edit cleaners

saved_model/config.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4b4c6cae3dad062f31b0d87a4601d9266a24767f9f41949ec2eab3a13824dcd
-size 1794
+oid sha256:284f7d38e892008e195482b8490359d503f634fbc8b4b92ffa333b56848e6678
+size 1781

text/cleaners.py

@@ -1,40 +1,17 @@
 import re
-import pyopenjtalk
-from unidecode import unidecode
-from text.japanese import _japanese_marks
 
+def japanese_cleaners(text):
+    from text.japanese import japanese_to_romaji_with_accent
+    text = japanese_to_romaji_with_accent(text)
+    if len(text) == 0 or re.match('[A-Za-z]', text[-1]):
+        text += '.'
+    return text
 
-def japanese_triphone_cleaners(text):
-    sentences = re.split(_japanese_marks, text)
-    marks = re.findall(_japanese_marks, text)
-    text = ''
-    for i, sentence in enumerate(sentences):
-        phones = pyopenjtalk.g2p(sentence, kana=False)
-        phones = phones.replace(' ','')
-        phones = phones.replace('A', 'a').replace('I', 'i').replace('U', 'u').replace('E', 'e').replace('O', 'o')
-        phones = phones.replace('ch','ʧ').replace('sh','ʃ').replace('cl','Q')
-        triphones = []
-        length = len(phones)
-
-        for j, phone in enumerate(phones):
-            if length == 1:
-                triphone = phone
-            else:
-                if j == 0:
-                    triphone = f'{phone}+{phones[j+1]}'
-                elif j == length - 1:
-                    triphone = f'{phones[j-1]}-{phone}'
-                else:
-                    triphone = f'{phones[j-1]}-{phone}+{phones[j+1]}'
-
-            triphones.append(triphone)
-
-        subtext = ' '.join(triphones)
-        text += subtext
-        if i < len(marks):
-            text += unidecode(marks[i]).replace(' ', '')
 
-    if len(text) > 0  and re.match('[A-Za-z]',text[-1]):
-        text += '.'
-    
+def japanese_cleaners2(text):
+    text = text.replace('・・・', '…').replace('・', ' ')
+    text = japanese_cleaners(text).replace('ts', 'ʦ').replace('...', '…') \
+                                    .replace('(', '').replace(')', '') \
+                                    .replace('[', '').replace(']', '') \
+                                    .replace('*', ' ').replace('{', '').replace('}', '')
     return text