Update app.py

app.py

@@ -7,63 +7,63 @@ import json
 # Initialize sentiment analysis pipeline
 sentiment_analyzer = pipeline("sentiment-analysis", model="distilbert-base-uncased-finetuned-sst-2-english")
 
-# FastAPI app for API endpoint
+# FastAPI app
 app = FastAPI()
 
 @app.post("/calculate_scores")
 async def calculate_scores(request: Request):
-    logs = await request.json()
-    quality_score = 0
-    timeliness_score = 0
-    safety_score = 0
-    communication_score = 0
-    quality_count = 0
-    delay_count = 0
-    incident_count = 0
-    feedback_count = 0
+    try:
+        logs = await request.json()
+        quality_score = 0
+        timeliness_score = 0
+        safety_score = 0
+        communication_score = 0
+        quality_count = 0
+        delay_count = 0
+        incident_count = 0
+        feedback_count = 0
+
+        for log in logs:
+            log_type = log.get("log_type")
+            if log_type == "Quality" and "quality_score" in log and log["quality_score"] is not None:
+                try:
+                    quality_score += float(log["quality_score"])
+                    quality_count += 1
+                except (ValueError, TypeError):
+                    pass
+            if log_type == "Delay" and "delay_percentage" in log and log["delay_percentage"] is not None:
+                try:
+                    timeliness_score += (100 - float(log["delay_percentage"]))
+                    delay_count += 1
+                except (ValueError, TypeError):
+                    pass
+            if log_type == "Incident":
+                safety_score += (100 if log.get("safety_compliance", False) else 0)
+                incident_count += 1
+            if "feedback" in log and log["feedback"]:
+                feedback = log["feedback"]
+                sentiment = sentiment_analyzer(feedback)[0]
+                sentiment_score = sentiment["score"] * 100 if sentiment["label"] == "POSITIVE" else 50
+                base_score = 90 if len(feedback) > 50 else 70
+                communication_score += base_score * (sentiment_score / 100)
+                feedback_count += 1
 
-    for log in logs:
-        log_type = log.get("log_type")
-        # Safely handle quality_score
-        quality_value = log.get("quality_score")
-        if log_type == "Quality" and quality_value != None:
-            try:
-                quality_score += float(quality_value)
-                quality_count += 1
-            except (ValueError, TypeError):
-                pass
-        # Safely handle delay_percentage
-        delay_value = log.get("delay_percentage")
-        if log_type == "Delay" and delay_value != None:
-            try:
-                timeliness_score += (100 - float(delay_value))
-                delay_count += 1
-            except (ValueError, TypeError):
-                pass
-        if log_type == "Incident":
-            safety_score += (100 if log.get("safety_compliance", False) else 0)
-            incident_count += 1
-        if log.get("feedback"):
-            feedback = log["feedback"]
-            sentiment = sentiment_analyzer(feedback)[0]
-            sentiment_score = sentiment["score"] * 100 if sentiment["label"] == "POSITIVE" else 50
-            base_score = 90 if len(feedback) > 50 else 70
-            communication_score += base_score * (sentiment_score / 100)
-            feedback_count += 1
+        quality_score = quality_score / quality_count if quality_count > 0 else 0
+        timeliness_score = timeliness_score / delay_count if delay_count > 0 else 0
+        safety_score = safety_score / incident_count if incident_count > 0 else 0
+        communication_score = communication_score / feedback_count if feedback_count > 0 else 0
 
-    # Calculate averages
-    quality_score = quality_score / quality_count if quality_count > 0 else 0
-    timeliness_score = timeliness_score / delay_count if delay_count > 0 else 0
-    safety_score = safety_score / incident_count if incident_count > 0 else 0
-    communication_score = communication_score / feedback_count if feedback_count > 0 else 0
+        alert_flag = quality_score < 50 or timeliness_score < 50
 
-    return {
-        "quality_score": round(quality_score, 2),
-        "timeliness_score": round(timeliness_score, 2),
-        "safety_score": round(safety_score, 2),
-        "communication_score": round(communication_score, 2),
-        "alert_flag": quality_score < 50 or timeliness_score < 50
-    }
+        return {
+            "quality_score": round(quality_score, 2),
+            "timeliness_score": round(timeliness_score, 2),
+            "safety_score": round(safety_score, 2),
+            "communication_score": round(communication_score, 2),
+            "alert_flag": alert_flag
+        }
+    except Exception as e:
+        return {"error": str(e)}, 400
 
 # Gradio interface for testing
 def test_api(logs_json):
@@ -74,7 +74,6 @@ def test_api(logs_json):
     except Exception as e:
         return f"Error: {str(e)}"
 
-# Launch Gradio UI
 gr.Interface(
     fn=test_api,
     inputs=gr.Textbox(label="Enter Logs JSON", placeholder='[{"log_type": "Quality", "quality_score": 80, "feedback": "Great work!"}]'),
@@ -82,6 +81,6 @@ gr.Interface(
     title="Subcontractor Scoring API Tester"
 ).launch()
 
-# Run FastAPI with Uvicorn
 if __name__ == "__main__":
     uvicorn.run(app, host="0.0.0.0", port=7860)
+