Update app.py

app.py

@@ -220,9 +220,9 @@ async def transcript_to_speech(transcript_text, voice, rate, pitch):
                 intended_start_time = start_time
 
                 if i > 0:
-                    previous_line_match = re.match(r'(\d{2}):(\d{2}):(\d{2}),(\d{3})\s+.*', lines[i-1])
-                    if previous_line_match:
-                        prev_h, prev_m, prev_s, prev_ms = previous_line_match.groups()
+                    prev_line_match = re.match(r'(\d{2}):(\d{2}):(\d{2}),(\d{3})\s+.*', lines[i-1])
+                    if prev_line_match:
+                        prev_h, prev_m, prev_s, prev_ms = prev_line_match.groups()
                         previous_start_time_ms = (
                             int(prev_h) * 3600000 +
                             int(prev_m) * 60000 +
@@ -230,12 +230,13 @@ async def transcript_to_speech(transcript_text, voice, rate, pitch):
                             int(prev_ms)
                         )
                         time_difference = start_time - previous_start_time_ms
-
-                        if current_audio_duration > time_difference and timed_audio_segments:
-                            intended_start_time = previous_end_time_ms # Append to the previous audio
+                        # print(f"current_audio_duration:{current_audio_duration}, time_difference:{time_difference}, previous_end_time_ms:{previous_end_time_ms}, start_time:{start_time}")
+                        if current_audio_duration > time_difference:
+                            intended_start_time = previous_end_time_ms
+                        # print(f"intended_start_time:{intended_start_time}")
 
                 timed_audio_segments.append({'start': intended_start_time, 'audio': combined_line_audio})
-                previous_end_time_ms = intended_start_time + current_audio_duration
+                previous_end_time_ms = max(previous_end_time_ms, intended_start_time + current_audio_duration)
                 max_end_time_ms = max(max_end_time_ms, previous_end_time_ms)
         elif audio_paths:
             for path in audio_paths:
@@ -310,4 +311,3 @@ async def create_demo():
 if __name__ == "__main__":
     demo = asyncio.run(create_demo())
     demo.launch()
-