Update app.py

app.py

@@ -1,3 +1,4 @@
+
 from fastapi import FastAPI, UploadFile, File, Form, HTTPException, Request
 from fastapi.staticfiles import StaticFiles
 from fastapi.responses import RedirectResponse, JSONResponse, HTMLResponse
@@ -60,7 +61,7 @@ app.mount("/static", StaticFiles(directory="static"), name="static")
 app.mount("/images", StaticFiles(directory="images"), name="images")
 
 # Gemini API Configuration
-API_KEY = "AIzaSyCwmgD8KxzWiuivtySNtcZF_rfTvx9s9sY"
+API_KEY = "AIzaSyCwmgD8KxzWiuivtySNtcZF_rfTvx9s9sY"  # Replace with your actual API key
 genai.configure(api_key=API_KEY)
 
 # Model configurations
@@ -192,21 +193,22 @@ def translate_text(text: str, target_language: str):
             early_stopping=True
         )
         translated_text = translation_tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)[0]
-        logger.info(f"Translation took {time.time() - start_time:.2f} seconds")
-        
-        return translated_text
-        
-    except Exception as e:
-        logger.error(f"Translation error: {str(e)}", exc_info=True)
-        return f"Translation error: {str(e)}"
+        logger
 
 def detect_intent(text: str = None, file: UploadFile = None) -> tuple[str, str]:
     """Enhanced intent detection with dynamic translation and file translation support"""
     target_language = "English"  # Default
     
+    # Handle image captioning first: if a file is an image and no text is provided, assume image-to-text
+    if file:
+        content_type = file.content_type.lower() if file.content_type else ""
+        filename = file.filename.lower() if file.filename else ""
+        
+        if content_type.startswith('image/') and not text:
+            return "image-to-text", target_language
+    
     if file and text:
         text_lower = text.lower()
-        filename = file.filename.lower() if file.filename else ""
         
         # Check for file translation intent
         translate_patterns = [
@@ -223,7 +225,6 @@ def detect_intent(text: str = None, file: UploadFile = None) -> tuple[str, str]:
                     return "file-translate", target_language
         
         # Image-related intents
-        content_type = file.content_type.lower() if file.content_type else ""
         if content_type.startswith('image/') and text:
             if "what’s this" in text_lower or "does this fly" in text_lower or ("fly" in text_lower and any(q in text_lower for q in ['does', 'can', 'will'])):
                 return "visual-qa", target_language
@@ -378,7 +379,7 @@ async def process_input(
             
             content = await extract_text_from_file(file)
             if not content.strip():
-                raise HTTPException(status_code=400, detail="No text could be extracted from the file")
+                raise HTTPException(status_code=400, detail=" voisins du texte could be extracted from the file")
             
             # Split content into chunks to handle large files
             max_chunk_size = 512