Extracting specific repeating lines of text after a heading using fgetl and textscan

Here is an example of the data I am working with. I would like to extract the line directly following each KEY tag. The files have many thousands of these, so I need to create a loop with textscan or something similar.

> <NAME>
mary

> <AGE>
30

> <KEY>
RDHQFKQIGNG

> <NAME>
john

> <AGE>
56

> <KEY>
JFJNNFNFKFNN

Desired result:

RDHQFKQIGNG
JFJNNFNFKFNN

Here is where I am at (adapted from a similar question in the past), the code does not seem to be moving the cursor, and instead works for the first one, and then grabs all data after it, instead of just the data following the KEY line.

f = fopen('data.txt', 'rt'); 

tline = fgetl(f);
while isempty(strfind(tline, '> <KEY>'))
    if tline == -1 
        break;
    end
    line = fgetl(f);
end

if tline ~= -1
    data = textscan(f,'%s','Delimiter','\r\n');
else
    disp('not found');
end

fclose(f);

Thanks!

Answer 1

[Stephen23]

>> str = fileread('temp1.txt');
>> C = regexp(str,'(?<=> <KEY>\s+)\S+','match')
C = 
  'RDHQFKQIGNG'    'JFJNNFNFKFNN'

Tested on this file:

[Vincent Scalfani]

Thanks so much, it seems to work great with small text files, but in a file with about 20,000 of these tags, it is very slow. It is currently 4 hours in and still running. Any ideas how to adjust the code to make it more efficient?

[Stephen23]

Try this:

  E = regexp(str,'^> <KEY>\s+\S+','match','lineanchors');
  E = strtrim(strrep(E,'> <KEY>',''));

And have a play with this script:

[Vincent Scalfani]

Amazing!!! PERFECT. It took 1 second to process over 4 million lines of text. Thanks so much for your time.