Add optional *spoken_form* param to NatlinkDictationContainer method

Re: #387.

This allows formatting Dragon's dictation words using the spoken
form instead.

I've also mentioned this in the documentation of the Dictation
element.

dragonfly/engines/backend_natlink/dictation.py

@@ -56,8 +56,16 @@ def __init__(self, words, methods):
         DictationContainerBase.__init__(self, words=unicode_words,
                                         methods=methods)
 
-    def format(self):
-        """ Format and return this dictation. """
+    def format(self, use_spoken=False):
+        """
+        Format and return this dictation.
+
+        Arguments:
+         - *use_spoken* (*bool*, default: *False*) --
+           whether to use the spoken form of dictated words in the result
+           instead of the written form
+
+        """
         formatter = WordFormatter()
-        formatted = formatter.format_dictation(self._words)
+        formatted = formatter.format_dictation(self._words, use_spoken)
         return self.apply_methods(formatted)

dragonfly/grammar/elements_basic.py

@@ -1129,6 +1129,12 @@ class Dictation(ElementBase):
             Dictation("formattedtext").apply(str_func)
             Dictation("snake_text").lower().replace(" ", "_")
             Dictation("camelText").camel()
+
+        If you are using Dragon/Natlink and want the spoken form of the
+        dictated words instead of the written form, this can be achieved by
+        passing ``False`` when calling the ``format()`` method of the
+        ``NatlinkDictationContainer`` object given as this element's value.
+
     """
 
     # pylint: disable=redefined-builtin