dplyr::coalesce throws error when column name is missing

zed <- zed %>%
    dplyr::mutate(Email = coalesce(`Email Address`, Email, `E-mail Address`, `E-MAIL`, email, `E-Mail`, `E-mail`, `E-Mail Address`, `EMAIL ADDRESS`, EMAIL))

Error in `mutate_cols()`:
! Problem with `mutate()` column `Email`.
ℹ `Email = coalesce(...)`.
✖ object 'E-mail Address' not found
Caused by error in `list2()`:
! object 'E-mail Address' not found
Run `rlang::last_error()` to see where the error occurred.

How can we handle using coalesce to consolidate all of these columns into one column, while also handling the potential for one or more of these columns to be missing. Simply removing E-mail Address is not great because in other scenarios used by our function, E-mail Address will in fact exist.

You can do this with the help of do.call -

library(dplyr)

cols <- c("Email Address", "Email", "E-mail Address", "E-MAIL", "email")
# Sample data
set.seed(12442)
dat <- data.frame(matrix(sample(c(1:10, NA), 25, replace = TRUE), 5, 5, 
                  dimnames = list(NULL, cols)), check.names = FALSE)

dat %>%
  mutate(Email_new = do.call(coalesce, pick(any_of(cols))))

#  Email Address Email E-mail Address E-MAIL email Email_new
#1             7     7              8      4    10         7
#2             7     7              4      1     3         7
#3             2     7             NA      3     9         2
#4            NA     4              6      9     9         4
#5             6     6              1     10     8         6

Now let's delete one of the column and try again with the same code.

dat$email <- NULL

dat %>%
   mutate(Email_new = do.call(coalesce, pick(any_of(cols))))

#  Email Address Email E-mail Address E-MAIL Email_new
#1             7     7              8      4         7
#2             7     7              4      1         7
#3             2     7             NA      3         2
#4            NA     4              6      9         4
#5             6     6              1     10         6