Convert calibre modules to ebook_converter.

Here is the first batch of modules, which are needed for converting
several formats to LRF. Some of the logic has been change, more cleanups
will follow.

ebook_converter/spell/__init__.py

@@ -7,7 +7,7 @@ __copyright__ = '2014, Kovid Goyal <kovid at kovidgoyal.net>'
 
 from collections import namedtuple
 
-from calibre.utils.localization import canonicalize_lang
+from ebook_converter.utils.localization import canonicalize_lang
 
 DictionaryLocale = namedtuple('DictionaryLocale', 'langcode countrycode')
 
@@ -17,7 +17,7 @@ ccodes, ccodemap, country_names = None, None, None
 def get_codes():
     global ccodes, ccodemap, country_names
     if ccodes is None:
-        from calibre.utils.serialize import msgpack_loads
+        from ebook_converter.utils.serialize import msgpack_loads
         data = msgpack_loads(P('localization/iso3166.calibre_msgpack', allow_user_override=False, data=True))
         ccodes, ccodemap, country_names = data['codes'], data['three_map'], data['names']
     return ccodes, ccodemap