Extract Cyrillic letters from string

Question

The function below extracts any letters from the english alphabet from a string.

is.letter <- function(x) grepl("[:alpha:]", x) 

I want to build a similar function that extracts only cyrlic letters from a given string.

Update:

With the code provided by Wiktor Stribiżew, I get the following results:

grepl("\\p{Cyrillic}", x, perl=TRUE)

test[, c(2, 11)]
      MOBILE_NUMBER contain_cyrlic
    1  НЕМА ТЕЛЕФОН          FALSE
    2      НЕПОЗНАТ          FALSE
    3  НЕМА ТЕЛЕФОН          FALSE
    4  НЕМА ТЕЛЕФОН          FALSE

Any ideas?

Answer 1

stringi might provide more cross-platform/system consistent results for you but both stri_detect_regex and grepl (in "perl" mode) should do the trick:

library(stringi)
library(dplyr)

data_frame(
  MOBILE_NUMBER = c("НЕМА ТЕЛЕФОН", "НЕПОЗНАТ", "НЕМА ТЕЛЕФОН", "НЕМА ТЕЛЕФОН")
) -> tst

tst
## # A tibble: 4 × 1
##   MOBILE_NUMBER
##           <chr>
## 1  НЕМА ТЕЛЕФОН
## 2      НЕПОЗНАТ
## 3  НЕМА ТЕЛЕФОН
## 4  НЕМА ТЕЛЕФОН

(t1 <- mutate(tst, is_cyrillic = grepl("\\p{Cyrillic}", MOBILE_NUMBER, perl=TRUE)))
## # A tibble: 4 × 2
##   MOBILE_NUMBER is_cyrillic
##           <chr>       <lgl>
## 1  НЕМА ТЕЛЕФОН        TRUE
## 2      НЕПОЗНАТ        TRUE
## 3  НЕМА ТЕЛЕФОН        TRUE
## 4  НЕМА ТЕЛЕФОН        TRUE

(t2 <- mutate(tst, is_cyrillic = stri_detect_regex(MOBILE_NUMBER, "\\p{Cyrillic}")))

## # A tibble: 4 × 2
##   MOBILE_NUMBER is_cyrillic
##           <chr>       <lgl>
## 1  НЕМА ТЕЛЕФОН        TRUE
## 2      НЕПОЗНАТ        TRUE
## 3  НЕМА ТЕЛЕФОН        TRUE
## 4  НЕМА ТЕЛЕФОН        TRUE

identical(t1, t2)
## [1] TRUE