fix #1 - UTF-8 decode problems when using en_US.

Apparently the en_US dictionary in ubuntu cannot be decoded with utf-8,
but works fine with iso-8859-1. As a workaround this adds another
configuration and command line option: encoding. If left empty it
autodetects from the environment, otherwise it should be an encoding
definition that python should use to read the dictionary.

bin/pwgen

@@ -22,6 +22,9 @@ def main():
     parser.add_argument(
             '--myspell-dir', '-i',
             help='Directory containing myspell dictionaries')
+    parser.add_argument(
+            '--encoding', '-e',
+            help="Character encoding of the directory")
 
     parser.add_argument(
             '--lang', '-l',