Adding support of all languages in Duckling 0.4.24

duckling/language.py

@@ -11,12 +11,21 @@ class Language(object):
     ESTONIAN = 'et$core'
     FRENCH = 'fr$core'
     IRISH = 'ga$core'
+    HEBREW = 'he$core'
+    CROATIAN = 'hr$core'
     INDONESIAN = 'id$core'
     ITALIAN = 'it$core'
     JAPANESE = 'ja$core'
     KOREAN = 'ko$core'
+    BIRMAN = 'my$core'
+    NORVEGIAN = 'nb$core'
+    DUTCH = 'nl$core'
+    POLISH = 'pl$core'
     PORTUGUESE = 'pt$core'
+    ROMANIAN = 'ro$core'
     RUSSIAN = 'ru$core'
+    SWEDISH = 'sv$core'
+    TURKISH = 'tr$core'
     UKRAINIAN = 'uk$core'
     VIETNAMESE = 'vi$core'
     CHINESE = 'zh$core'
@@ -30,12 +39,21 @@ class Language(object):
         ESTONIAN,
         FRENCH,
         IRISH,
+        HEBREW,
+        CROATIAN,
         INDONESIAN,
         ITALIAN,
         JAPANESE,
         KOREAN,
+        BIRMAN,
+        NORVEGIAN,
+        DUTCH,
+        POLISH,
         PORTUGUESE,
+        ROMANIAN,
         RUSSIAN,
+        SWEDISH,
+        TURKISH,
         UKRAINIAN,
         VIETNAMESE,
         CHINESE,

duckling/test/test_duckling.py

@@ -5,6 +5,8 @@
 from dateutil.tz import tzlocal
 from duckling import Duckling, Dim, Language
 
+all_langauges = Language.SUPPORTED_LANGUAGES
+all_langauges.remove(Language.BIRMAN) # does not support arabic digits
 
 @pytest.fixture
 def test_input():
@@ -387,3 +389,16 @@ def test_multiple_dims(duckling_loaded):
     assert len(result) == 2
     assert result[0][u'value']['value'] == float(test_input)
     assert result[1][u'value']['value'] == float(test_input)
+
+@pytest.mark.parametrize("language", all_langauges)
+def test_supported_languages_with_number_dim(duckling_loaded, language):
+    test_input = '4242'
+    result = duckling_loaded.parse(test_input, dim_filter=Dim.NUMBER, language=language)
+
+    print(result)
+
+    assert len(result) >= 1
+    assert result[0][u'dim'] == Dim.NUMBER
+
+    result_val = result[0][u'value'][u'value']
+    assert str(int(result_val)) == test_input