Migrate to ukrainian-word-stress

requirements.txt

@@ -1 +1,2 @@
 TTS==0.7.1
+ukrainian-word-stress==1.0.0

stress.py

@@ -1,62 +1,25 @@
-from unittest import skip
-from gruut import sentences
-import torch
+from ukrainian_word_stress import Stressifier, StressSymbol
 
-importer = torch.package.PackageImporter("ukrainian-accentor/accentor-lite.pt")
-accentor = importer.load_pickle("uk-accentor", "model")
-replace_accents = importer.load_pickle("uk-accentor", "replace_accents")
+stressify = Stressifier(stress_symbol=StressSymbol.CombiningAcuteAccent)
 
-# Using GPU
-# accentor.cuda()
-# Back to CPU
-# accentor.cpu()
-
-vowels = "аеєиіїоуюя"
-consonants = "бвгґджзйклмнпрстфхцчшщь"
-special = "'"
-alphabet = vowels + consonants + special
-
-def accent_word(word):
-    with torch.no_grad():
-        stressed_words = accentor.predict([word], mode='stress')
-    plused_words = [replace_accents(x) for x in stressed_words]
-    return plused_words[0]
 
 def sentence_to_stress(sentence):
-    words = sentence.lower().split()
-    words = sum([[word, " "] for word in words], start=[])
-    new_list = []
-    for word in words:
-        first_word_sep = list(map(lambda letter: letter in alphabet, word.lower()))
-        if all(first_word_sep):
-            new_list.append(word)
+    stressed = stressify(sentence).replace(StressSymbol.CombiningAcuteAccent, "+")
+    new_stressed = ""
+    start = 0
+    last = 0
+    while True:
+        plus_position = stressed.find("+", start)
+        if plus_position != -1:
+            new_stressed += (
+                stressed[last : plus_position - 1] + "+" + stressed[plus_position - 1]
+            )
+            start = plus_position + 1
+            last = start
         else:
-            current_index = 0
-            past_index = 0
-            for letter in first_word_sep:
-                if letter == False:
-                    new_list.append(word[past_index:current_index])
-                    new_list.append(word[current_index])
-                    past_index = current_index + 1
-                current_index += 1
-            new_list.append(word[past_index:current_index])
-    #print(list(filter(lambda x: len(x) > 0, new_list)))
-    for word_index in range(0, len(new_list)):
-        element = new_list[word_index]
-        first_word_sep = list(map(lambda letter: letter in alphabet, element.lower()))
-        if not all(first_word_sep) or len(element) == 0:
-            continue
-        else:
-            vowels_in_words = list(map(lambda letter: letter in vowels, new_list[word_index]))
-            if vowels_in_words.count(True) == 0:
-                continue
-            elif vowels_in_words.count(True) == 1:
-                vowel_index = vowels_in_words.index(True)
-                new_list[word_index] = new_list[word_index][0:vowel_index] + "+" + new_list[word_index][vowel_index::]
-            else:
-                new_list[word_index] = accent_word(new_list[word_index])
-
-    return "".join(new_list)
+            new_stressed += stressed[last:]
+            break
+    return new_stressed
 
 
 if __name__ == "__main__":
@@ -68,10 +31,7 @@ if __name__ == "__main__":
     print(sentence_to_stress(sentence))
     sentence = "Не тільки в Україні таке може бути."
     print(sentence_to_stress(sentence))
-    #test_words1 = ["словотворення", "архаїчний", "програма", "а-ля-фуршет"]
-
-    stressed_words = accentor.predict(["привіт"], mode='stress')
-    plused_words = [replace_accents(x) for x in stressed_words]
-
-    print('With stress:', stressed_words)
-    print('With pluses:', plused_words)
+    sentence = "Н тльк в крн тк мж бт."
+    print(sentence_to_stress(sentence))
+    sentence = "Н тльк в крн тк мж бт."
+    print(sentence_to_stress(sentence))

ukrainian-accentor

@@ -1 +0,0 @@
-Subproject commit a3dd2cf9341db200853cfd19df142224a47749b2