How can I carry out overlap string match

Hi,

I have a string as such:

seq1 = 'QEAFEISKXXXXXXX'

I want to find all occurrences of two Xs separated by a single character (including X). By inspection, there are seven Xs side-by-side in the sequence so the answer i am looking for in terms of indices is: 9 10 11 12 13

I have used regexp as follows:

seq1 = 'QEAFEISKXXXXXXX';

seq2 = 'X.X';        % Dot indicates any character to match with

regexp(seq1, seq2)

The answer i get is: 9 12

How can i modify the code to get every occurrence?

Thanks.

Answer 1

[Walter Roberson]

seq2 = 'X(?=.X)'
regexp(seq1, seq2)

[Abel Chandra]

Thank you very much. That's what i was looking for. I have another question, if i now have to find X _ _ X (Xs separated by two characters)then how would i modify your solution?

[Abel Chandra]

I got it. seq2 = 'X(?=..X)'. Thanks again.

Answer 2

[Andrei Bobrov]

strfind('QEAFEISKXXXXXXX','XX')

[Abel Chandra]

Thank you for your answer. But in your answer, the two Xs are not separated by any character. What i am interested in is 'X_X' where _ can be character. E.g. 'XAX' 'XAX' 'XXX' etc. In my code i have used X.X but seems the regexp does not overlap through the sequence hence not giving the answer i am looking for.

[Andrei Bobrov]

exotics:

ii = strfind('QEAFEIXKXXXXXXX','X');
out = ii(any(bsxfun(@minus,ii(:),ii(:)') == -2,2));