The only difference between reading a binary file versus a text file is to translate end-of-line characters. Other than that (and implicitly, the types of characters you get in text versus binary), there's no difference. I would recommend you read the file as binary.
Unless you're saying that it is ACTUALLY a text file but just has a different file extension. Then I would recommend opening in text mode as per usual.
To open in binary mode:
fid = fopen( 'myfile.dat', 'rb' );
allbytes = fread(fid); % this call reads all characters from the file
fclose(fid);
To open in text mode:
fid = fopen('myfile.dat', 'rt' );
line = fgetl(fid); % this call reads a single line from the file
fclose(fid);
Have a look at this:
[EDIT] From your example...
infile = fopen( 'myfile.$PJ', 'rt' );
outfile = fopen( 'myfile_processed.$PJ', 'wt' );
while ~feof(infile)
line = fgetl(infile);
toks = textscan(line, '%s%f', 1);
if ~isempty(toks)
switch toks{1}
case {'HEIGHT', 'HOFFST', 'TOWHT'}
line = sprintf( '%s %d', toks{1}, fix(toks{2}+1) );
end
end
fprintf( outfile, '%s\n', line );
end
fclose(infile);
fclose(outfile);
