Re: I know this code could be better...

As petdance pointed out, your file reading routine can be simplified:
@lines = <NODE>; </CODE>
Further, you can directly insert into a hash from your split, because in a sense a hash is just a fancy list:
%known_files = map { my($k,$v)=split(/=/) } <LINKS>;

Post-processing would allow you to figure out the 'top_level' stuff:

     foreach (keys %known_files)
     {
          if(/announcements|links|news|info/i)
          {
                $top_level{$_} = $known_files{$_};
          }
     }
[download]

When you want to print an array, you just print it. You don't need to do anything special.
```
     sub middle
     {
          print @lines;
     }
[download]
```

Comment on Re: I know this code could be better... Select or Download Code