Yes, it's safe to do that, if what you want is what it happens to do.

What you get with a foreach loop is that it does the first element in the array, then the second, then the third, and so on, until it runs out. If you change the elements as you go, it doesn't bother Perl; on the second iteration it gives you whatever the second element is at that moment, regardless of what it was when you started the loop. So for example:

    @a = (0,1,2);
    foreach $element (@a) {
      print $element, "\n";
      $a[1] = 'foo';
    }
[download]

This prints out 0, foo, 2.

Similarly, the loop ends only when Perl runs out of elements, and if you add elements in the middle of the loop, Perl will iterate over the new elements also, until it reaches the current end of the array. This code is an infinite loop:

	@a = (1,2,3);
	foreach $element (@a) {
	  print $element, "\n";
	  push @a, $element+3;
	} 
	  

This prints 1 2 3 4 5 6 7... and so on forever, because it keeps extending the array, and Perl never gets to the end.

You want to use the File::Find module that comes with Perl. With File::Find you tell it what directory you want to search, and you supply a wanted subroutine that it invokes on every file that it finds. Inside of wanted, you get to do whetever you want, or you can just return immediately if you're not interested in a particular file. Your wanted might look like this:

	sub wanted {
	  return unless /\.mp3$/ || /\.pl$/;
	  # do something with the .mp3 or .pl file
	}

$location = `locate $name`;

It's a hell of a lot faster than traversing the file tree manually as it uses an updated database of file locations rather than tromping all over the file system. (remember to update the file database with 'updatedb' as root before doing intensive searches.

If on WinWhatever, ignore the above.

From the File::Recurse docs:

The File::Recurse module is designed for performing an operation on a tree of directories and files. The basic usage is simmilar to the find.pl library. Once one uses the File::Recurse module, you need only call the recurse function in order to perform recursive directory operations.

The function takes two parameters a function reference and a directory. The function referenced by the first parameter should expect to take one parameter: the full path to the file currently being operated on. This function is called once for every file and directory under the directory named by the second parameter.

For example:

recurse(\&func, "/");

would start at the top level of the filesystem and call "func" for every file and directory found (not including "/").

Perl allows a second form of calling this function which can be useful for situations where you want to do something simple in the function. In these cases, you can define an anonymous function by using braces like so:

recurse {print $_[0]} "/";

This would print every file and directory in the filesystem. However, as an added convenience y

sub parse_file ($) {
  #Do something with the file here.
}

sub traverse (@) {
  my @files = @_; 
  local *DIRH;

  #If @files is empty, why bother?
  return unless @files;

  foreach my $file (@files) {
    #Skip unreadable files.
    next unless -r $file;
    if (-d $file) {
      #If it's a directory ...
      chomp(my $cwd = system('/bin/pwd')); #save current directory, ..
+.
      chdir($file) || next;                #change to the new director
+y, ...
      opendir(DIRH,$file) || next;         #open it...
      traverse((readdir(DIRH));            #and recursively parse it.
      closedir(DIRH);                      #Close the directory.
      chdir($cwd);                         #Change directories back to
+ where we were.
    } else {
      #If it's not a directory, then parse it.
      parse_file($file);                    #This is where the actual 
+work is done.
    }
  }
}

traverse(@ARGV);
[download]