Add leading decimal places to numbers in text file

I have data in a text file and would like to change some numbers while keeping the rest of the line the same.

For instance I have:

Name,proj2;
Edition,59;
Start;
Variable,1;
    DataReg,0.00,0.00;
    Area,520.00;
300,221.7467,424.1668,801.0146;
390,117.4175,507.8583,29.2203;
    Area,530.00;
300,488.6090,963.0885,488.8977;
390,578.5251,546.8057,624.0601;
    Datareg,00.00,30.00;
    Area,520.00;
300,367.4366,913.2868,335.3568;
390,987.9820,796.1839,679.7280;
    Area,530.00;
300,106.7619,715.0371,698.7458;
390,653.7573,903.7206,197.8098;
    DataReg,00.00,160.00;
    Area,520.00;
300,291.9841,167.1684,489.6876;
390,431.6512,106.2163,339.4934;
    Area,530.00;
300,522.6770,5147.8709,3201.4549;
390,7137.8581,9142.7370,7101.0988;
    DataReg,30.00,00.00;
    Area,520.00;
300,2121.7467,4214.1668,8201.0146;
390,1217.4175,5107.8583,219.2203;
    Area,530.00;
300,4188.6090,9163.0885,4288.8977;
390,5178.5251,5146.8057,6224.0601;
    Datareg,30.00,30.00;
    Area,520.00;
300,3167.4366,9113.2868,3135.3568;
390,9287.9820,7296.1839,6279.7280;
    Area,530.00;
300,1106.7619,1715.0371,1698.7458;
390,1653.7573,1903.7206,1197.8098;
    DataReg,30.00,160.00;
    Area,520.00;
300,2921.9841,1672.1684,2489.6876;
390,4321.6512,1206.2163,2339.4934;
    Area,530.00;
300,522.6770,57.8709,3021.4549;
390,7327.8581,9422.7370,7021.0988;
    DataReg,160.00,0.00;
    Area,520.00;
300,316.4366,911.2868,313.3568;
390,92.9820,729.1839,627.7280;
    Area,530.00;
300,110.7619,171.0371,169.7458;
390,165.7573,190.7206,119.8098;
    DataReg,160.00,30.00;
    Area,520.00;
300,292.9841,167.1684,248.6876;
390,432.6512,120.2163,233.4934;
    Area,530.00;
300,52.6770,5.8709,302.4549;
390,732.8581,942.7370,702.0988;
DataReg,160.00,160.00;
    Area,520.00;
300,7292.9841,7167.1684,7248.6876;
390,7432.6512,7120.2163,7233.4934;
    Area,530.00;
300,752.6770,75.8709,7302.4549;
390,7732.8581,7942.7370,7702.0988;

(Sorry it's so long, just wanted to paint a vivid picture)

Now, what I want to do is change the numbers in lines containing "DataReg" into the same numbers except with three leading spaces in front of decimal while still containing the two trailing decimal places. For example, I want to change this:

    DataReg,0.00,0.00;
    Datareg,0.00,30.00;
    DataReg,0.00,160.00;
    DataReg,30.00,0.00;
    Datareg,30.00,30.00;
    DataReg,30.00,160.00;
    DataReg,160.00,0.00;
    DataReg,160.00,30.00;
    DataReg,160.00,160.00;

Into this:

    DataReg,000.00,000.00;
    Datareg,000.00,030.00;
    DataReg,000.00,160.00;
    DataReg,030.00,000.00;
    Datareg,030.00,030.00;
    DataReg,030.00,160.00;
    DataReg,160.00,000.00;
    DataReg,160.00,030.00;
    DataReg,160.00,160.00;

Notice how some values (i.e.,160.00) already have three leading places and hence don't need change. keep in mind this will be for a large .txt file and other DataReg numbers such as 170.00, 180.00, 190.00 will exist, but nothing above 999.00 exists. Lots of comments please!

Answer 1

[dpb]

Air code--caution!!!

c=readcell('YourTextFile.ext');         % bring file in as cellstr() array
isDR=startsWith(c,'DataReg');           % logical addressing array of wanted lines
fmt='DataReg,%07.3f,%07.3f;';           % desired output format
c(isDR)=compose(fmt,str2double(split(extractAfter(c(isDR),'DataReg,'),',')));   % rewrite as wanted
writecell(c,'YourNewTextFile.ext','QuoteStrings',0)  % put in new file (can overwrite original once sure is as wanted)

In short, don't try to fix up based on what is there by counting string lengths or such, just write out in the desired format the values that are there.

ADDENDUM:

Above patched to incorporate fixes outlined in Comments...the doc in R2019b is wrong on default behavior of quoted strings in "writecell'; and the release OP is using is same default behavior; whether the doc matches now or not, I don't know. Worthy of reporting the documentation error if doesn't.

[Timbo]

I'm getting an error

"Error using writecell (line 158)

Unsupported type 'char'. Use writematrix instead."

[Timbo]

writematrix doesn't fix either...

[dpb]

No, see the above fix -- as said, it was "air code"; typos not unheard of.

Missed the output variable in

writecell(c,'YourNewTextFile.ext') 

in the original.

[Timbo]

Okay so I have it working now, but one newfound issue is that every line begins and ends with quotation marks. For example, a snippet of the data now looks like this:

"DataReg,160.00,030.00"

"Area,520.00"

"300,292.9841,167.1684,248.6876"

I am now interested in knowing how to get rid of these quotation marks. Similarly, I still need each line to end with a semicolon. In other words, I want the data to look like this:

DataReg,160.00,030.00;

Area,520.00;

300,292.9841,167.1684,248.6876;

[dpb]

Again, that's air code (which it is/was because you didn't attach a sample file so didn't have easy way to test) -- "help us help you"

It appears that contrary to the doc writecell uses 'QuoteStrings',1 as default instead of as false.

To add the trailing semicolon, fix up the format string

fmt='DataReg,%07.3f,%07.3f;';  

It's not hard to figure this stuff out, read the doc and try some things out at command line...be much faster than waiting on us in future. (Now I will grant that thinking of doing in memory this way is a result of more experience than you undoubtedly have; hence the illustration of the concept. Once see that, however, then debugging the little nits to clean things up is an excellent mechanism for learning...and also illustrates why to not overwrite the original file(s) the first time! :) )

>> fmt='DataReg,%07.3f,%07.3f;';
>> compose(fmt,str2double(split(extractAfter(lower(c(1:3)),'datareg,'),',')))
ans =
  3×1 cell array
    {'DataReg,000.000,000.000;'}
    {'DataReg,000.000,030.000;'}
    {'DataReg,000.000,160.000;'}
>> writecell(ans,'tryit.csv')
>> type tryit.csv

"DataReg,000.000,000.000;"
"DataReg,000.000,030.000;"
"DataReg,000.000,160.000;"
>> writecell(ans,'tryit.csv',"QuoteStrings",0)
>> type tryit.csv

DataReg,000.000,000.000;
DataReg,000.000,030.000;
DataReg,000.000,160.000;
>> 

NB: the lower() in the above is owing to the data cut'n pasted from your original posting is not all capitalized regularly -- hence the match for the split string didn't always work so I patched it to convert to lower case and match on that string instead. I presumed in the answer that was a fignewton of the posting and not real in the file.

[Timbo]

The script works great, but now there is an unrelated issue I am having. For some odd reason my version of matlab doesn't seem to take the "readcell" function. I tried using "importdata" instead, but when I use it the .txt file displays NaN for the second DataReg #'s.

For example when I use importdata I get,

"DataReg,000.00, NaN;"

Is there anywhere I can find the readcell function to write myself? Or is there any way to get importdata to work? I've attached a snippet of the text file for convenience

[dpb]

readcell and friends is pretty recent, introduced in R2019a. If you can't update, the venerable textread or textscan will work. While deprecated, I still use textread on occasion as don't need the file handle as does textscan You just need a format string to input a string and only read by line ignoring whitespace. textread internally handles the return as a cell array...

There even is an example for textread -- they show reading an m-file

file=textread('fft.m', '%s', 'delimiter', '\n', 'whitespace', '');

Just swap out file name and return variable name to match.

[dpb]

Oh. You're going to have same problem with writecell, too, then, it was also not introduced until R2019a.

There's no other way then but to revert to low-level i/o -- this has been an Achilles' heel problem for as long as have had cell arrays...

fid=fopen('YourFile.ext','w');
for r=1:size(c,1)
  fprintf(fid,'%s\n',c{r});
end
fclose(fid);

[Timbo]

The work-arounds worked! Managed to make the changes I needed. Thank you kindly!