Let's say I have these strings:
row - 1: First_s - end
row - 2: Second_s - end
...
row - 400: string_400 - end
I'd like to use the matlab function regexp to select any non-white character that is between : and - end. So in my example it would select First_s Second_s string_400, etc. What would an appropriate regex look like?
Instead of lookaheads and lookbehinds, you match the strings that lead and trail the tokens:
>> Cs = regexp(strings,':\ (.*)\ - end','tokens');
>> Cs{1}{1}
ans =
'First_s'
>> Cs{2}{1}
ans =
'Second_s'
>> Cs{3}{1}
ans =
'string_400'
Using a strict lookbehind and lookahead: (?<=: ).+?(?= - end).
Uses regexprep to replace unwanted parts by nothing:
strings = {
'row - 1: First_s - end';
'row - 2: Second_s - end';
'row - 400: string_400 - end'}; %// example data. Assumed to be a cell array
strings = regexprep(strings,'.*:','') %// remove start
strings = regexprep(strings,'- end$','') %// remove end
strings = regexprep(strings,'\s','') %// remove spaces
The result in this example is:
strings =
'First_s'
'Second_s'
'string_400'
Here is the regex that works with ruby: row\s\-\s[\d]+\:\s(.*)\s-\send which would probably work with matlab as well.
You can see how this example works here: http://rubular.com/r/FuPwNcBy1I
