search multiple files same extension.

[Zuddie]

Hi All,

I have to move my mail server, up to now I've been using catch all forwarding... Now I have to set up a separate account for each of the names I have been using over the years.

The task seems hopeless, unless I can come up with a way to search all of my Eudora in boxes (and I have a lot of them) for all mail sent to *@domain.com (where * is a wild card) I have used perl before, and I know it's the right application to use for this, but I'm a bit rusty. Can anyone out there help me get started?

Zuddie

[KevinADC]

something like:

use strict;
use warnings;
@ARGV = <*.ext>;
open (OUT, '>mailboxes.txt') or die "$!";
while (<>) {
   print OUT "$1\n" if (/(\S+@domain\.com)/);
}
close OUT;

[Zuddie]

Got the following errors:

Possible unintended interpolation of @domain in string at C:\Perl\getemailnames.pl line 6.
Global symbol "@domain" requires explicit package name at C:\Perl\getemailnames.pl line 6.
Execution of C:\Perl\getemailnames.pl aborted due to compilation errors.

I changed:

print OUT "$1\n" if (/(\S+@domain\.com)/);

to:

print OUT "$1\n" if (/(\S+\@domain\.com)/);

Then it ran, and the output was as expected... Thanks

[numberwhun]

No worries. This is because the array @domain was not defined with my before first being used. Simply change the code to be like this:

use strict;
use warnings;
 
@ARGV = <*.ext>;
 
open (OUT, '>mailboxes.txt') or die "$!";
my @domain;
 
while (<>) {
print OUT "$1\n" if (/(\S+@domain\.com)/);
}
 
close OUT;
 

That should resolve that issue.

Regards,

Jeff

[KevinADC]

Actually, since he is searching for the literal string '[e-mail]' in the input files escaping the @ symbol is the correct thing to do. I should have had that in the code I posted, that was my mistake.

[numberwhun]

He he, sorry about that. I though it was an array. Wasn't thinking. All day watching two kids will do that to you.

Regards,

Jeff

[Zuddie]

Another useful modification would be to eliminate all text before the word just ahead of the @ symbol. For example if there is a space coma quote colon or equal sign before the name (blah blah:[e-mail] blah blah), it would be nice to get only [e-mail] and not the text ahead or following.

I'm thinking something along the lines of:

^(.*)([\s\,\"\;\:\=])(.*)(/(\S+\@domain\.com)/)(.*)

But I don't have the complete thought.

[KevinADC]

a simpler approach:

/([\w-]+\@domain\.com)/

The character class \w is: 'a-zA-Z0-9_' and I added a dash '-' which I think is legal for email addresses. So if you have:

joe blow:[e-mail]

it will return:

[e-mail]

There is probably a module for parsing email address out of text/files.

[Zuddie]

So the final version looks like this:

use strict;
use warnings;
@ARGV = <*.mbx>;
open (OUT, '>mailboxes.txt') or die "$!";
while (<>) {
    print OUT "$1\n" if (/([\w-]+\@domain\.com)/);
}
close OUT;

And it works... Thanks.

Yah you're probably right, there could be a module, but getting regular expressions right is so rewarding.