| from typing import Dict, List |
| import json |
| import google.generativeai as genai |
| from fastapi import HTTPException |
| from ..utils.prompt_templates import generate_traditional_prompt, generate_body_based_prompt |
|
|
| class SymptomAnalyzer: |
| def __init__(self): |
| self.model = genai.GenerativeModel('gemini-2.5-flash') |
| |
| async def analyze_traditional(self, symptoms: List[str], age: int, gender: str, duration: str) -> Dict: |
| try: |
| prompt = generate_traditional_prompt(symptoms, age, gender, duration) |
| response = self.model.generate_content(prompt) |
| |
| try: |
| parsed_response = json.loads(response.text) |
| return parsed_response |
| except json.JSONDecodeError as e: |
| cleaned_response = response.text.strip().strip('`').strip('json') |
| try: |
| parsed_response = json.loads(cleaned_response) |
| return parsed_response |
| except json.JSONDecodeError: |
| raise HTTPException( |
| status_code=500, |
| detail="Failed to parse response into valid JSON" |
| ) |
| |
| except Exception as e: |
| raise HTTPException(status_code=500, detail=f"Analysis error: {str(e)}") |
|
|
| async def analyze_body_based(self, data: Dict) -> Dict: |
| try: |
| prompt = generate_body_based_prompt(data) |
| response = self.model.generate_content(prompt) |
| |
| try: |
| parsed_response = json.loads(response.text) |
| return parsed_response |
| except json.JSONDecodeError as e: |
| cleaned_response = response.text.strip().strip('`').strip('json') |
| try: |
| parsed_response = json.loads(cleaned_response) |
| return parsed_response |
| except json.JSONDecodeError: |
| raise HTTPException( |
| status_code=500, |
| detail="Failed to parse response into valid JSON" |
| ) |
| |
| except Exception as e: |
| raise HTTPException(status_code=500, detail=f"Analysis error: {str(e)}") |
