Erm, I see no "weird box characters" in that (which could be an artifact of how you copied the text into your post) so I can't really tell how to match between them. At any rate, many terminal emulators will show control characters or otherwise unprintable characters as boxen. Some ideas to consider:

• Make sure you're using the correct encoding for the text in question (perhaps it's UTF-8 or the like)
• Maybe they're control characters (I've seen several formats which use \x1C "FS" as a field separator)
• Something like the *NIX od utility or a hex editor can help track down exactly what octets you've got