Major changes: - New frontend (platform/web/): Vite + React 18 + TypeScript + Tailwind - 4-module navigation: 数据送标 / 模型管理 / 车队管理 / 系统管理 - Data catalog with charts (DMS/ADAS/Lane 3-tab view) - Quality review workflow (标注质检): Good/Fine/Bad scoring with auto-advance - Audit enhancements: batch operations, rejection categories, Feishu notifications - Operation audit log (操作日志) - World model simulation studio (仿真工坊) - Dataset version management with snapshots and diff - ADAS 7-class dataset integration (138K images organized + compressed) - User management with Feishu integration and pagination - CRUD/search/filter on all pages, card layout redesign - PIL-optimized image overlay rendering - Auto-snapshot on build, in_review workflow stage - Removed embedded algorithm code (now in workspace)
1.7 KiB
1.7 KiB
HSAP 验收与缺口报告
生成方式:bash HSAP/scripts/smoke_platform_api.sh + 对照 BROWSER_QA_CHECKLIST.md
最近 smoke: 通过(ALL_OK),health / labeling batches(7) / fleet live(3) / export-jobs / bootstrap / export / ml-predict 均正常。
API 自动化
| 项 | 状态 | 备注 |
|---|---|---|
| Health + DB/Redis | 通过 | |
| labeling/batches | 通过 | 7 条(含 registry_only 占位批次) |
| dam bootstrap | 通过 | dam_15cls.xml 已写入 campaign |
| export / ml-predict | 通过 | ExportJob 与 jobs 队列联动 |
| fleet map-config / live | 通过 | 需 AS_FLEET_MAP_ENABLED=1 |
浏览器(需人工勾选)
见 BROWSER_QA_CHECKLIST.md。自动化无法覆盖 UI 布局与瓦片加载。
已确认缺口
| 缺口 | 影响 | 缓解 |
|---|---|---|
| inbox 无样例图片 | tasks 为空,画布无法 E2E 框选保存 |
往 datasets/dms/inbox/dam/batch_0516/ 放 jpg;或 reseed 演示批次 |
| export worker 占位 | job message 为记录型,未真跑 YOLO 转换 |
见 runner labeling_export;后续对接 as.py build |
| ML 预标 | 已停用 | 无 UI / 无 POST .../ml/predict |
| 无 vendor 回传 | 第三方离线标无法导入 | 本里程碑新增 import-vendor API |
| 无独立 Export/ML 页 | 仅在 Campaign 行操作 | 本里程碑新增 /labeling/export、/labeling/ml |
| 车队无 SSE | 15s 轮询 | 本里程碑新增 /fleet/stream |
| 数据入湖 staging | checklist A~E 未全实现 | 见 DATA_LAKE_GAP.md |
建议复测命令
bash HSAP/scripts/smoke_platform_api.sh
bash HSAP/scripts/smoke_labeling_api.sh
cd HSAP && docker compose up -d --build platform