modelsum(): catch the error

[kekec3778]

A short reprex:

a <- factor(c(rep("a", 5), rep("b", 5)), levels = c("a", "b"))
b <- factor(c(rep(NA_character_, 10)), levels = c("a", "b"))
c <- factor(c(sample(x = c("x", "y", "z"), size = 10, replace = TRUE)), levels = c("x", "y", "z"))
df <- tibble(a, b, c)

mod1 <- glm(formula = a ~ c, data = df, family = "binomial")  # works fine
mod2 <- glm(formula = a ~ b, data = df, family = "binomial") # error

Now, try to get this results with modelsum():

tab.mod <-
  arsenal::modelsum(
    formula = a ~ b + c,
    data = df, 
    family = "binomial",  
    adjust = ~ 1,   
    control = modelsum.control(
      show.intercept = FALSE,
      binomial.stats = c("OR", "CI.lower.OR", "CI.upper.OR", "p.value", "Nmiss")
    )
  )

Of course it fails, because a ~ b.
Is it possible to catch such errors and anyway show the results with default values for example "NA" or similar ?