chore: release v3.3.1.0

Version/3.3.1.0

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -3,17 +3,6 @@ description: 报错或漏洞请使用这个模板创建，不使用此模板创
 title: "[Bug]: "
 labels: ["bug?"]
 body:
-  - type: dropdown
-    attributes:
-      label: 部署方式
-      description: "主程序使用的部署方式"
-      options:
-        - 手动部署
-        - 安装器部署
-        - 一键安装包部署
-        - Docker部署
-    validations:
-      required: true
   - type: dropdown
     attributes:
       label: 消息平台适配器
@@ -27,37 +16,24 @@ body:
       required: false
   - type: input
     attributes:
-      label: 系统环境
-      description: 操作系统、系统架构、**主机地理位置**，地理位置最好写清楚，涉及网络问题排查。
-      placeholder: 例如： CentOS x64 中国大陆、Windows11 美国
-    validations:
-      required: true
-  - type: input
-    attributes:
-      label: Python环境
-      description: 运行程序的Python版本
-      placeholder: 例如： Python 3.10
+      label: 运行环境
+      description: 操作系统、系统架构、**Python版本**、**主机地理位置**
+      placeholder: 例如： CentOS x64 Python 3.10.3、Docker 的直接写 Docker 就行
     validations:
       required: true
   - type: input
     attributes:
       label: QChatGPT版本
       description: QChatGPT版本号
-      placeholder: 例如： v2.6.0，可以使用`!version`命令查看
+      placeholder: 例如：v3.3.0，可以使用`!version`命令查看，或者到 pkg/utils/constants.py 查看
     validations:
       required: true
   - type: textarea
     attributes:
       label: 异常情况
-      description: 完整描述异常情况，什么时候发生的、发生了什么，尽可能详细
+      description: 完整描述异常情况，什么时候发生的、发生了什么。**请附带日志信息。**
     validations:
       required: true
-  - type: textarea
-    attributes:
-      label: 日志信息
-      description: 请提供完整的 **登录框架 和 QChatGPT控制台**的相关日志信息（若有），不提供日志信息**无法**为您排查问题，请尽可能详细
-    validations:
-      required: false
   - type: textarea
     attributes:
       label: 启用的插件

.github/pull_request_template.md

@@ -2,24 +2,16 @@
 
 实现/解决/优化的内容: 
 
-### 事务
+## 检查清单
 
-- [ ] 已阅读仓库[贡献指引](https://github.com/RockChinQ/QChatGPT/blob/master/CONTRIBUTING.md)
-- [ ] 已与维护者在issues或其他平台沟通此PR大致内容
+### PR 作者完成
 
-## 以下内容可在起草PR后、合并PR前逐步完成
+- [ ] 阅读仓库[贡献指引](https://github.com/RockChinQ/QChatGPT/blob/master/CONTRIBUTING.md)了吗？
+- [ ] 与项目所有者沟通过了吗？
 
-### 功能
+### 项目所有者完成
 
-- [ ] 已编写完善的配置文件字段说明（若有新增）
-- [ ] 已编写面向用户的新功能说明（若有必要）
-- [ ] 已测试新功能或更改
-
-### 兼容性
-
-- [ ] 已处理版本兼容性
-- [ ] 已处理插件兼容问题
-
-### 风险
-
-可能导致或已知的问题: 
+- [ ] 相关 issues 链接了吗？
+- [ ] 配置项写好了吗？迁移写好了吗？生效了吗？
+- [ ] 依赖写到 requirements.txt 和 core/bootutils/deps.py 了吗
+- [ ] 文档编写了吗？

README.md

@@ -1,6 +1,6 @@
 
 <p align="center">
-<img src="https://qchatgpt.rockchin.top/logo.png" alt="QChatGPT" width="180" />
+<img src="https://qchatgpt.rockchin.top/chrome-512.png" alt="QChatGPT" width="180" />
 </p>
 <div align="center">
 
@@ -19,7 +19,7 @@
 <a href="http://qm.qq.com/cgi-bin/qm/qr?_wv=1027&k=66-aWvn8cbP4c1ut_1YYkvvGVeEtyTH8&authKey=pTaKBK5C%2B8dFzQ4XlENf6MHTCLaHnlKcCRx7c14EeVVlpX2nRSaS8lJm8YeM4mCU&noverify=0&group_code=195992197">
 <img alt="Static Badge" src="https://img.shields.io/badge/%E5%AE%98%E6%96%B9%E7%BE%A4-195992197-purple">
 </a>
-<a href="https://qm.qq.com/q/1yxEaIgXMA">
+<a href="https://qm.qq.com/q/PClALFK242">
 <img alt="Static Badge" src="https://img.shields.io/badge/%E7%A4%BE%E5%8C%BA%E7%BE%A4-619154800-purple">
 </a>
 <a href="https://codecov.io/gh/RockChinQ/QChatGPT" > 
@@ -45,11 +45,11 @@
 <hr/>
 
 <div align="center">
-京东云4090单卡15C90G实例 <br/>
+京东云 4090 单卡 15C90G 实例 <br/>
     仅需1.89/小时，包月1225元起  <br/>
     可选预装Stable Diffusion等应用，随用随停，计费透明，欢迎首选支持 <br/>
-https://3.cn/1ZOi6-Gj
+https://3.cn/24A-2NXd
 </div>
 
-<img alt="回复效果（带有联网插件）" src="https://qchatgpt.rockchin.top/assets/image/QChatGPT-0516.png" width="500px"/>
+<img alt="回复效果（带有联网插件）" src="https://qchatgpt.rockchin.top/QChatGPT-0516.png" width="500px"/>
 </div>

pkg/command/cmdmgr.py

@@ -8,7 +8,7 @@ from . import entities, operator, errors
 from ..config import manager as cfg_mgr
 
 # 引入所有算子以便注册
-from .operators import func, plugin, default, reset, list as list_cmd, last, next, delc, resend, prompt, cmd, help, version, update
+from .operators import func, plugin, default, reset, list as list_cmd, last, next, delc, resend, prompt, cmd, help, version, update, ollama, model
 
 
 class CommandManager:

pkg/command/operators/model.py

@@ -0,0 +1,86 @@
+from __future__ import annotations
+
+import typing
+
+from .. import operator, entities, cmdmgr, errors
+
+@operator.operator_class(
+    name="model",
+    help='显示和切换模型列表',
+    usage='!model\n!model show <模型名>\n!model set <模型名>',
+    privilege=2
+)
+class ModelOperator(operator.CommandOperator):
+    """Model命令"""
+    
+    async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        content = '模型列表:\n'
+
+        model_list = self.ap.model_mgr.model_list
+
+        for model in model_list:
+            content += f"\n名称: {model.name}\n"
+            content += f"请求器: {model.requester.name}\n"
+
+        content += f"\n当前对话使用模型: {context.query.use_model.name}\n"
+        content += f"新对话默认使用模型: {self.ap.provider_cfg.data.get('model')}\n"
+
+        yield entities.CommandReturn(text=content.strip())
+
+
+@operator.operator_class(
+    name="show",
+    help='显示模型详情',
+    privilege=2,
+    parent_class=ModelOperator
+)
+class ModelShowOperator(operator.CommandOperator):
+    """Model Show命令"""
+    
+    async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        model_name = context.crt_params[0]
+
+        model = None
+        for _model in self.ap.model_mgr.model_list:
+            if model_name == _model.name:
+                model = _model
+                break
+
+        if model is None:
+            yield entities.CommandReturn(error=errors.CommandError(f"未找到模型 {model_name}"))
+        else:
+            content = f"模型详情\n"
+            content += f"名称: {model.name}\n"
+            if model.model_name is not None:
+                content += f"请求模型名称: {model.model_name}\n"
+            content += f"请求器: {model.requester.name}\n"
+            content += f"密钥组: {model.token_mgr.provider}\n"
+            content += f"支持视觉: {model.vision_supported}\n"
+            content += f"支持工具: {model.tool_call_supported}\n"
+
+            yield entities.CommandReturn(text=content.strip())
+
+@operator.operator_class(
+    name="set",
+    help='设置默认使用模型',
+    privilege=2,
+    parent_class=ModelOperator
+)
+class ModelSetOperator(operator.CommandOperator):
+    """Model Set命令"""
+    
+    async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        model_name = context.crt_params[0]
+
+        model = None
+        for _model in self.ap.model_mgr.model_list:
+            if model_name == _model.name:
+                model = _model
+                break
+
+        if model is None:
+            yield entities.CommandReturn(error=errors.CommandError(f"未找到模型 {model_name}"))
+        else:
+            self.ap.provider_cfg.data['model'] = model_name
+            await self.ap.provider_cfg.dump_config()
+            yield entities.CommandReturn(text=f"已设置当前使用模型为 {model_name}，重置会话以生效")

pkg/command/operators/ollama.py

@@ -0,0 +1,121 @@
+from __future__ import annotations
+
+import json
+import typing
+import traceback
+
+import ollama
+from .. import operator, entities, errors
+
+
+@operator.operator_class(
+    name="ollama",
+    help="ollama平台操作",
+    usage="!ollama\n!ollama show <模型名>\n!ollama pull <模型名>\n!ollama del <模型名>"
+)
+class OllamaOperator(operator.CommandOperator):
+    async def execute(
+            self, context: entities.ExecuteContext
+    ) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        try:
+            content: str = '模型列表:\n'
+            model_list: list = ollama.list().get('models', [])
+            for model in model_list:
+                content += f"名称: {model['name']}\n"
+                content += f"修改时间: {model['modified_at']}\n"
+                content += f"大小: {bytes_to_mb(model['size'])}MB\n\n"
+            yield entities.CommandReturn(text=f"{content.strip()}")
+        except ollama.ResponseError as e:
+            yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型列表，请确认 Ollama 服务正常"))
+
+
+def bytes_to_mb(num_bytes):
+    mb: float = num_bytes / 1024 / 1024
+    return format(mb, '.2f')
+
+
+@operator.operator_class(
+    name="show",
+    help="ollama模型详情",
+    privilege=2,
+    parent_class=OllamaOperator
+)
+class OllamaShowOperator(operator.CommandOperator):
+    async def execute(
+            self, context: entities.ExecuteContext
+    ) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        content: str = '模型详情:\n'
+        try:
+            show: dict = ollama.show(model=context.crt_params[0])
+            model_info: dict = show.get('model_info', {})
+            ignore_show: str = 'too long to show...'
+
+            for key in ['license', 'modelfile']:
+                show[key] = ignore_show
+
+            for key in ['tokenizer.chat_template.rag', 'tokenizer.chat_template.tool_use']:
+                model_info[key] = ignore_show
+
+            content += json.dumps(show, indent=4)
+            yield entities.CommandReturn(text=content.strip())
+        except ollama.ResponseError as e:
+            yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型详情，请确认 Ollama 服务正常"))
+
+@operator.operator_class(
+    name="pull",
+    help="ollama模型拉取",
+    privilege=2,
+    parent_class=OllamaOperator
+)
+class OllamaPullOperator(operator.CommandOperator):
+    async def execute(
+            self, context: entities.ExecuteContext
+    ) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        try:
+            model_list: list = ollama.list().get('models', [])
+            if context.crt_params[0] in [model['name'] for model in model_list]:
+                yield entities.CommandReturn(text="模型已存在")
+                return
+        except ollama.ResponseError as e:
+            yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型列表，请确认 Ollama 服务正常"))
+            return
+
+        on_progress: bool = False
+        progress_count: int = 0
+        try:
+            for resp in ollama.pull(model=context.crt_params[0], stream=True):
+                total: typing.Any = resp.get('total')
+                if not on_progress:
+                    if total is not None:
+                        on_progress = True
+                    yield entities.CommandReturn(text=resp.get('status'))
+                else:
+                    if total is None:
+                        on_progress = False
+
+                    completed: typing.Any = resp.get('completed')
+                    if isinstance(completed, int) and isinstance(total, int):
+                        percentage_completed = (completed / total) * 100
+                        if percentage_completed > progress_count:
+                            progress_count += 10
+                            yield entities.CommandReturn(
+                                text=f"下载进度: {completed}/{total} ({percentage_completed:.2f}%)")
+        except ollama.ResponseError as e:
+            yield entities.CommandReturn(text=f"拉取失败: {e.error}")
+
+
+@operator.operator_class(
+    name="del",
+    help="ollama模型删除",
+    privilege=2,
+    parent_class=OllamaOperator
+)
+class OllamaDelOperator(operator.CommandOperator):
+    async def execute(
+            self, context: entities.ExecuteContext
+    ) -> typing.AsyncGenerator[entities.CommandReturn, None]:
+        try:
+            ret: str = ollama.delete(model=context.crt_params[0])['status']
+        except ollama.ResponseError as e:
+            ret = f"{e.error}"
+        yield entities.CommandReturn(text=ret)

pkg/command/operators/version.py

@@ -20,7 +20,7 @@ class VersionCommand(operator.CommandOperator):
 
         try:
             if await self.ap.ver_mgr.is_new_version_available():
-                reply_str += "\n\n有新版本可用, 使用 !update 更新"
+                reply_str += "\n\n有新版本可用。"
         except:
             pass
 

pkg/config/impls/yaml.py

@@ -0,0 +1,59 @@
+import os
+import shutil
+import yaml
+
+from .. import model as file_model
+
+
+class YAMLConfigFile(file_model.ConfigFile):
+    """YAML配置文件"""
+
+    def __init__(
+        self, config_file_name: str, template_file_name: str = None, template_data: dict = None
+    ) -> None:
+        self.config_file_name = config_file_name
+        self.template_file_name = template_file_name
+        self.template_data = template_data
+
+    def exists(self) -> bool:
+        return os.path.exists(self.config_file_name)
+
+    async def create(self):
+        if self.template_file_name is not None:
+            shutil.copyfile(self.template_file_name, self.config_file_name)
+        elif self.template_data is not None:
+            with open(self.config_file_name, "w", encoding="utf-8") as f:
+                yaml.dump(self.template_data, f, indent=4, allow_unicode=True)
+        else:
+            raise ValueError("template_file_name or template_data must be provided")
+
+    async def load(self, completion: bool=True) -> dict:
+
+        if not self.exists():
+            await self.create()
+
+        if self.template_file_name is not None:
+            with open(self.template_file_name, "r", encoding="utf-8") as f:
+                self.template_data = yaml.load(f, Loader=yaml.FullLoader)
+
+        with open(self.config_file_name, "r", encoding="utf-8") as f:
+            try:
+                cfg = yaml.load(f, Loader=yaml.FullLoader)
+            except yaml.YAMLError as e:
+                raise Exception(f"配置文件 {self.config_file_name} 语法错误: {e}")
+
+        if completion:
+
+            for key in self.template_data:
+                if key not in cfg:
+                    cfg[key] = self.template_data[key]
+
+        return cfg
+
+    async def save(self, cfg: dict):
+        with open(self.config_file_name, "w", encoding="utf-8") as f:
+            yaml.dump(cfg, f, indent=4, allow_unicode=True)
+
+    def save_sync(self, cfg: dict):
+        with open(self.config_file_name, "w", encoding="utf-8") as f:
+            yaml.dump(cfg, f, indent=4, allow_unicode=True)

pkg/config/manager.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 from . import model as file_model
-from .impls import pymodule, json as json_file
+from .impls import pymodule, json as json_file, yaml as yaml_file
 
 
 managers: ConfigManager = []
@@ -31,7 +31,16 @@ class ConfigManager:
 
 
 async def load_python_module_config(config_name: str, template_name: str, completion: bool=True) -> ConfigManager:
-    """加载Python模块配置文件"""
+    """加载Python模块配置文件
+    
+    Args:
+        config_name (str): 配置文件名
+        template_name (str): 模板文件名
+        completion (bool): 是否自动补全内存中的配置文件
+
+    Returns:
+        ConfigManager: 配置文件管理器
+    """
     cfg_inst = pymodule.PythonModuleConfigFile(
         config_name,
         template_name
@@ -44,7 +53,14 @@ async def load_python_module_config(config_name: str, template_name: str, comple
 
 
 async def load_json_config(config_name: str, template_name: str=None, template_data: dict=None, completion: bool=True) -> ConfigManager:
-    """加载JSON配置文件"""
+    """加载JSON配置文件
+    
+    Args:
+        config_name (str): 配置文件名
+        template_name (str): 模板文件名
+        template_data (dict): 模板数据
+        completion (bool): 是否自动补全内存中的配置文件
+    """
     cfg_inst = json_file.JSONConfigFile(
         config_name,
         template_name,
@@ -54,4 +70,28 @@ async def load_json_config(config_name: str, template_name: str=None, template_d
     cfg_mgr = ConfigManager(cfg_inst)
     await cfg_mgr.load_config(completion=completion)
 
-    return cfg_mgr
+    return cfg_mgr
+
+
+async def load_yaml_config(config_name: str, template_name: str=None, template_data: dict=None, completion: bool=True) -> ConfigManager:
+    """加载YAML配置文件
+    
+    Args:
+        config_name (str): 配置文件名
+        template_name (str): 模板文件名
+        template_data (dict): 模板数据
+        completion (bool): 是否自动补全内存中的配置文件
+
+    Returns:
+        ConfigManager: 配置文件管理器
+    """
+    cfg_inst = yaml_file.YAMLConfigFile(
+        config_name,
+        template_name,
+        template_data
+    )
+
+    cfg_mgr = ConfigManager(cfg_inst)
+    await cfg_mgr.load_config(completion=completion)
+
+    return cfg_mgr

pkg/core/app.py

@@ -9,13 +9,14 @@ from ..provider.session import sessionmgr as llm_session_mgr
 from ..provider.modelmgr import modelmgr as llm_model_mgr
 from ..provider.sysprompt import sysprompt as llm_prompt_mgr
 from ..provider.tools import toolmgr as llm_tool_mgr
+from ..provider import runnermgr
 from ..config import manager as config_mgr
 from ..audit.center import v2 as center_mgr
 from ..command import cmdmgr
 from ..plugin import manager as plugin_mgr
 from ..pipeline import pool
 from ..pipeline import controller, stagemgr
-from ..utils import version as version_mgr, proxy as proxy_mgr
+from ..utils import version as version_mgr, proxy as proxy_mgr, announce as announce_mgr
 
 
 class Application:
@@ -33,6 +34,8 @@ class Application:
 
     tool_mgr: llm_tool_mgr.ToolManager = None
 
+    runner_mgr: runnermgr.RunnerManager = None
+
     # ======= 配置管理器 =======
 
     command_cfg: config_mgr.ConfigManager = None
@@ -69,6 +72,8 @@ class Application:
 
     ver_mgr: version_mgr.VersionManager = None
 
+    ann_mgr: announce_mgr.AnnouncementManager = None
+
     proxy_mgr: proxy_mgr.ProxyManager = None
 
     logger: logging.Logger = None

pkg/core/boot.py

@@ -7,14 +7,15 @@ from ..audit import identifier
 from . import stage
 
 # 引入启动阶段实现以便注册
-from .stages import load_config, setup_logger, build_app, migrate
+from .stages import load_config, setup_logger, build_app, migrate, show_notes
 
 
 stage_order = [
     "LoadConfigStage",
     "MigrationStage",
     "SetupLoggerStage",
-    "BuildAppStage"
+    "BuildAppStage",
+    "ShowNotesStage"
 ]
 
 

pkg/core/bootutils/deps.py

@@ -15,6 +15,7 @@ required_deps = {
     "aiohttp": "aiohttp",
     "psutil": "psutil",
     "async_lru": "async-lru",
+    "ollama": "ollama",
 }
 
 

pkg/core/entities.py

@@ -73,6 +73,9 @@ class Query(pydantic.BaseModel):
     resp_message_chain: typing.Optional[list[mirai.MessageChain]] = None
     """回复消息链，从resp_messages包装而得"""
 
+    # ======= 内部保留 =======
+    current_stage: "pkg.pipeline.stagemgr.StageInstContainer" = None
+
     class Config:
         arbitrary_types_allowed = True
 

pkg/core/migration.py

@@ -3,7 +3,7 @@ from __future__ import annotations
 import abc
 import typing
 
-from ..core import app
+from . import app
 
 
 preregistered_migrations: list[typing.Type[Migration]] = []

pkg/core/migrations/m010_ollama_requester_config.py

@@ -0,0 +1,23 @@
+from __future__ import annotations
+
+from .. import migration
+
+
+@migration.migration_class("ollama-requester-config", 10)
+class MsgTruncatorConfigMigration(migration.Migration):
+    """迁移"""
+
+    async def need_migrate(self) -> bool:
+        """判断当前环境是否需要运行此迁移"""
+        return 'ollama-chat' not in self.ap.provider_cfg.data['requester']
+
+    async def run(self):
+        """执行迁移"""
+        
+        self.ap.provider_cfg.data['requester']['ollama-chat'] = {
+            "base-url": "http://127.0.0.1:11434",
+            "args": {},
+            "timeout": 600
+        }
+
+        await self.ap.provider_cfg.dump_config()

pkg/core/migrations/m011_command_prefix_config.py

@@ -0,0 +1,21 @@
+from __future__ import annotations
+
+from .. import migration
+
+
+@migration.migration_class("command-prefix-config", 11)
+class CommandPrefixConfigMigration(migration.Migration):
+    """迁移"""
+
+    async def need_migrate(self) -> bool:
+        """判断当前环境是否需要运行此迁移"""
+        return 'command-prefix' not in self.ap.command_cfg.data
+
+    async def run(self):
+        """执行迁移"""
+        
+        self.ap.command_cfg.data['command-prefix'] = [
+            "!", "！"
+        ]
+
+        await self.ap.command_cfg.dump_config()

pkg/core/migrations/m012_runner_config.py

@@ -0,0 +1,19 @@
+from __future__ import annotations
+
+from .. import migration
+
+
+@migration.migration_class("runner-config", 12)
+class RunnerConfigMigration(migration.Migration):
+    """迁移"""
+
+    async def need_migrate(self) -> bool:
+        """判断当前环境是否需要运行此迁移"""
+        return 'runner' not in self.ap.provider_cfg.data
+
+    async def run(self):
+        """执行迁移"""
+        
+        self.ap.provider_cfg.data['runner'] = 'local-agent'
+
+        await self.ap.provider_cfg.dump_config()

pkg/core/note.py

@@ -0,0 +1,44 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from . import app
+
+preregistered_notes: list[typing.Type[LaunchNote]] = []
+
+def note_class(name: str, number: int):
+    """注册一个启动信息
+    """
+    def decorator(cls: typing.Type[LaunchNote]) -> typing.Type[LaunchNote]:
+        cls.name = name
+        cls.number = number
+        preregistered_notes.append(cls)
+        return cls
+
+    return decorator
+
+
+class LaunchNote(abc.ABC):
+    """启动信息
+    """
+    name: str
+
+    number: int
+
+    ap: app.Application
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+
+    @abc.abstractmethod
+    async def need_show(self) -> bool:
+        """判断当前环境是否需要显示此启动信息
+        """
+        pass
+
+    @abc.abstractmethod
+    async def yield_note(self) -> typing.AsyncGenerator[typing.Tuple[str, int], None]:
+        """生成启动信息
+        """
+        pass

pkg/core/notes/n001_classic_msgs.py

@@ -0,0 +1,20 @@
+from __future__ import annotations
+
+import typing
+
+from .. import note, app
+
+
+@note.note_class("ClassicNotes", 1)
+class ClassicNotes(note.LaunchNote):
+    """经典启动信息
+    """
+
+    async def need_show(self) -> bool:
+        return True
+
+    async def yield_note(self) -> typing.AsyncGenerator[typing.Tuple[str, int], None]:
+        
+        yield await self.ap.ann_mgr.show_announcements()
+
+        yield await self.ap.ver_mgr.show_version_update()

pkg/core/notes/n002_selection_mode_on_windows.py

@@ -0,0 +1,21 @@
+from __future__ import annotations
+
+import typing
+import os
+import sys
+import logging
+
+from .. import note, app
+
+
+@note.note_class("SelectionModeOnWindows", 2)
+class SelectionModeOnWindows(note.LaunchNote):
+    """Windows 上的选择模式提示信息
+    """
+
+    async def need_show(self) -> bool:
+        return os.name == 'nt'
+
+    async def yield_note(self) -> typing.AsyncGenerator[typing.Tuple[str, int], None]:
+
+        yield """您正在使用 Windows 系统，若窗口左上角显示处于”选择“模式，程序将被暂停运行，此时请右键窗口中空白区域退出选择模式。""", logging.INFO

pkg/core/stages/build_app.py

@@ -13,6 +13,7 @@ from ...provider.session import sessionmgr as llm_session_mgr
 from ...provider.modelmgr import modelmgr as llm_model_mgr
 from ...provider.sysprompt import sysprompt as llm_prompt_mgr
 from ...provider.tools import toolmgr as llm_tool_mgr
+from ...provider import runnermgr
 from ...platform import manager as im_mgr
 
 @stage.stage_class("BuildAppStage")
@@ -53,12 +54,10 @@ class BuildAppStage(stage.BootingStage):
 
         # 发送公告
         ann_mgr = announce.AnnouncementManager(ap)
-        await ann_mgr.show_announcements()
+        ap.ann_mgr = ann_mgr
 
         ap.query_pool = pool.QueryPool()
 
-        await ap.ver_mgr.show_version_update()
-
         plugin_mgr_inst = plugin_mgr.PluginManager(ap)
         await plugin_mgr_inst.initialize()
         ap.plugin_mgr = plugin_mgr_inst
@@ -83,6 +82,11 @@ class BuildAppStage(stage.BootingStage):
         llm_tool_mgr_inst = llm_tool_mgr.ToolManager(ap)
         await llm_tool_mgr_inst.initialize()
         ap.tool_mgr = llm_tool_mgr_inst
+
+        runner_mgr_inst = runnermgr.RunnerManager(ap)
+        await runner_mgr_inst.initialize()
+        ap.runner_mgr = runner_mgr_inst
+
         im_mgr_inst = im_mgr.PlatformManager(ap=ap)
         await im_mgr_inst.initialize()
         ap.platform_mgr = im_mgr_inst

pkg/core/stages/migrate.py

@@ -3,9 +3,10 @@ from __future__ import annotations
 import importlib
 
 from .. import stage, app
-from ...config import migration
-from ...config.migrations import m001_sensitive_word_migration, m002_openai_config_migration, m003_anthropic_requester_cfg_completion, m004_moonshot_cfg_completion
-from ...config.migrations import m005_deepseek_cfg_completion, m006_vision_config, m007_qcg_center_url, m008_ad_fixwin_config_migrate, m009_msg_truncator_cfg
+from .. import migration
+from ..migrations import m001_sensitive_word_migration, m002_openai_config_migration, m003_anthropic_requester_cfg_completion, m004_moonshot_cfg_completion
+from ..migrations import m005_deepseek_cfg_completion, m006_vision_config, m007_qcg_center_url, m008_ad_fixwin_config_migrate, m009_msg_truncator_cfg
+from ..migrations import m010_ollama_requester_config, m011_command_prefix_config, m012_runner_config
 
 
 @stage.stage_class("MigrationStage")

pkg/core/stages/show_notes.py

@@ -0,0 +1,28 @@
+from __future__ import annotations
+
+from .. import stage, app, note
+from ..notes import n001_classic_msgs, n002_selection_mode_on_windows
+
+
+@stage.stage_class("ShowNotesStage")
+class ShowNotesStage(stage.BootingStage):
+    """显示启动信息阶段
+    """
+
+    async def run(self, ap: app.Application):
+
+        # 排序
+        note.preregistered_notes.sort(key=lambda x: x.number)
+
+        for note_cls in note.preregistered_notes:
+            try:
+                note_inst = note_cls(ap)
+                if await note_inst.need_show():
+                    async for ret in note_inst.yield_note():
+                        if not ret:
+                            continue
+                        msg, level = ret
+                        if msg:
+                            ap.logger.log(level, msg)
+            except Exception as e:
+                continue

pkg/pipeline/cntfilter/cntfilter.py

@@ -1,6 +1,8 @@
 from __future__ import annotations
 
 import mirai
+import mirai.models
+import mirai.models.message
 
 from ...core import app
 
@@ -63,6 +65,7 @@ class ContentFilterStage(stage.PipelineStage):
         """请求llm前处理消息
         只要有一个不通过就不放行，只放行 PASS 的消息
         """
+
         if not self.ap.pipeline_cfg.data['income-msg-check']:
             return entities.StageProcessResult(
                 result_type=entities.ResultType.CONTINUE,
@@ -145,11 +148,13 @@ class ContentFilterStage(stage.PipelineStage):
             
             contain_non_text = False
 
+            text_components = [mirai.Plain, mirai.models.message.Source]
+
             for me in query.message_chain:
-                if not isinstance(me, mirai.Plain):
+                if type(me) not in text_components:
                     contain_non_text = True
                     break
-            
+
             if contain_non_text:
                 self.ap.logger.debug(f"消息中包含非文本消息，跳过内容过滤器检查。")
                 return entities.StageProcessResult(

pkg/pipeline/controller.py

@@ -4,6 +4,8 @@ import asyncio
 import typing
 import traceback
 
+import mirai
+
 from ..core import app, entities
 from . import entities as pipeline_entities
 from ..plugin import events
@@ -68,6 +70,17 @@ class Controller:
         """检查输出
         """
         if result.user_notice:
+            # 处理str类型
+
+            if isinstance(result.user_notice, str):
+                result.user_notice = mirai.MessageChain(
+                    mirai.Plain(result.user_notice)
+                )
+            elif isinstance(result.user_notice, list):
+                result.user_notice = mirai.MessageChain(
+                    *result.user_notice
+                )
+
             await self.ap.platform_mgr.send(
                 query.message_event,
                 result.user_notice,
@@ -109,6 +122,8 @@ class Controller:
 
         while i < len(self.ap.stage_mgr.stage_containers):
             stage_container = self.ap.stage_mgr.stage_containers[i]
+
+            query.current_stage = stage_container  # 标记到 Query 对象里
             
             result = stage_container.inst.process(query, stage_container.inst_name)
 
@@ -149,7 +164,7 @@ class Controller:
         try:
             await self._execute_from_stage(0, query)
         except Exception as e:
-            self.ap.logger.error(f"处理请求时出错 query_id={query.query_id}: {e}")
+            self.ap.logger.error(f"处理请求时出错 query_id={query.query_id} stage={query.current_stage.inst_name} : {e}")
             self.ap.logger.debug(f"Traceback: {traceback.format_exc()}")
             # traceback.print_exc()
         finally:

pkg/pipeline/process/handlers/chat.py

@@ -10,7 +10,7 @@ import mirai
 from .. import handler
 from ... import entities
 from ....core import entities as core_entities
-from ....provider import entities as llm_entities
+from ....provider import entities as llm_entities, runnermgr
 from ....plugin import events
 
 
@@ -62,9 +62,8 @@ class ChatMessageHandler(handler.MessageHandler):
                 )
 
             if event_ctx.event.alter is not None:
-                query.message_chain = mirai.MessageChain([
-                    mirai.Plain(event_ctx.event.alter)
-                ])
+                # if isinstance(event_ctx.event, str):  # 现在暂时不考虑多模态alter
+                query.user_message.content = event_ctx.event.alter
 
             text_length = 0
 
@@ -72,7 +71,9 @@ class ChatMessageHandler(handler.MessageHandler):
 
             try:
 
-                async for result in self.runner(query):
+                runner = self.ap.runner_mgr.get_runner()
+
+                async for result in runner.run(query):
                     query.resp_messages.append(result)
 
                     self.ap.logger.info(f'对话({query.query_id})响应: {self.cut_str(result.readable_str())}')
@@ -109,64 +110,3 @@ class ChatMessageHandler(handler.MessageHandler):
                     response_seconds=int(time.time() - start_time),
                     retry_times=-1,
                 )
-
-    async def runner(
-        self,
-        query: core_entities.Query,
-    ) -> typing.AsyncGenerator[llm_entities.Message, None]:
-        """执行一个请求处理过程中的LLM接口请求、函数调用的循环
-        
-        这是临时处理方案，后续可能改为使用LangChain或者自研的工作流处理器
-        """
-        await query.use_model.requester.preprocess(query)
-
-        pending_tool_calls = []
-
-        req_messages = query.prompt.messages.copy() + query.messages.copy() + [query.user_message]
-
-        # 首次请求
-        msg = await query.use_model.requester.call(query.use_model, req_messages, query.use_funcs)
-
-        yield msg
-
-        pending_tool_calls = msg.tool_calls
-
-        req_messages.append(msg)
-
-        # 持续请求，只要还有待处理的工具调用就继续处理调用
-        while pending_tool_calls:
-            for tool_call in pending_tool_calls:
-                try:
-                    func = tool_call.function
-                    
-                    parameters = json.loads(func.arguments)
-
-                    func_ret = await self.ap.tool_mgr.execute_func_call(
-                        query, func.name, parameters
-                    )
-
-                    msg = llm_entities.Message(
-                        role="tool", content=json.dumps(func_ret, ensure_ascii=False), tool_call_id=tool_call.id
-                    )
-
-                    yield msg
-
-                    req_messages.append(msg)
-                except Exception as e:
-                    # 工具调用出错，添加一个报错信息到 req_messages
-                    err_msg = llm_entities.Message(
-                        role="tool", content=f"err: {e}", tool_call_id=tool_call.id
-                    )
-
-                    yield err_msg
-
-                    req_messages.append(err_msg)
-
-            # 处理完所有调用，再次请求
-            msg = await query.use_model.requester.call(query.use_model, req_messages, query.use_funcs)
-
-            yield msg
-
-            pending_tool_calls = msg.tool_calls
-
-            req_messages.append(msg)

pkg/pipeline/process/process.py

@@ -42,7 +42,9 @@ class Processor(stage.PipelineStage):
         self.ap.logger.info(f"处理 {query.launcher_type.value}_{query.launcher_id} 的请求({query.query_id}): {message_text}")
 
         async def generator():
-            if message_text.startswith('!') or message_text.startswith('！'):
+            cmd_prefix = self.ap.command_cfg.data['command-prefix']
+
+            if any(message_text.startswith(prefix) for prefix in cmd_prefix):
                 async for result in self.cmd_handler.handle(query):
                     yield result
             else:

pkg/pipeline/resprule/resprule.py

@@ -44,8 +44,8 @@ class GroupRespondRuleCheckStage(stage.PipelineStage):
 
         use_rule = rules['default']
 
-        if str(query.launcher_id) in use_rule:
-            use_rule = use_rule[str(query.launcher_id)]
+        if str(query.launcher_id) in rules:
+            use_rule = rules[str(query.launcher_id)]
 
         for rule_matcher in self.rule_matchers:  # 任意一个匹配就放行
             res = await rule_matcher.match(str(query.message_chain), query.message_chain, use_rule, query)

pkg/pipeline/resprule/rules/prefix.py

@@ -20,11 +20,14 @@ class PrefixRule(rule_model.GroupRespondRule):
         for prefix in prefixes:
             if message_text.startswith(prefix):
 
+                # 查找第一个plain元素
+                for me in message_chain:
+                    if isinstance(me, mirai.Plain):
+                        me.text = me.text[len(prefix):]
+
                 return entities.RuleJudgeResult(
                     matching=True,
-                    replacement=mirai.MessageChain([
-                        mirai.Plain(message_text[len(prefix):])
-                    ]),
+                    replacement=message_chain,
                 )
 
         return entities.RuleJudgeResult(

pkg/platform/manager.py

@@ -146,9 +146,9 @@ class PlatformManager:
         if len(self.adapters) == 0:
             self.ap.logger.warning('未运行平台适配器，请根据文档配置并启用平台适配器。')
 
-    async def send(self, event: mirai.MessageEvent, msg: mirai.MessageChain, adapter: msadapter.MessageSourceAdapter, check_quote=True, check_at_sender=True):
+    async def send(self, event: mirai.MessageEvent, msg: mirai.MessageChain, adapter: msadapter.MessageSourceAdapter):
         
-        if check_at_sender and self.ap.platform_cfg.data['at-sender'] and isinstance(event, GroupMessage):
+        if self.ap.platform_cfg.data['at-sender'] and isinstance(event, GroupMessage):
 
             msg.insert(
                 0,
@@ -160,7 +160,7 @@ class PlatformManager:
         await adapter.reply_message(
             event,
             msg,
-            quote_origin=True if self.ap.platform_cfg.data['quote-origin'] and check_quote else False
+            quote_origin=True if self.ap.platform_cfg.data['quote-origin'] else False
         )
 
     async def run(self):

pkg/platform/sources/aiocqhttp.py

@@ -47,7 +47,16 @@ class AiocqhttpMessageConverter(adapter.MessageConverter):
             elif type(msg) is mirai.Face:
                 msg_list.append(aiocqhttp.MessageSegment.face(msg.face_id))
             elif type(msg) is mirai.Voice:
-                msg_list.append(aiocqhttp.MessageSegment.record(msg.path))
+                arg = ''
+                if msg.base64:
+                    arg = msg.base64
+                    msg_list.append(aiocqhttp.MessageSegment.record(f"base64://{arg}"))
+                elif msg.url:
+                    arg = msg.url
+                    msg_list.append(aiocqhttp.MessageSegment.record(arg))
+                elif msg.path:
+                    arg = msg.path
+                    msg_list.append(aiocqhttp.MessageSegment.record(msg.path))
             elif type(msg) is forward.Forward:
 
                 for node in msg.node_list:

pkg/platform/sources/qqbotpy.py

@@ -3,16 +3,15 @@ from __future__ import annotations
 import logging
 import typing
 import datetime
-import asyncio
 import re
 import traceback
-import json
-import threading
 
 import mirai
 import botpy
 import botpy.message as botpy_message
 import botpy.types.message as botpy_message_type
+import pydantic
+import pydantic.networks
 
 from .. import adapter as adapter_model
 from ...pipeline.longtext.strategies import forward
@@ -23,10 +22,12 @@ from ...config import manager as cfg_mgr
 class OfficialGroupMessage(mirai.GroupMessage):
     pass
 
+class OfficialFriendMessage(mirai.FriendMessage):
+    pass
 
 event_handler_mapping = {
     mirai.GroupMessage: ["on_at_message_create", "on_group_at_message_create"],
-    mirai.FriendMessage: ["on_direct_message_create"],
+    mirai.FriendMessage: ["on_direct_message_create", "on_c2c_message_create"],
 }
 
 
@@ -193,12 +194,11 @@ class OfficialMessageConverter(adapter_model.MessageConverter):
 
     @staticmethod
     def extract_message_chain_from_obj(
-        message: typing.Union[botpy_message.Message, botpy_message.DirectMessage],
+        message: typing.Union[botpy_message.Message, botpy_message.DirectMessage, botpy_message.GroupMessage, botpy_message.C2CMessage],
         message_id: str = None,
         bot_account_id: int = 0,
     ) -> mirai.MessageChain:
         yiri_msg_list = []
-
         # 存id
 
         yiri_msg_list.append(
@@ -207,7 +207,7 @@ class OfficialMessageConverter(adapter_model.MessageConverter):
             )
         )
 
-        if type(message) is not botpy_message.DirectMessage:
+        if type(message) not in [botpy_message.DirectMessage, botpy_message.C2CMessage]:
             yiri_msg_list.append(mirai.At(target=bot_account_id))
 
         if hasattr(message, "mentions"):
@@ -218,7 +218,7 @@ class OfficialMessageConverter(adapter_model.MessageConverter):
                 yiri_msg_list.append(mirai.At(target=mention.id))
 
         for attachment in message.attachments:
-            if attachment.content_type == "image":
+            if attachment.content_type.startswith("image"):
                 yiri_msg_list.append(mirai.Image(url=attachment.url))
             else:
                 logging.warning(
@@ -256,7 +256,7 @@ class OfficialEventConverter(adapter_model.EventConverter):
 
     def target2yiri(
         self,
-        event: typing.Union[botpy_message.Message, botpy_message.DirectMessage]
+        event: typing.Union[botpy_message.Message, botpy_message.DirectMessage, botpy_message.GroupMessage, botpy_message.C2CMessage],
     ) -> mirai.Event:
         import mirai.models.entities as mirai_entities
 
@@ -296,7 +296,7 @@ class OfficialEventConverter(adapter_model.EventConverter):
                     ).timestamp()
                 ),
             )
-        elif type(event) == botpy_message.DirectMessage:  # 私聊，转私聊事件
+        elif type(event) == botpy_message.DirectMessage:  # 频道私聊，转私聊事件
             return mirai.FriendMessage(
                 sender=mirai_entities.Friend(
                     id=event.guild_id,
@@ -312,7 +312,7 @@ class OfficialEventConverter(adapter_model.EventConverter):
                     ).timestamp()
                 ),
             )
-        elif type(event) == botpy_message.GroupMessage:
+        elif type(event) == botpy_message.GroupMessage:  # 群聊，转群聊事件
 
             replacing_member_id = self.member_openid_mapping.save_openid(event.author.member_openid)
 
@@ -340,6 +340,25 @@ class OfficialEventConverter(adapter_model.EventConverter):
                     ).timestamp()
                 ),
             )
+        elif type(event) == botpy_message.C2CMessage:  # 私聊，转私聊事件
+
+            user_id_alter = self.member_openid_mapping.save_openid(event.author.user_openid)  # 实测这里的user_openid与group的member_openid是一样的
+
+            return OfficialFriendMessage(
+                sender=mirai_entities.Friend(
+                    id=user_id_alter,
+                    nickname=user_id_alter,
+                    remark=user_id_alter,
+                ),
+                message_chain=OfficialMessageConverter.extract_message_chain_from_obj(
+                    event, event.id
+                ),
+                time=int(
+                    datetime.datetime.strptime(
+                        event.timestamp, "%Y-%m-%dT%H:%M:%S%z"
+                    ).timestamp()
+                ),
+            )
 
 
 @adapter_model.adapter_class("qq-botpy")
@@ -369,6 +388,7 @@ class OfficialAdapter(adapter_model.MessageSourceAdapter):
     group_openid_mapping: OpenIDMapping[str, int] = None
 
     group_msg_seq = None
+    c2c_msg_seq = None
 
     def __init__(self, cfg: dict, ap: app.Application):
         """初始化适配器"""
@@ -376,6 +396,7 @@ class OfficialAdapter(adapter_model.MessageSourceAdapter):
         self.ap = ap
 
         self.group_msg_seq = 1
+        self.c2c_msg_seq = 1
 
         switchs = {}
 
@@ -455,18 +476,58 @@ class OfficialAdapter(adapter_model.MessageSourceAdapter):
                 ]
                 await self.bot.api.post_dms(**args)
             elif type(message_source) == OfficialGroupMessage:
-                if "image" in args or "file_image" in args:
+
+                if "file_image" in args:  # 暂不支持发送文件图片
                     continue
+
                 args["group_openid"] = self.group_openid_mapping.getkey(
                     message_source.sender.group.id
                 )
 
+                if "image" in args:
+                    uploadMedia = await self.bot.api.post_group_file(
+                        group_openid=args["group_openid"],
+                        file_type=1,
+                        url=str(args['image'])
+                    )
+
+                    del args['image']
+                    args['media'] = uploadMedia
+                    args['msg_type'] = 7
+
                 args["msg_id"] = cached_message_ids[
                     str(message_source.message_chain.message_id)
                 ]
                 args["msg_seq"] = self.group_msg_seq
                 self.group_msg_seq += 1
+
                 await self.bot.api.post_group_message(**args)
+            elif type(message_source) == OfficialFriendMessage:
+                if "file_image" in args:
+                    continue
+                args["openid"] = self.member_openid_mapping.getkey(
+                    message_source.sender.id
+                )
+
+                if "image" in args:
+                    uploadMedia = await self.bot.api.post_c2c_file(
+                        openid=args["openid"],
+                        file_type=1,
+                        url=str(args['image'])
+                    )
+
+                    del args['image']
+                    args['media'] = uploadMedia
+                    args['msg_type'] = 7
+
+                args["msg_id"] = cached_message_ids[
+                    str(message_source.message_chain.message_id)
+                ]
+
+                args["msg_seq"] = self.c2c_msg_seq
+                self.c2c_msg_seq += 1
+
+                await self.bot.api.post_c2c_message(**args)
 
     async def is_muted(self, group_id: int) -> bool:
         return False

pkg/plugin/context.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import typing
 import abc
 import pydantic
+import mirai
 
 from . import events
 from ..provider.tools import entities as tools_entities
@@ -165,11 +166,54 @@ class EventContext:
     }
     """
 
+    # ========== 插件可调用的 API ==========
+
     def add_return(self, key: str, ret):
         """添加返回值"""
         if key not in self.__return_value__:
             self.__return_value__[key] = []
         self.__return_value__[key].append(ret)
+    
+    async def reply(self, message_chain: mirai.MessageChain):
+        """回复此次消息请求
+        
+        Args:
+            message_chain (mirai.MessageChain): YiriMirai库的消息链，若用户使用的不是 YiriMirai 适配器，程序也能自动转换为目标消息链
+        """
+        await self.host.ap.platform_mgr.send(
+            event=self.event.query.message_event,
+            msg=message_chain,
+            adapter=self.event.query.adapter,
+        )
+    
+    async def send_message(
+        self,
+        target_type: str,
+        target_id: str,
+        message: mirai.MessageChain
+    ):
+        """主动发送消息
+        
+        Args:
+            target_type (str): 目标类型，`person`或`group`
+            target_id (str): 目标ID
+            message (mirai.MessageChain): YiriMirai库的消息链，若用户使用的不是 YiriMirai 适配器，程序也能自动转换为目标消息链
+        """
+        await self.event.query.adapter.send_message(
+            target_type=target_type,
+            target_id=target_id,
+            message=message
+        )
+
+    def prevent_postorder(self):
+        """阻止后续插件执行"""
+        self.__prevent_postorder__ = True
+
+    def prevent_default(self):
+        """阻止默认行为"""
+        self.__prevent_default__ = True
+
+    # ========== 以下是内部保留方法，插件不应调用 ==========
 
     def get_return(self, key: str) -> list:
         """获取key的所有返回值"""
@@ -183,14 +227,6 @@ class EventContext:
             return self.__return_value__[key][0]
         return None
 
-    def prevent_default(self):
-        """阻止默认行为"""
-        self.__prevent_default__ = True
-
-    def prevent_postorder(self):
-        """阻止后续插件执行"""
-        self.__prevent_postorder__ = True
-
     def is_prevented_default(self):
         """是否阻止默认行为"""
         return self.__prevent_default__
@@ -198,6 +234,7 @@ class EventContext:
     def is_prevented_postorder(self):
         """是否阻止后序插件执行"""
         return self.__prevent_postorder__
+    
 
     def __init__(self, host: APIHost, event: events.BaseEventModel):
 

pkg/plugin/manager.py

@@ -140,7 +140,7 @@ class PluginManager:
         for plugin in self.plugins:
             if plugin.enabled:
                 if event.__class__ in plugin.event_handlers:
-                    self.ap.logger.debug(f'插件 {plugin.plugin_name} 触发事件 {event.__class__.__name__}')
+                    self.ap.logger.debug(f'插件 {plugin.plugin_name} 处理事件 {event.__class__.__name__}')
                     
                     is_prevented_default_before_call = ctx.is_prevented_default()
 
@@ -150,7 +150,7 @@ class PluginManager:
                             ctx
                         )
                     except Exception as e:
-                        self.ap.logger.error(f'插件 {plugin.plugin_name} 触发事件 {event.__class__.__name__} 时发生错误: {e}')
+                        self.ap.logger.error(f'插件 {plugin.plugin_name} 处理事件 {event.__class__.__name__} 时发生错误: {e}')
                         self.ap.logger.debug(f"Traceback: {traceback.format_exc()}")
                     
                     emitted_plugins.append(plugin)

pkg/provider/entities.py

@@ -95,7 +95,7 @@ class Message(pydantic.BaseModel):
             for ce in self.content:
                 if ce.type == 'text':
                     mc.append(mirai.Plain(ce.text))
-                elif ce.type == 'image':
+                elif ce.type == 'image_url':
                     if ce.image_url.url.startswith("http"):
                         mc.append(mirai.Image(url=ce.image_url.url))
                     else:  # base64

pkg/provider/modelmgr/apis/anthropicmsgs.py

@@ -72,12 +72,13 @@ class AnthropicMessages(api.LLMAPIRequester):
 
                 for i, ce in enumerate(m.content):
                     if ce.type == "image_url":
+                        base64_image, image_format = await image.qq_image_url_to_base64(ce.image_url.url)
                         alter_image_ele = {
                             "type": "image",
                             "source": {
                                 "type": "base64",
-                                "media_type": "image/jpeg",
-                                "data": await image.qq_image_url_to_base64(ce.image_url.url)
+                                "media_type": f"image/{image_format}",
+                                "data": base64_image
                             }
                         }
                         msg_dict["content"][i] = alter_image_ele

pkg/provider/modelmgr/apis/chatcmpl.py

@@ -102,9 +102,16 @@ class OpenAIChatCompletions(api.LLMAPIRequester):
         messages: typing.List[llm_entities.Message],
         funcs: typing.List[tools_entities.LLMFunction] = None,
     ) -> llm_entities.Message:
-        req_messages = [  # req_messages 仅用于类内，外部同步由 query.messages 进行
-            m.dict(exclude_none=True) for m in messages
-        ]
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get("content")
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get("type") == "text" for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict["content"] = "\n".join(part["text"] for part in content)
+            req_messages.append(msg_dict)
 
         try:
             return await self._closure(req_messages, model, funcs)
@@ -129,7 +136,5 @@ class OpenAIChatCompletions(api.LLMAPIRequester):
         self,
         original_url: str,
     ) -> str:
-        
-        base64_image = await image.qq_image_url_to_base64(original_url)
-
-        return f"data:image/jpeg;base64,{base64_image}"
+        base64_image, image_format = await image.qq_image_url_to_base64(original_url)
+        return f"data:image/{image_format};base64,{base64_image}"

pkg/provider/modelmgr/apis/ollamachat.py

@@ -0,0 +1,105 @@
+from __future__ import annotations
+
+import asyncio
+import os
+import typing
+from typing import Union, Mapping, Any, AsyncIterator
+
+import async_lru
+import ollama
+
+from .. import api, entities, errors
+from ... import entities as llm_entities
+from ...tools import entities as tools_entities
+from ....core import app
+from ....utils import image
+
+REQUESTER_NAME: str = "ollama-chat"
+
+
+@api.requester_class(REQUESTER_NAME)
+class OllamaChatCompletions(api.LLMAPIRequester):
+    """Ollama平台 ChatCompletion API请求器"""
+    client: ollama.AsyncClient
+    request_cfg: dict
+
+    def __init__(self, ap: app.Application):
+        super().__init__(ap)
+        self.ap = ap
+        self.request_cfg = self.ap.provider_cfg.data['requester'][REQUESTER_NAME]
+
+    async def initialize(self):
+        os.environ['OLLAMA_HOST'] = self.request_cfg['base-url']
+        self.client = ollama.AsyncClient(
+            timeout=self.request_cfg['timeout']
+        )
+
+    async def _req(self,
+                   args: dict,
+                   ) -> Union[Mapping[str, Any], AsyncIterator[Mapping[str, Any]]]:
+        return await self.client.chat(
+            **args
+        )
+
+    async def _closure(self, req_messages: list[dict], use_model: entities.LLMModelInfo,
+                       user_funcs: list[tools_entities.LLMFunction] = None) -> (
+            llm_entities.Message):
+        args: Any = self.request_cfg['args'].copy()
+        args["model"] = use_model.name if use_model.model_name is None else use_model.model_name
+
+        messages: list[dict] = req_messages.copy()
+        for msg in messages:
+            if 'content' in msg and isinstance(msg["content"], list):
+                text_content: list = []
+                image_urls: list = []
+                for me in msg["content"]:
+                    if me["type"] == "text":
+                        text_content.append(me["text"])
+                    elif me["type"] == "image_url":
+                        image_url = await self.get_base64_str(me["image_url"]['url'])
+                        image_urls.append(image_url)
+                msg["content"] = "\n".join(text_content)
+                msg["images"] = [url.split(',')[1] for url in image_urls]
+        args["messages"] = messages
+
+        resp: Mapping[str, Any] | AsyncIterator[Mapping[str, Any]] = await self._req(args)
+        message: llm_entities.Message = await self._make_msg(resp)
+        return message
+
+    async def _make_msg(
+            self,
+            chat_completions: Union[Mapping[str, Any], AsyncIterator[Mapping[str, Any]]]) -> llm_entities.Message:
+        message: Any = chat_completions.pop('message', None)
+        if message is None:
+            raise ValueError("chat_completions must contain a 'message' field")
+
+        message.update(chat_completions)
+        ret_msg: llm_entities.Message = llm_entities.Message(**message)
+        return ret_msg
+
+    async def call(
+            self,
+            model: entities.LLMModelInfo,
+            messages: typing.List[llm_entities.Message],
+            funcs: typing.List[tools_entities.LLMFunction] = None,
+    ) -> llm_entities.Message:
+        req_messages: list = []
+        for m in messages:
+            msg_dict: dict = m.dict(exclude_none=True)
+            content: Any = msg_dict.get("content")
+            if isinstance(content, list):
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    msg_dict["content"] = "\n".join(part["text"] for part in content)
+            req_messages.append(msg_dict)
+        try:
+            return await self._closure(req_messages, model)
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+
+    @async_lru.alru_cache(maxsize=128)
+    async def get_base64_str(
+            self,
+            original_url: str,
+    ) -> str:
+        base64_image, image_format = await image.qq_image_url_to_base64(original_url)
+        return f"data:image/{image_format};base64,{base64_image}"

pkg/provider/modelmgr/modelmgr.py

@@ -6,7 +6,7 @@ from . import entities
 from ...core import app
 
 from . import token, api
-from .apis import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl
+from .apis import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl, ollamachat
 
 FETCH_MODEL_LIST_URL = "https://api.qchatgpt.rockchin.top/api/v2/fetch/model_list"
 

pkg/provider/runner.py

@@ -0,0 +1,40 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from ..core import app, entities as core_entities
+from . import entities as llm_entities
+
+
+preregistered_runners: list[typing.Type[RequestRunner]] = []
+
+def runner_class(name: str):
+    """注册一个请求运行器
+    """
+    def decorator(cls: typing.Type[RequestRunner]) -> typing.Type[RequestRunner]:
+        cls.name = name
+        preregistered_runners.append(cls)
+        return cls
+
+    return decorator
+
+
+class RequestRunner(abc.ABC):
+    """请求运行器
+    """
+    name: str = None
+
+    ap: app.Application
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+
+    async def initialize(self):
+        pass
+
+    @abc.abstractmethod
+    async def run(self, query: core_entities.Query) -> typing.AsyncGenerator[llm_entities.Message, None]:
+        """运行请求
+        """
+        pass

pkg/provider/runnermgr.py

@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+from . import runner
+from ..core import app
+
+from .runners import localagent
+
+
+class RunnerManager:
+
+    ap: app.Application
+
+    using_runner: runner.RequestRunner
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+
+    async def initialize(self):
+
+        for r in runner.preregistered_runners:
+            if r.name == self.ap.provider_cfg.data['runner']:
+                self.using_runner = r(self.ap)
+                await self.using_runner.initialize()
+                break
+
+    def get_runner(self) -> runner.RequestRunner:
+        return self.using_runner

pkg/provider/runners/localagent.py

@@ -0,0 +1,70 @@
+from __future__ import annotations
+
+import json
+import typing
+
+from .. import runner
+from ...core import app, entities as core_entities
+from .. import entities as llm_entities
+
+
+@runner.runner_class("local-agent")
+class LocalAgentRunner(runner.RequestRunner):
+    """本地Agent请求运行器
+    """
+
+    async def run(self, query: core_entities.Query) -> typing.AsyncGenerator[llm_entities.Message, None]:
+        """运行请求
+        """
+        await query.use_model.requester.preprocess(query)
+
+        pending_tool_calls = []
+
+        req_messages = query.prompt.messages.copy() + query.messages.copy() + [query.user_message]
+
+        # 首次请求
+        msg = await query.use_model.requester.call(query.use_model, req_messages, query.use_funcs)
+
+        yield msg
+
+        pending_tool_calls = msg.tool_calls
+
+        req_messages.append(msg)
+
+        # 持续请求，只要还有待处理的工具调用就继续处理调用
+        while pending_tool_calls:
+            for tool_call in pending_tool_calls:
+                try:
+                    func = tool_call.function
+                    
+                    parameters = json.loads(func.arguments)
+
+                    func_ret = await self.ap.tool_mgr.execute_func_call(
+                        query, func.name, parameters
+                    )
+
+                    msg = llm_entities.Message(
+                        role="tool", content=json.dumps(func_ret, ensure_ascii=False), tool_call_id=tool_call.id
+                    )
+
+                    yield msg
+
+                    req_messages.append(msg)
+                except Exception as e:
+                    # 工具调用出错，添加一个报错信息到 req_messages
+                    err_msg = llm_entities.Message(
+                        role="tool", content=f"err: {e}", tool_call_id=tool_call.id
+                    )
+
+                    yield err_msg
+
+                    req_messages.append(err_msg)
+
+            # 处理完所有调用，再次请求
+            msg = await query.use_model.requester.call(query.use_model, req_messages, query.use_funcs)
+
+            yield msg
+
+            pending_tool_calls = msg.tool_calls
+
+            req_messages.append(msg)

pkg/utils/announce.py

@@ -4,6 +4,7 @@ import json
 import typing
 import os
 import base64
+import logging
 
 import pydantic
 import requests
@@ -107,17 +108,20 @@ class AnnouncementManager:
 
     async def show_announcements(
         self
-    ):
+    ) -> typing.Tuple[str, int]:
         """显示公告"""
         try:
             announcements = await self.fetch_new()
+            ann_text = ""
             for ann in announcements:
-                self.ap.logger.info(f'[公告] {ann.time}: {ann.content}')
+                ann_text += f"[公告] {ann.time}: {ann.content}\n"
 
             if announcements:
 
                 await self.ap.ctr_mgr.main.post_announcement_showed(
                     ids=[item.id for item in announcements]
                 )
+
+            return ann_text, logging.INFO
         except Exception as e:
-            self.ap.logger.warning(f'获取公告时出错: {e}')
+            return f'获取公告时出错: {e}', logging.WARNING

pkg/utils/constants.py

@@ -1 +1 @@
-semantic_version = "v3.2.2"
+semantic_version = "v3.3.1.0"

pkg/utils/image.py

@@ -8,14 +8,14 @@ import aiohttp
 
 async def qq_image_url_to_base64(
     image_url: str
-) -> str:
-    """将QQ图片URL转为base64
+) -> typing.Tuple[str, str]:
+    """将QQ图片URL转为base64，并返回图片格式
 
     Args:
         image_url (str): QQ图片URL
 
     Returns:
-        str: base64编码
+        typing.Tuple[str, str]: base64编码和图片格式
     """
     parsed = urlparse(image_url)
     query = parse_qs(parsed.query)
@@ -35,7 +35,12 @@ async def qq_image_url_to_base64(
         ) as resp:
             resp.raise_for_status()  # 检查HTTP错误
             file_bytes = await resp.read()
+            content_type = resp.headers.get('Content-Type')
+            if not content_type or not content_type.startswith('image/'):
+                image_format = 'jpeg'
+            else: 
+                image_format = content_type.split('/')[-1]
 
     base64_str = base64.b64encode(file_bytes).decode()
 
-    return base64_str
+    return base64_str, image_format

pkg/utils/version.py

@@ -1,6 +1,8 @@
 from __future__ import annotations
 
 import os
+import typing
+import logging
 import time
 
 import requests
@@ -213,11 +215,11 @@ class VersionManager:
 
     async def show_version_update(
         self
-    ):
+    ) -> typing.Tuple[str, int]:
         try:
 
             if await self.ap.ver_mgr.is_new_version_available():
-                self.ap.logger.info("有新版本可用，请使用 !update 命令更新")
+                return "有新版本可用，请使用管理员账号发送 !update 命令更新", logging.INFO
         
         except Exception as e:
-            self.ap.logger.warning(f"检查版本更新时出错: {e}")
+            return f"检查版本更新时出错: {e}", logging.WARNING

requirements.txt

@@ -14,4 +14,5 @@ pydantic
 websockets
 urllib3
 psutil
-async-lru
+async-lru
+ollama

res/wiki/1-功能使用.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 ## 功能点列举
 
 <details>

res/wiki/2-功能常见问题.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 使用过程中的一些疑问，这里不是解决异常的地方，遇到异常请见`常见错误`页
 
 ### ❓ 如何更新代码到最新版本？

res/wiki/3-常见错误.md

@@ -1 +1,4 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 搜索[主仓库issue](https://github.com/RockChinQ/QChatGPT/issues)和[安装器issue](https://github.com/RockChinQ/qcg-installer/issues)

res/wiki/4-技术信息.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 以下是QChatGPT实现原理等技术信息，贡献之前请仔细阅读
 
 > 太久没更了，过时了，建议读源码，~~注释还挺全的~~  

res/wiki/5-插件使用.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 QChatGPT 插件使用Wiki
 
 ## 简介

res/wiki/6-插件使用-内容函数.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 > 说白了就是ChatGPT官方插件那种东西
 
 内容函数是基于[GPT的Function Calling能力](https://platform.openai.com/docs/guides/gpt/function-calling)实现的，这是一种嵌入对话中，由GPT自动调用的函数。  

res/wiki/7-插件开发.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 QChatGPT 插件开发Wiki
 
 > 请先阅读[插件使用页](https://github.com/RockChinQ/QChatGPT/wiki/5-%E6%8F%92%E4%BB%B6%E4%BD%BF%E7%94%A8)  

res/wiki/8-官方接口、ChatGPT网页版、ChatGPT-API区别.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 ## 多个对话接口有何区别？
 
 出于对稳定性的高要求，本项目主线接入的是GPT-3模型接口，此接口由OpenAI官方开放，稳定性强。  

res/wiki/9-go-cqhttp配置.md

@@ -1,3 +1,6 @@
+> [!WARNING]
+> 此 Wiki 已弃用，所有文档已迁移到 [项目主页](https://qchatgpt.rockchin.top)
+
 # 配置go-cqhttp用于登录QQ
 
 > 若您是从旧版本升级到此版本以使用go-cqhttp的用户，请您按照`config-template.py`的内容修改`config.py`，添加`msg_source_adapter`配置项并将其设为`nakuru`，同时添加`nakuru_config`字段按照说明配置。

templates/command.json

@@ -1,3 +1,7 @@
 {
-    "privilege": {}
+    "privilege": {},
+    "command-prefix": [
+        "!",
+        "！"
+    ]
 }

templates/provider.json

@@ -37,11 +37,17 @@
             "base-url": "https://api.deepseek.com",
             "args": {},
             "timeout": 120
+        },
+        "ollama-chat": {
+            "base-url": "http://127.0.0.1:11434",
+            "args": {},
+            "timeout": 600
         }
     },
     "model": "gpt-3.5-turbo",
     "prompt-mode": "normal",
     "prompt": {
         "default": "" 
-    }
+    },
+    "runner": "local-agent"
 }