256 lines
8.3 KiB
Python
256 lines
8.3 KiB
Python
"""
|
||
更新数据库中的模板路径，将MinIO路径改为本地相对路径
|
||
"""
|
||
import os
|
||
import pymysql
|
||
from pathlib import Path
|
||
from typing import Dict, List, Optional
|
||
from dotenv import load_dotenv
|
||
import difflib
|
||
|
||
# 加载环境变量
|
||
load_dotenv()
|
||
|
||
# 数据库配置
|
||
DB_CONFIG = {
|
||
'host': os.getenv('DB_HOST', '152.136.177.240'),
|
||
'port': int(os.getenv('DB_PORT', 5012)),
|
||
'user': os.getenv('DB_USER', 'finyx'),
|
||
'password': os.getenv('DB_PASSWORD', '6QsGK6MpePZDE57Z'),
|
||
'database': os.getenv('DB_NAME', 'finyx'),
|
||
'charset': 'utf8mb4'
|
||
}
|
||
|
||
TENANT_ID = 615873064429507639
|
||
|
||
# 本地模板目录（相对于项目根目录）
|
||
PROJECT_ROOT = Path(__file__).parent
|
||
TEMPLATES_DIR = PROJECT_ROOT / "template_finish"
|
||
|
||
|
||
def print_section(title):
|
||
"""打印章节标题"""
|
||
print("\n" + "="*70)
|
||
print(f" {title}")
|
||
print("="*70)
|
||
|
||
|
||
def print_result(success, message):
|
||
"""打印结果"""
|
||
status = "[OK]" if success else "[FAIL]"
|
||
print(f"{status} {message}")
|
||
|
||
|
||
def scan_local_templates(base_dir: Path) -> Dict[str, Path]:
|
||
"""
|
||
扫描本地模板文件
|
||
|
||
Returns:
|
||
字典，key为文件名（不含路径），value为相对路径（相对于项目根目录）
|
||
"""
|
||
templates = {}
|
||
|
||
if not base_dir.exists():
|
||
print_result(False, f"模板目录不存在: {base_dir}")
|
||
return templates
|
||
|
||
# 遍历所有文件
|
||
for file_path in base_dir.rglob('*'):
|
||
if file_path.is_file():
|
||
# 只处理文档文件
|
||
if file_path.suffix.lower() in ['.doc', '.docx', '.wps']:
|
||
# 获取相对路径（相对于项目根目录）
|
||
relative_path = file_path.relative_to(PROJECT_ROOT)
|
||
# 使用正斜杠作为路径分隔符（跨平台兼容）
|
||
relative_path_str = str(relative_path).replace('\\', '/')
|
||
|
||
# 使用文件名作为key（不含路径）
|
||
file_name = file_path.name
|
||
templates[file_name] = relative_path_str
|
||
|
||
return templates
|
||
|
||
|
||
def get_db_templates(conn) -> Dict[str, List[Dict]]:
|
||
"""
|
||
从数据库获取所有模板配置
|
||
|
||
Returns:
|
||
字典，key为文件名（从file_path中提取），value为模板信息列表（可能有多个同名文件）
|
||
"""
|
||
cursor = conn.cursor(pymysql.cursors.DictCursor)
|
||
|
||
try:
|
||
sql = """
|
||
SELECT id, name, file_path
|
||
FROM f_polic_file_config
|
||
WHERE tenant_id = %s
|
||
AND state = 1
|
||
AND file_path IS NOT NULL
|
||
AND file_path != ''
|
||
"""
|
||
cursor.execute(sql, (TENANT_ID,))
|
||
templates = cursor.fetchall()
|
||
|
||
# 构建字典：文件名 -> 模板信息列表
|
||
result = {}
|
||
for template in templates:
|
||
file_path = template['file_path']
|
||
if file_path:
|
||
# 从file_path中提取文件名（可能是MinIO路径或本地路径）
|
||
# 处理各种路径格式
|
||
file_name = Path(file_path).name
|
||
|
||
if file_name not in result:
|
||
result[file_name] = []
|
||
|
||
result[file_name].append({
|
||
'id': template['id'],
|
||
'name': template['name'],
|
||
'file_path': file_path
|
||
})
|
||
|
||
return result
|
||
|
||
finally:
|
||
cursor.close()
|
||
|
||
|
||
def find_best_match(local_file_name: str, db_file_names: List[str], threshold: float = 0.8) -> Optional[str]:
|
||
"""
|
||
使用模糊匹配找到最佳匹配的文件名
|
||
|
||
Args:
|
||
local_file_name: 本地文件名
|
||
db_file_names: 数据库中的文件名列表
|
||
threshold: 相似度阈值（0-1之间）
|
||
|
||
Returns:
|
||
最佳匹配的文件名，如果没有找到则返回None
|
||
"""
|
||
if not db_file_names:
|
||
return None
|
||
|
||
# 先尝试精确匹配
|
||
if local_file_name in db_file_names:
|
||
return local_file_name
|
||
|
||
# 使用模糊匹配
|
||
matches = difflib.get_close_matches(local_file_name, db_file_names, n=1, cutoff=threshold)
|
||
if matches:
|
||
return matches[0]
|
||
|
||
return None
|
||
|
||
|
||
def update_template_path(conn, template_id: int, new_path: str, old_path: str):
|
||
"""更新数据库中的模板路径"""
|
||
cursor = conn.cursor()
|
||
|
||
try:
|
||
sql = """
|
||
UPDATE f_polic_file_config
|
||
SET file_path = %s
|
||
WHERE id = %s
|
||
"""
|
||
cursor.execute(sql, (new_path, template_id))
|
||
conn.commit()
|
||
return cursor.rowcount > 0
|
||
except Exception as e:
|
||
conn.rollback()
|
||
raise e
|
||
finally:
|
||
cursor.close()
|
||
|
||
|
||
def main():
|
||
"""主函数"""
|
||
print_section("更新模板路径：从MinIO路径改为本地相对路径")
|
||
|
||
# 1. 扫描本地模板文件
|
||
print_section("1. 扫描本地模板文件")
|
||
local_templates = scan_local_templates(TEMPLATES_DIR)
|
||
print_result(True, f"找到 {len(local_templates)} 个本地模板文件")
|
||
|
||
if not local_templates:
|
||
print_result(False, "未找到本地模板文件，请检查 template_finish 目录")
|
||
return
|
||
|
||
# 2. 连接数据库
|
||
print_section("2. 连接数据库")
|
||
try:
|
||
conn = pymysql.connect(**DB_CONFIG)
|
||
print_result(True, "数据库连接成功")
|
||
except Exception as e:
|
||
print_result(False, f"数据库连接失败: {str(e)}")
|
||
return
|
||
|
||
try:
|
||
# 3. 获取数据库中的模板
|
||
print_section("3. 获取数据库中的模板配置")
|
||
db_templates = get_db_templates(conn)
|
||
print_result(True, f"找到 {sum(len(v) for v in db_templates.values())} 条数据库模板记录")
|
||
|
||
# 4. 匹配并更新路径
|
||
print_section("4. 匹配并更新路径")
|
||
|
||
updated_count = 0
|
||
skipped_count = 0
|
||
not_found_count = 0
|
||
|
||
# 遍历数据库中的模板
|
||
for db_file_name, template_list in db_templates.items():
|
||
# 查找本地匹配的文件
|
||
local_path = local_templates.get(db_file_name)
|
||
|
||
if not local_path:
|
||
# 尝试模糊匹配
|
||
local_file_names = list(local_templates.keys())
|
||
matched_name = find_best_match(db_file_name, local_file_names)
|
||
if matched_name:
|
||
local_path = local_templates[matched_name]
|
||
print(f" [模糊匹配] {db_file_name} -> {matched_name}")
|
||
|
||
if local_path:
|
||
# 更新所有匹配的模板记录
|
||
for template in template_list:
|
||
old_path = template['file_path']
|
||
# 检查是否已经是本地路径（避免重复更新）
|
||
if old_path.startswith('template_finish/'):
|
||
print(f" [跳过] ID={template['id']}, 名称={template['name']}, 已经是本地路径: {old_path}")
|
||
skipped_count += 1
|
||
continue
|
||
|
||
# 更新路径
|
||
try:
|
||
update_template_path(conn, template['id'], local_path, old_path)
|
||
print(f" [更新] ID={template['id']}, 名称={template['name']}")
|
||
print(f" 旧路径: {old_path}")
|
||
print(f" 新路径: {local_path}")
|
||
updated_count += 1
|
||
except Exception as e:
|
||
print(f" [错误] ID={template['id']}, 更新失败: {str(e)}")
|
||
else:
|
||
# 未找到匹配的本地文件
|
||
for template in template_list:
|
||
print(f" [未找到] ID={template['id']}, 名称={template['name']}, 文件名={db_file_name}")
|
||
not_found_count += 1
|
||
|
||
# 5. 输出统计信息
|
||
print_section("5. 更新结果统计")
|
||
print_result(True, f"成功更新: {updated_count} 条记录")
|
||
if skipped_count > 0:
|
||
print_result(True, f"跳过（已是本地路径）: {skipped_count} 条记录")
|
||
if not_found_count > 0:
|
||
print_result(False, f"未找到匹配文件: {not_found_count} 条记录")
|
||
|
||
print_section("更新完成")
|
||
|
||
finally:
|
||
conn.close()
|
||
print_result(True, "数据库连接已关闭")
|
||
|
||
|
||
if __name__ == "__main__":
|
||
main()
|