Read a specific number in a text file containing a mixture of strings and numbers

Hi there,

I need to read "a specific number" (highlighted below) in 1998 different text files (text1.txt, text2.txt, ..., text1998.txt) simultaneously and print that numbers in another file (preferably, an excel file). All the files are a mixture of strings and numbers and look like as follows:

----------------------------------------------------------------------------------------------------------------------------------------------------------------

Structure: Biaxial House model with 1 story.

Maximum responses---------------------------

Displacements: X Y R Story--1: 1.267323e+000 1.251265e-016 1.574376e-011

Maximum Inter-story responses---------------------------

Inter-story Drifts: dX dY dR Story--1: 1.267323e+000 1.251265e-016 1.574376e-011

Maximum story forces---------------------------

Total force at story: Fx Fy Mr Story--1: 6.152071e+003 1.364242e-012 4.009894e-003

Maximum story Accelerations---------------------------

Global Accel at story Centroid: Ax Ay Ar Story--1: 3.860713e+001 9.869805e-017 3.364166e-010

The average acceleration is calculated using a 0.01sec time interval Note: If the specified time interval is smaller than the time interval of input excitation, the excitation time interval is used to calculate the average Accel. Maximum story Average Accelerations---------------------------

Global Accel at story Centroids:(average) Ax Ay Ar Story--1: 8.247387e-001 9.865127e-017 2.810327e-010

Building Damage index calculated based on Hysteretic elements: 8.239350e-001

----------------------------------------------------------------------------------------------------------------------------------------------------------------

I really don't know how to do that (I know how to read a simple excel file, or how to read a *.txt file that only contains numbers, but really got into trouble with reading that specific number in a mixed file, and also don't know how to use a 'for' loop that reads all the 1998 files for me!). So, could any of you do me a favor and help me with this please (a sample file is also attached below. Please note that the suffix is *.gen but I use notepad to read these files)?!

Thanks a lot

Answer 1

[Geoff Hayes]

Mamali - if you can assume that all text files are the same and so follow the same pattern which is that you are interested in the first number of the line that follows the one that contains the word Drifts then you can use the following to grab that value.

numFiles = 1998;
myData   = zeros(numFiles,1);
for k=1:numFiles

      % create the file name
      filename = sprintf('text%d.txt',k);

      % try to open the file
      fid = fopen(filename,'r');

      if fid > 0

          % loop over each line until the end of file (eof) is reached
          while ~feof(fid)

             % grab the next line in the file
              lineTxt = fgetl(fid);

              % look for the text Drifts which indicates that the next line
              % is where we want to grab the number from
              if strfind(lineTxt,'Drifts')

                  % get the next line
                  lineTxt = fgetl(fid);

                  % split the line of text into a cell array using the space
                  % character as the delimiter
                  cellArray = strsplit(lineTxt,' ');

                  % grab the second element of the array
                  numOfInterest = str2double(cellArray{2});

                  % do something with this number
                  myData(k) = numOfInterest;

                  % exit from file
                  break;
              end
          end

          % close the file
          fclose(fid);
      end
  end

Try the above and see what happens!

[Antonio]

Dear Geoff - Thanks a lot for your great assistance. Your code looks like awesome, but I still have two problems! First, it overwrites the results, i.e., after running the code I only get the result for the last file, i.e. text1998.txt! Second, the results are not numbers! They are strings and cannot be used in post calculations. I think the reason is that they're in quotations like this: '1.267323e+000'. Could you tell me please how to solve these issues?

[Geoff Hayes]

Mamali - try creating an array for each of those values read from the file. See the edited code above. Note that I'm not clear why results are strings especially after the str2double conversion.

[Antonio]

Geoff - your code works perfectly. I appreciate your kind assistance. :) About the results, they became strings when I copied them from MATLAB to excel! I solved this issue anyway. Thanks a lot ...

Answer 2

[per isakson]

An alternative using regular expressions to parse one file

tic
str = fileread('MM1E1I1.txt');
xpr_num = [ 
    '\d+'               ... one or more digits
    '(\.\d+)?'          ... zero or one group of one dot and one or more digits     
    '([eE][\+-]\d{3})?' ... zero or one group of the letter E, "+" or "-", and three digits
    ]; 
cac = regexp( str                                                      ... 
    , ['(?<=Inter-story Drifts: +dX +dY +dR\r?\nStory--1: +)',xpr_num] ...
    , 'match', 'once'                                                  );   
toc

and

>> str2double( cac )
ans =
    1.2673

With R2013a,64bit on an old vanilla desktop: Elapsed time is 0.002555 seconds.

[Antonio]

per isakson - Thanks for your comment :)