How to select columns in R based on exclusive shared values?

I have a dataframe that contains the following 8 values: 0, 1, 2, 3, 4, 5, ?, and -.
The values from 0 to 5 are characters and not numeric. For a given set of rows, I want to know what column values they share exclusively. For example, in the table below, I want to select the columns that have the same values for rows 3-5, but only when their values are exclusive to those three rows. So I want to select x6, x7, and x8, but not x5 (because the '2' value is found in row 2 as well). How do I do this in R?

 x1 x2 x3 x4 x5 x6 x7 x8 x9
 1  0  5  1  1  ?  0  5  5
 2  1  ?  1  2  5  1  -  5
 3  2  1  3  2  1  3  ?  4
 4  3  ?  4  2  1  3  ?  4
 5  4  0  1  2  1  3  ?  2

Here's a function that you can try -

select_exclusive_values <- function(x, rows) {
  #Select only interested value
  sub_x <- x[rows]
  #get all unique values of it
  unq_sub_x <- unique(sub_x)
  #check if all the interested values are the same and 
  #none of them occur elsewhere
  length(unq_sub_x) == 1 && all(unq_sub_x != x[-rows])
}

This checks for one column

select_exclusive_values(df$x1, 3:5)
#[1] FALSE

You may use sapply to apply it for every column in the dataframe.

rows <- 3:5
res <- sapply(df, select_exclusive_values, rows)

res
#   x1    x2    x3    x4    x5    x6    x7    x8    x9 
#FALSE FALSE FALSE FALSE FALSE  TRUE  TRUE  TRUE FALSE 

To select those columns -

df[res]
#  x6 x7 x8
#1  ?  0  5
#2  5  1  -
#3  1  3  ?
#4  1  3  ?
#5  1  3  ?

With dplyr, select columns where all values are equal to the first one, among the selected rows; but also the first value is different from any of the rows' value.

library(dplyr)
r = 3:5
df %>% 
  select(where(~ all(.x[r] == first(.x[r])) && all(first(.x[r]) != .x[-r])))
  
  x6 x7 x8
1  ?  0  5
2  5  1  -
3  1  3  ?
4  1  3  ?
5  1  3  ?