Update app.py

app.py

@@ -33,22 +33,23 @@ async def generate_accurate_srt(batch_text, batch_num, start_offset):
     # Initialize SRT content
     srt_content = ""
     words = batch_text.split()
-    num_segments = max(1, len(words) // 10)  # Calculate number of segments based on ~10 words per segment
+    num_segments = max(1, len(words) // 15)  # Group words into segments of ~15 words each
     segment_duration = actual_length / num_segments  # Duration for each segment
     start_time = start_offset
+    min_display_duration = 1.5  # Set a minimum display time of 1.5 seconds per subtitle
 
     # Build SRT content with accurate timing
-    for i in range(0, len(words), 10):
-        segment_words = words[i:i+10]
-        end_time = start_time + segment_duration
-        srt_content += f"{i // 10 + 1 + (batch_num * 100)}\n"
+    for i in range(0, len(words), 15):  # Process ~15 words per subtitle
+        segment_words = words[i:i+15]
+        end_time = start_time + max(segment_duration, min_display_duration)  # Ensure each subtitle shows at least min_display_duration
+        srt_content += f"{i // 15 + 1 + (batch_num * 100)}\n"
         srt_content += f"{format_time(start_time)} --> {format_time(end_time)}\n"
         srt_content += " ".join(segment_words) + "\n\n"
         start_time = end_time
 
-    return srt_content, audio_file, start_time
+    return srt_content, audio_file, start_time  # Return updated start time for cumulative tracking
 
-# Batch processing function with concurrent processing and progress indicator
+# Batch processing function with cumulative timing and progress indicator
 async def batch_process_srt_and_audio(script_text, progress=gr.Progress()):
     batches = [script_text[i:i+500] for i in range(0, len(script_text), 500)]
     all_srt_content = ""
@@ -56,16 +57,8 @@ async def batch_process_srt_and_audio(script_text, progress=gr.Progress()):
     start_offset = 0.0  # Track cumulative time offset for SRT timing
 
     # Prepare tasks for concurrent batch processing
-    tasks = [
-        generate_accurate_srt(batch_text, batch_num, start_offset)
-        for batch_num, batch_text in enumerate(batches)
-    ]
-    
-    # Execute tasks concurrently and update progress after each completion
-    completed_tasks = 0
-    total_tasks = len(tasks)
-    for task in asyncio.as_completed(tasks):
-        srt_content, audio_file, end_offset = await task
+    for batch_num, batch_text in enumerate(batches):
+        srt_content, audio_file, end_offset = await generate_accurate_srt(batch_text, batch_num, start_offset)
         all_srt_content += srt_content
 
         # Append the audio of each batch to the combined audio
@@ -77,8 +70,7 @@ async def batch_process_srt_and_audio(script_text, progress=gr.Progress()):
         os.remove(audio_file)
 
         # Update progress
-        completed_tasks += 1
-        progress(completed_tasks / total_tasks)
+        progress((batch_num + 1) / len(batches))
 
     # Generate unique names for the final files
     unique_id = uuid.uuid4()