Create app.py

app.py

@@ -0,0 +1,281 @@
+import gradio as gr
+import os
+import time
+import uuid
+import tempfile
+from PIL import Image, ImageDraw, ImageFont
+import base64
+import mimetypes
+from io import BytesIO
+from google import genai
+from google.genai import types
+
+def generate(text, images, api_key, model="gemini-2.5-flash-image-preview"):
+    """Generate image using Gemini API"""
+    # Initialize client using provided api_key (or fallback to env variable)
+    client = genai.Client(api_key=(api_key.strip() if api_key and api_key.strip() != ""
+                                     else os.environ.get("GEMINI_API_KEY")))
+    
+    # Prepare contents with images first, then text
+    contents = images + [text]
+    
+    response = client.models.generate_content(
+        model=model,
+        contents=contents,
+    )
+
+    text_response = ""
+    image_path = None
+    
+    for part in response.candidates[0].content.parts:
+        if part.text is not None:
+            text_response += part.text + "\n"
+        elif part.inline_data is not None:
+            # Create a temporary file to store the generated image
+            with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmp:
+                temp_path = tmp.name
+                generated_image = Image.open(BytesIO(part.inline_data.data))
+                generated_image.save(temp_path)
+                image_path = temp_path
+                print(f"Generated image saved to: {temp_path} with prompt: {text}")
+    
+    return image_path, text_response
+
+def print_like_dislike(x: gr.LikeData):
+    """Handle like/dislike interactions"""
+    print(x.index, x.value, x.liked)
+
+def add_message(history, message, api_key):
+    """Add user message to chat history"""
+    # Handle file uploads
+    if message["files"]:
+        for file_path in message["files"]:
+            # Check if file_path is a string (direct path) or dict (with path key)
+            if isinstance(file_path, str):
+                history.append(((file_path,), None))
+            elif isinstance(file_path, dict) and "path" in file_path:
+                history.append(((file_path["path"],), None))
+    
+    # Handle text message
+    if message["text"] is not None and message["text"].strip():
+        history.append((message["text"], None))
+    
+    return history, gr.MultimodalTextbox(value=None, interactive=False, file_types=["image"])
+
+def bot(history, api_key):
+    """Bot response function that processes images and text"""
+    if not history:
+        return history
+    
+    # Check if there are images in the history
+    images = []
+    uploaded_image_paths = []
+    for msg in history:
+        if isinstance(msg[0], tuple) and len(msg[0]) > 0:
+            # This is an image message
+            img_path = msg[0][0]
+            if os.path.exists(img_path):
+                try:
+                    img = Image.open(img_path)
+                    if img.mode == "RGBA":
+                        img = img.convert("RGBA")
+                    images.append(img)
+                    uploaded_image_paths.append(img_path)
+                except Exception as e:
+                    print(f"Error loading image {img_path}: {e}")
+    
+    # Get the text prompt from the last text message
+    text_prompt = None
+    for msg in reversed(history):
+        if isinstance(msg[0], str):
+            text_prompt = msg[0]
+            break
+    
+    if not images:
+        response = "**Please upload an image first!** I can help you edit images using Gemini AI."
+        history[-1][1] = ""
+        for character in response:
+            history[-1][1] += character
+            time.sleep(0.02)
+            yield history
+        return
+    
+    if not text_prompt:
+        response = f"**I see {len(images)} image(s) uploaded!** Please provide a text prompt telling me what you'd like to do with the image(s)."
+        history[-1][1] = ""
+        for character in response:
+            history[-1][1] += character
+            time.sleep(0.02)
+            yield history
+        return
+    
+    try:
+        # Generate image using Gemini
+        image_path, text_response = generate(text=text_prompt, images=images, api_key=api_key)
+        
+        if image_path:
+            # Add the generated image to chat
+            history.append((None, (image_path,)))
+            yield history
+            
+            # Add text response
+            response = f"**Image generated successfully!** Here's your edited image based on the prompt: '{text_prompt}'"
+            history.append((None, response))
+            yield history
+        else:
+            # Only text response
+            response = f"**Gemini Response:** {text_response}"
+            history[-1][1] = ""
+            for character in response:
+                history[-1][1] += character
+                time.sleep(0.02)
+                yield history
+                
+    except Exception as e:
+        response = f"**Error:** {str(e)}"
+        history[-1][1] = ""
+        for character in response:
+            history[-1][1] += character
+            time.sleep(0.02)
+            yield history
+
+# Create the Gradio interface
+with gr.Blocks(css="""
+    .header-container {
+        text-align: center;
+        padding: 20px;
+        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+        color: white;
+        border-radius: 10px;
+        margin-bottom: 20px;
+    }
+    .header-container img {
+        width: 50px;
+        height: 50px;
+        margin-bottom: 10px;
+    }
+    .header-container h1 {
+        margin: 10px 0;
+        font-size: 2.5em;
+    }
+    .header-container p {
+        margin: 5px 0;
+        font-size: 1.1em;
+    }
+    .header-container a {
+        color: #ffd700;
+        text-decoration: none;
+    }
+    .header-container a:hover {
+        text-decoration: underline;
+    }
+    .api-key-section {
+        background: #f8f9fa;
+        padding: 15px;
+        border-radius: 8px;
+        margin-bottom: 20px;
+        border: 1px solid #dee2e6;
+    }
+    .chat-container {
+        border: 1px solid #dee2e6;
+        border-radius: 10px;
+        padding: 20px;
+        background: white;
+    }
+""") as demo:
+    
+    # Custom HTML header
+    gr.HTML(
+    """
+    <div class="header-container">
+      <div>
+          <img src="https://www.gstatic.com/lamda/images/gemini_favicon_f069958c85030456e93de685481c559f160ea06b.png" alt="Gemini logo">
+      </div>
+      <div>
+          <h1>Chat with Gemini Image Editor</h1>
+          <p>Upload images and chat with AI to edit them! | 
+          <a href="https://aistudio.google.com/apikey">Get API Key</a> | 
+          Powered by <a href="https://gradio.app/">Gradio</a>⚡️</p>
+      </div>
+    </div>
+    """
+    )
+    
+    # API Key configuration
+    with gr.Row():
+        with gr.Column(scale=1):
+            api_key_input = gr.Textbox(
+                lines=1,
+                placeholder="Enter your Gemini API Key",
+                label="Gemini API Key",
+                type="password",
+                elem_classes="api-key-section"
+            )
+        with gr.Column(scale=2):
+            gr.Markdown("""
+            **Instructions:**
+            - Upload one or more images using the file upload button
+            - Type your editing instructions in the chat
+            - The AI will process your images based on your text prompt
+            - You can upload multiple images at once and chat about them
+            - Supported formats: PNG, JPG, JPEG, WEBP
+            """)
+    
+    # Chat interface
+    with gr.Row(elem_classes="chat-container"):
+        chatbot = gr.Chatbot(
+            [],
+            elem_id="chatbot",
+            bubble_full_width=False,
+            height=500,
+            show_label=False
+        )
+
+    # Chat input with file upload
+    chat_input = gr.MultimodalTextbox(
+        interactive=True, 
+        file_types=["image"], 
+        file_count="multiple",
+        placeholder="Upload one or more images and type your editing instructions...", 
+        show_label=False
+    )
+    
+    # Set up chat interactions
+    chat_msg = chat_input.submit(
+        add_message, 
+        [chatbot, chat_input, api_key_input], 
+        [chatbot, chat_input], 
+        queue=False
+    ).then(
+        bot, 
+        [chatbot, api_key_input], 
+        chatbot, 
+        api_name="bot_response"
+    )
+    
+    # Re-enable input after response
+    chat_msg.then(
+        lambda: gr.MultimodalTextbox(interactive=True), 
+        None, 
+        [chat_input], 
+        queue=False
+    )
+    
+    # Like/dislike functionality
+    chatbot.like(print_like_dislike, None, None)
+    
+    # Examples section
+    gr.Markdown("## Try these examples")
+    gr.Markdown("""
+    1. Upload an image and type: "change text to 'HELLO WORLD'"
+    2. Upload an image and type: "remove the background"
+    3. Upload an image and type: "add sunglasses to the person"
+    4. Upload an image and type: "make it look like a painting"
+    5. Upload multiple images and type: "apply the same style to all images"
+    6. Upload multiple images and type: "create a collage of these images"
+    """)
+
+# Launch the demo
+demo.queue()
+if __name__ == "__main__":
+    demo.launch(share=True)