feat(rich): 备注/工单说明/知识正文支持 Markdown 与图片附件

写一件事的时候,光有纯文本不够:要贴截图、要分段、要标重点。
这次把「正文」这条路补全,思路是**存 Markdown 而不是 HTML**——
能导出、能全文检索、能直接喂给 LLM,也不必每次渲染都重新清洗一遍 HTML。

后端
- attachment 表 + RLS + 迁移 0005;本地磁盘存储,key 由内容 sha256 决定,
  换对象存储只改 services/storage.py 一个文件
- 上传在**读流时**卡 20MB,不先落盘再判断;磁盘名由服务端生成,用户文件名只进库和响应头
- 列表带 attachment_count、详情带 attachments,避免一屏 N 次请求
- raw 直链给 <img src> 用(图片带不了 Authorization 头),只允许本工作区成员,
  非成员一律 8001 不区分「不存在」与「没权限」
- 5 个用例:上传下载、超限、跨租户、删除确认与权限、raw 鉴权

前端
- 自研零依赖 Markdown 渲染器:先转义再上标签,链接协议白名单
- RichTextEditor:工具栏 + Ctrl+B/I + 预览,粘贴或拖入图片自动上传并插入引用
- AttachmentPanel:上传/下载/删除,下载走带 Authorization 的 fetch 转 blob
- 任务备注、工单说明与验收结论、知识正文、评论全部接入

顺手修的
- TaskDetail.hydrate 漏了 setAttachments,导致附件面板永远是空的(e2e 抓到的)

71 passed;web tsc 0 错、构建 255.07 kB │ gzip 80.85 kB;
e2e:richtext 9 项 + smoke 7 + tickets 8 + knowledge 7,全绿。
This commit is contained in:
dadao
2026-10-29 01:53:46 +08:00
parent 0a3b0f3ca0
commit a24a493403
24 changed files with 1843 additions and 25 deletions
+13 -2
View File
@@ -31,6 +31,7 @@ from ..models_task import Activity, Comment, Item, ItemTag, Tag
from ..models_ticket import TicketExt
from ..schemas import Pagination, encode_cursor, ok
from ..serializers import item_out, tag_out, tags_of
from .attachments import count_for_items, list_for_item
router = APIRouter(prefix="/workspaces/{ws_id}/items", tags=["任务"])
tag_router = APIRouter(prefix="/workspaces/{ws_id}/tags", tags=["标签"])
@@ -294,9 +295,13 @@ async def list_items(
.all()
)
tags = await _tags_of(session, ctx.ws_id, [r.id for r in rows])
counts = await count_for_items(session, ctx.ws_id, [r.id for r in rows])
return ok(
{
"items": [_item_out(r, tags.get(r.id)) for r in rows],
"items": [
dict(_item_out(r, tags.get(r.id)), attachment_count=counts.get(str(r.id), 0))
for r in rows
],
"next_cursor": None,
"has_more": False,
}
@@ -325,9 +330,13 @@ async def list_items(
encode_cursor(getattr(rows[-1], sort_attr), rows[-1].id) if has_more and rows else None
)
tags = await _tags_of(session, ctx.ws_id, [r.id for r in rows])
counts = await count_for_items(session, ctx.ws_id, [r.id for r in rows])
return ok(
{
"items": [_item_out(r, tags.get(r.id)) for r in rows],
"items": [
dict(_item_out(r, tags.get(r.id)), attachment_count=counts.get(str(r.id), 0))
for r in rows
],
"next_cursor": next_cursor,
"has_more": has_more,
}
@@ -405,6 +414,8 @@ async def get_item(item_id: uuid.UUID, ctx: MemberDep, session: SessionDep) -> d
.all()
)
data = _item_out(item, tags.get(item.id))
data["attachments"] = await list_for_item(session, ctx.ws_id, item.id)
data["attachment_count"] = len(data["attachments"])
data["activities"] = [
{
"id": str(a.id),