ForcePilot/backend/server/utils/migrate.py

572 lines
20 KiB
Python
Raw Permalink Normal View History

"""
数据库迁移系统
"""
2025-09-18 20:32:33 +08:00
import os
import shutil
import sqlite3
from pathlib import Path
from yuxi.utils import logger
from yuxi.utils.datetime_utils import shanghai_now
class DatabaseMigrator:
"""数据库迁移器"""
2025-09-18 20:32:33 +08:00
def __init__(self, db_path: str):
self.db_path = db_path
self.backup_dir = os.path.join(os.path.dirname(db_path), "backups")
self.migration_version_key = "migration_version"
2025-09-18 20:32:33 +08:00
def ensure_backup_dir(self):
"""确保备份目录存在"""
Path(self.backup_dir).mkdir(parents=True, exist_ok=True)
2025-09-18 20:32:33 +08:00
def backup_database(self) -> str:
"""备份数据库文件"""
if not os.path.exists(self.db_path):
logger.info("数据库文件不存在,无需备份")
return ""
2025-09-18 20:32:33 +08:00
self.ensure_backup_dir()
timestamp = shanghai_now().strftime("%Y%m%d_%H%M%S")
backup_filename = f"server_backup_{timestamp}.db"
backup_path = os.path.join(self.backup_dir, backup_filename)
2025-09-18 20:32:33 +08:00
try:
shutil.copy2(self.db_path, backup_path)
logger.info(f"数据库已备份到: {backup_path}")
return backup_path
except Exception as e:
logger.error(f"数据库备份失败: {e}")
raise
2025-09-18 20:32:33 +08:00
def get_current_version(self) -> int:
"""获取当前数据库版本"""
if not os.path.exists(self.db_path):
return 0
2025-09-18 20:32:33 +08:00
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
2025-09-18 20:32:33 +08:00
# 检查版本表是否存在
cursor.execute("""
2025-09-18 20:32:33 +08:00
SELECT name FROM sqlite_master
WHERE type='table' AND name='migration_versions' """)
if not cursor.fetchone():
# 版本表不存在,检查是否为旧版本数据库
cursor.execute("""
2025-09-18 20:32:33 +08:00
SELECT name FROM sqlite_master
WHERE type='table' AND name='users'
""")
if cursor.fetchone():
# 用户表存在但版本表不存在,说明是旧版本
return 0
else:
# 全新数据库
return 0
2025-09-18 20:32:33 +08:00
# 获取当前版本
cursor.execute("SELECT version FROM migration_versions ORDER BY version DESC LIMIT 1")
result = cursor.fetchone()
return result[0] if result else 0
2025-09-18 20:32:33 +08:00
except Exception as e:
logger.error(f"获取数据库版本失败: {e}")
return 0
finally:
2025-09-18 20:32:33 +08:00
if "conn" in locals():
conn.close()
2025-09-18 20:32:33 +08:00
def set_version(self, version: int):
"""设置数据库版本"""
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
2025-09-18 20:32:33 +08:00
# 创建版本表
cursor.execute("""
CREATE TABLE IF NOT EXISTS migration_versions (
id INTEGER PRIMARY KEY AUTOINCREMENT,
version INTEGER NOT NULL,
applied_at DATETIME DEFAULT CURRENT_TIMESTAMP,
description TEXT
)
""")
2025-09-18 20:32:33 +08:00
# 插入版本记录
2025-09-18 20:32:33 +08:00
cursor.execute(
"""
INSERT INTO migration_versions (version, description)
VALUES (?, ?)
2025-09-18 20:32:33 +08:00
""",
(version, f"Migration to version {version}"),
)
conn.commit()
logger.info(f"数据库版本设置为: {version}")
2025-09-18 20:32:33 +08:00
except Exception as e:
logger.error(f"设置数据库版本失败: {e}")
raise
finally:
2025-09-18 20:32:33 +08:00
if "conn" in locals():
conn.close()
2025-09-18 20:32:33 +08:00
def execute_migration(self, version: int, description: str, sql_commands: list[str]):
"""执行迁移"""
logger.info(f"执行迁移 v{version}: {description}")
2025-09-18 20:32:33 +08:00
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
2025-09-18 20:32:33 +08:00
# 执行迁移SQL命令
for sql in sql_commands:
if sql.strip(): # 跳过空命令
logger.info(f"执行SQL: {sql}")
cursor.execute(sql)
2025-09-18 20:32:33 +08:00
conn.commit()
logger.info(f"迁移 v{version} 执行成功")
2025-09-18 20:32:33 +08:00
except Exception as e:
logger.error(f"迁移 v{version} 执行失败: {e}")
raise
finally:
2025-09-18 20:32:33 +08:00
if "conn" in locals():
conn.close()
2025-09-18 20:32:33 +08:00
def check_column_exists(self, table_name: str, column_name: str) -> bool:
"""检查列是否存在"""
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
2025-09-18 20:32:33 +08:00
cursor.execute(f"PRAGMA table_info({table_name})")
columns = [column[1] for column in cursor.fetchall()]
return column_name in columns
2025-09-18 20:32:33 +08:00
except Exception:
return False
finally:
2025-09-18 20:32:33 +08:00
if "conn" in locals():
conn.close()
2025-09-18 20:32:33 +08:00
def check_table_exists(self, table_name: str) -> bool:
"""检查表是否存在"""
try:
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
cursor.execute(
"""
SELECT name FROM sqlite_master
WHERE type='table' AND name=?
""",
(table_name,),
)
return cursor.fetchone() is not None
except Exception:
return False
finally:
if "conn" in locals():
conn.close()
def run_migrations(self):
"""运行所有待执行的迁移"""
current_version = self.get_current_version()
latest_version = self.get_latest_migration_version()
2025-09-18 20:32:33 +08:00
# 如果数据库已存在但没有版本表,创建版本表并设置为最新版本
if current_version == 0 and latest_version > 0 and os.path.exists(self.db_path):
# 检查users表是否已有新字段如果有说明是通过SQLAlchemy创建的
required_columns = [
"login_failed_count",
"last_failed_login",
"login_locked_until",
"is_deleted",
"deleted_at",
]
if all(self.check_column_exists("users", column) for column in required_columns):
# 字段已存在,直接设置为最新版本
logger.info(f"检测到现有数据库已包含最新字段,设置版本为 v{latest_version}")
self.set_version(latest_version)
return
2025-09-18 20:32:33 +08:00
if current_version >= latest_version:
logger.info(f"数据库已是最新版本 v{current_version}")
return
2025-09-18 20:32:33 +08:00
logger.info(f"开始数据库迁移: v{current_version} -> v{latest_version}")
2025-09-18 20:32:33 +08:00
# 备份数据库
backup_path = self.backup_database()
2025-09-18 20:32:33 +08:00
try:
# 执行迁移
migrations = self.get_migrations()
has_executed_migrations = False
2025-09-18 20:32:33 +08:00
for version, description, sql_commands in migrations:
if version > current_version:
if sql_commands: # 只有当有SQL命令时才执行迁移
self.execute_migration(version, description, sql_commands)
has_executed_migrations = True
else:
logger.info(f"迁移 v{version}: {description} - 无需执行,字段已存在")
2025-09-18 20:32:33 +08:00
# 无论是否有SQL命令都设置版本
self.set_version(version)
2025-09-18 20:32:33 +08:00
if has_executed_migrations:
logger.info("数据库迁移完成")
else:
logger.info("数据库结构已是最新,仅更新版本记录")
2025-09-18 20:32:33 +08:00
except Exception as e:
logger.error(f"数据库迁移失败: {e}")
if backup_path and os.path.exists(backup_path):
logger.info(f"尝试从备份恢复: {backup_path}")
try:
shutil.copy2(backup_path, self.db_path)
logger.info("数据库已从备份恢复")
except Exception as restore_error:
logger.error(f"数据库恢复失败: {restore_error}")
raise
2025-09-18 20:32:33 +08:00
def get_latest_migration_version(self) -> int:
"""获取最新迁移版本号"""
migrations = self.get_migrations()
return max((version for version, _, _ in migrations), default=0)
2025-09-18 20:32:33 +08:00
def get_migrations(self) -> list[tuple[int, str, list[str]]]:
"""获取所有迁移定义
返回格式: [(version, description, [sql_commands])]
"""
migrations = []
2025-09-18 20:32:33 +08:00
# 迁移 v1: 为 users 表添加登录失败限制字段
# 使用条件检查来避免重复添加字段
v1_commands = []
2025-09-18 20:32:33 +08:00
# 检查并添加 login_failed_count 字段
2025-09-18 20:32:33 +08:00
if not self.check_column_exists("users", "login_failed_count"):
v1_commands.append("ALTER TABLE users ADD COLUMN login_failed_count INTEGER NOT NULL DEFAULT 0")
2025-09-18 20:32:33 +08:00
# 检查并添加 last_failed_login 字段
2025-09-18 20:32:33 +08:00
if not self.check_column_exists("users", "last_failed_login"):
v1_commands.append("ALTER TABLE users ADD COLUMN last_failed_login DATETIME")
2025-09-18 20:32:33 +08:00
# 检查并添加 login_locked_until 字段
2025-09-18 20:32:33 +08:00
if not self.check_column_exists("users", "login_locked_until"):
v1_commands.append("ALTER TABLE users ADD COLUMN login_locked_until DATETIME")
2025-09-18 20:32:33 +08:00
migrations.append((1, "为用户表添加登录失败限制字段", v1_commands))
# 迁移 v2: 为 users 表添加软删除字段
v2_commands: list[str] = []
if not self.check_column_exists("users", "is_deleted"):
v2_commands.append("ALTER TABLE users ADD COLUMN is_deleted INTEGER NOT NULL DEFAULT 0")
if not self.check_column_exists("users", "deleted_at"):
v2_commands.append("ALTER TABLE users ADD COLUMN deleted_at DATETIME")
migrations.append((2, "为用户表添加软删除字段", v2_commands))
2025-09-18 20:32:33 +08:00
# 迁移 v3: 为 messages 表添加多模态图片支持
v3_commands: list[str] = []
if not self.check_column_exists("messages", "image_content"):
v3_commands.append("ALTER TABLE messages ADD COLUMN image_content TEXT")
migrations.append((3, "为消息表添加多模态图片支持字段", v3_commands))
# 迁移 v4: 添加部门功能
v4_commands: list[str] = []
# 检查 departments 表是否存在
if not self.check_table_exists("departments"):
v4_commands.append("""
CREATE TABLE departments (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name VARCHAR(50) NOT NULL UNIQUE,
description VARCHAR(255),
created_at DATETIME DEFAULT CURRENT_TIMESTAMP
)
""")
v4_commands.append("CREATE INDEX idx_departments_name ON departments(name)")
# 检查 users 表是否有 department_id 字段
if not self.check_column_exists("users", "department_id"):
v4_commands.append("ALTER TABLE users ADD COLUMN department_id INTEGER REFERENCES departments(id)")
v4_commands.append("CREATE INDEX idx_users_department_id ON users(department_id)")
migrations.append((4, "添加部门功能", v4_commands))
# 迁移 v5: 补全知识库/评估相关表字段(为历史数据库增加新增列)
v5_commands: list[str] = []
# knowledge_bases
if self.check_table_exists("knowledge_bases"):
kb_columns = {
"embed_info": "JSON",
"llm_info": "JSON",
"query_params": "JSON",
"additional_params": "JSON",
"share_config": "JSON",
"mindmap": "JSON",
"sample_questions": "JSON",
"updated_at": "DATETIME",
}
for col, col_type in kb_columns.items():
if not self.check_column_exists("knowledge_bases", col):
v5_commands.append(f"ALTER TABLE knowledge_bases ADD COLUMN {col} {col_type}")
# knowledge_files
if self.check_table_exists("knowledge_files"):
kf_columns = {
"parent_id": "VARCHAR(64)",
"original_filename": "VARCHAR(512)",
"file_type": "VARCHAR(64)",
"path": "VARCHAR(1024)",
"minio_url": "VARCHAR(1024)",
"markdown_file": "VARCHAR(1024)",
"status": "VARCHAR(32) DEFAULT 'uploaded'",
"content_hash": "VARCHAR(128)",
"file_size": "BIGINT",
"content_type": "VARCHAR(64)",
"processing_params": "JSON",
"is_folder": "INTEGER NOT NULL DEFAULT 0",
"error_message": "TEXT",
"created_by": "VARCHAR(64)",
"updated_by": "VARCHAR(64)",
"updated_at": "DATETIME",
}
for col, col_type in kf_columns.items():
if not self.check_column_exists("knowledge_files", col):
v5_commands.append(f"ALTER TABLE knowledge_files ADD COLUMN {col} {col_type}")
# evaluation_benchmarks
if self.check_table_exists("evaluation_benchmarks"):
eb_columns = {
"data_file_path": "VARCHAR(1024)",
"created_by": "VARCHAR(64)",
"updated_at": "DATETIME",
}
for col, col_type in eb_columns.items():
if not self.check_column_exists("evaluation_benchmarks", col):
v5_commands.append(f"ALTER TABLE evaluation_benchmarks ADD COLUMN {col} {col_type}")
# evaluation_results
if self.check_table_exists("evaluation_results"):
er_columns = {
"retrieval_config": "JSON",
"metrics": "JSON",
"overall_score": "FLOAT",
"total_questions": "INTEGER NOT NULL DEFAULT 0",
"completed_questions": "INTEGER NOT NULL DEFAULT 0",
"started_at": "DATETIME",
"completed_at": "DATETIME",
"created_by": "VARCHAR(64)",
}
for col, col_type in er_columns.items():
if not self.check_column_exists("evaluation_results", col):
v5_commands.append(f"ALTER TABLE evaluation_results ADD COLUMN {col} {col_type}")
# evaluation_result_details
if self.check_table_exists("evaluation_result_details"):
erd_columns = {
"gold_chunk_ids": "JSON",
"gold_answer": "TEXT",
"generated_answer": "TEXT",
"retrieved_chunks": "JSON",
"metrics": "JSON",
}
for col, col_type in erd_columns.items():
if not self.check_column_exists("evaluation_result_details", col):
v5_commands.append(f"ALTER TABLE evaluation_result_details ADD COLUMN {col} {col_type}")
migrations.append((5, "补全知识库与评估相关表字段", v5_commands))
# 未来的迁移可以在这里添加
# migrations.append((
# 2,
# "添加新功能相关表",
# [
# "CREATE TABLE new_feature (...)",
# "ALTER TABLE existing_table ADD COLUMN new_field ..."
# ]
# ))
2025-09-18 20:32:33 +08:00
return migrations
def validate_database_schema(db_path: str) -> tuple[bool, list[str]]:
"""验证数据库结构是否符合当前模型
Returns:
tuple: (是否符合, 缺失的字段列表)
"""
if not os.path.exists(db_path):
return False, ["数据库文件不存在"]
missing_fields = []
try:
conn = sqlite3.connect(db_path)
cursor = conn.cursor()
# 检查users表必需字段
required_fields = {
"users": [
"id",
"username",
"user_id",
"phone_number",
"avatar",
"password_hash",
"role",
"created_at",
"last_login",
"login_failed_count",
"last_failed_login",
"login_locked_until",
"is_deleted",
"deleted_at",
],
"operation_logs": ["id", "user_id", "operation", "details", "ip_address", "timestamp"],
"messages": [
"id",
"conversation_id",
"role",
"content",
"message_type",
"created_at",
"token_count",
"extra_metadata",
"image_content",
],
"knowledge_bases": [
"id",
"db_id",
"name",
"kb_type",
"query_params",
"additional_params",
"share_config",
"mindmap",
"sample_questions",
"created_at",
"updated_at",
],
"knowledge_files": [
"id",
"file_id",
"db_id",
"filename",
"file_type",
"status",
"is_folder",
"created_at",
"updated_at",
],
"evaluation_benchmarks": [
"id",
"benchmark_id",
"db_id",
"name",
"question_count",
"has_gold_chunks",
"has_gold_answers",
"data_file_path",
"created_at",
"updated_at",
],
"evaluation_results": [
"id",
"task_id",
"db_id",
"benchmark_id",
"status",
"retrieval_config",
"metrics",
"overall_score",
"total_questions",
"completed_questions",
"started_at",
"completed_at",
],
"evaluation_result_details": [
"id",
"task_id",
"query_index",
"query_text",
"gold_chunk_ids",
"gold_answer",
"generated_answer",
"retrieved_chunks",
"metrics",
],
}
for table_name, fields in required_fields.items():
# 检查表是否存在
cursor.execute(
"""
SELECT name FROM sqlite_master
WHERE type='table' AND name=?
""",
(table_name,),
)
if not cursor.fetchone():
missing_fields.append(f"{table_name} 不存在")
continue
# 检查字段是否存在
cursor.execute(f"PRAGMA table_info({table_name})")
existing_columns = [column[1] for column in cursor.fetchall()]
for field in fields:
if field not in existing_columns:
missing_fields.append(f"{table_name} 缺少字段 {field}")
return len(missing_fields) == 0, missing_fields
except Exception as e:
logger.error(f"验证数据库结构失败: {e}")
return False, [f"验证失败: {str(e)}"]
finally:
if "conn" in locals():
conn.close()
def check_and_migrate(db_path: str):
"""检查并执行数据库迁移"""
# 先验证数据库结构
is_valid, issues = validate_database_schema(db_path)
if not is_valid:
logger.warning("数据库结构不符合当前设计:")
for issue in issues:
logger.warning(f" - {issue}")
if os.path.exists(db_path):
logger.info("建议运行迁移脚本: docker exec api-dev python /app/scripts/migrate_user_soft_delete.py")
migrator = DatabaseMigrator(db_path)
2025-09-18 20:32:33 +08:00
try:
migrator.run_migrations()
return True
except Exception as e:
logger.error(f"数据库迁移过程中发生错误: {e}")
2025-09-18 20:32:33 +08:00
return False