use multiple hotwords for Alexa/Assistant, fixed crash on runtime

README.md

@@ -99,11 +99,13 @@ To change the hotwords (currently *Alexa* and *Google*), either change both thes
 
 or `/etc/opt/AlexaPi/config.yaml` and `/opt/AlexaPi/src/keyphrase.list` **after** setup.
 
+In the `config.yaml`, you may use a single String or an Array of Strings for your **phrase_assistant** (only). Every string there will trigger Assistant, everything else in your `keyphrase.list` will trigger Alexa.
+
 In the `keyphrase.list`, you can also tweak the sensitivity of the hotword recognition. From the [CMUSphinx Wiki](https://cmusphinx.github.io/wiki/faq/#q-how-to-implement-hot-word-listening):
 > Threshold must be tuned for every keyphrase on a test data to get the right balance missed detections and false alarms. You can try values like 1e-5 to 1e-50.
 > For the best accuracy it is better to have keyphrase with 3-4 syllables. Too short phrases are easily confused.
 
-You can also combine wakewords, e.g. "ok google". Also make sure that your new hotwords are included in the language model. Check the following directory for a file with `.dict` or `.dic` extension and add your hotwords if not already there: `/usr/local/lib/python2.7/dist-packages/pocketsphinx/model/`
+You can also combine wakewords, e.g. "ok google" and/or "hey google". Also make sure that your new hotwords are included in the language model. Check the following directory for a file with `.dict` or `.dic` extension and add your hotwords if not already there: `/usr/local/lib/python2.7/dist-packages/pocketsphinx/model/`
 
 
 ### Change Hotword language

src/alexapi/triggers/pocketsphinxtrigger.py

@@ -92,7 +92,10 @@ def thread(self):
 
 			if triggered:
 				### Assistant Starts Here
-				voice_command = self._decoder.hyp().hypstr
+				try:
+					voice_command = self._decoder.hyp().hypstr
+				except:
+					voice_command = ""
 				self._trigger_callback(self, voice_command)
 				###
 

src/config.template.yaml

@@ -90,9 +90,15 @@ triggers:
     enabled: true
     voice_confirm: true
 
+    # add your Hotwords to ./keyphrase.list too (in lowercase)!
+    ## 'phrase' string only
     phrase: "alexa"
+    ## 'phrase_assistant' can be a string or string[] (lowercase)
+    ## all howtwords here will trigger Assistant, everything else in your ./keyphrase.lst will trigger Alexa
+    ## phrase_assistant = ["ok google", "hey google"]
     phrase_assistant: "google"
     threshold: 1e-10 # unused, change in ./keyphrase.list
+
     language: "en-us"
     dictionary: "cmudict-en-us.dict"
 

src/main.py

@@ -337,11 +337,14 @@ def alexa_speech_recognizer_generate_data(audio, boundary):
 ### AssistantPi
 def assistant_handler(voice_command):
     # clean voice_command
-    voice_command = voice_command.rstrip()
+    if voice_command is None or voice_command is "":
+        return False
+    else:
+        voice_command = voice_command.rstrip()
     logger.debug('Pocketsphinx triggered with hotword: **' + voice_command + '**')
     # compare to phrase_assistant from config
     voice_command_assistant = config['triggers']['pocketsphinx']['phrase_assistant']
-    if voice_command == voice_command_assistant:
+    if voice_command in voice_command_assistant:
         global p
         if p is not None:
             # SDK is ready, start recording