Subsetting `[` tibble row with `"0"` seems buggy

[eitsupi]

It seems to behave unexpectedly or generate errors.

tbl <- tibble::tibble(a = 1:3)

tbl[as.character(0:1), ]
#> # A tibble: 2 × 1
#>       a
#>   <int>
#> 1     1
#> 2     1
tbl[as.character(-1:0), ]
#> # A tibble: 2 × 1
#>       a
#>   <int>
#> 1    NA
#> 2    NA
tbl[as.character(0:2), ]
#> # A tibble: 3 × 1
#> Error in `[<-`:
#> ! Assigned data `map(.subset(x, unname), vectbl_set_names, NULL)` must
#>   be compatible with existing data.
#> ✖ Existing data has 3 rows.
#> ✖ Assigned data has 2 rows.
#> ℹ Only vectors of size 1 are recycled.
#> Caused by error in `vectbl_recycle_rhs_rows()`:
#> ! Can't recycle input of size 2 to size 3.
tbl["0", ]
#> # A tibble: 1 × 1
#> Error in `[<-`:
#> ! Assigned data `map(.subset(x, unname), vectbl_set_names, NULL)` must
#>   be compatible with existing data.
#> ✖ Existing data has 1 row.
#> ✖ Assigned data has 0 rows.
#> ℹ Row updates require a list value. Do you need `list()` or `as.list()`?
#> Caused by error in `vectbl_recycle_rhs_rows()`:
#> ! Can't recycle input of size 0 to size 1.

^{Created on 2025-05-03 with reprex v2.1.1}

[krlmlr]

Thanks, interesting. For reference:

tbl <- data.frame(a = 1:3)

tbl[as.character(0:1), ]
#> [1] NA  1
tbl[as.character(-1:0), ]
#> [1] NA NA
tbl[as.character(0:2), ]
#> [1] NA  1  2
tbl["0", ]
#> [1] NA

^{Created on 2025-05-03 with reprex v2.1.1}

[krlmlr]

And a bunch of simpler working cases:

tbl <- tibble::tibble(a = 1:3)

tbl["-1", ]
#> # A tibble: 1 × 1
#>       a
#>   <int>
#> 1    NA
tbl["0", ]
#> # A tibble: 1 × 1
#> Error in `[<-` at tibble/R/tbl_df.R:75:3:
#> ! Assigned data `map(.subset(x, unname), vectbl_set_names, NULL)` must
#>   be compatible with existing data.
#> ✖ Existing data has 1 row.
#> ✖ Assigned data has 0 rows.
#> ℹ Row updates require a list value. Do you need `list()` or `as.list()`?
#> Caused by error in `vectbl_recycle_rhs_rows()` at tibble/R/subassign-backend.R:35:5:
#> ! Can't recycle input of size 0 to size 1.
tbl["1", ]
#> # A tibble: 1 × 1
#>       a
#>   <int>
#> 1     1
tbl["2", ]
#> # A tibble: 1 × 1
#>       a
#>   <int>
#> 1     2
tbl["3", ]
#> # A tibble: 1 × 1
#>       a
#>   <int>
#> 1     3
tbl["4", ]
#> # A tibble: 1 × 1
#>       a
#>   <int>
#> 1    NA

^{Created on 2025-05-03 with reprex v2.1.1}

[moodymudskipper]

The object is corrupted when "0" is part of the index, and print() is either misleading or failing. All the weirdness comes from recycling or attempting to when printing.

tbl <- tibble::tibble(a = 1:3)

constructive::construct(tbl[c("0", "1"), ])
#> list(a = 1L) |>
#>   structure(row.names = c(NA, -2L), class = c("tbl_df", "tbl", "data.frame"))

constructive::construct(tbl["0", ])
#> list(a = integer(0)) |>
#>   structure(row.names = c(NA, -1L), class = c("tbl_df", "tbl", "data.frame"))

^{Created on 2025-06-04 with reprex v2.1.0}

The "row.names" attribute is correct but the data isn't.

I think this is the culprit, "0" should map to NA

tibble:::string_to_indices(c("-1", "0", "1"))
#> [1] NA  0  1

^{Created on 2025-06-04 with reprex v2.1.0}

[eitsupi]

Thanks @moodymudskipper, that seems true #1637.