Step 02 - Remove dates

Many titles carry start/end dates for a position. remove_dates() detects them, sets a DATE.X flag, and strips the date fragment.

remove_dates()
|
+- convert_ordinal()  1ST/2ND/3RD -> FIRST/SECOND/THIRD
+- has_date()         TRUE/FALSE: does the string contain a date?
+- remove_date()      strip 'YY, YY-YY, mm/dd/yyyy, month names, trailing numbers

Examples

x <- c("TREASURER (ENDED 3/16/23)",
       "DIRECTOR THRU 12/23",
       "SECRETARY UNTIL JAN 2023",
       "TRUSTEE (RESIGNED 1/24/23)",
       "PRESIDENT")

data.frame(title = x,
           has_date = has_date(x),
           cleaned  = remove_date(x))
#>                        title has_date          cleaned
#> 1  TREASURER (ENDED 3/16/23)     TRUE  TREASURER ENDED
#> 2        DIRECTOR THRU 12/23     TRUE    DIRECTOR THRU
#> 3   SECRETARY UNTIL JAN 2023     TRUE  SECRETARY UNTIL
#> 4 TRUSTEE (RESIGNED 1/24/23)     TRUE TRUSTEE RESIGNED
#> 5                  PRESIDENT    FALSE        PRESIDENT

Ordinals are spelled out first so later steps treat them consistently:

convert_ordinal(c("1ST VICE PRESIDENT", "2ND LIEUTENANT GOVERNOR"))
#> [1] "FIRST VICE PRESIDENT"       "SECOND LIEUTENANT GOVERNOR"

Input: TITLE_RAW / cleaned title. Output: TitleTxt2 (date-free text) and the DATE.X flag. Continue with step 03.