#!/usr/bin/env python3 """ AI功能演示脚本 展示改进后的智谱AI集成功能 """ import os import sys import time # 确保项目根目录在Python路径中 from pathlib import Path project_root = Path(__file__).parent.parent.absolute() if str(project_root) not in sys.path: sys.path.insert(0, str(project_root)) def demo_ai_setup_guide(): """演示AI设置指南""" print("🤖 智谱AI功能演示") print("=" * 60) from server.config.ai_config import print_ai_setup_guide print_ai_setup_guide() def demo_ai_config(): """演示AI配置功能""" print("\n📋 AI配置演示") print("-" * 40) from server.config.ai_config import get_ai_config config = get_ai_config() print(f"当前模型: {config.model}") print(f"API密钥已设置: {config.is_enabled()}") print(f"可用模型: {', '.join(config.get_available_models())}") # 演示模型切换 print("\n🔄 模型切换演示:") models_to_test = ["glm-4-flash", "glm-4", "glm-4-plus"] for model in models_to_test: if config.set_model(model): print(f"✅ 成功切换到: {model}") else: print(f"❌ 切换失败: {model}") # 恢复默认模型 config.set_model("glm-4-flash") print(f"🔄 恢复默认模型: {config.model}") def demo_zhipu_client(): """演示智谱AI客户端功能""" print("\n🔌 智谱AI客户端演示") print("-" * 40) api_key = os.getenv('ZHIPU_API_KEY') if not api_key: print("⚠️ 未设置ZHIPU_API_KEY环境变量") print("💡 请设置API密钥以体验完整功能") return try: from server.ai.zhipu_client import ZhipuClient client = ZhipuClient(api_key) # 显示客户端信息 model_info = client.get_model_info() print(f"📋 客户端信息:") for key, value in model_info.items(): print(f" {key}: {value}") # 测试连接 print("\n🔗 测试API连接...") if client.test_connection(): print("✅ API连接成功") # 演示简单对话 print("\n💬 简单对话演示:") test_messages = [ "你好，请简短介绍一下自己", "你能做什么？", "请用一句话总结人工智能的作用" ] for i, message in enumerate(test_messages, 1): print(f"\n用户 {i}: {message}") response = client.simple_chat(message, "你是一个友好的AI助手，请简洁回复。") if response: print(f"AI: {response}") else: print("AI: [无回复]") time.sleep(1) # 避免请求过快 else: print("❌ API连接失败") except Exception as e: print(f"❌ 客户端演示失败: {e}") def demo_ai_manager(): """演示AI管理器功能""" print("\n🧠 AI管理器演示") print("-" * 40) api_key = os.getenv('ZHIPU_API_KEY') try: from server.ai.ai_manager import AIManager manager = AIManager(api_key) # 显示AI状态 status = manager.get_ai_status() print(f"📊 AI状态:") for key, value in status.items(): print(f" {key}: {value}") if not manager.is_enabled(): print("⚠️ AI功能未启用，跳过对话演示") return # 演示消息响应判断 print("\n🎯 消息响应判断演示:") test_messages = [ ("你好", "群聊"), ("@AI 你好", "群聊"), ("这是一个问题吗？", "群聊"), ("普通消息", "群聊"), ("AI能帮我吗", "群聊"), ("私聊消息", "私聊"), ] for message, chat_type in test_messages: is_group = chat_type == "群聊" should_respond = manager.should_respond_to_message(message, is_group) status = "会回复" if should_respond else "不回复" print(f" {chat_type} '{message}' -> {status}") # 演示AI对话处理 print("\n💬 AI对话处理演示:") test_conversations = [ (1, "Alice", "你好，AI助手！", 1), # 群聊 (2, "Bob", "你能帮我解释一下Python吗？", None), # 私聊 ] for user_id, username, message, chat_group_id in test_conversations: chat_type = "群聊" if chat_group_id else "私聊" print(f"\n{chat_type} - {username}: {message}") response = manager.process_message(user_id, username, message, chat_group_id) if response: print(f"AI: {response}") else: print("AI: [不回复]") time.sleep(1) # 避免请求过快 except Exception as e: print(f"❌ AI管理器演示失败: {e}") def demo_context_management(): """演示上下文管理功能""" print("\n📚 上下文管理演示") print("-" * 40) try: from server.ai.context_manager import ContextManager manager = ContextManager() # 模拟群聊对话 print("🗣️ 模拟群聊对话:") group_id = "1" conversations = [ ("user", "Alice: 你好，AI助手！"), ("assistant", "你好Alice！很高兴见到你，有什么可以帮助你的吗？"), ("user", "Bob: 我想了解Python编程"), ("assistant", "Python是一门很棒的编程语言！它简洁易学，应用广泛。"), ("user", "Alice: 能推荐一些学习资源吗？"), ] for role, content in conversations: manager.add_message(group_id, role, content, is_group=True) print(f" {content}") # 获取上下文 context = manager.get_context(group_id, is_group=True) print(f"\n📋 群聊上下文包含 {len(context)} 条消息") # 模拟私聊对话 print("\n💬 模拟私聊对话:") user_id = "100" private_conversations = [ ("user", "用户私聊: 这是私密问题"), ("assistant", "我理解这是私聊，我会保护您的隐私。"), ] for role, content in private_conversations: manager.add_message(user_id, role, content, is_group=False) print(f" {content}") # 获取私聊上下文 private_context = manager.get_context(user_id, is_group=False) print(f"\n📋 私聊上下文包含 {len(private_context)} 条消息") # 显示上下文摘要 summary = manager.get_context_summary() print(f"\n📊 上下文摘要:") for key, value in summary.items(): print(f" {key}: {value}") # 显示系统提示词 group_prompt = manager.get_system_prompt("group") private_prompt = manager.get_system_prompt("private") print(f"\n📝 系统提示词:") print(f" 群聊提示词: {group_prompt[:50]}...") print(f" 私聊提示词: {private_prompt[:50]}...") except Exception as e: print(f"❌ 上下文管理演示失败: {e}") def demo_model_comparison(): """演示不同模型的对比""" print("\n🔬 模型对比演示") print("-" * 40) api_key = os.getenv('ZHIPU_API_KEY') if not api_key: print("⚠️ 未设置ZHIPU_API_KEY环境变量，跳过模型对比") return try: from server.ai.zhipu_client import ZhipuClient # 测试不同模型 models_to_test = ["glm-4-flash", "glm-4"] test_question = "请用一句话解释什么是人工智能" for model in models_to_test: print(f"\n🤖 测试模型: {model}") client = ZhipuClient(api_key) if client.set_model(model): response = client.simple_chat(test_question, "请简洁回复") if response: print(f"回复: {response}") else: print("回复: [无回复]") else: print(f"❌ 模型 {model} 不可用") time.sleep(2) # 避免请求过快 except Exception as e: print(f"❌ 模型对比演示失败: {e}") def main(): """主演示函数""" print("🎭 智谱AI功能完整演示") print("=" * 60) # 检查API密钥 api_key = os.getenv('ZHIPU_API_KEY') if api_key: print(f"✅ 检测到API密钥: {api_key[:8]}...") print("🚀 将演示完整功能") else: print("⚠️ 未检测到API密钥") print("🔧 将演示基础功能和配置") # 运行各个演示 demos = [ demo_ai_setup_guide, demo_ai_config, demo_zhipu_client, demo_ai_manager, demo_context_management, demo_model_comparison, ] for demo_func in demos: try: demo_func() except KeyboardInterrupt: print("\n\n👋 演示被用户中断") break except Exception as e: print(f"\n❌ 演示 {demo_func.__name__} 出现错误: {e}") print("\n" + "=" * 60) print("🎉 AI功能演示完成！") print("\n💡 要在聊天室中使用AI功能:") print(" 1. 设置环境变量: export ZHIPU_API_KEY='your_key'") print(" 2. 启动服务器: python -m server.main") print(" 3. 启动客户端: python -m client.main") print(" 4. 在群聊中@AI或私聊AI用户") if __name__ == "__main__": try: main() except KeyboardInterrupt: print("\n\n👋 演示被用户中断") except Exception as e: print(f"\n❌ 演示出现错误: {e}")