Paper Search MCP Server

# paper_search_mcp/paper.py """ Paper 数据模型 - 学术论文标准化格式 2025 最佳实践版本： - 使用 Pydantic V2 提供运行时类型验证 - 自动类型转换和友好错误提示 - 内置 JSON 序列化支持 """ from pydantic import BaseModel, Field, field_validator, ConfigDict from datetime import datetime from typing import List, Dict, Optional, Any class Paper(BaseModel): """学术论文标准化数据模型 使用 Pydantic V2 提供: - 运行时类型验证 - 自动类型转换（如字符串日期 -> datetime） - JSON 序列化/反序列化 - 友好的验证错误信息 Example: >>> paper = Paper( ... paper_id="2106.12345", ... title="Attention Is All You Need", ... source="arxiv" ... ) >>> paper.model_dump() {'paper_id': '2106.12345', 'title': 'Attention Is All You Need', ...} """ # 配置 model_config = ConfigDict( # 允许额外字段（兼容性） extra='ignore', # 验证赋值 validate_assignment=True, ) # ======================================== # 核心字段（必填） # ======================================== paper_id: str = Field( ..., min_length=1, description="唯一标识符 (如 arXiv ID, PMID, DOI)" ) title: str = Field( ..., min_length=1, description="论文标题" ) source: str = Field( ..., description="来源平台 (如 'arxiv', 'pubmed', 'semantic')" ) # ======================================== # 核心字段（可选，有默认值） # ======================================== authors: List[str] = Field( default_factory=list, description="作者列表" ) abstract: str = Field( default="", description="摘要文本" ) doi: str = Field( default="", description="数字对象标识符 (DOI)" ) published_date: Optional[datetime] = Field( default=None, description="发布日期" ) pdf_url: str = Field( default="", description="PDF 直接下载链接" ) url: str = Field( default="", description="论文页面 URL" ) # ======================================== # 扩展字段（可选） # ======================================== updated_date: Optional[datetime] = Field( default=None, description="最后更新日期" ) categories: List[str] = Field( default_factory=list, description="学科分类" ) keywords: List[str] = Field( default_factory=list, description="关键词" ) citations: int = Field( default=0, ge=0, description="被引用次数" ) references: List[str] = Field( default_factory=list, description="参考文献 ID/DOI 列表" ) extra: Dict[str, Any] = Field( default_factory=dict, description="平台特定的额外元数据" ) # ======================================== # 字段验证器 # ======================================== @field_validator('title', 'abstract', mode='before') @classmethod def clean_whitespace(cls, v: Any) -> str: """清理标题和摘要中的多余空白和换行符""" if v is None: return "" if isinstance(v, str): # 替换换行为空格，合并多个空格 return ' '.join(v.split()) return str(v) @field_validator('authors', mode='before') @classmethod def ensure_authors_list(cls, v: Any) -> List[str]: """确保作者字段是列表""" if v is None: return [] if isinstance(v, str): # 支持分号或逗号分隔的字符串 if ';' in v: return [a.strip() for a in v.split(';') if a.strip()] elif ',' in v: return [a.strip() for a in v.split(',') if a.strip()] return [v.strip()] if v.strip() else [] return list(v) @field_validator('citations', mode='before') @classmethod def ensure_citations_int(cls, v: Any) -> int: """确保引用数是整数""" if v is None: return 0 if isinstance(v, str): try: return int(v) except ValueError: return 0 return int(v) if v else 0 # ======================================== # 序列化方法 # ======================================== def to_dict(self) -> Dict[str, Any]: """转换为字典格式（兼容旧 API） 此方法保持与旧 dataclass 版本的兼容性， 输出格式与之前相同（分号分隔的字符串等） Returns: Dict: 序列化的论文数据 """ return { 'paper_id': self.paper_id, 'title': self.title, 'authors': '; '.join(self.authors) if self.authors else '', 'abstract': self.abstract, 'doi': self.doi, 'published_date': self.published_date.isoformat() if self.published_date else '', 'pdf_url': self.pdf_url, 'url': self.url, 'source': self.source, 'updated_date': self.updated_date.isoformat() if self.updated_date else '', 'categories': '; '.join(self.categories) if self.categories else '', 'keywords': '; '.join(self.keywords) if self.keywords else '', 'citations': self.citations, 'references': '; '.join(self.references) if self.references else '', 'extra': str(self.extra) if self.extra else '' } def to_json_dict(self) -> Dict[str, Any]: """转换为 JSON 友好的字典格式 使用 Pydantic 的 model_dump，保持列表格式， 日期自动转换为 ISO 格式字符串 Returns: Dict: JSON 友好的论文数据 """ return self.model_dump(mode='json') # ======================================== # 测试代码 # ======================================== if __name__ == "__main__": # 测试基本创建 print("=" * 60) print("1. Testing basic Paper creation...") print("=" * 60) paper = Paper( paper_id="2106.12345", title="Attention Is\n All You Need ", # 会自动清理 source="arxiv", authors=["John Doe", "Jane Smith"], abstract="This is a test\n\nabstract.", doi="10.1234/example", published_date=datetime.now(), ) print(f"Title (cleaned): '{paper.title}'") print(f"Abstract (cleaned): '{paper.abstract}'") print(f"Authors: {paper.authors}") # 测试 to_dict() 兼容方法 print("\n" + "=" * 60) print("2. Testing to_dict() compatibility...") print("=" * 60) d = paper.to_dict() print(f"Authors (semicolon): '{d['authors']}'") print(f"Published date: '{d['published_date']}'") # 测试 model_dump() print("\n" + "=" * 60) print("3. Testing model_dump() (Pydantic native)...") print("=" * 60) json_dict = paper.to_json_dict() print(f"Authors (list): {json_dict['authors']}") # 测试验证 print("\n" + "=" * 60) print("4. Testing validation...") print("=" * 60) try: # 这应该失败：paper_id 不能为空 invalid = Paper(paper_id="", title="Test", source="test") except Exception as e: print(f"Validation error (expected): {e}") # 测试作者字符串解析 print("\n" + "=" * 60) print("5. Testing authors string parsing...") print("=" * 60) paper2 = Paper( paper_id="test", title="Test", source="test", authors="Alice; Bob; Charlie" # 字符串会被自动解析 ) print(f"Parsed authors: {paper2.authors}") print("\n✅ All tests passed!")