updated analyzer.py

analyzer.py

@@ -192,6 +192,7 @@
 #             "error_type": type(e).__name__
 #         }
 
+# analyzer.py
 # analyzer.py
 
 import torch
@@ -207,42 +208,44 @@ formatter = logging.Formatter("[%(asctime)s] [%(levelname)s] - %(message)s")
 handler.setFormatter(formatter)
 logger.addHandler(handler)
 
-
 def analyze_code(tokenizer, model, language, code):
+    """
+    Analyze and fix buggy code using CodeT5+ model with 'fix:' prompt prefix.
+    Works across multiple programming languages.
+    """
     start_time = time.time()
 
-    prompt = (
-        f"Language: {language}\n"
-        f"Task: Fix the following buggy code and explain the bugs.\n"
-        f"Input Code:\n{code.strip()}\n\n"
-        f"Respond with a JSON like this:\n"
-        f"{{\n"
-        f"  \"bug_analysis\": [{{\"line_number\": X, \"error_message\": \"...\", \"explanation\": \"...\", \"fix_suggestion\": \"...\"}}],\n"
-        f"  \"corrected_code\": \"...\"\n"
-        f"}}"
-    )
+    # Prepare prompt in CodeT5+ style
+    prompt = f"fix: {code.strip()}"
+
+    logger.info(f"🔍 Starting analysis for language: {language}")
+    logger.info(f"🧾 Prompt: {prompt[:80]}...")
 
     try:
+        # Tokenize and generate response
         inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=512).to(model.device)
-        output = model.generate(**inputs, max_new_tokens=1024)
-        response = tokenizer.decode(output[0], skip_special_tokens=True)
+        with torch.no_grad():
+            output = model.generate(**inputs, max_new_tokens=1024)
+
+        # Decode output
+        response = tokenizer.decode(output[0], skip_special_tokens=True).strip()
 
-        # Clean response if needed
-        cleaned = response.strip().strip("```json").strip("```").strip()
-        json_output = json.loads(cleaned)
+        elapsed = round(time.time() - start_time, 2)
+        logger.info(f"✅ Inference completed in {elapsed}s")
 
         return {
-            "bug_analysis": json_output.get("bug_analysis", []),
-            "corrected_code": json_output.get("corrected_code", code)
+            "bug_analysis": [],  # Optional: You could add heuristics here
+            "corrected_code": response
         }
 
     except Exception as e:
+        logger.error(f"❌ Error during analysis: {e}")
         return {
             "bug_analysis": [{
                 "line_number": 0,
-                "error_message": "Failed to parse",
+                "error_message": "Inference failed",
                 "explanation": str(e),
-                "fix_suggestion": "Try simplifying the code."
+                "fix_suggestion": "Try again with simpler code or retry later"
             }],
             "corrected_code": code
         }