take out numbers from string

I attached file which contains as this

'10.0.7.4:22->10.0.8.5:26856'
'10.0.13.4:22->10.0.12.5:9997'
'10.0.9.4:22->10.0.10.5:47576'
'10.0.14.4:443->10.0.10.5:4922'
'10.0.14.4:443->10.0.10.5:20113'
'10.0.11.4:80->10.0.13.5:61779'
'10.0.10.4:22->10.0.9.5:64788'
'10.0.9.4:80->10.0.10.5:25283'

I want to take out numbers appears after 1st and 2nd colon in two different variable. i.e. A=[22 22 443 443 80 22 80] and B=[26856 9997 47576 4922....] but A and B should be in column format. how to do it.

Answer 1

[Guillaume]

There's no point in creating two different variables when it's much easier to have the output as two columns of the same variable. Here's one way to do it:

suffix = regexp(VarName6, '(?<=:)\d+', 'match');  %match sequence of numbers following a :
suffix = str2double(vertcat(suffix{:}))           %convert cell array of cell arrays into one big cell array and convert strings to numeric

[José-Luis]

I really should work on my regexes... +1

[Jay Hanuman]

works Thank you.

Answer 2

[José-Luis]

I'll get you started:

 data = [{'10.0.7.4:22->10.0.8.5:26856'};{'10.0.13.4:22->10.0.12.5:9997'}];
 result = cellfun(@(x) regexp(x,':[0-9]+','match'),data,'UniformOutput',false);
 a = cellfun(@(x) sscanf(x{1},':%d'),result);
 b = cellfun(@(x) sscanf(x{2},':%d'),result);

There are other, possibly better, ways.

[Guillaume]

Well, I'd certainly use captures (tokens in matlab) or look-behind to avoid returning the : as part of the match.

Capture:

regexp(x, ':([0-9]+)', 'tokens')

Look-behind:

regexp(x, '(?<=:)[0-9]+', 'match')

[Jay Hanuman]

works Thank you.