"""content_db.py — SQLite 内容数据库操作类""" import json import sqlite3 from pathlib import Path from typing import Optional class ContentDB: """SQLite 内容数据库操作类 每个文档对应一个独立的 SQLite 数据库文件,存储文档的所有 Blocks。 """ def __init__(self, db_path: str): self.db_path = Path(db_path) self.conn: Optional[sqlite3.Connection] = None def connect(self): """连接数据库""" self.conn = sqlite3.connect(str(self.db_path)) self.conn.row_factory = sqlite3.Row return self def close(self): """关闭连接""" if self.conn: self.conn.close() self.conn = None def __enter__(self): """上下文管理器入口""" return self.connect() def __exit__(self, exc_type, exc_val, exc_tb): """上下文管理器退出""" self.close() def create_tables(self): """创建 document_blocks 表""" self.conn.execute(""" CREATE TABLE IF NOT EXISTS document_blocks ( id TEXT PRIMARY KEY, block_order INTEGER NOT NULL, type TEXT NOT NULL, level INTEGER DEFAULT 0, "index" INTEGER DEFAULT 0, content TEXT NOT NULL, word_style TEXT DEFAULT '', style TEXT DEFAULT '{}', metadata TEXT DEFAULT '{}' ) """) # 创建索引 self.conn.execute( 'CREATE INDEX IF NOT EXISTS idx_block_order ON document_blocks(block_order)' ) self.conn.execute( 'CREATE INDEX IF NOT EXISTS idx_type ON document_blocks(type)' ) self.conn.execute( 'CREATE INDEX IF NOT EXISTS idx_level ON document_blocks(level)' ) self.conn.commit() def insert_blocks(self, blocks: list[dict]): """批量插入 blocks""" for block in blocks: content = block['content'] if isinstance(content, (dict, list)): content = json.dumps(content, ensure_ascii=False) self.conn.execute(""" INSERT INTO document_blocks (id, block_order, type, level, "index", content, word_style, style, metadata) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?) """, ( block['id'], block['block_order'], block['type'], block.get('level', 0), block.get('index', 0), content, block.get('word_style', ''), json.dumps(block.get('style', {}), ensure_ascii=False), json.dumps(block.get('metadata', {}), ensure_ascii=False) )) self.conn.commit() def get_blocks(self, order_by: str = 'block_order') -> list[dict]: """查询所有 blocks""" cursor = self.conn.execute(f""" SELECT * FROM document_blocks ORDER BY {order_by} """) rows = cursor.fetchall() return [self._row_to_dict(row) for row in rows] def get_block_by_id(self, block_id: str) -> Optional[dict]: """按 ID 查询单个 block""" cursor = self.conn.execute(""" SELECT * FROM document_blocks WHERE id = ? """, (block_id,)) row = cursor.fetchone() return self._row_to_dict(row) if row else None def update_block(self, block_id: str, updates: dict): """更新单个 block""" set_clauses = [] params = [] if 'content' in updates: content = updates['content'] if isinstance(content, (dict, list)): content = json.dumps(content, ensure_ascii=False) set_clauses.append('content = ?') params.append(content) if 'style' in updates: set_clauses.append('style = ?') params.append(json.dumps(updates['style'], ensure_ascii=False)) if 'word_style' in updates: set_clauses.append('word_style = ?') params.append(updates['word_style']) if 'metadata' in updates: set_clauses.append('metadata = ?') params.append(json.dumps(updates['metadata'], ensure_ascii=False)) if not set_clauses: return params.append(block_id) sql = f"UPDATE document_blocks SET {', '.join(set_clauses)} WHERE id = ?" self.conn.execute(sql, params) self.conn.commit() def delete_block(self, block_id: str): """删除单个 block""" self.conn.execute('DELETE FROM document_blocks WHERE id = ?', (block_id,)) self.conn.commit() def search_blocks(self, query: str, block_type: Optional[str] = None) -> list[dict]: """搜索 blocks""" sql = "SELECT * FROM document_blocks WHERE content LIKE ?" params = [f'%{query}%'] if block_type: sql += " AND type = ?" params.append(block_type) sql += " ORDER BY block_order" cursor = self.conn.execute(sql, params) rows = cursor.fetchall() return [self._row_to_dict(row) for row in rows] def get_headings(self) -> list[dict]: """获取所有标题块""" cursor = self.conn.execute(""" SELECT * FROM document_blocks WHERE type = 'heading' ORDER BY block_order """) rows = cursor.fetchall() return [self._row_to_dict(row) for row in rows] def get_stats(self) -> dict: """获取统计信息""" cursor = self.conn.execute(""" SELECT type, COUNT(*) as count FROM document_blocks GROUP BY type """) stats = {row['type']: row['count'] for row in cursor.fetchall()} cursor = self.conn.execute("SELECT COUNT(*) as total FROM document_blocks") total = cursor.fetchone()['total'] return { 'total': total, 'by_type': stats } def _row_to_dict(self, row) -> dict: """将 sqlite3.Row 转换为字典""" d = dict(row) # 解析 JSON 字段 if d.get('style'): try: d['style'] = json.loads(d['style']) except (json.JSONDecodeError, TypeError): d['style'] = {} else: d['style'] = {} if d.get('metadata'): try: d['metadata'] = json.loads(d['metadata']) except (json.JSONDecodeError, TypeError): d['metadata'] = {} else: d['metadata'] = {} # content 可能是 JSON(富文本或表格) if d.get('content'): try: d['content'] = json.loads(d['content']) except (json.JSONDecodeError, TypeError): pass # 保持原字符串 return d