[R] Update the docs to show how to avoid situations like data loss with leading zero in partition column

### Describe the bug, including details regarding any error messages, version, and platform.

Arrow does not correctly interpret some partition values as character columns, resulting in data loss.

``` r
library(arrow)
library(dplyr)

# character col with leading zero
mtcars2 <- mutate(mtcars, test = paste0(0, mtcars$carb)) |> as_tibble()

tmp <- tempfile()

write_dataset(mtcars2, tmp, partitioning = "test")

mtcars3 <- open_dataset(tmp) |> collect()

unique(mtcars2$test)
#> [1] "04" "01" "02" "03" "06" "08"
unique(mtcars3$test)
#> [1] 1 2 3 4 6 8
```

^{Created on 2024-01-17 with [reprex v2.0.2](https://reprex.tidyverse.org)}

### Component(s)

R