Update app.py

app.py

@@ -1,226 +1,117 @@
 import gradio as gr
 import cv2
 import numpy as np
-import tempfile
-import os
 import json
-from PIL import Image
-import time
-import math
 
-# ============================================================================
-# 📋 OCR FUNCTION (Video Hardsub Extractor)
-# ============================================================================
-
-def extract_subtitles(video_file, language, start_time, end_time, 
-                      box_x, box_y, box_w, box_h, aspect_lock):
+def extract_subtitles(video, language, start, end, x, y, w, h, lock):
     """
-    Video ထဲက hardcoded subtitles တွေကို ထုတ်ပေးမယ့် function
+    Extract hardcoded subtitles from video
     """
-    # Video ဖိုင်ကိုဖွင့်မယ်
-    cap = cv2.VideoCapture(video_file)
+    if video is None:
+        return None, None, None, [], "No video uploaded"
     
-    # Video info
+    # Open video
+    cap = cv2.VideoCapture(video)
     fps = cap.get(cv2.CAP_PROP_FPS)
-    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
     width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
     height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
     
-    # Start/End time ကို frame number အဖြစ်ပြောင်းမယ်
-    start_frame = max(0, int(start_time * fps))
-    end_frame = min(total_frames, int(end_time * fps))
+    # Calculate frames
+    start_frame = int(start * fps)
+    end_frame = int(end * fps)
+    
+    # Sample every 1 second
+    sample_interval = int(fps)
+    cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
     
-    # စာတန်းတွေသိမ်းမယ့် list
     subtitles = []
     last_text = ""
-    current_start = start_time
-    frame_count = 0
-    total_frames_to_process = end_frame - start_frame
-    
-    # Progress updates
-    progress = []
-    
-    # Frame အလိုက်ဖတ်မယ် (1 FPS sampling)
-    sample_interval = int(fps)  # 1 frame per second
-    
-    cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
+    current_start = start
     
+    frame_count = 0
     for frame_num in range(start_frame, end_frame, sample_interval):
         ret, frame = cap.read()
         if not ret:
             break
-            
+        
         frame_count += 1
         current_time = frame_num / fps
         
-        # Progress calculation
-        progress_percent = (frame_count / (total_frames_to_process / sample_interval)) * 100
-        
-        # ရွေးထားတဲ့နေရာကိုဖြတ်မယ်
-        x1 = int((box_x / 100) * width)
-        y1 = int((box_y / 100) * height)
-        x2 = int(((box_x + box_w) / 100) * width)
-        y2 = int(((box_y + box_h) / 100) * height)
-        
-        # Ensure coordinates are within frame
-        x1 = max(0, min(x1, width-1))
-        y1 = max(0, min(y1, height-1))
-        x2 = max(x1+1, min(x2, width))
-        y2 = max(y1+1, min(y2, height))
-        
-        crop = frame[y1:y2, x1:x2]
+        # Crop subtitle area
+        x1 = int((x / 100) * width)
+        y1 = int((y / 100) * height)
+        x2 = int(((x + w) / 100) * width)
+        y2 = int(((y + h) / 100) * height)
         
-        # OCR Simulation - ဒီနေရာမှာ တကယ့် OCR engine ထည့်လို့ရတယ်
-        text = simulate_ocr(crop, language, current_time)
+        # Simulate OCR (replace with actual OCR later)
+        sample_texts = ["Hello", "Welcome", "ဟယ်လို", "สวัสดี", "こんにちは", "안녕하세요", "你好"]
+        text = sample_texts[frame_count % len(sample_texts)]
         
-        # စာသားတူရင် merge လုပ်မယ်
+        # Merge duplicates
         if text and text != last_text:
             if last_text:
                 subtitles.append({
-                    'start': current_start,
-                    'end': current_time,
-                    'text': last_text,
-                    'status': 'exited'
+                    "start": current_start,
+                    "end": current_time,
+                    "text": last_text
                 })
             current_start = current_time
             last_text = text
-            
-            # Add entered status
-            subtitles.append({
-                'start': current_time,
-                'end': current_time,
-                'text': text,
-                'status': 'entered'
-            })
     
-    # နောက်ဆုံးစာတန်းကိုထည့်မယ်
+    # Add last subtitle
     if last_text:
         subtitles.append({
-            'start': current_start,
-            'end': end_time,
-            'text': last_text,
-            'status': 'exited'
+            "start": current_start,
+            "end": end,
+            "text": last_text
         })
     
     cap.release()
     
-    # SRT format ပြောင်းမယ်
-    srt_content = generate_srt(subtitles)
+    # Generate SRT
+    srt_content = ""
+    for i, sub in enumerate(subtitles, 1):
+        srt_content += f"{i}\n{format_time(sub['start'])} --> {format_time(sub['end'])}\n{sub['text']}\n\n"
     
-    # JSON format ပြောင်းမယ်
-    json_content = generate_json(subtitles)
+    # Generate JSON
+    json_content = json.dumps(subtitles, indent=2, ensure_ascii=False)
     
-    # TXT format ပြောင်းမယ်
-    txt_content = generate_txt(subtitles)
+    # Generate TXT
+    txt_content = ""
+    for sub in subtitles:
+        txt_content += f"[{format_time(sub['start'])} --> {format_time(sub['end'])}] {sub['text']}\n"
     
-    # ဖိုင်တွေသိမ်းမယ်
-    srt_path = "extracted_subtitles.srt"
-    json_path = "extracted_subtitles.json"
-    txt_path = "extracted_subtitles.txt"
+    # Save files
+    srt_path = "subtitles.srt"
+    json_path = "subtitles.json"
+    txt_path = "subtitles.txt"
     
     with open(srt_path, "w", encoding="utf-8") as f:
         f.write(srt_content)
-    
     with open(json_path, "w", encoding="utf-8") as f:
         f.write(json_content)
-    
     with open(txt_path, "w", encoding="utf-8") as f:
         f.write(txt_content)
     
-    # စာတန်းစာရင်းကို Dataframe အတွက်ပြင်ဆင်မယ်
-    subtitle_list = []
-    for s in subtitles:
-        if s['status'] != 'entered':  # Only show actual subtitle lines
-            subtitle_list.append([
-                format_time(s['start']), 
-                format_time(s['end']), 
-                s['text'][:50] + "..." if len(s['text']) > 50 else s['text']
-            ])
+    # Format for display
+    display_list = [[format_time(s["start"]), format_time(s["end"]), s["text"]] for s in subtitles]
     
-    return srt_path, json_path, txt_path, subtitle_list, f"Processed {frame_count} frames"
-
-def simulate_ocr(image, language, current_time):
-    """
-    OCR Simulation - တကယ့် OCR အတွက် placeholder
-    """
-    # Sample texts for demonstration
-    samples = [
-        "Hello and welcome",
-        "နမူနာစာသား", 
-        "ตัวอย่างข้อความ",
-        "サンプルテキスト",
-        "샘플 텍스트",
-        "示例文本",
-        "ข้อความตัวอย่าง"
-    ]
-    
-    # Return sample text based on time
-    idx = int(current_time) % len(samples)
-    return samples[idx] if idx < len(samples) else ""
-
-def generate_srt(subtitles):
-    """SRT format ပြောင်းမယ်"""
-    srt_content = ""
-    counter = 1
-    
-    for i, sub in enumerate(subtitles):
-        if sub['status'] != 'entered':
-            start = format_time(sub['start'])
-            end = format_time(sub['end'])
-            srt_content += f"{counter}\n{start} --> {end}\n{sub['text']}\n\n"
-            counter += 1
-    
-    return srt_content
-
-def generate_json(subtitles):
-    """JSON format ပြောင်းမယ်"""
-    json_list = []
-    for sub in subtitles:
-        if sub['status'] != 'entered':
-            json_list.append({
-                'start': sub['start'],
-                'start_formatted': format_time(sub['start']),
-                'end': sub['end'],
-                'end_formatted': format_time(sub['end']),
-                'text': sub['text']
-            })
-    return json.dumps(json_list, indent=2, ensure_ascii=False)
-
-def generate_txt(subtitles):
-    """TXT format ပြောင်းမယ်"""
-    txt_content = ""
-    for sub in subtitles:
-        if sub['status'] != 'entered':
-            txt_content += f"[{format_time(sub['start'])} --> {format_time(sub['end'])}] {sub['text']}\n"
-    return txt_content
+    return srt_path, json_path, txt_path, display_list, f"Processed {frame_count} frames"
 
 def format_time(seconds):
-    """အချိန်ကို SRT format (HH:MM:SS,mmm) ပြောင်းမယ်"""
     h = int(seconds // 3600)
     m = int((seconds % 3600) // 60)
     s = int(seconds % 60)
     ms = int((seconds - int(seconds)) * 1000)
     return f"{h:02d}:{m:02d}:{s:02d},{ms:03d}"
 
-def set_start_from_video(current_time):
-    return current_time
-
-def set_end_from_video(current_time):
-    return current_time
-
-def reset_adjustments():
-    return 10, 80, 80, 15
-
-def toggle_aspect_lock(current_lock):
-    return not current_lock
-
 def update_duration(start, end):
     return max(0, end - start)
 
 def set_full_video(duration):
     return 0, duration
 
-def update_video_info(video):
+def get_video_duration(video):
     if video:
         cap = cv2.VideoCapture(video)
         fps = cap.get(cv2.CAP_PROP_FPS)
@@ -230,148 +121,75 @@ def update_video_info(video):
         return duration, duration, 0
     return 0, 0, 0
 
-# ============================================================================
-# 📋 GRADIO INTERFACE
-# ============================================================================
-
-custom_css = """
-.highlight-subtitle {
-    background-color: #fff3c9 !important;
-    border-left: 4px solid #f5b342 !important;
-    font-weight: bold !important;
-    box-shadow: 0 0 15px #f5b342 !important;
-    transition: all 0.3s ease;
-}
-.entered-subtitle {
-    background-color: #e6f7ff !important;
-    border-left: 4px solid #1890ff !important;
-}
-.exited-subtitle {
-    opacity: 0.7;
-}
-"""
-
-with gr.Blocks(theme=gr.themes.Soft(), title="Video Hardsub OCR", css=custom_css) as demo:
-    gr.Markdown("""
-    # 🎬 Video Hardsub OCR Extractor
-    ### Video ထဲက hardcoded စာတန်းတွေကို ထုတ်ယူပါ။
-    """)
-    
-    current_time_state = gr.State(0)
-    video_duration_state = gr.State(0)
-    aspect_lock_state = gr.State(False)
-    
-    with gr.Tabs():
-        # ====================================================================
-        # PAGE 1: SETUP PAGE
-        # ====================================================================
-        with gr.TabItem("🎥 Setup", id="setup"):
+# Create Gradio interface
+with gr.Blocks(title="Video Hardsub OCR") as demo:
+    gr.Markdown("# 🎬 Video Hardsub OCR Extractor")
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            # Inputs
+            video = gr.Video(label="Upload Video")
+            lang = gr.Dropdown(
+                choices=["English", "Japanese", "Korean", "Chinese (Simplified)", "Chinese (Traditional)", "Thai"],
+                label="Language", value="English"
+            )
+            
+            # Subtitle area
+            gr.Markdown("### 📐 Subtitle Area")
+            x_pos = gr.Slider(0, 100, value=10, label="X Position %")
+            y_pos = gr.Slider(0, 100, value=80, label="Y Position %")
+            width = gr.Slider(0, 100, value=80, label="Width %")
+            height = gr.Slider(0, 100, value=15, label="Height %")
+            
+            # Time range
+            gr.Markdown("### ⏱️ Time Range")
+            start = gr.Number(label="Start (s)", value=0)
+            end = gr.Number(label="End (s)", value=10)
+            duration_disp = gr.Number(label="Duration", value=10, interactive=False)
+            
             with gr.Row():
-                with gr.Column(scale=2):
-                    video_input = gr.Video(label="Upload Video", interactive=True)
-                    
-                    with gr.Row():
-                        play_btn = gr.Button("▶ Play", variant="secondary")
-                        pause_btn = gr.Button("⏸ Pause", variant="secondary")
-                        current_time_display = gr.Number(label="Current Time (s)", value=0, interactive=False)
-                        duration_display = gr.Number(label="Duration (s)", value=0, interactive=False)
-                
-                with gr.Column(scale=1):
-                    gr.Markdown("### 🌐 Language Selection")
-                    language = gr.Dropdown(
-                        choices=["English", "Japanese", "Korean", "Chinese (Simplified)", "Chinese (Traditional)", "Thai"],
-                        label="Language", value="English"
-                    )
-                    
-                    gr.Markdown("### 📐 Adjust Subtitle Area")
-                    with gr.Row():
-                        box_x = gr.Slider(0, 100, value=10, step=0.5, label="X Position %")
-                        box_y = gr.Slider(0, 100, value=80, step=0.5, label="Y Position %")
-                    with gr.Row():
-                        box_w = gr.Slider(0, 100, value=80, step=0.5, label="Width %")
-                        box_h = gr.Slider(0, 100, value=15, step=0.5, label="Height %")
-                    
-                    with gr.Row():
-                        reset_btn = gr.Button("↺ Reset", variant="secondary")
-                        aspect_lock_btn = gr.Button("🔒 Aspect Lock", variant="secondary")
-                    
-                    gr.Markdown("### ⏱️ Time Range")
-                    with gr.Row():
-                        start_time = gr.Number(label="Start (s)", value=0, step=0.1)
-                        set_start_btn = gr.Button("Set Start", variant="secondary")
-                    with gr.Row():
-                        end_time = gr.Number(label="End (s)", value=10, step=0.1)
-                        set_end_btn = gr.Button("Set End", variant="secondary")
-                    
-                    duration_display_small = gr.Number(label="Duration", value=10, interactive=False)
-                    full_video_btn = gr.Button("📽️ Full Video", variant="secondary")
-                    
-                    process_btn = gr.Button("⚡ PROCESS VIDEO", variant="primary", size="lg")
-        
-        # ====================================================================
-        # PAGE 2: RESULTS PAGE
-        # ====================================================================
-        with gr.TabItem("📄 Results", id="results"):
-            gr.Markdown("## 📊 Extraction Results")
+                set_start = gr.Button("Set Start")
+                set_end = gr.Button("Set End")
+                full_video = gr.Button("Full Video")
             
+            # Process button
+            process = gr.Button("Process Video", variant="primary")
+        
+        with gr.Column(scale=1):
+            # Outputs
+            video_out = gr.Video(label="Video")
+            subtitles_out = gr.Dataframe(
+                headers=["Start", "End", "Text"],
+                label="Extracted Subtitles"
+            )
             with gr.Row():
-                with gr.Column(scale=1):
-                    video_output = gr.Video(label="Video Playback", interactive=False)
-                
-                with gr.Column(scale=1):
-                    subtitle_output = gr.Dataframe(
-                        headers=["Start", "End", "Text"],
-                        label="Extracted Subtitles",
-                        wrap=True
-                    )
-                    
-                    with gr.Row():
-                        srt_download = gr.File(label="SRT")
-                        json_download = gr.File(label="JSON")
-                        txt_download = gr.File(label="TXT")
-                    
-                    back_btn = gr.Button("↩ Back to Setup", variant="secondary")
-    
-    # ========================================================================
-    # EVENT HANDLERS
-    # ========================================================================
+                srt_file = gr.File(label="SRT")
+                json_file = gr.File(label="JSON")
+                txt_file = gr.File(label="TXT")
+    
+    # Current time display
+    current_time = gr.Number(value=0, visible=False)
+    duration = gr.Number(value=0, visible=False)
+    
+    # Event handlers
+    video.change(
+        fn=get_video_duration,
+        inputs=[video],
+        outputs=[duration, end, current_time]
+    )
     
-    set_start_btn.click(fn=set_start_from_video, inputs=[current_time_display], outputs=[start_time])
-    set_end_btn.click(fn=set_end_from_video, inputs=[current_time_display], outputs=[end_time])
-    reset_btn.click(fn=reset_adjustments, inputs=[], outputs=[box_x, box_y, box_w, box_h])
-    aspect_lock_btn.click(fn=toggle_aspect_lock, inputs=[aspect_lock_state], outputs=[aspect_lock_state])
+    set_start.click(fn=lambda t: t, inputs=[current_time], outputs=[start])
+    set_end.click(fn=lambda t: t, inputs=[current_time], outputs=[end])
     
-    start_time.change(fn=update_duration, inputs=[start_time, end_time], outputs=[duration_display_small])
-    end_time.change(fn=update_duration, inputs=[start_time, end_time], outputs=[duration_display_small])
+    start.change(fn=update_duration, inputs=[start, end], outputs=[duration_disp])
+    end.change(fn=update_duration, inputs=[start, end], outputs=[duration_disp])
     
-    full_video_btn.click(fn=set_full_video, inputs=[duration_display], outputs=[start_time, end_time])
+    full_video.click(fn=set_full_video, inputs=[duration], outputs=[start, end])
     
-    process_btn.click(
+    process.click(
         fn=extract_subtitles,
-        inputs=[video_input, language, start_time, end_time, box_x, box_y, box_w, box_h, aspect_lock_state],
-        outputs=[srt_download, json_download, txt_download, subtitle_output, current_time_display]
-    ).then(
-        fn=lambda: gr.Tabs(selected="results"),
-        inputs=[],
-        outputs=None,
-        _js="() => { document.querySelector('#results-tab').click(); }"
-    )
-    
-    back_btn.click(
-        fn=lambda: gr.Tabs(selected="setup"),
-        inputs=[],
-        outputs=None,
-        _js="() => { document.querySelector('#setup-tab').click(); }"
+        inputs=[video, lang, start, end, x_pos, y_pos, width, height, gr.State(False)],
+        outputs=[srt_file, json_file, txt_file, subtitles_out, current_time]
     )
-    
-    video_input.change(
-        fn=update_video_info,
-        inputs=[video_input],
-        outputs=[duration_display, end_time, current_time_display]
-    )
-
-# ============================================================================
-# 📋 LAUNCH
-# ============================================================================
 
 demo.launch()