And/Or Detection

library(strex)

How it works

strex offers easy and/or versions of stringr::str_detect() via str_detect_all() and str_detect_any(). These are vectorized over string but not pattern. stringr::fixed() and stringr::coll()) are handled correctly. Otherwise, stringr regular expressions are used. For str_detect_all(), a pattern argument c("x", "y") is converted to "(?=.*x)(?=.*y)". For str_detect_any(), a pattern argument c("x", "y") is converted to "x|y".

Examples

str_detect_all("quick brown fox", c("x", "y", "z"))
#> [1] FALSE
str_detect_all(c(".", "-"), ".")
#> [1] TRUE TRUE
str_detect_all(c(".", "-"), coll("."))
#> [1]  TRUE FALSE
str_detect_all(c(".", "-"), coll("."), negate = TRUE)
#> [1] FALSE  TRUE
str_detect_all(c(".", "-"), c(".", ":"))
#> [1] FALSE FALSE
str_detect_all(c(".", "-"), coll(c(".", ":")))
#> [1] FALSE FALSE
str_detect_all("xyzabc", c("a", "c", "z"))
#> [1] TRUE
str_detect_all(c("xyzabc", "abcxyz"), c(".b", "^x"))
#> [1]  TRUE FALSE
str_detect_any("quick brown fox", c("x", "y", "z"))
#> [1] TRUE
str_detect_any(c(".", "-"), ".")
#> [1] TRUE TRUE
str_detect_any(c(".", "-"), coll("."))
#> [1]  TRUE FALSE
str_detect_any(c(".", "-"), coll("."), negate = TRUE)
#> [1] FALSE  TRUE
str_detect_any(c(".", "-"), c(".", ":"))
#> [1] TRUE TRUE
str_detect_any(c(".", "-"), coll(c(".", ":")))
#> [1]  TRUE FALSE
str_detect_any(c("xyzabc", "abcxyz"), c(".b", "^x"))
#> [1] TRUE TRUE

Performance

Unless you’re doing a huge amount of computation, it won’t matter, but FWIW, it’s faster to convert to regex using str_escape() rather than using coll().

bench::mark(
  str_detect_all(rep("*", 1000), rep(str_escape("*"), 555)),
  str_detect_all(rep("*", 1000), coll(rep("*", 555))),
  min_iterations = 100
)
#> # A tibble: 2 × 6
#>   expression                             min median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>                           <bch> <bch:>     <dbl> <bch:byt>    <dbl>
#> 1 "str_detect_all(rep(\"*\", 1000), r… 171ms  184ms      5.49  162.22KB    0    
#> 2 "str_detect_all(rep(\"*\", 1000), c… 159ms  161ms      6.19    4.31MB    0.688