Files
ssw-ontology/scripts/fix_clearance_and_classification.py
adminandClaude Code aef8b83e24 docs(release): v1.8.0 发布说明 + 本体导入设计约束落地记录 + 对标分析
- 新增 docs/ONTOLOGY_IMPORT.md:本体 schema 导入的硬约束(fail-fast 闸门、整事务、
  幂等键、IRI 双空间映射、不做范围对账删除),并记录实现落点与逐条偏差
- 新增 docs/auto-ontology-benchmark-20261106.md:NVIDIA auto-ontology 对标分析与
  Top-8 路线图,本版已落地项逐条标注
- docs/README.md 同步 API 清单与 MCP 工具表(54 个:只读 27 / 写入 27)
- 项目 CLAUDE.md 补「本体 schema 导入」小节:闸门顺序、事务与串行理由、幂等表、
  RDF Schema 命名空间年份坑
- README.md 镜像标签更新到 1.8.0,本地构建镜像补上 Linux 静态二进制交叉编译步骤,
  健康检查端点改回 /api/v1/version
- scripts/ 收拢本体 v2 迁移与分级校正脚本(一次性运维工具,不参与构建)

Co-Authored-By: Claude Code <[email protected]>
2026-11-06 13:53:25 +08:00

148 lines
5.9 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
按规划修正本体分级与主体 clearance(2026-10-13 续):
规划:L0 公众 / L1 客户 / L2 员工 / L3 经理以上 / L4 公司管理层
问题:上一轮误把 75 名员工 clearance 提到 L0-L3,且把本该员工可见的数据定在 L3,
导致"员工该看到的很多数据反而靠提权才看得到"。正确做法:
① 员工 clearance 回落 L0-L2;
② 把本该员工可见的数据类型默认级从 L3 降到 L2;
③ UsageControlPolicy 等经理级数据保持 L3(员工不可见,正确);
④ 仅「国家秘密」等真正涉密实例保留 L4 覆盖层。
幂等、可复跑;默认 dry-run 仅预览,--apply 才落库。
"""
import sys, json, argparse
sys.path.insert(0, 'scripts')
import raise_employee_clearance as r
BUILTIN = {'public', 'analyst', 'connector', 'TradingAssistant', 'manager', 'ingest'}
EMP_LEVELS = ['L0', 'L1', 'L2']
LEVELS = ['L0', 'L1', 'L2', 'L3', 'L4']
# 类型默认级重分类:
# L3 -> L2(员工可见的内部/运营/政策类,有实例的才是"员工访问不了"的根因)
TO_L2 = ['PolicyDocument', 'SecurityRequirement', 'SpaceManager',
'TrustedCryptographyCard', 'TrustCredential', 'TDSInterconnection',
'UsageAction', 'UsageConstraint']
# 保持 L3(经理以上,含 UsageControlPolicy 36 实例 + 两个 0 实例契约/计费类)
KEEP_L3 = ['DigitalContract', 'MeteringBill', 'UsageControlPolicy']
# 实例覆盖层修正:UsageControlPolicy 是否从 L4 回落 L3 交由用户确认(默认保留 L4,
# 它属「公司管理层」治理范畴,且不影响员工可见性;员工本就不该看访问控制策略)。
# PolicyDocument 的 cs_20260918_124(国家秘密)保持 L4 不动。
REVERT_UCP_L4 = False
def ensure_classification_property(name):
# 补建 classification 属性(幂等 ON CONFLICT),否则 setObjectTypeClassification 校验会拒
body = {
'api_name': 'classification',
'display_name': '数据分级',
'data_type': 'string',
'enum_values': LEVELS,
'description': '数据分级 L0公众/L1客户/L2员工/L3经理以上/L4公司管理层',
'indexable': False,
}
return r.api('POST', f'/ontology/object-types/{name}/properties', body)
def set_type_default(name, default):
# 先确保 classification 属性存在(校验依赖它)
ensure_classification_property(name)
body = {'classification_property': 'classification',
'classification_levels': LEVELS,
'classification_default': default}
return r.api('PUT', f'/ontology/object-types/{name}/classification', body)
def get_clearance(name):
code, data = r.api('GET', f'/access/principals/{name}/clearances')
if isinstance(data, dict):
return data.get('markings') or data.get('clearances')
return data
def set_clearance(name, levels):
# 关键:端点只认 markings 字段(handlers_access.go:318),传 clearances 会被清空
return r.api('PUT', f'/access/principals/{name}/clearances', {'markings': levels})
def get_markings(typ, pk):
code, data = r.api('GET', f'/access/objects/{typ}/{pk}/markings')
if isinstance(data, dict):
return data.get('markings') or []
return []
def set_markings(typ, pk, levels):
return r.api('PUT', f'/access/objects/{typ}/{pk}/markings', {'markings': levels})
def list_instances(typ, limit=2000):
code, data = r.api('GET', f'/objects?type={typ}&limit={limit}')
if isinstance(data, dict):
return data.get('items', data.get('objects', []))
return data or []
def main():
ap = argparse.ArgumentParser()
ap.add_argument('--apply', action='store_true', help='落库(默认 dry-run)')
args = ap.parse_args()
A = args.apply
print(f'== 模式: {"APPLY" if A else "DRY-RUN"} ==')
# ---- Phase 1: 员工 clearance 回落 L0-L2 ----
print('\n[Phase 1] 员工 clearance -> L0-L2')
code, pdata = r.api('GET', '/access/principals')
ps = pdata if isinstance(pdata, list) else pdata.get('principals', pdata)
emp = [p['name'] for p in ps if p['name'] not in BUILTIN]
print(f' 员工主体数: {len(emp)}')
for n in emp:
cur = get_clearance(n)
if cur == EMP_LEVELS:
continue
print(f' {n}: {cur} -> {EMP_LEVELS}')
if A:
set_clearance(n, EMP_LEVELS)
# ---- Phase 2: 类型默认级重分类 ----
print('\n[Phase 2] 类型默认级 L3->L2(员工可见类)')
for n in TO_L2:
print(f' {n}: default L3 -> L2')
if A:
set_type_default(n, 'L2')
for n in KEEP_L3:
print(f' {n}: 保持 L3(经理以上)')
# ---- Phase 3: recompute(仅写 classification 列,不碰 object_markings)----
if A:
print('\n[Phase 3] recompute classification(apply=true)')
code, data = r.api('POST', '/ontology/classification/recompute',
{'apply': True, 'dry_run': False, 'limit': 20000})
print(' ', json.dumps(data, ensure_ascii=False)[:400])
else:
print('\n[Phase 3] recompute 将在 --apply 时执行(materialize 类型默认级)')
# ---- Phase 4: 实例覆盖层修正 ----
print('\n[Phase 4] 实例覆盖层')
if REVERT_UCP_L4:
ucp = list_instances('UsageControlPolicy', 200)
print(f' UsageControlPolicy 实例 {len(ucp)} 条:回落被误标的 L4 -> (移除覆盖, 落类型默认 L3)')
for o in ucp:
pk = o.get('pk')
cur = get_markings('UsageControlPolicy', pk)
if cur == ['L4']:
print(f' {pk}: L4 -> [] (回落 L3)')
if A:
set_markings('UsageControlPolicy', pk, [])
print(' PolicyDocument.cs_20260918_124(国家秘密)保持 L4 覆盖层(不动)')
print('\n== 完成 ==')
if __name__ == '__main__':
main()