Fix BOM removal when a classlist file is read.

This was broken when the UTF-8 encoding method was changed in #1831.
Since the file contents are now UTF-8 decoded, a different regular
expression is needed to match a BOM.

This replaces #2305.  That pull request will be closed.

lib/WeBWorK/File/Classlist.pm

@@ -54,7 +54,7 @@ sub parse_classlist($) {
 
 		# Remove a byte order mark from the beginning of the file if present.  Excel inserts this on some systems, and
 		# the presence of this multibyte character causes a classlist import to fail.
-		s/^\xEF\xBB\xBF//;
+		s/^\x{FEFF}//;
 
 		s/^\s*//;
 		s/\s*$//;