Add additional Sphinx engine functionality and tests

- Add wake up/sleep functionality with options in the default
  config module.
- Add start asleep config option (default is on).
- Add set_keyphrase & unset_keyphrase engine methods for adding
  additional key phrases (issue #3).
- Extract some code from recognise_forever method into a new
  process_buffer method that can be used for streaming audio from
  other sources (issue #6).

dragonfly/engines/backend_sphinx/config.py

@@ -1,3 +1,5 @@
+# This Python file uses the following encoding: utf-8
+
 """
 Default CMU Pocket Sphinx engine class configuration module
 """
@@ -8,9 +10,19 @@
 
 # Configuration for the Pocket Sphinx decoder.
 DECODER_CONFIG = DefaultConfig()
+
 # Silence the decoder output by default.
 DECODER_CONFIG.set_string("-logfn", os.devnull)
 
+# Configuration for wake/sleep phrases
+# Note that the following CMU Sphinx tutorial has some advice on keyword thresholds
+# values: https://cmusphinx.github.io/wiki/tutoriallm/#keyword-lists
+START_ASLEEP = True
+WAKE_PHRASE = "wake up"
+WAKE_PHRASE_THRESHOLD = 1e-20
+SLEEP_PHRASE = "go to sleep"
+SLEEP_PHRASE_THRESHOLD = 1e-40
+
 # Keyword arguments given to the PyAudio.open method for opening a stream from a
 # microphone. PyAudio streams are used by the engine to recognise speech from audio.
 # You may wish to change the values to be optimal for the models you are using;