finding e-mail address which begins with 2 character and different domain names as array

hello every one; i have 3 arrays and they are declared as the following:

 dom_lists={'000' 'hotmal.com';'001' 'gmail.com';'010' 'yahoo.com'; '011' 'mail.com';'100' 'live.com';'101' 'myspace.com';'110' 'msn.com';'111' 'mynet.com'};
sub_g1={'aj';'ih';'vn';'hu';'eg';'is';'rd';'nt';'me';'ah';'zb';'en';'mm'};
sub_g2={'001';'101'; '101';'101'; '000'; '110'; '001'; '111'; '101'; '000'; '110'; '000'; '000'};
list_emialadre={'[e-mail]';'[e-mail]';...............}; 5408 emails which are based on domain names, e.g. hotmail.com have 676 email address and also gmail.com have 676....

what i want is; generating the sub_g2's meaning or equivalent strings from domain_lists vector. after that, sub_g1's data is also used for finding from the email address which begins with the sub_g1's dual characters and ends sub_2 domain_lis. so help me for solving this problem.

Answer 1

[Stephen23]

If you split the problem into parts then it is much easier to solve. Here are the data call arrays, which I altered e.g. by adding one sample email address that actually matches the second pair+domain data (otherwise the two email addresses given do not match any, so we would not have a positive result):

dom_lists = {'001' 'gmail.com'; '000' 'hotmal.com';'010' 'yahoo.com'; '011' 'mail.com';'100' 'live.com';'101' 'myspace.com';'110' 'msn.com';'111' 'mynet.com'};
sub_g1 = {'aj'; 'ih'; 'vn'; 'hu'; 'eg'; 'is'; 'rd'; 'nt'; 'me'; 'ah'; 'zb'; 'en'; 'mm'};
sub_g2 = {'001';'101'; '101';'101'; '000'; '110'; '001'; '111'; '101'; '000'; '110'; '000'; '000'};
email_list = {'[e-mail]'; '[e-mail]'; '[e-mail]'};

First convert the binary strings into numeric values with bin2dec, as this makes them easier to work with:

sub_N = bin2dec(cell2mat(sub_g2));
dom_N = bin2dec(cell2mat(dom_lists(:,1)));

Then simply match these numeric values using bsxfun, and extract the correct domain strings:

[row,~] = find(bsxfun(@eq,dom_N,sub_N.'));
dom_g2 = dom_lists(row,2);

Then create regexp regular expressions based on the character-pair and domain strings, and use these to locate the matching email addresses:

rgx = strcat('^',sub_g1,'.*@',strrep(dom_g2,'.','\.'),'$');
mtc = cellfun(@(s)regexp(email_list,s),rgx, 'UniformOutput',false);
out = ~cellfun('isempty',[mtc{:}]);

where out can be shown in the command window:

>> out
out =
   0     0     0     0     0     0     0     0     0     0     0     0     0
   0     0     0     0     0     0     0     0     0     0     0     0     0
   0     1     0     0     0     0     0     0     0     0     0     0     0

out is a logical array where each row corresponds to one of the given email addresses (only three in the sample data!) and each column corresponds to the pair+domain data of sub_g1 and sub_g2 (thus thirteen columns). From this array we can see that the third email address matches the data of the second pair+domain data, which is what was stated at the beginning, so the algorithm has successfully detected this positive test case.

[Mohamuud hassan]

thank you Stephen Cobeldick that was my objectives,but i have also one problem

    mtc = cellfun(@(s)regexp(emaillist,s),rgx1, 'UniformOutput',false);
    out = ~cellfun('isempty',[mtc{:}]);  
how i can benefit out vector's values to generate matched emial addresses and store in array with out loop;
e.g:

sub_N = bin2dec(cell2mat(sub_g2));
dom_N = bin2dec(cell2mat(dom_lists(:,1)));
[rowww,~] = find(bsxfun(@eq,dom_N,sub_N.'));
dom_g2 = dom_lists(rowww,2);
rgx1 = strcat(sub_g1,'@',dom_g2,'$');
[malisrow,malistcol]=size(emaillists);
 comlett=1;
 for mailout=1:malistcol
         for mailin=1:malisrow
         comresult=strncmpi(emaillists(mailin,mailout),rgx1(comlett,mailout),2);
         if(comresult>0) 
             k11(comlett,mailout)=emaillists(mailin,mailout)
             comlett=comlett+1;
         end
    end
end

but, this code was not running well, help be for generating founded email addresses from maillist vector N.B: the attache file is email address

[Stephen23]

I wrote and tested some fully functioning code that correctly locates and matches a list of email addresses to the provided data arrays. You changed that code by removing the core function regexp and several cellfun calls, and added a strncmpi call and several for-loops... and now it does not work?

My answer did not use any for loops, and it already searches all of the email addresses: did you read the part of the description "...each row corresponds to one of the given email addresses ... and each column corresponds to the pair+domain data..." ? See, I already explained that all email addresses are accounted for, as are the data that you want to match them with. No loops are required.

If you attach your mail-list data correctly (you will need to push both buttons: Choose file and Attach file) then we can see this working properly.

[Mohamuud hassan]

those are the mat fıle for the lısst emaıl and orıgınal fıle

[Mohamuud hassan]

i understand your explanation. but my point is how i can get the real email addresses after generating pair+domain data. that is why i used this loop or for loop. so help me again to extract the out vector's binary from original emial addresses according to those 5408 email addresses. thank you in advance.

[Stephen23]

Thank you for uploading your data. I ran my code and it works without error on your whole data list, exactly as I wrote it and without changing a single line of code.

You can simply use find on the output array out:

>> [idx,idy] = find(out)
idx =
       686
      4272
      4616
      4259
      3607
      1122
      5090
      4373
      4032
idy =
   1
   2
   3
   4
   6
   7
   8
   9
  11

Where each element of idx is the index for the list of email addresses, and the corresponding element of idy are the pair+domain list values. So we can check this by, for example by viewing the second value from idx and idy, first the pair+domain data:

>> sub_g1{idy(2)}
ans =
ih
>> dom_g2{idy(2)}
ans =
myspace.com

and then the email data:

>> email_list{idx(2)}
ans =
[e-mail]

Note that some of the pair+domain data do not match an email address!

[Mohamuud hassan]

thank you Stephen Cobeldick for your continuous support.

[Stephen23]

@abdulkarim hassan: I'm glad to help! On this forum it is also considered polite to accept answers that resolve your questions.