aigroup-econ-mcp 0.4.2__py3-none-any.whl → 1.4.3__py3-none-any.whl

aigroup_econ_mcp/tools/tool_registry.py

@@ -1,170 +0,0 @@
-"""
-工具注册器模块
-简化MCP工具的注册和管理
-"""
-
-from typing import Dict, Any, Optional, List, Callable
-from pydantic import Field
-from typing import Annotated
-
-from .base import with_file_support_decorator as econometric_tool
-
-
-# 标准文件输入参数定义
-FILE_INPUT_PARAMS = {
-    "file_content": Annotated[
-        Optional[str],
-        Field(
-            default=None,
-            description="""CSV或JSON文件内容
-
-📁 支持格式：
-- CSV: 带表头的列数据，自动检测分隔符
-- JSON: {"变量名": [数据], ...} 或 [{"变量1": 值, ...}, ...]
-
-💡 使用方式：
-- 提供文件内容字符串
-- 系统会自动解析并识别变量
-- 优先使用file_content，如果提供则忽略其他数据参数"""
-        )
-    ],
-    "file_format": Annotated[
-        str,
-        Field(
-            default="auto",
-            description="""文件格式
-
-可选值：
-- "auto": 自动检测（默认）
-- "csv": CSV格式
-- "json": JSON格式"""
-        )
-    ]
-}
-
-
-class ToolConfig:
-    """工具配置类"""
-    
-    def __init__(
-        self,
-        name: str,
-        impl_func: Callable,
-        tool_type: str,
-        description: str = "",
-        extra_params: Dict[str, Any] = None
-    ):
-        self.name = name
-        self.impl_func = impl_func
-        self.tool_type = tool_type
-        self.description = description
-        self.extra_params = extra_params or {}
-
-
-def create_tool_wrapper(config: ToolConfig):
-    """
-    创建工具包装器，自动添加文件输入支持
-    
-    Args:
-        config: 工具配置对象
-    
-    Returns:
-        包装后的工具函数
-    """
-    @econometric_tool(config.tool_type)
-    async def tool_wrapper(ctx, **kwargs):
-        """动态生成的工具包装器"""
-        # 调用实际的实现函数
-        return await config.impl_func(ctx, **kwargs)
-    
-    # 设置函数名和文档
-    tool_wrapper.__name__ = config.name
-    tool_wrapper.__doc__ = config.description
-    
-    return tool_wrapper
-
-
-# 工具类型到参数映射
-TOOL_TYPE_PARAMS = {
-    "multi_var_dict": {
-        "data": Annotated[
-            Optional[Dict[str, List[float]]],
-            Field(default=None, description="数据字典，格式为 {变量名: [数值列表]}")
-        ]
-    },
-    "regression": {
-        "y_data": Annotated[
-            Optional[List[float]],
-            Field(default=None, description="因变量数据")
-        ],
-        "x_data": Annotated[
-            Optional[List[List[float]]],
-            Field(default=None, description="自变量数据")
-        ],
-        "feature_names": Annotated[
-            Optional[List[str]],
-            Field(default=None, description="特征名称")
-        ]
-    },
-    "single_var": {
-        "data": Annotated[
-            Optional[List[float]],
-            Field(default=None, description="时间序列数据")
-        ]
-    },
-    "panel": {
-        "y_data": Annotated[
-            Optional[List[float]],
-            Field(default=None, description="因变量数据")
-        ],
-        "x_data": Annotated[
-            Optional[List[List[float]]],
-            Field(default=None, description="自变量数据")
-        ],
-        "entity_ids": Annotated[
-            Optional[List[str]],
-            Field(default=None, description="个体标识符")
-        ],
-        "time_periods": Annotated[
-            Optional[List[str]],
-            Field(default=None, description="时间标识符")
-        ],
-        "feature_names": Annotated[
-            Optional[List[str]],
-            Field(default=None, description="特征名称")
-        ]
-    },
-    "time_series": {
-        "data": Annotated[
-            Optional[Dict[str, List[float]]],
-            Field(default=None, description="多变量时间序列数据")
-        ]
-    }
-}
-
-
-def get_tool_params(tool_type: str, extra_params: Dict[str, Any] = None) -> Dict[str, Any]:
-    """
-    获取工具的完整参数定义
-    
-    Args:
-        tool_type: 工具类型
-        extra_params: 额外的参数定义
-    
-    Returns:
-        完整的参数字典
-    """
-    params = {}
-    
-    # 添加基础参数
-    if tool_type in TOOL_TYPE_PARAMS:
-        params.update(TOOL_TYPE_PARAMS[tool_type])
-    
-    # 添加文件输入参数
-    params.update(FILE_INPUT_PARAMS)
-    
-    # 添加额外参数
-    if extra_params:
-        params.update(extra_params)
-    
-    return params

aigroup_econ_mcp/tools/validation.py

@@ -1,482 +0,0 @@
-"""
-统一的参数验证系统
-提供类型检查、数据验证和参数验证功能
-"""
-
-import re
-import numpy as np
-import pandas as pd
-from typing import Any, Dict, List, Optional, Union, Tuple, Callable
-from pydantic import BaseModel, Field, validator
-from enum import Enum
-import warnings
-
-
-class ValidationError(Exception):
-    """参数验证错误"""
-    def __init__(self, message: str, field: str = None, value: Any = None):
-        self.message = message
-        self.field = field
-        self.value = value
-        super().__init__(self.message)
-
-
-class DataType(Enum):
-    """支持的数据类型"""
-    NUMERIC = "numeric"
-    INTEGER = "integer"
-    FLOAT = "float"
-    STRING = "string"
-    BOOLEAN = "boolean"
-    LIST = "list"
-    DICT = "dict"
-    DATAFRAME = "dataframe"
-    SERIES = "series"
-
-
-class ValidationRule(BaseModel):
-    """验证规则定义"""
-    required: bool = Field(default=True, description="是否必需")
-    data_type: DataType = Field(description="数据类型")
-    min_value: Optional[float] = Field(default=None, description="最小值")
-    max_value: Optional[float] = Field(default=None, description="最大值")
-    min_length: Optional[int] = Field(default=None, description="最小长度")
-    max_length: Optional[int] = Field(default=None, description="最大长度")
-    pattern: Optional[str] = Field(default=None, description="正则表达式模式")
-    allowed_values: Optional[List[Any]] = Field(default=None, description="允许的值")
-    custom_validator: Optional[Callable] = Field(default=None, description="自定义验证函数")
-
-
-class ParameterValidator:
-    """
-    参数验证器
-    提供统一的参数验证和类型检查功能
-    """
-    
-    def __init__(self, strict_mode: bool = True):
-        """
-        初始化参数验证器
-        
-        Args:
-            strict_mode: 严格模式，如果为True，验证失败时抛出异常
-        """
-        self.strict_mode = strict_mode
-        self._validation_rules: Dict[str, ValidationRule] = {}
-    
-    def add_rule(self, field_name: str, rule: ValidationRule):
-        """
-        添加验证规则
-        
-        Args:
-            field_name: 字段名称
-            rule: 验证规则
-        """
-        self._validation_rules[field_name] = rule
-    
-    def validate_parameter(self, field_name: str, value: Any) -> Tuple[bool, Optional[str]]:
-        """
-        验证单个参数
-        
-        Args:
-            field_name: 字段名称
-            value: 参数值
-            
-        Returns:
-            Tuple[bool, Optional[str]]: (是否验证通过, 错误信息)
-        """
-        if field_name not in self._validation_rules:
-            return True, None
-        
-        rule = self._validation_rules[field_name]
-        
-        # 检查必需性
-        if rule.required and value is None:
-            return False, f"参数 '{field_name}' 是必需的"
-        
-        if value is None:
-            return True, None
-        
-        # 检查数据类型
-        type_valid, type_error = self._validate_data_type(value, rule.data_type)
-        if not type_valid:
-            return False, type_error
-        
-        # 检查数值范围
-        if rule.min_value is not None or rule.max_value is not None:
-            range_valid, range_error = self._validate_numeric_range(value, rule.min_value, rule.max_value)
-            if not range_valid:
-                return False, range_error
-        
-        # 检查长度范围
-        if rule.min_length is not None or rule.max_length is not None:
-            length_valid, length_error = self._validate_length(value, rule.min_length, rule.max_length)
-            if not length_valid:
-                return False, length_error
-        
-        # 检查正则表达式模式
-        if rule.pattern is not None:
-            pattern_valid, pattern_error = self._validate_pattern(value, rule.pattern)
-            if not pattern_valid:
-                return False, pattern_error
-        
-        # 检查允许的值
-        if rule.allowed_values is not None:
-            allowed_valid, allowed_error = self._validate_allowed_values(value, rule.allowed_values)
-            if not allowed_valid:
-                return False, allowed_error
-        
-        # 自定义验证
-        if rule.custom_validator is not None:
-            try:
-                custom_valid = rule.custom_validator(value)
-                if not custom_valid:
-                    return False, f"参数 '{field_name}' 未通过自定义验证"
-            except Exception as e:
-                return False, f"参数 '{field_name}' 自定义验证失败: {str(e)}"
-        
-        return True, None
-    
-    def _validate_data_type(self, value: Any, data_type: DataType) -> Tuple[bool, Optional[str]]:
-        """验证数据类型"""
-        try:
-            if data_type == DataType.NUMERIC:
-                if not isinstance(value, (int, float, np.number)):
-                    return False, f"期望数值类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.INTEGER:
-                if not isinstance(value, (int, np.integer)):
-                    return False, f"期望整数类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.FLOAT:
-                if not isinstance(value, (float, np.floating)):
-                    return False, f"期望浮点数类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.STRING:
-                if not isinstance(value, str):
-                    return False, f"期望字符串类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.BOOLEAN:
-                if not isinstance(value, bool):
-                    return False, f"期望布尔类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.LIST:
-                if not isinstance(value, (list, tuple, np.ndarray)):
-                    return False, f"期望列表类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.DICT:
-                if not isinstance(value, dict):
-                    return False, f"期望字典类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.DATAFRAME:
-                if not isinstance(value, pd.DataFrame):
-                    return False, f"期望DataFrame类型，实际类型: {type(value).__name__}"
-            
-            elif data_type == DataType.SERIES:
-                if not isinstance(value, pd.Series):
-                    return False, f"期望Series类型，实际类型: {type(value).__name__}"
-            
-            return True, None
-            
-        except Exception as e:
-            return False, f"数据类型验证失败: {str(e)}"
-    
-    def _validate_numeric_range(self, value: Any, min_val: Optional[float], max_val: Optional[float]) -> Tuple[bool, Optional[str]]:
-        """验证数值范围"""
-        try:
-            numeric_value = float(value)
-            
-            if min_val is not None and numeric_value < min_val:
-                return False, f"数值 {numeric_value} 小于最小值 {min_val}"
-            
-            if max_val is not None and numeric_value > max_val:
-                return False, f"数值 {numeric_value} 大于最大值 {max_val}"
-            
-            return True, None
-            
-        except (ValueError, TypeError):
-            return False, "无法转换为数值进行范围验证"
-    
-    def _validate_length(self, value: Any, min_len: Optional[int], max_len: Optional[int]) -> Tuple[bool, Optional[str]]:
-        """验证长度范围"""
-        try:
-            length = len(value)
-            
-            if min_len is not None and length < min_len:
-                return False, f"长度 {length} 小于最小长度 {min_len}"
-            
-            if max_len is not None and length > max_len:
-                return False, f"长度 {length} 大于最大长度 {max_len}"
-            
-            return True, None
-            
-        except TypeError:
-            return False, "无法获取长度信息"
-    
-    def _validate_pattern(self, value: Any, pattern: str) -> Tuple[bool, Optional[str]]:
-        """验证正则表达式模式"""
-        try:
-            if not isinstance(value, str):
-                return False, "模式验证仅适用于字符串类型"
-            
-            if not re.match(pattern, value):
-                return False, f"字符串 '{value}' 不匹配模式 '{pattern}'"
-            
-            return True, None
-            
-        except re.error as e:
-            return False, f"正则表达式模式错误: {str(e)}"
-    
-    def _validate_allowed_values(self, value: Any, allowed_values: List[Any]) -> Tuple[bool, Optional[str]]:
-        """验证允许的值"""
-        if value not in allowed_values:
-            return False, f"值 '{value}' 不在允许的值列表中: {allowed_values}"
-        
-        return True, None
-    
-    def validate_all(self, parameters: Dict[str, Any]) -> Dict[str, Any]:
-        """
-        验证所有参数
-        
-        Args:
-            parameters: 参数字典
-            
-        Returns:
-            Dict[str, Any]: 验证后的参数（包含类型转换）
-            
-        Raises:
-            ValidationError: 验证失败时抛出
-        """
-        validated_params = {}
-        errors = []
-        
-        for field_name, rule in self._validation_rules.items():
-            value = parameters.get(field_name)
-            
-            # 验证参数
-            is_valid, error_msg = self.validate_parameter(field_name, value)
-            
-            if not is_valid:
-                if self.strict_mode:
-                    raise ValidationError(error_msg, field_name, value)
-                else:
-                    errors.append(f"{field_name}: {error_msg}")
-                    continue
-            
-            # 类型转换
-            if value is not None and rule.data_type:
-                try:
-                    converted_value = self._convert_type(value, rule.data_type)
-                    validated_params[field_name] = converted_value
-                except Exception as e:
-                    error_msg = f"参数 '{field_name}' 类型转换失败: {str(e)}"
-                    if self.strict_mode:
-                        raise ValidationError(error_msg, field_name, value)
-                    else:
-                        errors.append(error_msg)
-            else:
-                validated_params[field_name] = value
-        
-        if errors and not self.strict_mode:
-            warnings.warn(f"参数验证警告: {', '.join(errors)}")
-        
-        return validated_params
-    
-    def _convert_type(self, value: Any, data_type: DataType) -> Any:
-        """类型转换"""
-        if data_type == DataType.INTEGER:
-            return int(value)
-        elif data_type == DataType.FLOAT:
-            return float(value)
-        elif data_type == DataType.STRING:
-            return str(value)
-        elif data_type == DataType.BOOLEAN:
-            return bool(value)
-        elif data_type == DataType.LIST:
-            if isinstance(value, (tuple, np.ndarray)):
-                return list(value)
-            return value
-        else:
-            return value
-
-
-# 预定义的验证器实例
-class EconometricValidator:
-    """计量经济学专用验证器"""
-    
-    @staticmethod
-    def create_data_validator() -> ParameterValidator:
-        """创建数据验证器"""
-        validator = ParameterValidator()
-        
-        # 数据验证规则
-        validator.add_rule("data", ValidationRule(
-            required=True,
-            data_type=DataType.DICT,
-            min_length=1,
-            custom_validator=lambda x: all(isinstance(v, (list, np.ndarray)) for v in x.values())
-        ))
-        
-        validator.add_rule("y_data", ValidationRule(
-            required=True,
-            data_type=DataType.LIST,
-            min_length=10,
-            custom_validator=lambda x: all(isinstance(v, (int, float)) for v in x)
-        ))
-        
-        validator.add_rule("x_data", ValidationRule(
-            required=True,
-            data_type=DataType.LIST,
-            min_length=1,
-            custom_validator=lambda x: all(isinstance(row, (list, np.ndarray)) and len(row) > 0 for row in x)
-        ))
-        
-        return validator
-    
-    @staticmethod
-    def create_model_validator() -> ParameterValidator:
-        """创建模型参数验证器"""
-        validator = ParameterValidator()
-        
-        # 模型参数验证规则
-        validator.add_rule("n_estimators", ValidationRule(
-            required=False,
-            data_type=DataType.INTEGER,
-            min_value=1,
-            max_value=10000
-        ))
-        
-        validator.add_rule("max_depth", ValidationRule(
-            required=False,
-            data_type=DataType.INTEGER,
-            min_value=1,
-            max_value=100
-        ))
-        
-        validator.add_rule("learning_rate", ValidationRule(
-            required=False,
-            data_type=DataType.FLOAT,
-            min_value=0.001,
-            max_value=1.0
-        ))
-        
-        validator.add_rule("alpha", ValidationRule(
-            required=False,
-            data_type=DataType.FLOAT,
-            min_value=0.0,
-            max_value=100.0
-        ))
-        
-        return validator
-
-
-# 便捷验证函数
-def validate_econometric_data(data: Dict[str, Any]) -> Dict[str, Any]:
-    """
-    验证计量经济学数据
-    
-    Args:
-        data: 输入数据，格式为 {变量名: [数值列表]}
-        
-    Returns:
-        Dict[str, Any]: 验证后的数据
-    """
-    # 创建专门的数据验证器
-    validator = ParameterValidator()
-    
-    # 添加数据验证规则
-    validator.add_rule("data", ValidationRule(
-        required=True,
-        data_type=DataType.DICT,
-        min_length=1,
-        custom_validator=lambda x: all(isinstance(v, (list, np.ndarray)) for v in x.values())
-    ))
-    
-    # 验证数据格式
-    validated = validator.validate_all({"data": data})
-    return validated["data"]
-
-
-def validate_model_parameters(parameters: Dict[str, Any]) -> Dict[str, Any]:
-    """
-    验证模型参数
-    
-    Args:
-        parameters: 模型参数
-        
-    Returns:
-        Dict[str, Any]: 验证后的参数
-    """
-    validator = EconometricValidator.create_model_validator()
-    return validator.validate_all(parameters)
-
-
-def validate_time_series_data(data: List[float], min_length: int = 10) -> List[float]:
-    """
-    验证时间序列数据
-    
-    Args:
-        data: 时间序列数据
-        min_length: 最小长度要求
-        
-    Returns:
-        List[float]: 验证后的数据
-        
-    Raises:
-        ValidationError: 验证失败
-    """
-    if not isinstance(data, (list, np.ndarray)):
-        raise ValidationError("时间序列数据必须是列表或数组")
-    
-    if len(data) < min_length:
-        raise ValidationError(f"时间序列数据长度不足，至少需要 {min_length} 个观测点")
-    
-    if not all(isinstance(x, (int, float)) for x in data):
-        raise ValidationError("时间序列数据必须全部为数值")
-    
-    # 检查缺失值
-    if any(pd.isna(x) for x in data):
-        raise ValidationError("时间序列数据包含缺失值")
-    
-    return list(data)
-
-def validate_numeric_data(data: List[float], data_name: str = "数据") -> bool:
-    """
-    验证数值数据
-    
-    Args:
-        data: 数值数据列表
-        data_name: 数据名称，用于错误信息
-        
-    Returns:
-        bool: 验证是否通过
-        
-    Raises:
-        ValidationError: 验证失败时抛出
-    """
-    if not isinstance(data, (list, np.ndarray)):
-        raise ValidationError(f"{data_name} 必须是列表或数组")
-    
-    if len(data) == 0:
-        raise ValidationError(f"{data_name} 不能为空")
-    
-    if not all(isinstance(x, (int, float)) for x in data):
-        raise ValidationError(f"{data_name} 必须全部为数值")
-    
-    # 检查缺失值
-    if any(pd.isna(x) for x in data):
-        raise ValidationError(f"{data_name} 包含缺失值")
-    
-    return True
-
-# 导出主要类和函数
-__all__ = [
-    "ValidationError",
-    "DataType",
-    "ValidationRule", 
-    "ParameterValidator",
-    "EconometricValidator",
-    "validate_econometric_data",
-    "validate_model_parameters",
-    "validate_time_series_data",
-    "validate_numeric_data"
-]