<FEATURE REQ>[`as_date()`] Would it be possible for `as_date` to only produce future dates?

[JoeMarino2021]

I was wondering if as_date() could have a future argument that would cause only future dates to be created? Consider the dates of the upcoming NFL season, as an example. Here are the last few dates of the season, which are 2026 and 2027 dates.

[1] "December 27" "December 28" "December 31" "January 3" "January 4" "January 10"

If I perform as_date(format = "%B %d") on these dates, it creates the following results.

[1] "2026-12-27" "2026-12-28" "2026-12-31" "2026-01-03" "2026-01-04" "2026-01-10"

With the future argument, we could have as_date(format = "%B %d", future = TRUE) and produce this output.

[1] "2026-12-27" "2026-12-28" "2026-12-31" "2027-01-03" "2027-01-04" "2027-01-10"

Maybe even as_date() could be smart enough to detect any future rollover points from December to January in a list that was longer and multiyear and add more than just one year to the output dates when needed. My example only spans the crossover from 2026 to 2027, but a longer list may, perhaps, also contain a crossover from 2027 to 2028 at some point.

It is not too much trouble to correct this now, with

mutate(Date = case_when(Date < today() ~ Date %m+% years(1), .default = Date))

but code could be so much cleaner, and possibly faster, if as_date() had a future argument.

Thanks for creating and maintaining such wonderful tools, and considering this request. Have a great day!

[JoeMarino2021]

The more I have thought about this, another couple of argument options might be useful.

min_date == The minimum/earliest date to be returned
sorted == The input dates are in order, so if at any point the vector appears to go "backwards", then add a year at this point to these dates

When someone provides the sorted argument, years are added as needed at any rollovers to the converted dates so that the output vector is also sorted with no "reversing" of the dates.

Thanks again for making R even more enjoyable! Have a good one!