Step 04 - Split multi-titles

With the separators reduced to &, split_titles() turns a row that holds several roles into one row per role. The person’s name, pay, hours and checkboxes repeat on each row; Num.Titles numbers the rows. A few common pairs are collapsed to one title rather than split: “PRESIDENT & CEO” is a CEO.

Before and after

person	before (TitleTxt3)	after (TitleTxt4)	Num.Titles
PERSON 01	SEC & TREAS	SEC	1
PERSON 01	SEC & TREAS	TREAS	2
PERSON 02	BOARD MEMBER & TREASURER	BOARD MEMBER	1
PERSON 02	BOARD MEMBER & TREASURER	TREASURER	2
PERSON 03	CHIEF EXEC OFFICER & FOUNDER	CHIEF EXEC OFFICER	1
PERSON 03	CHIEF EXEC OFFICER & FOUNDER	FOUNDER	2
PERSON 04	EVP & COO	EVP	1
PERSON 04	EVP & COO	COO	2
PERSON 05	PRESIDENT & CEO	CEO	1
PERSON 06	TRUSTEE & TREASURER	TRUSTEE	1
PERSON 06	TRUSTEE & TREASURER	TREASURER	2
PERSON 07	FOUNDER & EXECUTIVE DIRECTOR	FOUNDER	1
PERSON 07	FOUNDER & EXECUTIVE DIRECTOR	EXECUTIVE DIRECTOR	2
PERSON 08	VP FINANCE AND ADMINISTRATION	VP FINANCE AND ADMINISTRATION	1
PERSON 09	EXECUTIVE DIRECTOR	EXECUTIVE DIRECTOR	1

Nine people become 15 rows.

What it does

split_titles()
|
+- apply_misc_split_rules()  fixes for edge cases (PRESIDENT & CEO -> CEO, ...)
+- identify_split_num()      how many titles does the string hold?
+- remove_first_split()      peel off the text before the first "&", repeat

Later steps work on one title per row. The person-level view comes back in step 08, which counts each person once (title.count, title.order), and step 09, which gives each person one role.

Assets

Step 04 uses no crosswalk or word list: it relies on the & separators that step 03 wrote.

Try it

identify_split_num("CFO & TREASURER & DIRECTOR")   # 3
#> [1] 3
remove_first_split("CFO & TREASURER & DIRECTOR")   # " TREASURER & DIRECTOR"
#> [1] " TREASURER & DIRECTOR"

data(tinypartvii)
d <- dplyr::sample_n(tinypartvii, 120)
df <- d |> standardize_df() |> remove_dates() |> standardize_conj() |> split_titles()
#> [OK] standardize df step complete
#> [OK] remove dates step complete
#> [OK] standardize conjunctions step complete
#> [OK] split titles step complete
df |>
  dplyr::filter(grepl("&", TitleTxt3)) |>
  dplyr::select(F9_07_COMP_DTK_NAME_PERS, Num.Titles, TitleTxt3, TitleTxt4) |>
  head(8)
#>      F9_07_COMP_DTK_NAME_PERS Num.Titles             TitleTxt3 TitleTxt4
#> 51                                     1  TREASURER & YMCA CFO TREASURER
#> 51.1                                   2  TREASURER & YMCA CFO  YMCA CFO
#> 71              ANNA ALVARADO          1 SECRETARY & TREASURER SECRETARY
#> 71.1            ANNA ALVARADO          2 SECRETARY & TREASURER TREASURER
#> 96         RABBI DANIEL ZARET          1       PRESIDENT & CEO       CEO

Input: TitleTxt3. Output: one row per title, with the single title in TitleTxt4 and its position within the person’s entry in Num.Titles. Continue with step 05.