Undefined function 'regexp' for input arguments of type 'struct'

I want to extract keywords(which the word start with import.XXXX) from multiple files under annotation folder. I run my code like this, Y I get the error? Anybody can help?

r = dir(fullfile('C:\Users\ASUS\Desktop\AngryBirdTest\android\support\annotation', '*.java'));

for i=1:length(r)
    p = regexpi(FileList,'import.\w*.\w*.\w*','match');
    disp(p(i));
end

[Walter Roberson]

Are those things stored inside the files or are they part of the file names?

Answer 1

[Walter Roberson]

My guess:

projectdir = 'C:\Users\ASUS\Desktop\AngryBirdTest\android\support\annotation';

r = dir(fullfile(projectdir, '*.java'));
filenames = {r.name};
FileList = fullfile(projectdir, filenames);

for K = 1:length(FileList)
  thisfile = FileList{K};
  fprintf('\nExamining file "%s"\n', thisfile);     %if desired
  filecontent = fileread(thisfile);
  p = regexpi(filecontent, 'import.\w*.\w*.\w*', 'match');
  disp(p);
end

[ai ping Ng]

r = dir(fullfile('C:\Users\ASUS\Desktop\AngryBirdTest\API_ClassFileTest/', '*.java'));

for i=1:length(r)

     p = regexp(fileread(strcat('C:\Users\ASUS\Desktop\AngryBirdTest\API_ClassFileTest/',r(i).name)), 'import android.\w*.\w*.\w*','match');
     disp(p)

end

This is the final code I have figured out that day, thanks for your answer.I will use yr answer to learn it more.

[ai ping Ng]

But then why my workspace showing that my p have 0X0 cell only? How to show it in my workspace?

[Walter Roberson]

I need a sample file to test with.

[ai ping Ng]

I finally found out is because I didn't save it in matrix. I can output my p now after I save it in matrix form. Can I know how to output as a big feature vector like 000010000001.....01 like that? Mean if that feature exists, it shows as 1, if not it show it as 0. I use this "features3 = ~cellfun(@isempty,regexp(no_duplicates,r2));" but then it doesn't work. Below are my sample file and sample code. Hope you can help me, truly appreciated.

   r = dir(fullfile('C:\Users\ASUS\Desktop\AngryBirdTest\API_ClassFileTest/', '*.java'));
  r2 = regexp(fileread('Dataset3_API.txt'), '\r?\n', 'split');
  saveValueP = [];

for i=1:length(r)

     p = regexp(fileread(strcat('C:\Users\ASUS\Desktop\AngryBirdTest\API_ClassFileTest/',r(i).name)), 'import android.\w*.\w*.\w*','match');
     saveValueP = [saveValueP p];
     no_duplicates = unique(saveValueP);
     features3 = ~cellfun(@isempty,regexp(no_duplicates,r2));

end

 disp(features3)

[Walter Roberson]

As I wrote before:

If you have a cell array of strings to test, then you need to define whether you want one feature vector per cell member, or if you want the feature vector to reflect whether the substrings occur in any of the cell members.

[ai ping Ng]

Thanks. Get it. I change into a string, it works.

r = dir(fullfile('C:\Users\ASUS\Desktop\FYP\M1\ClassFileTest/', '*.java'));
r2 = regexp(fileread('Dataset3_API.txt'), '\r?\n', 'split');
saveValueP = [];

for i=1:length(r)

     p = regexp(fileread(strcat('C:\Users\ASUS\Desktop\FYP\M1\ClassFileTest/',r(i).name)), 'import android.\w*.\w*.\w*','match');
     saveValueP = [saveValueP p];
     no_duplicates = unique(saveValueP); %remove duplicated features

end

str = strjoin(no_duplicates);
features3 = ~cellfun(@isempty,regexp(str,r2,'match'));
disp(no_duplicates)

I still meet a question after that. Let say now I am examining the Class file which under M1 folder >> 'C:\Users\ASUS\Desktop\FYP\M1\ClassFileTest/' How if I have 'M1 folder' 'M2 folder' and 'M3 folder' ... 'Mx folder' to undergo? Is that a way to make it automatically examine and then can produce all of my folders' feature vectors?

[Walter Roberson]

https://www.mathworks.com/matlabcentral/answers/57446-faq-how-can-i-process-a-sequence-of-files

Answer 2

[Image Analyst]

You didn't show us how you got FileList. Evidently it thinks it's a structure.