How to extract only parts of a text file? not just skip a few lines but extracting segments found in all text of file.

Hello.

I'm been trying to extract only certain parts of a .txt file without success.

The file is way to big to post any part of, but for example. When you compile a arduino sketch and you copy the output shown in the in the bottom of the IDE window and paste it into a .txt file you get number of very long lines. Somewhere in those lines there is a path to the directory where the IDE use some files for example:

-IC:\Program Files (x86)\Arduino\hardware\arduino\sam\system/libsam -IC:\Program Files...

following that there are a few more paths in each line but before that there are a lot of other garble, commands and so.

I want to extract only the C:\ paths, I thought maybe I could specify to read everything starting with '-IC:\Program Files (x86)\' and ends with ' ', there is a blank space between /libsam and -IC... above.

But I can't find any straight way of doing that and my attempts so far has not yielded any desirable result.

So far I have gotten this:

fileName = 'DMA.txt';

data = fileread(fileName);

s = strfind(data, '-IC:');

then if I do, EDU>> whos data

Name      Size               Bytes  Class    Attributes

data       1x34585            69170  char

and s is a 1x140 double, then I got the starting point of every part of the text that I want and I have checked how many characters after each starting point is of interest and it varies but it is always less then 100.

I have almost a answer I think, the only problem is how to store 140 different strings. t = data(s(1):s(2)) gives the first occurrence so a for loop like

 for i = 1:length(s) 
   str = data(s(i):s(i+1);
 end

would almost do it, but I don't know how to handle the str variable? It needs to be an array of strings but nothing I do result in that.

I attach the file in question.

Regards

[David]

There are others C:\ paths so only -IC:\ paths should read.

[David]

I think I might have found a way using fileread() and strfind(), I'll post when/if I get it right.

[dpb]

How large are the files overall? It often is simplest to snarf up the whole file intact into memory and then work on it there.

How many lines/path occurrences are there and are they all or only a specific one or few of interest?

Regular expressions are often useful in such contexts...