GH-35534: [R] Ensure missing grouping variables are added to the beginning of the variable list

[paleolimbot]

Rationale for this change

As reported by @eitsupi, dplyr adds missing grouping variables to the beginning of the variable list; however, we add them to the end of the variable list. Our general policy is to match dplyr's behaviour everywhere.

Before this PR:

library(arrow, warn.conflicts = FALSE)
#> Some features are not enabled in this build of Arrow. Run `arrow_info()` for more information.
library(dplyr, warn.conflicts = FALSE)

tibble::tibble(int = 1:4, chr = letters[1:4]) |> 
  group_by(chr) |> 
  select(int) |> 
  collect()
#> Adding missing grouping variables: `chr`
#> # A tibble: 4 × 2
#> # Groups:   chr [4]
#>   chr     int
#>   <chr> <int>
#> 1 a         1
#> 2 b         2
#> 3 c         3
#> 4 d         4

arrow_table(int = 1:4, chr = letters[1:4]) |> 
  group_by(chr) |> 
  select(int) |> 
  collect()
#> # A tibble: 4 × 2
#> # Groups:   chr [4]
#>     int chr  
#>   <int> <chr>
#> 1     1 a    
#> 2     2 b    
#> 3     3 c    
#> 4     4 d

After this PR:

library(arrow, warn.conflicts = FALSE)
#> Some features are not enabled in this build of Arrow. Run `arrow_info()` for more information.
library(dplyr, warn.conflicts = FALSE)

tibble::tibble(int = 1:4, chr = letters[1:4]) |> 
  group_by(chr) |> 
  select(int) |> 
  collect()
#> Adding missing grouping variables: `chr`
#> # A tibble: 4 × 2
#> # Groups:   chr [4]
#>   chr     int
#>   <chr> <int>
#> 1 a         1
#> 2 b         2
#> 3 c         3
#> 4 d         4

arrow_table(int = 1:4, chr = letters[1:4]) |> 
  group_by(chr) |> 
  select(int) |> 
  collect()
#> # A tibble: 4 × 2
#> # Groups:   chr [4]
#>   chr     int
#>   <chr> <int>
#> 1 a         1
#> 2 b         2
#> 3 c         3
#> 4 d         4

Are these changes tested?

Yes, a test was added.

Are there any user-facing changes?

Yes, column ordering will be different. This could be a breaking change because existing code that refers to columns by index may change; however, referring to a column by name is much more common.

• Closes: [R] Column order after group_by(foo) |> select(...) is different from dplyr #35534

[github-actions]

⚠️ GitHub issue #35534 has been automatically assigned in GitHub to PR creator.

[thisisnic]

Great, thanks for fixing this!

[conbench-apache-arrow]

Conbench analyzed the 6 benchmark runs on commit 7de273b4.

There was 1 benchmark result indicating a performance regression:

• Commit Run on ursa-thinkcentre-m75q at 2023-06-28 18:14:26Z
  • params=<STATIC_VECTOR(std::shared_ptr)>, source=cpp-micro, suite=arrow-small-vector-benchmark

The full Conbench report has more details.