Split string and find unique characters for each row using data.table

Suppose I have a data table:

dt <- data.table(x = c("a,b,b,c,NA","b,b,NA,c","a,b,c,c,c,d"))
dt

             x
1:  a,b,b,c,NA
2:    b,b,NA,c
3: a,b,c,c,c,d

Now for each row I would like to split the x column, extract all unique characters and paste them together so I get:

             x
1:    a,b,c,NA
2:      b,NA,c
3:     a,b,c,d

I have tried this so far but after that I am stuck and another issue is that I have datasets with over two million observations.

dt[, tstrsplit(x, ",")]
   V1 V2 V3 V4   V5   V6
1:  a  b  b  c   NA <NA>
2:  b  b NA  c <NA> <NA>
3:  a  b  c  c    c    d

>Solution :

dt[, x2 := lapply(strsplit(x, ","), function(y) paste0(unique(y), collapse = ","))]
#              x       x2
# 1:  a,b,b,c,NA a,b,c,NA
# 2:    b,b,NA,c   b,NA,c
# 3: a,b,c,c,c,d  a,b,c,d