How do I read, change and write data back into the same txt file, at the same place?

I have many files in one folder like the two examples attached. I want to change the number formed in columns 14-21 of any row that starts with "SO" "R " "TS" or "SH" by "2.335". Then write the numbers back to the row in the same file. Below is what I've tried so far:

    clear; clc; close all;
    delta = 2.335;% delta is the change in elevation
    lst = dir('*.txt');

    for n = 1:size(lst,1)
       t = lst(1).name;
       fid = fopen(t);
       tline = fgets(fid);
         while ischar(tline)
             tline = fgetl(fid);
               if tline(1) == 'S'& tline(2)== 'O' 
               disp(tline(15:21));
               elev = str2num(tline(15:21));
               elev = elev + delta;
               elev = num2str(elev)

        end
    end
  end

Thank you.

[Daniel Price]

I have made progress with my code... I now am able to write the lines the way I want to. How do I write the lines back into the file where they were before?

 clear; clc; close all;
delta = 2.335;
lst = dir('*.txt'); % Collect list of .wsw files to convert
for n = 1:size(lst,1)
    t = lst(1).name;
    fid = fopen(t);
    tline = fgets(fid);
    count = 1;
    while ischar(tline)
        count = count + 1;
        tline = fgetl(fid);
        if tline(1) == 'S'& tline(2)== 'O'
            disp(tline(15:21));
            elev = str2num(tline(15:21));
            elev = elev + delta;
            elev = num2str(elev);
            tline(15:21)= elev(1:7);
            disp(tline);
            %             fprintf(tline,'%100s', fid(count,:));
        else if tline(1) == 'R'& tline(2)== ' '
                disp(tline(15:21));
                elev = str2num(tline(15:21));
                elev = elev + delta;
                elev = num2str(elev);
                tline(15:21)= elev(1:7);
                disp(tline);
                %             fprintf(tline,'%100s', fid(count,:));
            else if tline(1) == 'T'& tline(2)== 'S'
                    disp(tline(15:21));
                    elev = str2num(tline(15:21));
                    elev = elev + delta;
                    elev = num2str(elev);
                    tline(15:21)= elev(1:7);
                    disp(tline);
                    %             fprintf(tline,'%100s', fid(count,:));
                else if tline(1) == 'S'& tline(2)== 'H'
                        disp(tline(15:21));
                        elev = str2num(tline(15:21));
                        elev = elev + delta;
                        elev = num2str(elev);
                        tline(15:21)= elev(1:7);
                        disp(tline);
                        %             fprintf(tline,'%100s', fid(count,:));
                    end
                end
            end
        end
    end
end

Answer 1

[Kelly Kearney]

To replace a specific bit of text in a text file, you're going to have to rewrite the entire file. Assuming the files above are representative of your real files, this shouldn't be too time-consuming, because the files are small and easily fit in memory. So I recommend just reading in the entire file at once rather than looping over lines. Using either startsWith (in newer versions of Matlab) or strncmp (in older ones) will be a bit more efficient than your method of comparing individual characters.

The other key to replacing the text with your newly-calculated number is to make sure the new value has the same number of characters as the old one; you can make sure of that by using sprintf instead of num2str when converting your new value into a string.

% Read entire file as character array

txt = fileread(file);

% Split into lines

txt = regexp(txt, '\r\n', 'split');

% Find lines starting with certain letters

istarget = startsWith(txt, 'SO') | ...
           startsWith(txt, 'R') | ...
           startsWith(txt, 'TS') | ...
           startsWith(txt, 'SH');

% Replace value in columns 15-21 with new value, properly formatted

delta = 2.335;
for ii = find(istarget)
    elev = str2num(txt{ii}(15:21));
    txt{ii}(15:21) = sprintf('%7.3f', elev+delta);
end

% Write to new file

fid = fopen('newfile.txt', 'wt');
fprintf(fid, '%s\n', txt{:});
fclose(fid);

[Daniel Price]

Thank you, Kelly, for your excellent answer and additional guidance.

[Stephen23]

+1 clear advice

[Thiago de Aquino Costa Sousa]

Dear community,

I have many txt files that uses a tab delimiter. I have to change the content between the 17th and 18th tab delimiter of all my files. This is one the data inside the file:

_______

Data file generated on Tue Aug 30 12:40:29 2022

A B C D E F G H I J K L M N O P Q R S T U

1 2.00000 2.00000 0.00000 1.00000 28.00000 64.00000 88.00000 1.20000 275.00000 50.00000 30.00000 165.00000 850.00000 500.00000 0.00000 10.00000 10.00000 15.00000 0,0,0,0, 0,0,0,0, false

_______

So I have to change the value for the variable Q (in this case 10.0000) for a specific value. However, some of my files have a different pattern for numbers, instead of 10.00000 it is only 10, what makes the counting characters unfeasible. Furthermore, after some delimiters there are spaces. Then, I decided to count the tab delimeters, and change the content between the 17th and 18th delimiter to solve my problem, that I don't know if it will work. This is my code so far...

ThemeCopy

A = regexp(fileread(myfile), '\n', 'split'); %upload my file to a cell array

B = strfind(A{4}, sprintf('\t')); %takes the cell 4 in the array and finds where I have tab delimiters

C = B{4}(17)+1:B{4}(18)-1; %select the content between the 17th and 18th tab delimiters

Please, can somebody help me???