Update app.py

app.py

@@ -30,28 +30,30 @@ def call_openai_image_api(prompt: str, api_key: str, input_image: Image.Image |
 
     headers = {"Authorization": f"Bearer {api_key}"}
     # Hypothetical model name from the original code. Replace with "dall-e-2" or "dall-e-3" if needed.
-    model = "gpt-image-1"
+    model = "gpt-image-1" # Using the model specified in the original code
     size = "1024x1024"
+    response = None # Initialize response variable
 
-   try:
+    try:
         if input_image:
             # --- Image Editing ---
             if not isinstance(input_image, Image.Image):
-                return None, None, "Error: Invalid image provided for editing."
+                 return None, None, "Error: Invalid image provided for editing."
 
-            # CORRECTED INDENTATION for this block:
             # Convert PIL Image to bytes for the API request
             byte_stream = io.BytesIO()
-            input_image.save(byte_stream, format="PNG") # Save PIL image to bytes buffer
+            input_image.save(byte_stream, format="PNG") # Save PIL image to bytes buffer [[1]]
             byte_stream.seek(0) # Rewind buffer to the beginning
 
             files = {
                 "image": ("input_image.png", byte_stream, "image/png"),
             }
+            # CORRECTED data dictionary: removed 'response_format'
             data = {
                 "prompt": prompt,
                 "model": model,
                 "size": size,
+                # "response_format": "b64_json", # <-- THIS LINE IS REMOVED
             }
             api_url = "https://api.openai.com/v1/images/edits"
             print("Calling OpenAI Image Edit API...") # Debug print
@@ -59,14 +61,14 @@ def call_openai_image_api(prompt: str, api_key: str, input_image: Image.Image |
 
         else:
             # --- Image Generation ---
-            # (This part remains the same)
+            # (This part remains the same as it uses response_format correctly via json payload)
             headers["Content-Type"] = "application/json"
             payload = {
                 "prompt": prompt,
                 "model": model,
                 "response_format": "b64_json", # Keep this for generation
                 "size": size,
-                "n": 1,
+                "n": 1, # Generate one image
             }
             api_url = "https://api.openai.com/v1/images/generations"
             print("Calling OpenAI Image Generation API...") # Debug print
@@ -77,25 +79,31 @@ def call_openai_image_api(prompt: str, api_key: str, input_image: Image.Image |
 
         # Process successful response
         response_data = response.json()
-        if not response_data.get("data") or not response_data["data"][0].get("b64_json"):
-             return input_image, None, f"Error: Unexpected API response format: {response_data}"
+        # Ensure the expected data structure is present
+        if not response_data.get("data") or not isinstance(response_data["data"], list) or len(response_data["data"]) == 0:
+            return input_image, None, f"Error: Unexpected API response format - 'data' array missing or empty: {response_data}"
+        if not response_data["data"][0].get("b64_json"):
+             return input_image, None, f"Error: Unexpected API response format - 'b64_json' key missing: {response_data}"
 
         img_b64 = response_data["data"][0]["b64_json"]
-        img_bytes = base64.b64decode(img_b64)
-        result_image = Image.open(io.BytesIO(img_bytes))
+        img_bytes = base64.b64decode(img_b64) # Decode base64 string [[1]]
+        result_image = Image.open(io.BytesIO(img_bytes)) # Convert bytes to PIL Image [[1]]
         print("Image processed successfully.") # Debug print
         return input_image, result_image, "Success!"
 
     except requests.exceptions.RequestException as e:
         error_message = f"API Request Error: {e}"
-        try:
-            # Attempt to get more specific error from OpenAI response
-            error_detail = response.json()
-            error_message += f"\nAPI Error Details: {error_detail}"
-        except:
-             # Fallback if response is not JSON or parsing fails
-             if response:
-                 error_message += f"\nRaw Response Text: {response.text}"
+        # Check if response exists before trying to access its attributes/methods
+        if response is not None:
+            try:
+                # Attempt to get more specific error from OpenAI response
+                error_detail = response.json()
+                error_message += f"\nAPI Error Details: {error_detail}"
+            except requests.exceptions.JSONDecodeError:
+                # Fallback if response is not JSON
+                error_message += f"\nRaw Response Text: {response.text}"
+            except Exception as json_e:
+                 error_message += f"\nError parsing JSON response: {json_e}\nRaw Response Text: {response.text}"
         print(error_message) # Debug print
         return input_image, None, error_message
     except Exception as e:
@@ -106,7 +114,7 @@ def call_openai_image_api(prompt: str, api_key: str, input_image: Image.Image |
 
 # --- Gradio Interface Setup ---
 
-# Check for API key in environment variables [[3], [5]]
+# Check for API key in environment variables
 api_key_env = os.environ.get("OPENAI_API_KEY")
 api_key_present_info = "OpenAI API Key found in environment variables." if api_key_env else "OpenAI API Key not found in environment variables. Please enter it below."
 
@@ -128,23 +136,23 @@ def process_image_request(prompt_input, api_key_input, uploaded_image):
     return original_img, result_img, status
 
 
-# Build the Gradio interface using Blocks for more layout control
+# Build the Gradio interface using Blocks for more layout control [[7]]
 with gr.Blocks() as demo:
-    gr.Markdown("# OpenAI GPT-Image-1 Text-to-Image Demo") # App title [[7]]
+    gr.Markdown("# OpenAI GPT-Image-1 Text-to-Image Demo") # App title
     gr.Markdown("Enter a prompt to generate an image, or upload an image and enter a prompt to edit it.")
 
     with gr.Row():
         with gr.Column(scale=1):
-            prompt_input = gr.Textbox(label="Image Description (Prompt)", lines=3, placeholder="e.g., A futuristic cityscape at sunset") # Text input for prompt [[1], [5]]
+            prompt_input = gr.Textbox(label="Image Description (Prompt)", lines=3, placeholder="e.g., A futuristic cityscape at sunset") # Text input for prompt
             gr.Markdown(f"*{api_key_present_info}*")
-            api_key_input = gr.Textbox(label="OpenAI API Key", type="password", placeholder="Enter your key if not set in environment") # Password input for API key [[7]]
-            uploaded_image_input = gr.Image(type="pil", label="Upload Image to Edit (Optional)") # Image upload [[2]]
+            api_key_input = gr.Textbox(label="OpenAI API Key", type="password", placeholder="Enter your key if not set in environment") # Password input for API key
+            uploaded_image_input = gr.Image(type="pil", label="Upload Image to Edit (Optional)") # Image upload [[4]]
             submit_button = gr.Button("Generate / Edit Image")
         with gr.Column(scale=2):
             status_output = gr.Textbox(label="Status", interactive=False)
             with gr.Row():
                  original_image_output = gr.Image(label="Original Image", interactive=False)
-                 result_image_output = gr.Image(label="Generated / Edited Image", interactive=False) # Display output image [[6]]
+                 result_image_output = gr.Image(label="Generated / Edited Image", interactive=False) # Display output image
 
     # Connect the button click event to the processing function
     submit_button.click(
@@ -153,6 +161,6 @@ with gr.Blocks() as demo:
         outputs=[original_image_output, result_image_output, status_output]
     )
 
-# Launch the Gradio app
+# Launch the Gradio app [[2]]
 if __name__ == "__main__":
     demo.launch()