Find all possible substrings of length n

Question

I have an interesting (only for me, perhaps, :)) question. I have text like:

\"abbba\"

The question is to find all possible substrings of

Answer 1

We may use

x <- "abbba"
allsubstr <- function(x, n) unique(substring(x, 1:(nchar(x) - n + 1), n:nchar(x)))
allsubstr(x, 2)
# [1] "ab" "bb" "ba"
allsubstr(x, 3)
# [1] "abb" "bbb" "bba"

where substring extracts a substring from x starting and ending at specified positions. We exploit the fact that substring is vectorized and pass 1:(nchar(x) - n + 1) as starting positions and n:nchar(x) as ending positions.

Answer 2

With combn all combinations of the vector will be arranged by column. Splitting the vector prior and transposing the result will give the result as a matrix. It can then be combined with do.call(paste,...) on the matrix as a data frame:

mat <- unique(t(combn(strsplit(x, "")[[1]],2)))
do.call(paste0, as.data.frame(mat))
#[1] "ab" "aa" "bb" "ba"

Update

We can also specify the way combn treats the combinations with a shorter syntax (@docendo):

unique(combn(strsplit(x, "")[[1]],3, FUN=paste, collapse=""))

edit

Use this solution only if you are seeking all combinations. If you are only seeking a rolling split, use Julius' answer.