面向慢病管理的智能Skill记忆体系：跨轮次交互、结构化数据与健康图谱构建.146

# 安装依赖
# pip install json matplotlib datetime uuid
import json
import uuid
from datetime import datetime
import matplotlib.pyplot as plt
import os

def generate_user_id():
    """
    生成全局唯一用户ID，实现不同用户数据隔离
    是状态持久化的基础：一人一档案，数据不混淆
    """
    return str(uuid.uuid4())

# 示例：创建用户
user_id = generate_user_id()
print(f"慢病管理系统用户唯一ID：{user_id}")

def parse_diabetes_input(input_text: str, measure_time=None):
    """
    糖尿病Skill：解析自然语言为结构化血糖数据
    :param input_text: 用户输入文本
    :param measure_time: 测量时间，默认当前时间
    :return: 结构化血糖数据
    """
    if measure_time is None:
        measure_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
    
    # 简化版语义解析（工业级可对接大模型API）
    glucose_value = 0.0
    measure_type = "未知"
    
    if "空腹" in input_text:
        measure_type = "空腹"
    if "餐后2小时" in input_text:
        measure_type = "餐后2h"
    
    # 提取数值
    for word in input_text.split():
        word = word.replace("mmol/L", "")
        try:
            glucose_value = float(word)
            break
        except:
            continue
    
    # 医学规则判断血糖状态
    status = "正常"
    if measure_type == "空腹":
        if glucose_value > 6.1:
            status = "偏高"
        elif glucose_value < 3.9:
            status = "偏低"
    elif measure_type == "餐后2h":
        if glucose_value > 7.8:
            status = "偏高"
    
    return {
        "measure_time": measure_time,
        "glucose_value": glucose_value,
        "measure_type": measure_type,
        "status": status
    }

def save_user_skill_data(user_id: str, skill_type: str, data: dict):
    """
    状态持久化核心函数：将Skill数据保存到本地文件
    :param user_id: 用户唯一ID
    :param skill_type: 技能类型（diabetes/hypertension/diet）
    :param data: 结构化数据
    """
    filename = f"{user_id}_{skill_type}.json"
    
    # 读取历史数据（若存在）
    history_data = []
    if os.path.exists(filename):
        with open(filename, "r", encoding="utf-8") as f:
            history_data = json.load(f)
    
    # 追加新数据
    history_data.append(data)
    
    # 写入文件（持久化）
    with open(filename, "w", encoding="utf-8") as f:
        json.dump(history_data, f, ensure_ascii=False, indent=2)
    
    print(f"✅ 数据已持久化保存至：{filename}")
    return True

def get_user_skill_history(user_id: str, skill_type: str):
    """
    读取用户持久化的历史数据，用于病程追踪
    """
    filename = f"{user_id}_{skill_type}.json"
    if not os.path.exists(filename):
        return []
    with open(filename, "r", encoding="utf-8") as f:
        return json.load(f)

def analyze_glucose_trend(user_id: str):
    """
    糖尿病病程追踪：分析血糖趋势，生成统计报告
    """
    data_list = get_user_skill_history(user_id, "diabetes")
    if not data_list:
        return "暂无血糖数据"
    
    # 统计计算
    values = [item["glucose_value"] for item in data_list]
    times = [item["measure_time"][5:16] for item in data_list]
    avg_value = sum(values) / len(values)
    max_value = max(values)
    min_value = min(values)
    
    # 异常统计
    abnormal_count = len([item for item in data_list if item["status"] != "正常"])
    
    report = f"""
📊 糖尿病血糖病程分析报告
总记录次数：{len(data_list)}
平均血糖值：{avg_value:.2f} mmol/L
最高血糖值：{max_value:.2f} mmol/L
最低血糖值：{min_value:.2f} mmol/L
异常记录次数：{abnormal_count}
    """
    return report, times, values

def plot_glucose_trend(user_id: str, save_path=None):
    """
    绘制血糖趋势图，保存为图片（病程追踪可视化）
    """
    
    data_list = get_user_skill_history(user_id, "diabetes")
    
    if not data_list:
        print("暂无血糖数据")
        return
    
    if save_path is None:
        save_path = f"{user_id}_glucose_trend.png"
    
    plt.rcParams["font.sans-serif"] = ["SimHei", "DejaVu Sans"]
    plt.rcParams["axes.unicode_minus"] = False
    
    values = [item["glucose_value"] for item in data_list]
    times = [item["measure_time"][5:16] for item in data_list]
    
    plt.figure(figsize=(12, 6))
    plt.plot(times, values, marker="o", linewidth=2, color="#e63946", label="血糖值")
    plt.axhline(y=6.1, color="orange", linestyle="--", label="空腹血糖上限")
    plt.title(f"用户 {user_id[:8]}... 血糖趋势追踪图", fontsize=14)
    plt.xlabel("测量时间", fontsize=12)
    plt.ylabel("血糖值 (mmol/L)", fontsize=12)
    plt.xticks(rotation=45)
    plt.grid(alpha=0.3)
    plt.legend()
    plt.tight_layout()
    plt.savefig(save_path, dpi=150)
    plt.close()
    
    print(f"✅ 血糖趋势图已保存：{save_path}")
    return save_path

if __name__ == "__main__":
    # 1. 创建用户
    user_id = generate_user_id()
    print(f"新用户ID：{user_id}")

    # 2. 模拟用户连续输入血糖数据
    user_inputs = [
        "2025-12-20 08:00 空腹血糖6.5mmol/L",
        "2025-12-21 08:10 空腹血糖5.9mmol/L",
        "2025-12-22 08:05 空腹血糖7.2mmol/L",
        "2025-12-23 07:50 空腹血糖6.8mmol/L"
    ]

    # 3. 解析并持久化数据
    for text in user_inputs:
        data = parse_diabetes_input(text)
        save_user_skill_data(user_id, "diabetes", data)

    # 4. 生成病程报告
    report, _, _ = analyze_glucose_trend(user_id)
    print(report)

    # 5. 生成趋势图
    plot_glucose_trend(user_id)

[
  {
    "measure_time": "2025-12-20 08:00",
    "glucose_value": 6.5,
    "measure_type": "空腹",
    "status": "偏高"
  },
  {
    "measure_time": "2025-12-21 08:10",
    "glucose_value": 5.9,
    "measure_type": "空腹",
    "status": "正常"
  },
  {
    "measure_time": "2025-12-22 08:05",
    "glucose_value": 7.2,
    "measure_type": "空腹",
    "status": "偏高"
  },
  {
    "measure_time": "2025-12-23 07:50",
    "glucose_value": 6.8,
    "measure_type": "空腹",
    "status": "偏高"
  }
]