Update app.py

app.py

@@ -138,6 +138,7 @@ E2TTS_ema_model2 = load_custom(
 
 
 
+
 def chunk_text(text, max_chars=100):
     """
     Splits the input text into chunks, each with a maximum number of characters.
@@ -152,9 +153,9 @@ def chunk_text(text, max_chars=100):
     Returns:
         List[str]: A list of text chunks.
     """
+    split_after_space_chars = 135
     chunks = []
     current_chunk = ""
-    split_after_space_chars = 135
     
     # Split the text into sentences based on punctuation followed by whitespace
     sentences = re.split(r"(?<=[;:,.!?])\s+|(?<=[；：，。！？])", text)
@@ -170,11 +171,19 @@ def chunk_text(text, max_chars=100):
             current_chunk = sentence + " "
 
         # If current chunk exceeds split_after_space_chars and no punctuation, split at space
-        if len(current_chunk) > split_after_space_chars and re.search(r"\w", current_chunk):
-            split_index = current_chunk.rfind(" ")
-            if split_index != -1:
-                chunks.append(current_chunk[:split_index].strip())  # Add the chunk before the space
-                current_chunk = current_chunk[split_index:].strip()  # Start new chunk after the space
+        if len(current_chunk) > split_after_space_chars:
+            # Check if the chunk has punctuation
+            if not re.search(r"[;:,.!?；：，。！？]", current_chunk):
+                # No punctuation; split after the last space before 135 characters
+                split_index = current_chunk.rfind(" ", 0, split_after_space_chars)
+                if split_index != -1:
+                    chunks.append(current_chunk[:split_index].strip())  # Add the chunk before the space
+                    current_chunk = current_chunk[split_index:].strip()  # Start new chunk after the space
+            else:
+                # Chunk contains punctuation; just add it to chunks
+                if current_chunk:
+                    chunks.append(current_chunk.strip())
+                current_chunk = ""
 
     # Append any remaining text in current_chunk to chunks
     if current_chunk:
@@ -184,6 +193,7 @@ def chunk_text(text, max_chars=100):
 
 
 
+
 @gpu_decorator
 def infer_batch(ref_audio, ref_text, gen_text_batches, exp_name, remove_silence, cross_fade_duration=0.15, progress=gr.Progress()):
     if exp_name == "English":