. │ └─rlang::cnd_signal(cnd)
 29. │   └─rlang:::signal_abort(cnd)
 30. │     └─base::signalCondition(cnd)
 31. └─rlang (local) `<fn>`(`<vctrs___>`)
 32.   └─handlers[[1L]](cnd)
 33.     └─tidyr:::rethrow_id_cols_oob(...)
 34.       └─tidyr:::check_string(i, .internal = TRUE)
 35.         └─tidyr:::stop_input_type(...)
 36.           └─rlang::abort(message, ..., call = call, arg = arg)

[DavisVaughan]

Fixed with dev tidyr

df <- tibble(col1 = c("x", "y"), col2 = c(1, 2), col3 = c("a", "b"), col4 = c("c", "d"))

df %>% pivot_wider(id_cols = all_of(c("col1","col2")), names_from = col3, values_from = -all_of(c("col1", "col4")))
#> Error in `pivot_wider()`:
#> ℹ In argument: `all_of(c("col1", "col2"))`.
#> Caused by error in `all_of()`:
#> ! Can't subset elements that don't exist.
#> ✖ Element `col2` doesn't exist.

This is the correct error because values_from already selects col2, so it no longer exists for id_cols to select.