File size: 6,673 Bytes
d24a2f3
0a39518
8a409a5
21e6f34
d24a2f3
 
dd03a74
c901468
 
0a39518
08c3547
d24a2f3
dd03a74
08c3547
dd03a74
8a409a5
dd03a74
a2554b6
 
0a39518
dd03a74
 
0a39518
a2554b6
893e301
dd03a74
0a39518
 
 
 
 
 
dd03a74
 
0a39518
dd03a74
893e301
0a39518
 
893e301
21e6f34
0a39518
b3273f6
9807843
b3273f6
ab41f5c
 
 
b3273f6
c3c3d92
 
 
b3273f6
dd03a74
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8a409a5
dd03a74
 
 
 
 
 
 
 
 
 
 
 
 
 
c3c3d92
 
dd03a74
 
 
 
 
 
 
 
 
b3273f6
dd03a74
 
 
 
 
b3273f6
 
 
dd03a74
 
b3273f6
dd03a74
 
 
 
b3273f6
dd03a74
 
 
 
c3c3d92
dd03a74
b3273f6
dd03a74
b3273f6
 
c3c3d92
 
dd03a74
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0a39518
c3c3d92
dd03a74
 
c3c3d92
dd03a74
c3c3d92
dd03a74
c3c3d92
dd03a74
c3c3d92
 
dd03a74
 
 
c3c3d92
dd03a74
 
 
c3c3d92
dd03a74
 
 
 
 
c3c3d92
dd03a74
 
 
 
 
 
 
 
 
 
 
 
0a39518
 
 
c3c3d92
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
import os
import time
import tempfile
import uuid
import google.generativeai as genai
import requests
from flask import Flask, request, render_template, send_from_directory, url_for, flash, jsonify
from moviepy.video.io.VideoFileClip import VideoFileClip
from moviepy.audio.io.AudioFileClip import AudioFileClip
from werkzeug.utils import secure_filename
from dotenv import load_dotenv

# Initialize Flask app
load_dotenv()
app = Flask(__name__, template_folder='templates', static_folder='static')

# Configuration
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
TTS_API_URL = os.getenv("TTS_API_URL")

if not GEMINI_API_KEY or not TTS_API_URL:
    raise ValueError("Missing required environment variables")

genai.configure(api_key=GEMINI_API_KEY)

# File storage setup
UPLOAD_FOLDER = 'uploads'
DOWNLOAD_FOLDER = 'downloads'
os.makedirs(UPLOAD_FOLDER, exist_ok=True)
os.makedirs(DOWNLOAD_FOLDER, exist_ok=True)
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
app.config['DOWNLOAD_FOLDER'] = DOWNLOAD_FOLDER
app.config['MAX_CONTENT_LENGTH'] = 100 * 1024 * 1024  # 100MB
app.secret_key = os.urandom(24)

# Voice options
VOICE_CHOICES = {
    "Male (Charon)": "Charon",
    "Female (Zephyr)": "Zephyr"
}

GEMINI_PROMPT = """
You are an expert AI scriptwriter. Your task is to watch the provided video and transcribe ALL spoken dialogue into a SINGLE, CONTINUOUS block of modern, colloquial Tamil.

**CRITICAL INSTRUCTIONS:**
1.  **Single Script:** Combine all dialogue from all speakers into one continuous script.
2.  **NO Timestamps or Speaker Labels:** Do NOT include any timestamps or speaker identifiers.
3.  **Incorporate Performance:** Add English style prompts (e.g., `Say happily:`, `Whisper mysteriously:`) and performance tags (e.g., `[laugh]`, `[sigh]`) directly into the text for an expressive narration.

**EXAMPLE OUTPUT:**
Say happily: வணக்கம்! [laugh] எப்படி இருக்கீங்க? Whisper mysteriously: அந்த ரகசியம் எனக்கு மட்டும் தான் தெரியும்.
"""

def generate_tamil_script(video_path):
    """Generate Tamil script using Gemini"""
    try:
        print("Uploading video to Gemini...")
        video_file = genai.upload_file(video_path, mime_type="video/mp4")
        
        # Wait for processing
        while video_file.state.name == "PROCESSING":
            time.sleep(5)
            video_file = genai.get_file(video_file.name)
        
        if video_file.state.name != "ACTIVE":
            raise Exception(f"Gemini processing failed: {video_file.state.name}")
        
        print("Generating script...")
        model = genai.GenerativeModel(model_name="models/gemini-2.5-flash")
        response = model.generate_content([GEMINI_PROMPT, video_file])
        genai.delete_file(video_file.name)
        
        if hasattr(response, 'text') and response.text:
            return " ".join(response.text.strip().splitlines())
        raise Exception("No valid script generated")
    except Exception as e:
        print(f"Script generation failed: {str(e)}")
        raise

def generate_audio_track(text, voice, cheerful, output_path):
    """Generate audio using TTS API"""
    try:
        print("Generating audio track...")
        payload = {
            "text": text,
            "voice_name": voice,
            "cheerful": cheerful
        }
        
        response = requests.post(TTS_API_URL, json=payload, timeout=300)
        if response.status_code != 200:
            raise Exception(f"TTS API error: {response.status_code}")
        
        with open(output_path, "wb") as f:
            f.write(response.content)
        print("Audio track generated successfully")
    except Exception as e:
        print(f"Audio generation failed: {str(e)}")
        raise

def replace_video_audio(video_path, audio_path, output_path):
    """Replace video audio track"""
    video = None
    audio = None
    try:
        print("Replacing audio track...")
        video = VideoFileClip(video_path)
        audio = AudioFileClip(audio_path)
        video.audio = audio
        video.write_videofile(
            output_path,
            codec="libx264",
            audio_codec="aac",
            threads=4,
            verbose=False
        )
        print("Video processing complete!")
    except Exception as e:
        print(f"Video processing failed: {str(e)}")
        raise
    finally:
        if video:
            video.close()
        if audio:
            audio.close()

@app.route('/')
def index():
    """Main page"""
    return render_template('index.html', voices=VOICE_CHOICES)

@app.route('/process', methods=['POST'])
def process_video():
    """Handle video processing"""
    if 'video' not in request.files:
        flash("No video file uploaded", "error")
        return render_template('index.html', voices=VOICE_CHOICES)
    
    file = request.files['video']
    if file.filename == '':
        flash("No file selected", "error")
        return render_template('index.html', voices=VOICE_CHOICES)
    
    # Save uploaded file
    filename = secure_filename(f"{uuid.uuid4()}_{file.filename}")
    video_path = os.path.join(app.config['UPLOAD_FOLDER'], filename)
    file.save(video_path)
    
    # Get processing options
    voice = request.form.get('voice', 'Charon')
    cheerful = request.form.get('cheerful') == 'on'
    
    try:
        # Generate script
        script = generate_tamil_script(video_path)
        
        # Generate audio
        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_audio:
            audio_path = temp_audio.name
        
        generate_audio_track(script, voice, cheerful, audio_path)
        
        # Create dubbed video
        final_filename = f"dubbed_{filename}"
        final_path = os.path.join(app.config['DOWNLOAD_FOLDER'], final_filename)
        replace_video_audio(video_path, audio_path, final_path)
        
        # Cleanup
        os.unlink(audio_path)
        os.unlink(video_path)
        
        return jsonify({
            'status': 'success',
            'video_url': url_for('download', filename=final_filename),
            'script': script
        })
        
    except Exception as e:
        # Cleanup on error
        if os.path.exists(video_path):
            os.unlink(video_path)
        return jsonify({
            'status': 'error',
            'message': str(e)
        }), 500

@app.route('/download/<filename>')
def download(filename):
    """Serve processed video"""
    return send_from_directory(app.config['DOWNLOAD_FOLDER'], filename)

if __name__ == '__main__':
    app.run(host="0.0.0.0", port=7860)