I know this is not a perl based solution but in general, when there is "unprintable character weirdness" in any text file or program script I am working with, I find the unix utility "od" to be very helpful. It provides an octal dump (or decimal, or hexadecimal in some versions) of each and every byte of the argument file. You can use it like this:

$ od -c <filename>
[download]

which will produce output similar to:

$ od -c innkeeper.pl

00000   #   !   /   u   s   r   /   l   o   c   a   l   /   b   i   n
00020   /   p   e   r   l  \n  \n   u   s   e       F   i   l   e   :
00040   :   s   t   a   t   ;  \n   u   s   e       F   r   e   e   z
00060   e   T   h   a   w       q   w   (   f   r   e   e   z   e
00100   t   h   a   w   )   ;       #       I   m   p   o   r   t
00120   f   r   e   e   z   e   (   )       a   n   d       t   h   a
00140   w   (   )  \n  \n   %   l   i   s   t   =   (   )   ;  \n  \n
00160   $   A   R   G   V   [   0   ]       o   r       d   i   e
00200   "   u   s   a   g   e   :       $   0       d   a   t   a   f
00220   i   l   e   \   n   "   ;  \n  \n   o   p   e   n       (   I
00240   N   ,       s   h   i   f   t   )   ;  \n  \n   w   h   i   l
00260   e       (   $   o   b   j   =   <   I   N   >   )       {  \n
00300       c   h   o   m   p   (   $   o   b   j   )   ;  \n       (
00320   $   s   ,   $   s   o   )   =   t   h   a   w   (   $   o   b
00340   j   )   ;  \n       p   r   i   n   t       "   $   s       =
00360   >       "   ,       "   "       .       l   o   c   a   l   t
00400   i   m   e   (   $   s   o   -   >   m   t   i   m   e   )   ,
00420       "   \   n   "   ;  \n   }  \n
[download]

You can see the decimal value of unprintable characters, and some common ones are given the well-known names like \n and \t, etc. Hope this may help in the future.