Parse addresses with 'County Road xx' names

[rptetzloff]

Training for multi-letter County Road names, e.g. "County Road XX."

"County Road D" was parsing correctly:

>>> address1 = usaddress.tag("1234 COUNTY ROAD D, FRANKLIN, WI 54567")
>>> pprint(address1)
(OrderedDict([('AddressNumber', '1234'),
              ('StreetNamePreType', 'COUNTY ROAD'),
              ('StreetName', 'D'),
              ('PlaceName', 'FRANKLIN'),
              ('StateName', 'WI'),
              ('ZipCode', '54567')]),
 'Street Address')

"County Road DD" was not parsing correctly:

>>> address2 = usaddress.tag("1234 COUNTY ROAD DD, FRANKLIN, WI 54567")
>>> pprint(address2)
(OrderedDict([('AddressNumber', '1234'),
              ('StreetName', 'COUNTY ROAD'),
              ('StreetNamePostType', 'DD'),
              ('PlaceName', 'FRANKLIN'),
              ('StateName', 'WI'),
              ('ZipCode', '54567')]),
 'Street Address')

Training xml located at:

usaddress/training/county_road_xx.xml

Testing xml located at:

usaddress/measure_performance/test_data/county_road_xx.xml

Each contained 6 addresses.

All tests passed.

> nosetests . 
<snip>
----------------------------------------------------------------------
Ran 4929 tests in 1.329s

OK