Create app.py

app.py

@@ -0,0 +1,193 @@
+import gradio as gr
+import os
+import numpy as np
+from moviepy.editor import ImageClip, AudioFileClip, CompositeVideoClip, TextClip, ColorClip
+from gtts import gTTS
+import tempfile
+from PIL import Image
+
+# --- Helper: Resolution & Bitrate ---
+def get_specs(ratio, quality):
+    res_map = {
+        "144p": (256, 144, "300k"),
+        "240p": (426, 240, "600k"),
+        "360p": (640, 360, "1000k"),
+        "720p": (1280, 720, "2500k"),
+        "1080p": (1920, 1080, "5000k")
+    }
+    w, h, b = res_map.get(quality, res_map["360p"])
+    
+    if ratio == "9:16 (Shorts)": return h, w, b
+    elif ratio == "1:1 (Square)": return h, h, b
+    else: return w, h, b # 16:9
+
+# --- TTS Function with Language ---
+def text_to_audio(text, lang='en'):
+    if not text: return None
+    try:
+        t = tempfile.NamedTemporaryFile(delete=False, suffix='.mp3')
+        gTTS(text=text, lang=lang).save(t.name)
+        return t.name
+    except Exception as e:
+        print(f"TTS Error: {e}")
+        return None
+
+# --- Main Video Generation ---
+def create_controlled_ad(bg_img, char_img, prod_img, audio_file, text_input, lang_select, ratio, quality, overlay_text, 
+                         c_x, c_y, c_zoom, p_x, p_y, p_zoom):
+    
+    # 1. Audio Handling
+    audio_path = audio_file
+    if not audio_path and text_input:
+        audio_path = text_to_audio(text_input, lang=lang_select)
+    
+    if not audio_path:
+        return None, "Please provide Audio File or Text to Speak."
+
+    try:
+        W, H, bitrate = get_specs(ratio, quality)
+        audio_clip = AudioFileClip(audio_path)
+        duration = audio_clip.duration
+        
+        clips = []
+
+        # 2. Background Layer (Fixed for Gradio Array/Input handling)
+        if bg_img is not None:
+            if isinstance(bg_img, np.ndarray):
+                bg_pil = Image.fromarray(bg_img)
+            else:
+                bg_pil = Image.open(bg_img)
+            
+            bg_clip = ImageClip(np.array(bg_pil.resize((W, H)))).set_duration(duration)
+        else:
+            bg_clip = ColorClip(size=(W, H), color=(20, 20, 20)).set_duration(duration)
+        
+        clips.append(bg_clip)
+
+        # Helper to place images with Sliders
+        def place_image(img_input, pos_x, pos_y, zoom_factor, default_size_pct):
+            if img_input is None: return
+            
+            if isinstance(img_input, np.ndarray):
+                pil_img = Image.fromarray(img_input)
+            else:
+                pil_img = Image.open(img_input)
+                
+            base_h = int(H * default_size_pct)
+            base_w = int(base_h * (pil_img.width / pil_img.height))
+            
+            final_h = int(base_h * zoom_factor)
+            final_w = int(base_w * zoom_factor)
+            
+            if final_w <= 0: final_w = 1
+            if final_h <= 0: final_h = 1
+
+            clip = ImageClip(np.array(pil_img)).resize((final_w, final_h)).set_duration(duration)
+            
+            center_x = (W - final_w) // 2
+            center_y = (H - final_h) // 2
+            
+            offset_x = int((pos_x / 100) * W)
+            offset_y = int((pos_y / 100) * H)
+            
+            final_pos = (center_x + offset_x, center_y + offset_y)
+            clips.append(clip.set_position(final_pos))
+
+        # 3. Add Character
+        place_image(char_img, c_x, c_y, c_zoom, 0.50)
+        
+        # 4. Add Product
+        place_image(prod_img, p_x, p_y, p_zoom, 0.30)
+
+        # 5. Add Text Overlay
+        if overlay_text and len(str(overlay_text).strip()) > 0:
+            fontsize = max(12, int(W * 0.05))
+            txt_clip = TextClip(str(overlay_text), fontsize=fontsize, color='white', stroke_color='black', stroke_width=2, method='caption', size=(W-20, None))
+            txt_y = H - txt_clip.h - 20
+            clips.append(txt_clip.set_position(('center', txt_y)).set_duration(duration))
+
+        # 6. Render Video
+        final_video = CompositeVideoClip(clips, size=(W, H)).set_audio(audio_clip)
+        output_file = "controlled_ad.mp4"
+        
+        final_video.write_videofile(
+            output_file, 
+            fps=24, 
+            codec='libx264', 
+            audio_codec='aac',
+            preset='ultrafast', 
+            threads=4, 
+            bitrate=bitrate
+        )
+        
+        final_video.close()
+        audio_clip.close()
+        if not audio_file and audio_path and os.path.exists(audio_path):
+            os.remove(audio_path)
+            
+        return output_file, f"Video Generated in {lang_select.upper()}!"
+
+    except Exception as e:
+        import traceback
+        return None, f"Error: {str(e)}\n{traceback.format_exc()}"
+
+# --- UI Design ---
+with gr.Blocks(theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 🎛️ Pro Ad Maker: Multi-Language & Control")
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            bg_in = gr.Image(label="1. Background Image", type="numpy")
+            char_in = gr.Image(label="2. Character Image", type="numpy")
+            prod_in = gr.Image(label="3. Product Image", type="numpy")
+            
+            with gr.Accordion("Video & Voice Settings", open=True):
+                ratio_in = gr.Radio(["16:9 (YouTube)", "9:16 (Shorts)", "1:1 (Post)"], value="9:16 (Shorts)", label="Ratio")
+                qual_in = gr.Radio(["144p", "240p", "360p", "720p", "1080p"], value="360p", label="Quality")
+                
+                lang_in = gr.Dropdown(
+                    choices=[
+                        ("English", "en"),
+                        ("Hindi", "hi"),
+                        ("Spanish", "es"),
+                        ("French", "fr"),
+                        ("German", "de"),
+                        ("Japanese", "ja")
+                    ],
+                    value="hi",
+                    label="Select Voice Language"
+                )
+            
+            txt_in = gr.Textbox(lines=2, placeholder="Ad Script (e.g., 'Ye product best hai!')", label="Text to Speech")
+            aud_in = gr.Audio(label="OR Upload Audio File (Overrides Text)")
+            
+            generate_btn = gr.Button("🎥 Generate Video", variant="primary")
+
+        with gr.Column(scale=1):
+            gr.Markdown("### 🕹️ Alignment Controls")
+            
+            with gr.Tab("Character Adjust"):
+                c_zoom = gr.Slider(0.1, 2.0, value=1.0, step=0.1, label="Character Zoom (Size)")
+                c_x = gr.Slider(-50, 50, value=-20, step=1, label="Character Horizontal (Left/Right)")
+                c_y = gr.Slider(-50, 50, value=10, step=1, label="Character Vertical (Up/Down)")
+                
+            with gr.Tab("Product Adjust"):
+                p_zoom = gr.Slider(0.1, 2.0, value=1.0, step=0.1, label="Product Zoom (Size)")
+                p_x = gr.Slider(-50, 50, value=20, step=1, label="Product Horizontal (Left/Right)")
+                p_y = gr.Slider(-50, 50, value=0, step=1, label="Product Vertical (Up/Down)")
+
+            vid_out = gr.Video(label="Final Video")
+            status_out = gr.Textbox(label="Status")
+
+    generate_btn.click(
+        fn=create_controlled_ad,
+        inputs=[
+            bg_in, char_in, prod_in, aud_in, txt_in, lang_in, ratio_in, qual_in, txt_in,
+            c_x, c_y, c_zoom, 
+            p_x, p_y, p_zoom
+        ],
+        outputs=[vid_out, status_out]
+    )
+
+if __name__ == "__main__":
+    demo.launch()