Extracting mixed date from string in R-CodePudding

I have a vector of characters that looks like the table below, I would like to extract the dates from them and convert them as.Date. For example, row one would be 09-11-2021. The last number in the string is the number of columns and not part of the date.

   <chr>                                                                       
 1 By Leigh-Ann Butler, Shannon Cobb, Michael R. DonaldsonNov 9, 20213 Comments
 2 By Leigh-Ann Butler, Shannon Cobb, Michael R. DonaldsonNov 8, 20212 Comments
 3 By Rick AndersonNov 4, 202114 Comments                                      
 4 By Victoria Ficarra, Rob JohnsonNov 3, 20215 Comments                       
 5 By Roger C. SchonfeldNov 1, 202123 Comments                                 
 6 By Joseph EspositoOct 29, 20211 Comment                                     
 7 By Brigitte ShullOct 20, 20216 Comments

example.data <- c("By Leigh-Ann Butler, Shannon Cobb, Michael R. DonaldsonNov 9, 20213 Comments",
"By Leigh-Ann Butler, Shannon Cobb, Michael R. DonaldsonNov 8, 20212 Comments",
"By Rick AndersonNov 4, 202114 Comments",                                      
"By Victoria Ficarra, Rob JohnsonNov 3, 20215 Comments")

CodePudding user response：

You could use

as.Date(gsub(". (\\w{3}\\s\\d{1,2},\\s\\d{4}).*", "\\1", example.data), format = "%b %d, %Y")

#> [1] "2021-11-09" "2021-11-08" "2021-11-04" "2021-11-03"

CodePudding user response：

strcapture(".*(\\D{3})\\s (\\d{1,2}),\\s (\\d{4}).*",
           example.data, proto = list(mon="", day=0L, year=0L)) |>
  transform(date = as.Date(paste(mon, day, year), format = "%b %d %Y"))
#   mon day year       date
# 1 Nov   9 2021 2021-11-09
# 2 Nov   8 2021 2021-11-08
# 3 Nov   4 2021 2021-11-04
# 4 Nov   3 2021 2021-11-03