Extract text data without knowing the size of the data.

Hi, I have the following problems. I have a large amount of data that have similar format as follow. The problems are break into three parts. First part, I need to ignore the 3 headerlines without counting the number of headerlines and store MatlabStart,MatlabBin,MatlabVel and their corresponding data into matix. I have already done that by finding matching in '='. My problem now is how to recognize the data table( the data after MatlabVel) and store the header and its data into a different matrix/array. Again, without specifying the size of the table data. I know textscan can do sth like textscan(fid,%f%f%f%f 'delimiter', ' '). My point is how to let matlab recognize the data table without specifying the %f%f%f%f(e.g is there other way other than textscan?). And finally, how do i let matlab to stop scanning when it reaches the line 'threshold'. Since i do the '=' matching in the first part of the problem, so it is continue to store the remaining data after threshold. Any help or advice or hint will very much be appreciated.

This is just an matlab example\C:version.exe

Version: Matlab13.5

Identity: Default : 3.78

MatlabStart = 10;

MatlabBin = 125;

MatlabVel = 587;

Index Inbox Miscalculate Misinterprete

0 -1 -3000 5000

1 200 -1000 6788

2 500 800 400

Threshold

A

Fence1=100

Fence2=200

Fence3=400

Bet1= 2

Bet2= 3

Bet3= 4

B

Fence1=700

Fence2=150

Fence3=800

Bet1= 2

Bet2= 3

Bet3= 4

Answer 1

[Julius]

You can try a function called "readtext"

Download link: http://www.mathworks.com/matlabcentral/fileexchange/10946-readtext

Answer 2

[Jacky]

Should have known this function earlier. I did it the hard way using some codings. Thanks for referring the link. It definitely help me to simplify my code.

[Julius]

You're welcome. I had exactly the same problem as you one year ago and I spent quite a lot of time searching for solutions.

In my opinion Matlab should include functions of such purposes in its built-in function library.

Answer 3

[Jacky]

thanks