course-agent-od/backend/scripts/import_book_structure.py

"""
导入书籍结构到数据库
解析LaTeX文件并导入书籍、章节、节、小节（知识点）
层级结构：书籍 -> Chapter -> Section -> Subsection（知识点）
"""
import sys
import re
from pathlib import Path

# 添加项目根目录到路径
sys.path.insert(0, str(Path(__file__).parent.parent))

from sqlalchemy.orm import Session
from src.core.database import SessionLocal, create_tables
from src.core.config import get_settings
from src.models.book_structure import Book, Chapter, Section, Subsection
from src.services.latex_parser import LaTeXParser


def import_book_structure():
    """导入书籍结构"""
    settings = get_settings()
    book_dir = Path(settings.book_dir)
    
    # 如果相对路径不存在，尝试从项目根目录查找
    if not book_dir.exists():
        # 脚本位置: dofile/backend/scripts/import_book_structure.py
        # 项目根目录: 向上3级
        script_dir = Path(__file__).parent.parent.parent.parent  # 项目根目录
        book_dir = script_dir / "data" / "book"
        if not book_dir.exists():
            print(f"错误: 书籍目录不存在")
            print(f"尝试的路径1: {settings.book_dir}")
            print(f"尝试的路径2: {book_dir}")
            return
        print(f"使用书籍目录: {book_dir}")
    
    # 创建数据库表
    print("创建数据库表...")
    create_tables()
    
    # 创建数据库会话
    db: Session = SessionLocal()
    try:
        # 清空旧数据（如果存在）
        db.query(Subsection).delete()
        db.query(Section).delete()
        db.query(Chapter).delete()
        db.query(Book).delete()
        db.commit()
        print("已清空旧的书籍结构数据。")

        # 创建书籍
        book_title = "国土空间规划：概念、原理、方法"
        book_description = "基于LaTeX书籍内容构建的知识图谱系统"
        book = Book(title=book_title, description=book_description)
        db.add(book)
        db.commit()
        db.refresh(book)
        print(f"已创建书籍: {book.title}")

        parser = LaTeXParser(book_dir)
        
        # 解析主文件获取章节顺序
        main_tex_path = book_dir / "main.tex"
        if not main_tex_path.exists():
            print(f"错误: main.tex 文件不存在于 {book_dir}")
            return

        main_content = main_tex_path.read_text(encoding='utf-8')
        chapter_input_re = re.compile(r'^\\input{(chapter\d+)}', re.MULTILINE)
        
        chapter_files_in_order = []
        for match in chapter_input_re.finditer(main_content):
            chapter_files_in_order.append(f"{match.group(1)}.tex")

        if not chapter_files_in_order:
            print("未在 main.tex 中找到章节文件引用。")
            return

        total_chapters = 0
        total_sections = 0
        total_subsections = 0

        # 遍历章节文件并导入
        for chapter_idx, chapter_filename in enumerate(chapter_files_in_order):
            chapter_file_path = book_dir / chapter_filename
            if not chapter_file_path.exists():
                print(f"警告: 章节文件 {chapter_file_path} 不存在，跳过。")
                continue

            print(f"解析LaTeX文件: {chapter_file_path.name}...")
            chapter_structure = parser.parse_chapter_file(chapter_file_path)
            
            if not chapter_structure:
                print(f"警告: 文件 {chapter_file_path.name} 未解析出任何结构。")
                continue

            # 每个文件只包含一个 \chapter
            parsed_chapter_data = chapter_structure[0] 
            
            chapter_obj = Chapter(
                book_id=book.id,
                chapter_number=parsed_chapter_data["chapter_number"],
                title=parsed_chapter_data["title"],
                file_path=chapter_file_path.name,
                start_line=parsed_chapter_data["start_line"],
                end_line=parsed_chapter_data["end_line"],
                display_order=parsed_chapter_data["chapter_number"]
            )
            db.add(chapter_obj)
            db.flush() # Flush to get chapter_obj.id
            total_chapters += 1

            # 导入节（Section）
            for section_data in parsed_chapter_data.get("sections", []):
                section_obj = Section(
                    chapter_id=chapter_obj.id,
                    section_number=section_data["section_number"],
                    title=section_data["title"],
                    file_path=chapter_file_path.name,
                    start_line=section_data["start_line"],
                    end_line=section_data["end_line"],
                    display_order=section_data["section_number"]
                )
                db.add(section_obj)
                db.flush() # Flush to get section_obj.id
                total_sections += 1

                # 导入小节（Subsection，作为知识点）
                for subsection_data in section_data.get("subsections", []):
                    subsection_obj = Subsection(
                        section_id=section_obj.id,
                        subsection_number=subsection_data["subsection_number"],
                        title=subsection_data["title"],
                        file_path=chapter_file_path.name,
                        start_line=subsection_data["start_line"],
                        end_line=subsection_data["end_line"],
                        display_order=subsection_data["subsection_number"]
                    )
                    db.add(subsection_obj)
                    total_subsections += 1
        
        db.commit()
        print(f"\n[SUCCESS] 书籍结构导入成功!")
        print(f"  - 书籍: {book.title}")
        print(f"  - 章节: {total_chapters} 个")
        print(f"  - 节: {total_sections} 个")
        print(f"  - 小节（知识点）: {total_subsections} 个")

    except Exception as e:
        db.rollback()
        print(f"[ERROR] 导入失败: {e}")
        import traceback
        traceback.print_exc()
        raise
    finally:
        db.close()


if __name__ == "__main__":
    import_book_structure()