Parsing Complicated text file header

I have a complicated header to a .txt file that I'm having trouble reading and parsing. I want to pull out the Sample, the Size, and the data (roughly 13000 lines). I've included a simplified version below.

What is the best way to read this?

Thank you!

 Sample  Virgin 1050, Sec 6, Data Sample
Geometry  Rectangular (Length, Width, Thickness)
GeomFactor  27.573229 1/mm 0.041252 1/mm 1.137453 1/mm²
Size  24.2412  3.17500  0.276900  mm
SampParams  0.4000
Method  Creep - Sample Rate
Comment  37 Deg, 10 MPa
Xcomment  Air Bearing Gas: Air
Text
Multiplex  None
InstCalDate  Electronics:  2011-11-14 10:45:15 0.00006
AutoAnalysis  Off
MacroFile
OrgFile  C:\Folder\File
StartOfData
5.027167  37.02016  0.0000000  0.0000000  
5.078333  37.02214  1762.800  0.0000000  
5.127500  37.02239  2218.972  0.0000000

Answer 1

[Matt Tearle]

If the headers are uniformly formatted, you could simplify this a bit, but this should work:

    fid = fopen('data.txt');
    sample = fgetl(fid)
    foo = textscan(fid,'Size %f%f%f%*s','headerlines',2,'CollectOutput',true);
    sizeinfo = foo{1}
    foo = '';
    while ~strcmp(foo,'StartOfData')
        foo = fgetl(fid);
    end
    foo = textscan(fid,'%f%f%f%f','delimiter',',','CollectOutput',true);
    data = foo{1}
    fclose(fid);

I'm mixing-n-matching approaches. I'm assuming the first line is the "sample", and the fourth line is the "size". But for generality, rather than counting lines, I'm assuming the data starts after the line "StartOfData". Note the use of the HeaderLines property in textscan to skip fixed numbers of lines, and the use of fgetl and strcmp to hunt for a line with specific text.

[Evan Carlson]

Thank you for your help! This will work well.