dbplyr translation of pmax/pmin throws an error if it receives the na.rm = TRUE argument

[fh-jgutman]

With aggregate functions like max and min, the dbplyr:::check_na_rm gives a useful warning if the na.rm = TRUE option is not provided, and if na.rm = TRUE is provided, the option gets silently ignored. This behavior does not exist for the non-aggregate functions pmax() and pmin, which are now translated to GREATEST and LEAST in certain db backends. Instead, providing the option pmax(..., na.rm = TRUE) currently throws an error.

The desired behavior is that, similar to aggregate functions min/max, pmin and pmax also throw a warning via dbplyr:::check_na_rm if no na.rm argument is provided, and that no error is thrown if na.rm = TRUE is provided.

pg_con <- DBI::dbConnect(RPostgres::Postgres(), "test_data")
dplyr::copy_to(pg_con, mtcars)
db_mtcars <- dplyr::tbl(pg_con, "mtcars")

db_mtcars %>%
   dplyr::summarise(max_mpg = max(mpg, na.rm = TRUE)) # desired behavior

db_mtcars %>%
   dplyr::mutate(new_size = pmax(wt, drat, na.rm = TRUE)) # throws error

Output:

Error: Failed to prepare query: ERROR:  syntax error at or near "AS"
LINE 1: ..."am", "gear", "carb", GREATEST("wt", "drat", TRUE AS "na.rm"...
                                                             ^
In addition: Warning message:
Named arguments ignored for SQL GREATEST