select should be able to drop columns by number when duplicate names exist

[WilDoane]

When a data.frame contains columns with the same name, select can't be used to correct the problem by dropping columns.

Even if rename and other verbs are problematic in this regard, dropping duplicate columns should (in principle) be possible.

---

Consider these error cases:

tibble::tribble(
  ~a, ~a, ~b,
  1, 2, 3
) %>% 
  select(-1)

or

tibble::tribble(
  ~a, ~a, ~b,
  1, 2, 3
) %>% 
  select(starts_with("a"))

[batpigandme]

reprex with backtrace:

suppressPackageStartupMessages(library(tidyverse))
tibble::tribble(
  ~a, ~a, ~b,
  1, 2, 3
) %>% 
  select(-1)
#> Can't bind data because some arguments have the same name
rlang::last_error()
#> <error>
#> message: Can't bind data because some arguments have the same name
#> class:   `rlang_error`
#> backtrace:
#>   1. tibble::tribble(~a, ~a, ~b, 1, 2, 3)
#>  10. dplyr::select(., -1)
#>  11. tidyselect::vars_select(sel_vars(.data), !!!enquos(...))
#>  12. tidyselect:::vars_select_eval(.vars, quos)
#>  13. rlang::child_env(data_helpers_env, !!!data)
#>  14. rlang:::env_bind_impl(env, list2(...))
#> Call `rlang::last_trace()` to see the full backtrace

^{Created on 2019-06-06 by the reprex package (v0.3.0)}

[romainfrancois]

Moving this to tidyselect.

tidyselect::vars_select(c("a", "a", "b"), -1)
#> Can't bind data because some arguments have the same name

^{Created on 2019-06-07 by the reprex package (v0.2.1)}