Match All, Order-Sensitive and Duplicates-Sensitive

Find all indices of vector haystack that are equal to vector needles, taking into account the order of both vectors, and their duplicate values.

match_all() is essentially a much more efficient version of:

lapply(needles, \(i) which(haystack == i))

Like lapply(needles, \(i) which(haystack == i)), NAs are ignored.

match_all() internally calls collapse::fmatch and collapse::gsplit.
Core of the code is based on a suggestion by Sebastian Kranz (author of the 'collapse' package).

Usage

match_all(needles, haystack, unlist = TRUE)

Arguments

needles, haystack: vectors of the same type.
needles cannot contain NA/NaN.
Long vectors are not supported.
unlist: Boolean, indicating if the result should be a single unnamed integer vector (TRUE, default), or a named list of integer vectors (FALSE).

Value

An integer vector, or list of integer vectors.

If a list, each element of the list corresponds to each value of needles.

When needles and/or haystack is empty, or when haystack is fully NA, match_all() returns an empty integer vector (if unlist = TRUE), or an empty list (if unlist = FALSE).

Examples

n <- 200
haystack <- sample(letters, n, TRUE)
needles <- sample(letters, n/2, TRUE)
indices1 <- match_all(needles, haystack)
head(indices1)
#> [1]  26  32  71 108 113 116