reading part of a very large txt file

hi,

I have a text file looking like this:

x=1 y=2

1

2

3

4

*

x=4 y=2

4

2

*

x=3 y=1

2

44

*

and so on.

the data I need to read is, for example, all the numbers under x=4 y=2. the text file is very big (over 20 GB) and of course i don't want to read all of it.

Answer 1

[Guillaume]

Then read the file line by line until you find your pattern, then read the section you need until the next block of xy. Something like:

desiredxy = [4 2];

searchpattern = fprintf('x=%d y=%d', desiredxy);
fid = fopen('c:\somewhere\somefile.txt', 'rt');
assert(fid > 0, 'failed to open file');
datacolumn = [];
insection = false;
while true
   line = fgetl(fid);
   if ~ischar(line)
      error('reached end of file before pattern was found');
   end
   if insection
      %in the section of interest
      if line(1) == 'x'
         %beginning of next section. Done reading. Quit while loop
         break;
      else
         %read number
         datacolumn = [datacolumn; str2double(line)];  %#ok<AGROW>
      end
   elseif strcmp(line, searchpattern)
      %not in the section of interest but just found its start
      insection = true;
   end  
end

[yonatan s]

thank you.