add voc_list helper function (#54)

* add `voc_list` helper function

* ovos public method

ovos_workshop/skills/base.py

@@ -21,6 +21,7 @@
 from dataclasses import dataclass
 from hashlib import md5
 from inspect import signature
+from typing import List
 from itertools import chain
 from os.path import join, abspath, dirname, basename, isfile
 from threading import Event
@@ -133,7 +134,7 @@ def __init__(self, name=None, bus=None, resources_dir=None,
         self.reload_skill = True  #: allow reloading (default True)
 
         self.events = EventContainer(bus)
-        self.voc_match_cache = {}
+        self._voc_cache = {}
 
         # loaded lang file resources
         self._lang_resources = {}
@@ -1047,6 +1048,20 @@ def ask_selection(self, options, dialog='',
                 resp = match
         return resp
 
+    # method not present in mycroft-core
+    def _voc_list(self, voc_filename, lang=None) -> List[str]:
+
+        lang = lang or self.lang
+        cache_key = lang + voc_filename
+
+        if cache_key not in self._voc_cache:    
+            vocab = self._resources.load_vocabulary_file(voc_filename) or \
+                    CoreResources(lang).load_vocabulary_file(voc_filename)
+            if vocab:
+                self._voc_cache[cache_key] = list(chain(*vocab))
+
+        return self._voc_cache.get(cache_key) or []
+
     def voc_match(self, utt, voc_filename, lang=None, exact=False):
         """Determine if the given utterance contains the vocabulary provided.
 
@@ -1070,21 +1085,17 @@ def voc_match(self, utt, voc_filename, lang=None, exact=False):
             bool: True if the utterance has the given vocabulary it
         """
         match = False
-        lang = lang or self.lang
-        cache_key = lang + voc_filename
-        if cache_key not in self.voc_match_cache:
-            vocab = self._resources.load_vocabulary_file(voc_filename) or \
-                    CoreResources(lang).load_vocabulary_file(voc_filename)
-            self.voc_match_cache[cache_key] = list(chain(*vocab))
-        if utt:
+        _vocs = self._voc_list(voc_filename, lang)
+
+        if utt and _vocs:
             if exact:
                 # Check for exact match
                 match = any(i.strip() == utt
-                            for i in self.voc_match_cache[cache_key])
+                            for i in _vocs)
             else:
                 # Check for matches against complete words
                 match = any([re.match(r'.*\b' + i + r'\b.*', utt)
-                             for i in self.voc_match_cache[cache_key]])
+                             for i in _vocs])
 
         return match
 

ovos_workshop/skills/ovos.py

@@ -1,5 +1,6 @@
 import re
 import time
+from typing import List
 
 from ovos_utils.intents import IntentBuilder, Intent
 from ovos_utils.log import LOG
@@ -132,17 +133,25 @@ def voc_match(self, *args, **kwargs):
         except FileNotFoundError:
             return False
 
-    def remove_voc(self, utt, voc_filename, lang=None):
-        """ removes any entry in .voc file from the utterance """
-        lang = lang or self.lang
-        cache_key = lang + voc_filename
+    def voc_list(self, voc_filename, lang=None) -> List[str]:
+        """
+        Get vocabulary list and cache the results
+
+        Args:
+            voc_filename (str): Name of vocabulary file (e.g. 'yes' for
+                                'res/text/en-us/yes.voc')
+            lang (str): Language code, defaults to self.lang
 
-        if cache_key not in self.voc_match_cache:
-            self.voc_match(utt, voc_filename, lang)
+        Returns:
+            list: List of vocabulary found in voc_filename
+        """
+        return self._voc_list(voc_filename, lang)
 
+    def remove_voc(self, utt, voc_filename, lang=None):
+        """ removes any entry in .voc file from the utterance """
         if utt:
             # Check for matches against complete words
-            for i in self.voc_match_cache.get(cache_key) or []:
+            for i in self.voc_list(voc_filename, lang):
                 # Substitute only whole words matching the token
                 utt = re.sub(r'\b' + i + r"\b", "", utt)
 

test/unittests/skills/test_mycroft_skill.py

@@ -437,6 +437,15 @@ def test_voc_match_exact(self):
                                     exact=True))
         self.assertFalse(s.voc_match("would you please turn off the lights",
                                      "turn_off_test", exact=True))
+
+    def test_voc_list(self):
+        s = SimpleSkill1()
+        s.root_dir = abspath(dirname(__file__))
+
+        self.assertEqual(s._voc_list("turn_off_test"),
+                         ["turn off", "switch off"])
+        cache_key = s.lang+"turn_off_test"
+        self.assertIn(cache_key, s._voc_cache)
 
     def test_translate_locations(self):
         """Assert that the a translatable list can be loaded from dialog and