`str_detect_all` for multiple patterns?

[tvedebrink]

In case you wants to filter all rows matching some string patterns, it would be handy to have str_detect_all for this task, I have used purrr to implement this, but maybe there is already a better way available?

str_detect_all <- function(string, patterns, negate = FALSE){
  n_patterns <- length(patterns)
  if(n_patterns == 1 | n_patterns == length(string)){
    return(str_detect(string = string, pattern = patterns, negate = negate))
  }
  map_any <- function(x, y) purrr::map2_lgl(.x = x, .y =  y, .f = any)
  patterns %>% 
    purrr::map(~ str_detect(string = string, pattern = .x, negate = negate)) %>% 
    purrr::reduce(.f = map_any)
}

In case you wanted to filter the mtcars data for cars with c("Lotus", "Duster", "Drive") in their name, this could be achieved by

> mtcars %>% as_tibble(rownames = "car") %>% filter(str_detect_all(car, c("Lotus", "Duster", "Drive")))
# A tibble: 3 × 12
  car              mpg   cyl  disp    hp  drat    wt  qsec    vs    am  gear  carb
  <chr>          <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Hornet 4 Drive  21.4     6 258     110  3.08  3.22  19.4     1     0     3     1
2 Duster 360      14.3     8 360     245  3.21  3.57  15.8     0     0     3     4
3 Lotus Europa    30.4     4  95.1   113  3.77  1.51  16.9     1     1     5     2

[arnaudgallou]

FWIW, you can already achieve that with str_detect() and regular expressions:

mtcars |> 
  tibble::as_tibble(rownames = "car") |> 
  dplyr::filter(stringr::str_detect(car, "Lotus|Duster|Drive"))
#> # A tibble: 3 × 12
#>   car            mpg   cyl  disp    hp  drat    wt  qsec    vs    am  gear  carb
#>   <chr>        <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 Hornet 4 Dr…  21.4     6 258     110  3.08  3.22  19.4     1     0     3     1
#> 2 Duster 360    14.3     8 360     245  3.21  3.57  15.8     0     0     3     4
#> 3 Lotus Europa  30.4     4  95.1   113  3.77  1.51  16.9     1     1     5     2

^{Created on 2024-01-05 by the reprex package (v2.0.1)}

[hadley]

Yeah, this is the approach that I'd recommend since the regexp engine can efficiently compile. If you're accepting user provided strings, make sure to call str_escape() to escape any special characters that they might be using.