add dates to output of sliding_period

It might be helpful to include the dates of the start/end of the in/out sample periods in the output of sliding_period(). This would also allow users to join the rsampling table with the Slice01, Slice02, etc labels to see a time series.

df <- data.frame(x = vctrs::new_date(c(1, 3, 4, 6, 7, 10)))
x <- sliding_period(df, x, "day", lookback = 2, assess_stop = 2, step = 2)
x$in_start_date = df$x[purrr::map_vec(x$splits, .f = function(x) min(x$in_id) )]
x$in_end_date = df$x[purrr::map_vec(x$splits, .f = function(x) max(x$in_id) )]

# Sliding period resampling 
# A tibble: 2 × 4
  splits        id     in_start_date in_end_date
  <list>        <chr>  <date>        <date>     
1 <split [2/1]> Slice1 1970-01-02    1970-01-04 
2 <split [2/1]> Slice2 1970-01-05    1970-01-07 

Start with the sliding_period() example in the issue and inspect how its splits and id columns are produced. Add start and end dates for the in-sample and assessment periods to the output, then verify that the displayed Slice1 and Slice2 values can be joined to those dates.