Update app.py

app.py

@@ -437,7 +437,6 @@ import torch
 from TTS.api import TTS
 from nltk.tokenize import sent_tokenize
 from pydub import AudioSegment
-# Assuming split_long_sentence and wipe_folder are defined elsewhere in your code
 
 default_target_voice_path = "default_voice.wav"  # Ensure this is a valid path
 default_language_code = "en"
@@ -483,18 +482,30 @@ def combine_wav_files(input_directory, output_directory, file_name):
     print(f"Combined audio saved to {output_file_path}")
 
 # Function to split long strings into parts
-def split_long_sentence(sentence, max_length=249, max_pauses=10):
+# Modify the function to handle special cases for Chinese, Italian, and default for others
+def split_long_sentence(sentence, language='en', max_pauses=10):
     """
     Splits a sentence into parts based on length or number of pauses without recursion.
     
     :param sentence: The sentence to split.
-    :param max_length: Maximum allowed length of a sentence.
+    :param language: The language of the sentence (default is English).
     :param max_pauses: Maximum allowed number of pauses in a sentence.
     :return: A list of sentence parts that meet the criteria.
     """
+    # Adjust the max_length and punctuation symbols based on language
+    if language == 'zh-cn':
+        max_length = 84  # Chinese-specific max length
+        punctuation = ['，', '。', '；', '！', '？']  # Chinese-specific punctuation
+    elif language == 'it':
+        max_length = 213  # Italian-specific max length
+        punctuation = [',', ';', '.']  # Standard punctuation
+    else:
+        max_length = 249  # Default max length for other languages
+        punctuation = [',', ';', '.']  # Default punctuation
+
     parts = []
-    while len(sentence) > max_length or sentence.count(',') + sentence.count(';') + sentence.count('.') > max_pauses:
-        possible_splits = [i for i, char in enumerate(sentence) if char in ',;.' and i < max_length]
+    while len(sentence) > max_length or sum(sentence.count(p) for p in punctuation) > max_pauses:
+        possible_splits = [i for i, char in enumerate(sentence) if char in punctuation and i < max_length]
         if possible_splits:
             # Find the best place to split the sentence, preferring the last possible split to keep parts longer
             split_at = possible_splits[-1] + 1
@@ -559,7 +570,7 @@ def convert_chapters_to_audio_custom_model(chapters_dir, output_audio_dir, targe
                 chapter_text = file.read()
                 sentences = sent_tokenize(chapter_text, language='italian' if language == 'it' else 'english')
                 for sentence in tqdm(sentences, desc=f"Chapter {chapter_num}"):
-                    fragments = split_long_sentence(sentence, max_length=249 if language == "en" else 213, max_pauses=10)
+                    fragments = split_long_sentence(sentence, language=language)
                     for fragment in fragments:
                         if fragment != "":
                             print(f"Generating fragment: {fragment}...")
@@ -606,7 +617,7 @@ def convert_chapters_to_audio_standard_model(chapters_dir, output_audio_dir, tar
                 chapter_text = file.read()
                 sentences = sent_tokenize(chapter_text, language='italian' if language == 'it' else 'english')
                 for sentence in tqdm(sentences, desc=f"Chapter {chapter_num}"):
-                    fragments = split_long_sentence(sentence, max_length=249 if language == "en" else 213, max_pauses=10)
+                    fragments = split_long_sentence(sentence, language=language)
                     for fragment in fragments:
                         if fragment != "":
                             print(f"Generating fragment: {fragment}...")