"""
知识库 ORM 模型

对应数据库表 knowledge_documents 和 knowledge_chunks。
"""

import uuid

from sqlalchemy import Column, String, Integer, Text, DateTime, ForeignKey, func

from app.database import Base


class KnowledgeDocument(Base):
    """知识库文档表"""
    __tablename__ = "knowledge_documents"

    id = Column(
        String(36),
        primary_key=True,
        default=lambda: str(uuid.uuid4()),
        comment="文档唯一标识（UUID）",
    )
    filename = Column(String(255), nullable=False, comment="原始文件名")
    title = Column(String(500), default="", comment="文档标题")
    author = Column(String(255), default="", comment="作者")
    total_chunks = Column(Integer, default=0, comment="分块总数")
    file_size = Column(Integer, default=0, comment="文件大小（字节）")
    file_hash = Column(String(64), unique=True, nullable=False, comment="SHA256 哈希")
    status = Column(
        String(20), default="pending",
        comment="处理状态: pending/processing/ready/error"
    )
    error_message = Column(Text, default="", comment="错误信息")

    created_at = Column(DateTime, server_default=func.now(), comment="创建时间")
    updated_at = Column(
        DateTime, server_default=func.now(), onupdate=func.now(), comment="更新时间"
    )

    def __repr__(self):
        return f"<KnowledgeDocument(id={self.id}, filename={self.filename}, status={self.status})>"


class KnowledgeChunk(Base):
    """知识库文本分块表"""
    __tablename__ = "knowledge_chunks"

    id = Column(
        String(36),
        primary_key=True,
        default=lambda: str(uuid.uuid4()),
        comment="分块唯一标识（UUID）",
    )
    document_id = Column(
        String(36),
        ForeignKey("knowledge_documents.id", ondelete="CASCADE"),
        nullable=False,
        comment="所属文档 ID",
    )
    chunk_index = Column(Integer, nullable=False, comment="分块序号（从 0 开始）")
    content = Column(Text, nullable=False, comment="分块文本内容")
    token_count = Column(Integer, default=0, comment="Token 数量（估算）")
    embedding_id = Column(Integer, nullable=True, comment="FAISS 索引中的 ID")

    created_at = Column(DateTime, server_default=func.now(), comment="创建时间")

    def __repr__(self):
        return f"<KnowledgeChunk(id={self.id}, doc_id={self.document_id}, index={self.chunk_index})>"