Easiest way is probably to use File::Slurp.

Here's an untested example:

#!env perl

use strict;
use warnings;

use File::Slurp qw/read_dir/;

my $Dir_Prefix = '../../dir1/dir2';
my $File_Path = 'dir3/filename';


MAIN: {
    my @files = read_dir($Dir_Prefix);
    foreach my $file (@files) {
        -d $file or next;
        if (-e "$Dir_Prefix/$file/$File_Path") {
            print "$file\n";
        }
    }
}
[download]

Update:Updated from a File::Find example.

That's Informative and usefull. Thanks

Hi arunagiri,
"..How can I get only the folder names inside dir2? Is there any easy way to get it?.."

You could use basename subroutine from the module File::Basename.

Update:

    basename("/foo/bar"); # this return "bar"
   #OR
    for my $ele (glob(...)){
      print basename($ele), $/;
    }
[download]

> Is there any easy way to get it?

No, not what you intend with "easy"!

Use a regex or split or substr to isolate what you wanted from the full path

edit

  DB<106> $head="/tmp/";$tail="/foo/file" 
 => "/foo/file"

  DB<107> $path= <$head*$tail>
 => "/tmp/bla/foo/file"

  DB<108> ($middle) = $path =~ /^\Q$head\E(.*)\Q$tail\E$/
 => "bla"
[download]

Easy enough! (?)

update

FWIW a one liner:

  DB<109> ($middle) = <$head*$tail> =~ /^\Q$head\E(.*)\Q$tail\E$/
 => "bla"
[download]

update

update

Not sure if I misunderstood your question or if you phrasing is just fuzzy. What are folder names inside dir2 ?

Sorry if I didn't mention it clearly. There are many folders inside dir2. Eg: A1, B1, C1 etc: are the folder inside dir2. only A1 and B1 folders have a file names "hello.txt" Hence, I want only the names of A1 and B1. If I do glob from some other folder like glob ("../../dir1/dir2/*/hello.txt") How can I get the only folders A1 and B1. I am going through the various suggestions that you guys gave. I will try them out. Thanks

> Sorry if I didn't mention it clearly.

Sorry still not clear! (And unfortunately you didn't update your posts with code-tags either).

2 possibilities:

• If you want the whole sub-path which fills into the wild-card, then the demonstrated code is already perfect. Try it out.

• If you're only interested in the highest level folder do a @folders = split '/', $path for every path found by your glob and just take $folder[$index], where index is the number of folders in the "head" part.

  E.g 4 for "../../dir1/dir2/" will give you the next directory after "dir2".

HTH! =)

Cheers Rolf

( addicted to the Perl Programming Language)

use feature 'say';
use Cwd;

my $origdir = cwd();
chdir '../../dir1/dir2';
for my $element( glob '*/dir3/filename' ) {
   say $element; 
}
[download]

Alternately, can you do ...?

my $suffix = '/dir3/filename';
for my $element( glob '*' ) {
   say $element . $suffix; 
}
[download]

So you are looking for the names of directories in dir2 that have subdirectories called dir3 and contain files called filename? Here it is:

use strict; 
use warnings;
foreach my $element (glob "dir1/dir2/*/dir3/filename"){ 
    print "$element:  "; 
    $element =~ m!dir1/dir2/(.*)/dir3/filename!;
    print "$1\n";
}
[download]

Comment: I left the ../../ out as it would not fit my sample dir structure I had to create. Without testing it, I would still say that you do NOT need to change the regex to make it work for your situation.

Works like a charm. Thanks.