""" 数据模型定义 """ import os from dataclasses import dataclass from typing import List, Dict, Any, Literal, TypedDict, Optional from enum import Enum def _get_default_max_tokens() -> int: """ 从环境变量获取默认的 max_tokens 值 Returns: 默认的 max_tokens 值 """ try: # 尝试从环境变量读取 env_value = os.getenv("MAX_TOKENS") if env_value: return int(env_value) except (ValueError, TypeError): pass # 如果环境变量不存在或无效，返回默认值 return 4000 def _get_default_temperature() -> float: """ 从环境变量获取默认的 temperature 值 Returns: 默认的 temperature 值 """ try: # 尝试从环境变量读取 env_value = os.getenv("TEMPERATURE") if env_value: return float(env_value) except (ValueError, TypeError): pass # 如果环境变量不存在或无效，返回默认值 return 0.7 class ChunkStrategy(Enum): """文档分块策略""" PARAGRAPH = "paragraph" # 基于段落分块 SEMANTIC = "semantic" # 语义分块 RECURSIVE = "recursive" # 递归分块 HYBRID = "hybrid" # 混合策略 FAST = "fast" # 快速分块策略 @dataclass class SlideInfo: """幻灯片信息数据类""" page_number: int title: str content_points: List[str] slide_type: Literal["title", "content", "conclusion"] description: str chart_config: Optional[Dict[str, Any]] = None # 新增：图表配置 def to_dict(self) -> Dict[str, Any]: """转换为字典格式""" result = { "page_number": self.page_number, "title": self.title, "content_points": self.content_points, "slide_type": self.slide_type, "description": self.description, } # 只有当chart_config存在时才添加到字典中 if self.chart_config is not None: result["chart_config"] = self.chart_config return result @classmethod def from_dict(cls, data: Dict[str, Any]) -> "SlideInfo": """从字典创建实例""" return cls( page_number=data.get("page_number", 1), title=data.get("title", ""), content_points=data.get("content_points", []), slide_type=data.get("slide_type", "content"), description=data.get("description", ""), chart_config=data.get("chart_config"), # 新增：图表配置 ) class PPTState(TypedDict): """PPT生成状态""" document_chunks: List[str] current_index: int ppt_title: str slides: List[Dict[str, Any]] total_pages: int page_count_mode: str document_structure: Dict[str, Any] accumulated_context: str # 项目信息参数 project_topic: str project_scenario: str project_requirements: str target_audience: str custom_audience: str ppt_style: str custom_style_prompt: str # 页数设置参数 min_pages: Optional[int] max_pages: Optional[int] fixed_pages: Optional[int] @dataclass class DocumentInfo: """文档信息""" title: str content: str file_path: str file_type: str encoding: str size: int def to_dict(self) -> Dict[str, Any]: """转换为字典格式""" return { "title": self.title, "content": self.content, "file_path": self.file_path, "file_type": self.file_type, "encoding": self.encoding, "size": self.size, } @dataclass class ProcessingConfig: """处理配置""" max_slides: int = 25 min_slides: int = 5 # 新增：最小页数 chunk_size: int = 3000 chunk_overlap: int = 200 chunk_strategy: ChunkStrategy = ChunkStrategy.PARAGRAPH llm_model: str = "gpt-4o-mini" llm_provider: str = "openai" temperature: float = None # 将在 __post_init__ 中设置默认值 max_tokens: int = None # 将在 __post_init__ 中设置默认值 recursion_limit: Optional[int] = None # 工作流递归限制，None表示自动计算 target_language: str = "zh" # 新增：目标语言，由用户在表单中选择 # API配置（从数据库读取，不使用环境变量） api_key: Optional[str] = None base_url: Optional[str] = None use_responses_api: bool = False enable_reasoning: bool = False reasoning_effort: str = "medium" def __post_init__(self): """后处理验证和默认值设置""" # 如果 max_tokens 为 None，从环境变量获取默认值 if self.max_tokens is None: self.max_tokens = _get_default_max_tokens() # 如果 temperature 为 None，从环境变量获取默认值 if self.temperature is None: self.temperature = _get_default_temperature() # 验证逻辑 if self.min_slides > self.max_slides: raise ValueError(f"最小页数({self.min_slides})不能大于最大页数({self.max_slides})") if self.min_slides < 1: raise ValueError("最小页数不能小于1") if self.max_slides > 1000: raise ValueError("最大页数不能超过1000") if self.recursion_limit is not None and self.recursion_limit < 10: raise ValueError("递归限制不能小于10") @property def slides_range(self) -> str: """获取页数范围的字符串表示""" if self.min_slides == self.max_slides: return f"{self.min_slides}页" return f"{self.min_slides}-{self.max_slides}页" def to_dict(self) -> Dict[str, Any]: """转换为字典格式""" return { "max_slides": self.max_slides, "min_slides": self.min_slides, "slides_range": self.slides_range, "chunk_size": self.chunk_size, "chunk_overlap": self.chunk_overlap, "chunk_strategy": self.chunk_strategy.value, "llm_model": self.llm_model, "llm_provider": self.llm_provider, "temperature": self.temperature, "max_tokens": self.max_tokens, "recursion_limit": self.recursion_limit, "target_language": self.target_language, "use_responses_api": self.use_responses_api, "enable_reasoning": self.enable_reasoning, "reasoning_effort": self.reasoning_effort, } @dataclass class PPTOutline: """PPT大纲结果""" title: str total_pages: int page_count_mode: str slides: List[SlideInfo] def to_dict(self) -> Dict[str, Any]: """转换为字典格式""" return { "title": self.title, "total_pages": self.total_pages, "page_count_mode": self.page_count_mode, "slides": [slide.to_dict() for slide in self.slides], } @classmethod def from_dict(cls, data: Dict[str, Any]) -> "PPTOutline": """从字典创建实例""" slides = [SlideInfo.from_dict(slide_data) for slide_data in data.get("slides", [])] return cls( title=data.get("title", ""), total_pages=data.get("total_pages", 0), page_count_mode=data.get("page_count_mode", "final"), slides=slides, )