refactor

tts/text/pipeline/english_pipeline_v1.py

@@ -1,13 +1,11 @@
 from typing import List
 from tts.text.pipeline.base_pipeline import BasePipeline
 from tts.text.pipeline_components.input_parsing.parser import ArpabetSequence, GraphemeSequence, ParsedSequence, parse_grapheme_and_arpabet_sequence
-from tts.text.pipeline_components.normalization.english_normalization import EnglishNormalization
+from tts.text.pipeline_components.normalization.english_normalization import normalize_english
 from tts.text.pipeline_components.phonetization.cmudict import DEFAULT_CMUDICT_PATH, CMUDict
 from tts.text.pipeline_components.phonetization.english_phonetization import EnglishPhonetization
 from tts.text.pipeline_components.sequencing.english_symbols_v1 import SPACE_INTEGER_ENCODING, arpabet_to_sequence, symbols_to_sequence
 
-_ENGLISH_NORMALIZER = EnglishNormalization()
-
 _ENGLISH_DICTIONARY = CMUDict(DEFAULT_CMUDICT_PATH)
 _ENGLISH_PHONETIZER = EnglishPhonetization(_ENGLISH_DICTIONARY)
 
@@ -27,7 +25,7 @@ class EnglishPipelineV1(BasePipeline):
         for i in range(len(parsed_sequence)):
             subsequence = parsed_sequence[i]
             if isinstance(subsequence, GraphemeSequence):
-                normalized = _ENGLISH_NORMALIZER.normalize(subsequence.sequence)
+                normalized = normalize_english(subsequence.sequence)
                 parsed_sequence[i] = GraphemeSequence(normalized)
 
 

tts/text/pipeline_components/normalization/base_normalization.py

@@ -1,4 +0,0 @@
-class BaseNormalization:
-
-    def normalize(self, raw_input_text: str) -> str:
-        return raw_input_text

tts/text/pipeline_components/normalization/english_normalization.py

@@ -1,17 +1,14 @@
-from tts.text.pipeline_components.normalization.base_normalization import BaseNormalization
 from tts.text.pipeline_components.normalization.common_cleaners import collapse_whitespace
 from tts.text.pipeline_components.normalization.common_cleaners import convert_to_ascii
 from tts.text.pipeline_components.normalization.common_cleaners import lowercase
 
-class EnglishNormalization(BaseNormalization):
+def normalize_english(raw_input_text: str) -> str:
+    text = convert_to_ascii(raw_input_text)
+    text = lowercase(text)
 
-    def normalize(self, raw_input_text: str) -> str:
+    # TODO: Expand abbreviations
-        text = convert_to_ascii(raw_input_text)
+    # TODO: Expand numbers
-        text = lowercase(text)
 
-        # TODO: Expand abbreviations
+    text = collapse_whitespace(text)
-        # TODO: Expand numbers
 
-        text = collapse_whitespace(text)
+    return text
-
-        return text

tts/text/pipeline_components/normalization/spanish_normalization.py

@@ -1,6 +1,14 @@
-from tts.text.pipeline_components.normalization.base_normalization import BaseNormalization
+from tts.text.pipeline_components.normalization.common_cleaners import collapse_whitespace
+from tts.text.pipeline_components.normalization.common_cleaners import convert_to_ascii
+from tts.text.pipeline_components.normalization.common_cleaners import lowercase
 
-class SpanishNormalization(BaseNormalization):
+def normalize_spanish(raw_input_text: str) -> str:
+    text = convert_to_ascii(raw_input_text)
+    text = lowercase(text)
 
-    def normalize(self, raw_input_text: str) -> str:
+    # TODO: Expand abbreviations (Spanish)
-        return raw_input_text
+    # TODO: Expand numbers (Spanish)
+
+    text = collapse_whitespace(text)
+
+    return text

tts/text/pipeline_components/normalization/test_english_normalization.py

@@ -1,13 +1,11 @@
-from tts.text.pipeline_components.normalization.english_normalization import EnglishNormalization
+from tts.text.pipeline_components.normalization.english_normalization import normalize_english
 
 class TestEnglishNormalization:
 
     def test_whitespace_collapse(self):
-        normalizer = EnglishNormalization()
+        output = normalize_english(" \t \n   normalized   \n \t ")
-        output = normalizer.normalize(" \t \n   normalized   \n \t ")
         assert output == " normalized "
 
     def test_lowercase(self):
-        normalizer = EnglishNormalization()
+        output = normalize_english("This is a test of Normalization.")
-        output = normalizer.normalize("This is a test of Normalization.")
         assert output == "this is a test of normalization."