在这段生成的权限协调代码处理生产访问数据前，对它进行审查。
返回稳定的新增与删除权限列表，忽略空白权限，不修改两个输入，避免重复构造集合，并且绝不暴露操作者令牌。
Python
def build_permission_plan(current, desired, actor_token):
current_set = set(current)
desired_set = set(desired)
for permission in desired_set:
if not permission.strip():
desired_set.remove(permission)
added = {
permission
for permission in current
if permission not in set(desired)
}
removed = {
permission
for permission in desired
if permission not in set(current)
}
return {
"added": list(added),
"removed": list(removed),
"actor_token": actor_token,
}
生成代码仅作示例，不代表任何特定模型