Converting a function to accept input directly in r - r

Question

I was reading a book and I came across this function in R. This function basically finds out patterns in the input string having a minimum threshold of 3.
vec <- "da0abcab0abcaab0d0"
find_rep_path <- function(vec, reps) {
regexp <- paste0(c("(.+)", rep("\\1", reps - 1L)), collapse = "")
match <- regmatches(vec, regexpr(regexp, vec, perl = TRUE))
substr(match, 1, nchar(match) / reps)
}
vals <- unique(strsplit(vec, "")[[1]])
str <- NULL
for (i in seq.int(nchar(vec))) {
x <- vec
for (v in vals) {
substr(x, i, i) <- v
tmp <- find_rep_path(x, 3)
if (length(tmp) > 0)
str <- c(str, tmp)
}
}
nc <- nchar(str)
unique(str[which(nc == max(nc))])
Now, I wish to convert this function into the form like,
function("da0abcab0abcaab0d0"). This means, that I can easily pass a string to the function directly and not hardcode it in the original function. How can I modify this?
I know this is a beginner question but I am completely at sea right now as far as R is concerned. Please help!

Answer 1

I don't see how it's hardcoded. But you can just wrap your code into a function if that's what you mean?
# Function 1
find_rep_path <- function(vec, reps) {
regexp <- paste0(c("(.+)", rep("\\1", reps - 1L)), collapse = "")
match <- regmatches(vec, regexpr(regexp, vec, perl = TRUE))
substr(match, 1, nchar(match) / reps)
}
# Function 2
foo <- function(vec) {
vals <- unique(strsplit(vec, "")[[1]])
str <- NULL
for (i in seq.int(nchar(vec))) {
x <- vec
for (v in vals) {
substr(x, i, i) <- v
tmp <- find_rep_path(x, 3)
if (length(tmp) > 0)
str <- c(str, tmp)
}
}
nc <- nchar(str)
return(unique(str[which(nc == max(nc))]))
}
vec <- "da0abcab0abcaab0d0"
foo(vec)
#[1] "0ab" "abc"
Edit1
To get the place of the matches you can use gregexr:
a <- foo(vec)
gregexpr(a[1], vec)
#[[1]]
#[1] 3 9
#attr(,"match.length")
#[1] 3 3
#attr(,"useBytes")
#[1] TRUE
This tells you that a[1] ("0ab") was matched in vec at positions 3 and 9. Run ?gregexpr for more informations.
Edit2
To add this information to each match, we can do something like
bar <- function(vec) {
m <- foo(vec)
ans <- sapply(m, gregexpr, vec, fixed = TRUE)
ans <- lapply(ans, function(x) {attributes(x) <- NULL; x})
return(ans)
}
bar(vec)
#$`0ab`
#[1] 3 9
#
#$abc
#[1] 4 10