newlearner has asked for the wisdom of the Perl Monks concerning the following question:

Hi Monks,

I have a directory structure like shown below

------------------------------------------------- 
Directory: /proj/newdata/
 Files: alpha.seq, beta.pm, gamma.txt
 Sub Directories: dog, cat

  Sub-Directory:dog
   Files: batman0.seq, batman1.txt
   Sub Directories: D1, D2

     Sub-Directory: D1
      Files: joker0.seq, joker1.pm
      Sub Directories: C0
         
         Sub-Directory: C0
         Files: file0.seq, file1.pm, file.txt

  Sub-Directory:cat
   Files: man0.seq, man1.txt
   Sub Directories: D3, D4

     Sub-Directory: D3
      Files: jok0.seq, jok1.pm
      Sub Directories: C1
         
         Sub-Directory: C1
         Files: f0.seq, f1.pm, f.txt

[download]
---------------------------------------------------

Now I want a perl script which can take the input directory name '/proj/newdata' and find all the files that end only with .seq and .pm and print them according to their directory structure like this:

RESULT: ------------------------------------------------- 

Group: /proj/newdata {

  File: alpha.seq (/proj/newdata/alpha.seq)
  File: beta.pm (/proj/newdata/beta.pm)
  Group: dog {
    
    File: batman0.seq (/proj/newdata/dog/batman0.seq)
    Group: D1 {
      File: joker0.seq (/proj/newdata/dog/D1/joker0.seq)
      File: joker1.pm (/proj/newdata/dog/D1/joker1.pm)

      Group: C0 {
       File: file0.seq (/proj/newdata/dog/D1/C0/file0.seq)
       File: file11.pm (/proj/newdata/dog/D1/C0/file1.pm)
      } #Group C0
     } # Group D1
    } # Group Dog

  Group: cat {
    
    File: man0.seq (/proj/newdata/cat/man0.seq)
  
   Group: D3 {
      File: jok0.seq (/proj/newdata/cat/D3/jok0.seq)
      File: jok1.pm (/proj/newdata/cat/D3/jok1.pm)

      Group: C1 {
       File: f0.seq (/proj/newdata/dog/D3/C1/file0.seq)
       File: f1.pm (/proj/newdata/dog/D3/C1/file1.pm)

      } #Group C1
     } # Group D3
    } # Group Cat

} # Group /proj/newdata

[download]

-------------------------------------------------------

The words after "#" are comments and are not required.

Any help would be appreciated. I used File::Find and some other ways but could not crack the correct directory structuring printing and it's been some time I am working on this.

Thankyou

Replies are listed 'Best First'.
Re: Traverse through directory structure and print in required format
by GrandFather (Saint) on Mar 14, 2011 at 23:03 UTC

    So show us the code you have tried and tell us where the problem is. File::Find is likely to be an appropriate part of the puzzle so show us the code you tried using it.

    True laziness is hard work
[reply]
Re: Traverse through directory structure and print in required format
by Anonymous Monk on Mar 14, 2011 at 22:58 UTC
    printing and it's been some time I am working on this.

    Then you should show your effort

[reply]
Re: Traverse through directory structure and print in required format
by ikegami (Patriarch) on Mar 14, 2011 at 23:05 UTC 
    use File::Basename qw( basename );

sub visit {
   my ($dir, $level) = @_;

   $level ||= 0;
   my $group_indent = ' ' x (2*$level);

   ++$level;
   my $child_indent = ' ' x (2*$level);

   printf($group_indent."Group: %s {\n", $level ? basename($dir) : $di
+r);

   my @file_names = do {
      opendir(my $dh, $dir);
      grep !/^\.\.?\z/, readdir($dh)
   };

   for my $fn (@file_names) {
      my $qfn = "$dir/$fn";
      if (-d $qfn) {
         visit($qfn, $level);
      } else {
         print($child_indent."File: $fn ($qfn)\n");
      }
   }

   print($group_indent."}\n");
}

visit('/proj/newdata');

    [download]

    Possible improvements:

    • Error checking for opendir.
    • Error checking for -d (although it's not likely to return an error if you were just able to read the directory).
    • Make "$dir/$fn" portable. (It will fail if you give "C:" as the initial directory on a Windows machine if the current directory for drive C: isn't the root directory.)

    Update: Was missing top level Group. Fixed.

[reply]
[d/l]
[select]
      This includes all the possible improvements I mentioned: 
      use Path::Class qw( dir );

sub visit {
   my ($dir, $level) = @_;

   $level ||= 0;
   my $group_indent = ' ' x (2*$level);

   ++$level;
   my $child_indent = ' ' x (2*$level);

   printf($group_indent."Group: %s {\n", $level ? $dir->basename : $di
+r);

   for my $file (dir($dir)->children()) {
      if ($file->is_dir) {
         visit($file, $level);
      } else {
         printf($child_indent."File: %s (%s)\n", $file->basename, $fil
+e);
      }
   }

   print($group_indent."}\n");
}

visit('/proj/newdata');

      [download]

      Update: Was missing top level Group. Fixed.

[reply]
[d/l]

Back to Seekers of Perl Wisdom