Compare commits

..

1 Commits

Author SHA1 Message Date
agent_vscode 76e0c08cec fix(docker): install python3/make/g++ toolchain for re2 native build
The develop image build broke at `pnpm install --frozen-lockfile`: the new
native dependency re2@1.25.0 (packages/mcp, search_in_page #330) always
compiles from source under pnpm — its prebuilt-binary downloader
(install-artifact-from-github) cannot identify the GitHub repo because pnpm
does not populate npm_package_repository_*/npm_package_json env vars ("No
github repository was identified. Building locally ..."), and node:22-slim
ships no python3/make/g++ for the node-gyp fallback.

- builder stage: add a cache-friendly apt layer with python3 make g++
  before COPY; the stage is discarded so the toolchain may stay.
- installer stage: install the toolchain, run the prod install as the node
  user via `su node -c`, and purge the toolchain — all in one RUN layer so
  the final image stays slim and node_modules ownership needs no extra
  chown layer; USER node is restored right after.

Fixes the failed run 28715009124 (develop docker build); release.yml uses
the same Dockerfile and is covered too.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-04 22:09:40 +03:00
27 changed files with 97 additions and 1143 deletions
+16 -2
View File
@@ -5,6 +5,13 @@ RUN npm install -g pnpm@10.4.0
FROM base AS builder
# re2 (packages/mcp) always compiles from source under pnpm (the prebuilt-binary
# download cannot identify the GitHub repo), so node-gyp needs python3/make/g++.
# This stage is discarded, so the toolchain can stay installed.
RUN apt-get update \
&& apt-get install -y --no-install-recommends python3 make g++ \
&& rm -rf /var/lib/apt/lists/*
WORKDIR /app
COPY . .
@@ -57,9 +64,16 @@ COPY --from=builder /app/patches /app/patches
RUN chown -R node:node /app
USER node
# Toolchain is needed transiently to compile re2 during the prod install; install
# and purge it in one layer to keep the final image slim. The install itself runs
# as the node user via su to keep node_modules ownership without a costly chown layer.
RUN apt-get update \
&& apt-get install -y --no-install-recommends python3 make g++ \
&& su node -c "pnpm install --frozen-lockfile --prod" \
&& apt-get purge -y --auto-remove python3 make g++ \
&& rm -rf /var/lib/apt/lists/*
RUN pnpm install --frozen-lockfile --prod
USER node
RUN mkdir -p /app/data/storage
-1
View File
@@ -13,7 +13,6 @@
},
"dependencies": {
"@ai-sdk/react": "^3.0.208",
"@braintree/sanitize-url": "7.1.2",
"@atlaskit/pragmatic-drag-and-drop": "1.8.1",
"@atlaskit/pragmatic-drag-and-drop-auto-scroll": "2.1.5",
"@atlaskit/pragmatic-drag-and-drop-flourish": "2.0.15",
+24 -58
View File
@@ -1,72 +1,38 @@
import { lazy, Suspense } from "react";
import { Navigate, Route, Routes } from "react-router-dom";
import { Center, Loader } from "@mantine/core";
import { Error404 } from "@/components/ui/error-404.tsx";
import Layout from "@/components/layouts/global/layout.tsx";
import { useTrackOrigin } from "@/hooks/use-track-origin";
// ShareLayout is route-split: its ShareShell chrome pulls in the table of
// contents (and thus TipTap), so keeping it out of the eager graph removes the
// editor engine from startup for authenticated users too.
const ShareLayout = lazy(
() => import("@/features/share/components/share-layout.tsx"),
);
// Auth / entry pages stay eager: they are the first paint for an unauthenticated
// visitor (e.g. /login) and are already small, so code-splitting them would only
// add a cold-chunk round trip to the most common cold-start path.
import SetupWorkspace from "@/pages/auth/setup-workspace.tsx";
import LoginPage from "@/pages/auth/login";
import Home from "@/pages/dashboard/home";
import Page from "@/pages/page/page";
import AccountSettings from "@/pages/settings/account/account-settings";
import WorkspaceMembers from "@/pages/settings/workspace/workspace-members";
import WorkspaceSettings from "@/pages/settings/workspace/workspace-settings";
import AiSettings from "@/pages/settings/workspace/ai-settings";
import Groups from "@/pages/settings/group/groups";
import GroupInfo from "./pages/settings/group/group-info";
import Spaces from "@/pages/settings/space/spaces.tsx";
import { Error404 } from "@/components/ui/error-404.tsx";
import AccountPreferences from "@/pages/settings/account/account-preferences.tsx";
import SpaceHome from "@/pages/space/space-home.tsx";
import PageRedirect from "@/pages/page/page-redirect.tsx";
import Layout from "@/components/layouts/global/layout.tsx";
import InviteSignup from "@/pages/auth/invite-signup.tsx";
import ForgotPassword from "@/pages/auth/forgot-password.tsx";
import PasswordReset from "./pages/auth/password-reset";
import PageRedirect from "@/pages/page/page-redirect.tsx";
import SharedPage from "@/pages/share/shared-page.tsx";
import Shares from "@/pages/settings/shares/shares.tsx";
import ShareLayout from "@/features/share/components/share-layout.tsx";
import ShareRedirect from "@/pages/share/share-redirect.tsx";
// Heavy / leaf pages are route-split with React.lazy so their code (most
// importantly the whole TipTap editor + KaTeX + lowlight grammars + drawio that
// the page editor and the readonly share editor pull in) is fetched only when
// the matching route is actually visited. The <Suspense> boundaries live inside
// each Layout (around its <Outlet/>), so the app shell stays mounted while a
// route chunk loads.
const Home = lazy(() => import("@/pages/dashboard/home"));
const Page = lazy(() => import("@/pages/page/page"));
const SpaceHome = lazy(() => import("@/pages/space/space-home.tsx"));
const SpaceTrash = lazy(() => import("@/pages/space/space-trash.tsx"));
const SpacesPage = lazy(() => import("@/pages/spaces/spaces.tsx"));
const FavoritesPage = lazy(() => import("@/pages/favorites/favorites-page"));
const LabelPage = lazy(() => import("@/pages/label/label-page"));
const SharedPage = lazy(() => import("@/pages/share/shared-page.tsx"));
const AccountSettings = lazy(
() => import("@/pages/settings/account/account-settings"),
);
const AccountPreferences = lazy(
() => import("@/pages/settings/account/account-preferences.tsx"),
);
const WorkspaceSettings = lazy(
() => import("@/pages/settings/workspace/workspace-settings"),
);
const AiSettings = lazy(() => import("@/pages/settings/workspace/ai-settings"));
const WorkspaceMembers = lazy(
() => import("@/pages/settings/workspace/workspace-members"),
);
const Groups = lazy(() => import("@/pages/settings/group/groups"));
const GroupInfo = lazy(() => import("./pages/settings/group/group-info"));
const Spaces = lazy(() => import("@/pages/settings/space/spaces.tsx"));
const Shares = lazy(() => import("@/pages/settings/shares/shares.tsx"));
import { useTrackOrigin } from "@/hooks/use-track-origin";
import SpacesPage from "@/pages/spaces/spaces.tsx";
import SpaceTrash from "@/pages/space/space-trash.tsx";
import FavoritesPage from "@/pages/favorites/favorites-page";
import LabelPage from "@/pages/label/label-page";
export default function App() {
useTrackOrigin();
return (
<Suspense
fallback={
<Center h="100vh">
<Loader size="sm" />
</Center>
}
>
<>
<Routes>
<Route index element={<Navigate to="/home" />} />
<Route path={"/login"} element={<LoginPage />} />
@@ -117,6 +83,6 @@ export default function App() {
<Route path="*" element={<Error404 />} />
</Routes>
</Suspense>
</>
);
}
@@ -1,37 +0,0 @@
import { describe, it, expect } from "vitest";
import { isChunkLoadError } from "./chunk-load-error-boundary";
// The detector decides whether a caught render error is a stale-deploy chunk-404
// (→ auto-reload to fetch the new manifest) vs a genuine app error (→ generic
// recovery UI, no reload). A false negative on a real chunk failure re-blanks the
// app; a false positive would auto-reload on an ordinary error. Pin both sides.
describe("isChunkLoadError", () => {
it("detects the ChunkLoadError name", () => {
expect(isChunkLoadError({ name: "ChunkLoadError", message: "x" })).toBe(true);
});
it.each([
"Failed to fetch dynamically imported module: https://x/assets/index-abc.js",
"error loading dynamically imported module",
"Importing a module script failed.",
])("detects the dynamic-import failure message %#", (message) => {
expect(isChunkLoadError({ name: "TypeError", message })).toBe(true);
});
it("is case-insensitive on the message", () => {
expect(
isChunkLoadError({ message: "FAILED TO FETCH DYNAMICALLY IMPORTED MODULE" }),
).toBe(true);
});
it.each([
null,
undefined,
{},
{ name: "TypeError", message: "Cannot read properties of undefined" },
{ message: "Network request failed" },
new Error("some ordinary render error"),
])("returns false for a non-chunk error %#", (err) => {
expect(isChunkLoadError(err)).toBe(false);
});
});
@@ -1,71 +0,0 @@
import { ReactNode } from "react";
import { ErrorBoundary } from "react-error-boundary";
import { Button, Center, Stack, Text } from "@mantine/core";
const RELOAD_FLAG = "chunk-reload-attempted";
// Heuristic detection of a failed dynamic import. Since the code-splitting work,
// every route (plus Aside / AiChatWindow) is React.lazy: when a new deploy
// replaces the hashed chunks, a tab left open on the old index.html requests a
// chunk URL that now 404s, and React.lazy rejects. Browsers / Vite surface these
// with a ChunkLoadError name or one of these messages.
export function isChunkLoadError(error: unknown): boolean {
if (!error) return false;
const name = (error as { name?: string }).name ?? "";
const message = (error as { message?: string }).message ?? "";
return (
name === "ChunkLoadError" ||
/Failed to fetch dynamically imported module/i.test(message) ||
/error loading dynamically imported module/i.test(message) ||
/Importing a module script failed/i.test(message)
);
}
function handleError(error: unknown) {
if (!isChunkLoadError(error)) return;
// A stale-chunk 404 is cured by a full reload that re-fetches index.html and
// the new chunk manifest. Auto-reload once, guarding against a reload loop
// (e.g. a genuinely missing chunk) with a one-shot sessionStorage flag. If the
// flag is already set we fall through to the manual recovery UI below.
try {
if (sessionStorage.getItem(RELOAD_FLAG)) return;
sessionStorage.setItem(RELOAD_FLAG, "1");
} catch {
// sessionStorage unavailable (private mode / disabled): skip the automatic
// reload rather than risk an unguarded loop; the fallback UI still recovers.
return;
}
window.location.reload();
}
// Root-level boundary that sits ABOVE every route-level Suspense boundary so a
// lazy route/component chunk failure is caught here instead of unmounting the
// whole tree into a blank white screen. Per-feature ErrorBoundaries (page.tsx,
// transclusion, page-embed) remain in place underneath for their local errors.
export function ChunkLoadErrorBoundary({ children }: { children: ReactNode }) {
return (
<ErrorBoundary
onError={handleError}
fallbackRender={({ error }) => {
const chunk = isChunkLoadError(error);
return (
<Center h="100vh" p="md">
<Stack align="center" gap="sm" maw={420}>
<Text fw={600}>
{chunk ? "A new version is available" : "Something went wrong"}
</Text>
<Text size="sm" c="dimmed" ta="center">
{chunk
? "Please reload the page to load the latest version."
: "An unexpected error occurred. Reloading the page may help."}
</Text>
<Button onClick={() => window.location.reload()}>Reload</Button>
</Stack>
</Center>
);
}}
>
{children}
</ErrorBoundary>
);
}
@@ -1,10 +1,9 @@
import { AppShell, Container } from "@mantine/core";
import React, { Suspense, useEffect, useRef, useState } from "react";
import React, { useEffect, useRef, useState } from "react";
import { useLocation } from "react-router-dom";
import { useTranslation } from "react-i18next";
import SettingsSidebar from "@/components/settings/settings-sidebar.tsx";
import { useAtom, useAtomValue } from "jotai";
import { aiChatWindowOpenAtom } from "@/features/ai-chat/atoms/ai-chat-atom.ts";
import { useAtom } from "jotai";
import {
APP_NAVBAR_ID,
NAVBAR_COLLAPSE_BREAKPOINT,
@@ -15,6 +14,8 @@ import {
} from "@/components/layouts/global/hooks/atoms/sidebar-atom.ts";
import { SpaceSidebar } from "@/features/space/components/sidebar/space-sidebar.tsx";
import { AppHeader } from "@/components/layouts/global/app-header.tsx";
import Aside from "@/components/layouts/global/aside.tsx";
import AiChatWindow from "@/features/ai-chat/components/ai-chat-window.tsx";
import GitmostGlobalBridge from "@/features/editor/gitmost/gitmost-global-bridge.tsx";
import classes from "./app-shell.module.css";
import { useToggleSidebar } from "@/components/layouts/global/hooks/hooks/use-toggle-sidebar.ts";
@@ -22,21 +23,6 @@ import GlobalSidebar from "@/components/layouts/global/global-sidebar.tsx";
import { ASIDE_PANEL_ID } from "@/hooks/use-toggle-aside.tsx";
import { MAIN_CONTENT_ID, SkipToMain } from "@/components/ui/skip-to-main.tsx";
// Lazily load the AI chat window so the AI SDK runtime it pulls in is fetched
// only after the user first opens the chat, instead of for every authenticated
// user on load. The window itself renders null while closed, so there is no
// behavior difference — it simply is not mounted until first opened.
const AiChatWindow = React.lazy(
() => import("@/features/ai-chat/components/ai-chat-window.tsx"),
);
// The right aside hosts the comment panel and table of contents, both of which
// pull in TipTap. It only ever renders on page routes, so lazy-loading it keeps
// the whole editor engine out of the eager global-shell startup graph.
const Aside = React.lazy(
() => import("@/components/layouts/global/aside.tsx"),
);
export default function GlobalAppShell({
children,
}: {
@@ -51,15 +37,6 @@ export default function GlobalAppShell({
const [isResizing, setIsResizing] = useState(false);
const sidebarRef = useRef(null);
// Latch: once the AI chat window has been opened, keep it mounted so an
// in-flight stream is never torn down. Before the first open the AI chat chunk
// is never fetched.
const aiChatOpen = useAtomValue(aiChatWindowOpenAtom);
const [aiChatEverOpened, setAiChatEverOpened] = useState(false);
useEffect(() => {
if (aiChatOpen) setAiChatEverOpened(true);
}, [aiChatOpen]);
const startResizing = React.useCallback((mouseDownEvent) => {
mouseDownEvent.preventDefault();
setIsResizing(true);
@@ -183,21 +160,13 @@ export default function GlobalAppShell({
: undefined
}
>
<Suspense fallback={null}>
<Aside />
</Suspense>
<Aside />
</AppShell.Aside>
)}
</AppShell>
{/* Floating AI chat window. Mounted once globally on first open; it is
position: fixed and self-hides when closed, so its place in the tree is
not critical. Kept mounted after the first open so a live stream is not
aborted. */}
{aiChatEverOpened && (
<Suspense fallback={null}>
<AiChatWindow />
</Suspense>
)}
{/* Floating AI chat window. Mounted once globally; it is position: fixed
and self-hides when closed, so its place in the tree is not critical. */}
<AiChatWindow />
{/* Global gitmost native bridge: registers listSpaces / listPages /
createPageWithRecording on window.gitmost so the native host can
create a page with a recording even when no page editor is open. */}
@@ -1,7 +1,5 @@
import { Suspense, useEffect } from "react";
import { UserProvider } from "@/features/user/user-provider.tsx";
import { Outlet, useParams } from "react-router-dom";
import { Center, Loader } from "@mantine/core";
import GlobalAppShell from "@/components/layouts/global/global-app-shell.tsx";
import { SearchSpotlight } from "@/features/search/components/search-spotlight.tsx";
import { useGetSpaceBySlugQuery } from "@/features/space/queries/space-query.ts";
@@ -10,39 +8,10 @@ export default function Layout() {
const { spaceSlug } = useParams();
const { data: space } = useGetSpaceBySlugQuery(spaceSlug);
// Warm the (now route-split) editor chunk during idle time on authenticated
// routes, so the first navigation to a page renders from cache instead of a
// cold chunk fetch. Best-effort: gated on requestIdleCallback and never blocks
// startup — the dynamic import mirrors the App.tsx route lazy loader so both
// resolve to the same chunk.
useEffect(() => {
const ric =
typeof window !== "undefined" && (window as any).requestIdleCallback;
const warm = () => {
// Best-effort prefetch: a failed warm-up (offline, stale 404) is harmless
// and must not surface as an unhandledrejection.
void import("@/pages/page/page").catch(() => {});
};
if (ric) {
const id = ric(warm);
return () => (window as any).cancelIdleCallback?.(id);
}
const timer = setTimeout(warm, 2000);
return () => clearTimeout(timer);
}, []);
return (
<UserProvider>
<GlobalAppShell>
<Suspense
fallback={
<Center h="60vh">
<Loader size="sm" />
</Center>
}
>
<Outlet />
</Suspense>
<Outlet />
</GlobalAppShell>
<SearchSpotlight spaceId={space?.id} />
</UserProvider>
@@ -1,8 +1,5 @@
import { atom } from "jotai";
// Type-only: these atoms only hold an Editor reference for typing. A value
// import would drag the whole @tiptap/core engine into the eager graph of every
// shell component that reads one of these atoms.
import type { Editor } from "@tiptap/core";
import { Editor } from "@tiptap/core";
import { PageEditMode } from "@/features/user/types/user.types.ts";
import type { DictationUnavailableReason } from "@/features/dictation/dictation-status";
@@ -1,16 +0,0 @@
import { lazy, Suspense } from "react";
import { EditorMenuProps } from "@/features/editor/components/table/types/types.ts";
// Lazily load the drawio bubble menu so it is split out of the editor chunk and
// fetched only when an editable editor is mounted (mirrors excalidraw-menu-lazy).
const DrawioMenu = lazy(
() => import("@/features/editor/components/drawio/drawio-menu.tsx"),
);
export default function DrawioMenuLazy(props: EditorMenuProps) {
return (
<Suspense fallback={null}>
<DrawioMenu {...props} />
</Suspense>
);
}
@@ -1,17 +0,0 @@
import { lazy, Suspense } from "react";
import { NodeViewProps } from "@tiptap/react";
// Lazily load the drawio node view so the heavy react-drawio embed runtime is
// split into its own chunk and fetched only when a drawio diagram is actually
// rendered (mirrors excalidraw-view-lazy).
const DrawioView = lazy(
() => import("@/features/editor/components/drawio/drawio-view.tsx"),
);
export default function DrawioViewLazy(props: NodeViewProps) {
return (
<Suspense fallback={null}>
<DrawioView {...props} />
</Suspense>
);
}
@@ -1,19 +0,0 @@
import { lazy, Suspense } from "react";
import { NodeViewProps } from "@tiptap/react";
// Lazily load the KaTeX-backed block math view so the katex chunk is fetched
// only when a document actually contains a math node (mirrors the mermaid/
// excalidraw lazy pattern). The local Suspense keeps a slow katex chunk from
// crashing or blocking the whole editor: while it loads we render the raw
// LaTeX source as a node-sized placeholder.
const MathBlockView = lazy(
() => import("@/features/editor/components/math/math-block.tsx"),
);
export default function MathBlockViewLazy(props: NodeViewProps) {
return (
<Suspense fallback={<div data-katex="true">{props.node.attrs.text}</div>}>
<MathBlockView {...props} />
</Suspense>
);
}
@@ -1,19 +0,0 @@
import { lazy, Suspense } from "react";
import { NodeViewProps } from "@tiptap/react";
// Lazily load the KaTeX-backed inline math view so the katex chunk is fetched
// only when a document actually contains a math node (mirrors the mermaid/
// excalidraw lazy pattern). The local Suspense keeps a slow katex chunk from
// crashing or blocking the whole editor: while it loads we render the raw
// LaTeX source as a node-sized placeholder.
const MathInlineView = lazy(
() => import("@/features/editor/components/math/math-inline.tsx"),
);
export default function MathInlineViewLazy(props: NodeViewProps) {
return (
<Suspense fallback={<span data-katex="true">{props.node.attrs.text}</span>}>
<MathInlineView {...props} />
</Suspense>
);
}
@@ -81,8 +81,8 @@ import {
createResizeHandle,
buildResizeClasses,
} from "@/features/editor/components/common/node-resize-handles.ts";
import MathInlineView from "@/features/editor/components/math/math-inline-lazy.tsx";
import MathBlockView from "@/features/editor/components/math/math-block-lazy.tsx";
import MathInlineView from "@/features/editor/components/math/math-inline.tsx";
import MathBlockView from "@/features/editor/components/math/math-block.tsx";
import ImageView from "@/features/editor/components/image/image-view.tsx";
import CalloutView from "@/features/editor/components/callout/callout-view.tsx";
import StatusView from "@/features/editor/components/status/status-view.tsx";
@@ -90,7 +90,7 @@ import VideoView from "@/features/editor/components/video/video-view.tsx";
import AudioView from "@/features/editor/components/audio/audio-view.tsx";
import AttachmentView from "@/features/editor/components/attachment/attachment-view.tsx";
import CodeBlockView from "@/features/editor/components/code-block/code-block-view.tsx";
import DrawioView from "../components/drawio/drawio-view-lazy.tsx";
import DrawioView from "../components/drawio/drawio-view";
import ExcalidrawView from "@/features/editor/components/excalidraw/excalidraw-view-lazy.tsx";
import EmbedView from "@/features/editor/components/embed/embed-view.tsx";
import HtmlEmbedView from "@/features/editor/components/html-embed/html-embed-view.tsx";
@@ -1,17 +1,8 @@
import { useEffect, useRef } from "react";
import { useNavigate } from "react-router-dom";
import { getDefaultStore } from "jotai";
// Literal value of WebSocketStatus.Connected from @hocuspocus/provider. Inlined
// so this always-mounted global bridge does not statically import
// @hocuspocus/provider — that import pulls Yjs (and, through a shared chunk, the
// whole TipTap engine) into the eager startup graph. yjsConnectionStatusAtom
// already stores these raw status strings.
const YJS_STATUS_CONNECTED = "connected";
// Type-only: importing Editor as a type keeps @tiptap/core (the whole editor
// engine) out of the eager global-shell graph — the bridge only uses it for
// annotations/casts, never as a runtime value.
import type { Editor } from "@tiptap/core";
import { WebSocketStatus } from "@hocuspocus/provider";
import { Editor } from "@tiptap/core";
import {
pageEditorAtom,
yjsConnectionStatusAtom,
@@ -25,19 +16,15 @@ import {
getSidebarPages,
} from "@/features/page/services/page-service.ts";
import { buildPageUrl } from "@/features/page/page.utils.ts";
// Types are erased at build time, so importing them does not pull the module's
// runtime (which drags in @tiptap + the editor-ext barrel). The actual recording
// helpers are dynamically imported at call time inside createPageWithRecording,
// keeping the editor engine out of the eager global-shell startup graph — the
// bridge is mounted for every authenticated user but recording is a rare,
// native-host-driven action.
import type {
import {
GitmostBridge,
GitmostCreatePagePayload,
GitmostCreatePageResult,
GitmostListPagesPayload,
GitmostListPagesResult,
GitmostListSpacesResult,
gitmostDecodePayloadToFile,
gitmostUploadFileToEditor,
} from "@/features/editor/gitmost/gitmost-recording.ts";
// How long to wait for a freshly-navigated page's editor to mount, become
@@ -70,7 +57,7 @@ function gitmostWaitForEditor(
!editor.isDestroyed &&
editor.isEditable &&
editorPageId === pageId &&
yjsStatus === YJS_STATUS_CONNECTED;
yjsStatus === WebSocketStatus.Connected;
if (ready) {
resolve(editor);
return;
@@ -184,12 +171,6 @@ export default function GitmostGlobalBridge() {
};
}
// Load the recording helpers on demand (see the import note above). This
// is the only place they are needed, so the @tiptap/editor-ext code they
// pull in stays out of the eager startup graph.
const { gitmostDecodePayloadToFile, gitmostUploadFileToEditor } =
await import("@/features/editor/gitmost/gitmost-recording.ts");
// Validate/decode the recording BEFORE creating the page so a bad
// payload never leaves an empty junk page behind. Per the createPage
// error contract, any decode failure collapses to "insert-failed" (the
@@ -59,7 +59,7 @@ import {
handlePaste,
} from "@/features/editor/components/common/editor-paste-handler.tsx";
import ExcalidrawMenu from "./components/excalidraw/excalidraw-menu-lazy";
import DrawioMenu from "./components/drawio/drawio-menu-lazy";
import DrawioMenu from "./components/drawio/drawio-menu";
import { useCollabToken } from "@/features/auth/queries/auth-query.tsx";
import SearchAndReplaceDialog from "@/features/editor/components/search-and-replace/search-and-replace-dialog.tsx";
import { useDebouncedCallback, useDocumentVisibility } from "@mantine/hooks";
@@ -1,20 +1,10 @@
import { Suspense } from "react";
import { Outlet } from "react-router-dom";
import { Center, Loader } from "@mantine/core";
import ShareShell from "@/features/share/components/share-shell.tsx";
export default function ShareLayout() {
return (
<ShareShell>
<Suspense
fallback={
<Center h="60vh">
<Loader size="sm" />
</Center>
}
>
<Outlet />
</Suspense>
<Outlet />
</ShareShell>
);
}
+1 -1
View File
@@ -1,7 +1,7 @@
// Source: https://github.com/mantinedev/mantine/blob/master/packages/@mantine/hooks/src/use-clipboard/use-clipboard.ts
// polyfilled to support execCommand fallback
import { useState } from "react";
import { execCommandCopy } from "@/lib/copy-to-clipboard.ts";
import { execCommandCopy } from "@docmost/editor-ext";
export type UseClipboardOptions = {
timeout?: number;
+1 -1
View File
@@ -1,7 +1,7 @@
import bytes from "bytes";
import { castToBoolean } from "@/lib/utils.tsx";
import { AvatarIconType } from "@/features/attachments/types/attachment.types.ts";
import { sanitizeUrl } from "@/lib/sanitize-url.ts";
import { sanitizeUrl } from "@docmost/editor-ext";
declare global {
interface Window {
-16
View File
@@ -1,16 +0,0 @@
// Client-local execCommand copy fallback (previously imported from
// @docmost/editor-ext). It lives here so the ubiquitous useClipboard / CopyButton
// path does not pull in the editor-ext barrel — and with it the whole TipTap
// engine — through the eager startup graph. Behavior is identical to the
// editor-ext helper it replaces.
export function execCommandCopy(text: string): void {
const textarea = document.createElement("textarea");
textarea.value = text;
textarea.style.position = "fixed";
textarea.style.left = "-9999px";
textarea.style.top = "-9999px";
document.body.appendChild(textarea);
textarea.select();
document.execCommand("copy");
document.body.removeChild(textarea);
}
-31
View File
@@ -1,31 +0,0 @@
import { describe, it, expect } from "vitest";
import { sanitizeUrl } from "./sanitize-url";
// `sanitizeUrl` is a byte-identical client-local copy of editor-ext's wrapper
// around @braintree/sanitize-url: it maps the sanitizer's "about:blank" XSS
// sentinel to "". These assertions mirror editor-ext's own security-contract
// test so the extracted copy keeps the same guarantees.
describe("sanitizeUrl", () => {
it("blocks dangerous schemes (returns empty string)", () => {
expect(sanitizeUrl("javascript:alert(1)")).toBe("");
expect(sanitizeUrl("data:text/html,<script>alert(1)</script>")).toBe("");
expect(sanitizeUrl("vbscript:msgbox(1)")).toBe("");
// Case / whitespace obfuscation must not slip past the sanitizer.
expect(sanitizeUrl(" JaVaScRiPt:alert(1)")).toBe("");
});
it("returns empty string for empty / undefined input", () => {
expect(sanitizeUrl(undefined)).toBe("");
expect(sanitizeUrl("")).toBe("");
});
it("allows safe https, relative file and mailto URLs", () => {
expect(sanitizeUrl("https://example.com/page")).toMatch(
/^https:\/\/example\.com\/page/,
);
expect(sanitizeUrl("/api/files/abc-123")).toBe("/api/files/abc-123");
expect(sanitizeUrl("mailto:user@example.com")).toBe(
"mailto:user@example.com",
);
});
});
-15
View File
@@ -1,15 +0,0 @@
import { sanitizeUrl as braintreeSanitizeUrl } from "@braintree/sanitize-url";
// Client-local copy of editor-ext's sanitizeUrl wrapper. Importing it from the
// editor-ext barrel dragged the whole TipTap engine into the eager startup graph
// via the app-wide config module (getFileUrl). This keeps the exact same
// behavior (braintree sanitize + normalize "about:blank" -> "") without that
// dependency.
export function sanitizeUrl(url: string | undefined): string {
if (!url) return "";
const sanitized = braintreeSanitizeUrl(url);
// Return an empty string instead of "about:blank".
return sanitized === "about:blank" ? "" : sanitized;
}
+27 -60
View File
@@ -13,14 +13,15 @@ import { ModalsProvider } from "@mantine/modals";
import { Notifications } from "@mantine/notifications";
import { QueryClient, QueryClientProvider } from "@tanstack/react-query";
import { HelmetProvider } from "react-helmet-async";
import { ChunkLoadErrorBoundary } from "@/components/chunk-load-error-boundary.tsx";
import "./i18n";
import { PostHogProvider } from "posthog-js/react";
import {
getPostHogHost,
getPostHogKey,
isCloud,
isPostHogEnabled,
} from "@/lib/config.ts";
import posthog from "posthog-js";
export const queryClient = new QueryClient({
defaultOptions: {
@@ -33,65 +34,31 @@ export const queryClient = new QueryClient({
},
});
if (isCloud() && isPostHogEnabled) {
posthog.init(getPostHogKey(), {
api_host: getPostHogHost(),
defaults: "2025-05-24",
disable_session_recording: true,
capture_pageleave: false,
});
}
const container = document.getElementById("root") as HTMLElement;
const root = (container as any).__reactRoot ??= ReactDOM.createRoot(container);
function renderApp() {
root.render(
<BrowserRouter>
<MantineProvider theme={theme} cssVariablesResolver={mantineCssResolver}>
<ModalsProvider>
<QueryClientProvider client={queryClient}>
<Notifications position="bottom-center" limit={3} zIndex={10000} />
<HelmetProvider>
{/* Root boundary above every lazy route's Suspense: a stale-chunk
404 after a deploy is caught and recovered here instead of
blanking the whole app. */}
<ChunkLoadErrorBoundary>
<App />
</ChunkLoadErrorBoundary>
</HelmetProvider>
</QueryClientProvider>
</ModalsProvider>
</MantineProvider>
</BrowserRouter>,
);
}
async function initAnalytics() {
// posthog-js is only pulled in for cloud deployments with analytics enabled, so
// self-hosted builds never download it. The gate is kept identical to the
// previous eager code so cloud analytics behavior is unchanged; the import is
// simply deferred behind it.
//
// Crucially this runs AFTER the immediate first render below, so first paint is
// never gated on the analytics chunk. Any failure (network, stale 404, or an
// ad-blocker blocking a chunk named "posthog") is swallowed so the user keeps a
// working app without analytics instead of a permanently blank page.
//
// NOTE: we init the posthog SINGLETON only and do NOT wrap the tree in
// <PostHogProvider>. The app has zero consumers of the PostHog React context
// (no usePostHog / useFeatureFlag* / PostHogFeature), and PostHogProvider given
// an already-initialized `client` is a no-op — all capture goes through the
// singleton. Re-rendering to attach the provider would only REMOUNT the whole
// App (running every mount effect twice and dropping local state / focus /
// in-progress input on cloud cold-load) for no functional gain.
if (!(isCloud() && isPostHogEnabled)) return;
try {
const { default: posthog } = await import("posthog-js");
posthog.init(getPostHogKey(), {
api_host: getPostHogHost(),
defaults: "2025-05-24",
disable_session_recording: true,
capture_pageleave: false,
});
} catch {
// Analytics failed to load — degrade gracefully; the app already rendered.
}
}
// Paint immediately for everyone (self-hosted stays exactly as instant as before,
// cloud no longer blocks on the analytics import). The posthog singleton is
// initialized after, without re-rendering the tree.
renderApp();
void initAnalytics();
root.render(
<BrowserRouter>
<MantineProvider theme={theme} cssVariablesResolver={mantineCssResolver}>
<ModalsProvider>
<QueryClientProvider client={queryClient}>
<Notifications position="bottom-center" limit={3} zIndex={10000} />
<HelmetProvider>
<PostHogProvider client={posthog}>
<App />
</PostHogProvider>
</HelmetProvider>
</QueryClientProvider>
</ModalsProvider>
</MantineProvider>
</BrowserRouter>,
);
-14
View File
@@ -63,20 +63,6 @@ export default defineConfig(({ mode }) => {
name: "vendor-mantine",
test: /[\\/]node_modules[\\/]@mantine[\\/]/,
},
// NOTE: TipTap/ProseMirror/Yjs are intentionally NOT force-grouped
// into a single vendor chunk. Doing so backfires: rolldown co-locates
// a small module shared with the (eager) react-i18next runtime into
// that group chunk, which then drags the whole ~590KB editor engine
// into the eager modulepreload graph. Left to the default splitting,
// the editor engine stays in lazily-loaded chunks pulled only by the
// route-split editor/share pages. KaTeX is safe to group (nothing
// eager references it).
// KaTeX in its own stable chunk; loaded on demand by the lazy math
// node views (never in the startup path).
{
name: "vendor-katex",
test: /[\\/]node_modules[\\/]katex[\\/]/,
},
],
},
},
@@ -635,17 +635,13 @@ const Attachment = Node.create({
},
name: {
default: null,
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class): "" -> default.
parseHTML: (el: HTMLElement) =>
el.getAttribute("data-attachment-name") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("data-attachment-name"),
renderHTML: (attrs: Record<string, any>) =>
attrs.name ? { "data-attachment-name": attrs.name } : {},
},
mime: {
default: null,
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class): "" -> default.
parseHTML: (el: HTMLElement) =>
el.getAttribute("data-attachment-mime") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("data-attachment-mime"),
renderHTML: (attrs: Record<string, any>) =>
attrs.mime ? { "data-attachment-mime": attrs.mime } : {},
},
@@ -693,10 +689,7 @@ const Video = Node.create({
},
alt: {
default: null,
// Empty-string-vs-absent idempotency: coerce "" back to the default so a
// stray empty `aria-label` never materializes `alt: ""` on a video stored
// with no alt (same GS-EDIT-REVERT class as the image `alt` fix).
parseHTML: (el: HTMLElement) => el.getAttribute("aria-label") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("aria-label"),
renderHTML: (attrs: Record<string, any>) =>
attrs.alt ? { "aria-label": attrs.alt } : {},
},
@@ -871,15 +864,13 @@ const diagramAttributes = () => ({
},
title: {
default: null,
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class): "" -> default.
parseHTML: (el: HTMLElement) => el.getAttribute("data-title") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("data-title"),
renderHTML: (attrs: Record<string, any>) =>
attrs.title ? { "data-title": attrs.title } : {},
},
alt: {
default: null,
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class): "" -> default.
parseHTML: (el: HTMLElement) => el.getAttribute("data-alt") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("data-alt"),
renderHTML: (attrs: Record<string, any>) =>
attrs.alt ? { "data-alt": attrs.alt } : {},
},
@@ -1115,8 +1106,7 @@ const Pdf = Node.create({
},
name: {
default: null,
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class): "" -> default.
parseHTML: (el: HTMLElement) => el.getAttribute("data-name") || null,
parseHTML: (el: HTMLElement) => el.getAttribute("data-name"),
renderHTML: (attrs: Record<string, any>) =>
attrs.name ? { "data-name": attrs.name } : {},
},
@@ -1501,29 +1491,6 @@ export const docmostExtensions = [
...parent.height,
parseHTML: (el: HTMLElement) => el.getAttribute("height"),
},
// Empty-string-vs-absent idempotency (GS-EDIT-REVERT class). `marked`
// renders `![](src)` as `<img alt="">`, so the stock Image `alt`
// parseHTML (`getAttribute("alt")`) materializes `alt: ""` on an image
// that was stored with NO alt (attr absent). That is a false diff against
// the editor-stored form (a no-alt image has alt ABSENT, not ""), so a
// git-sync / ai-chat touch of a page with a plain image produced phantom
// churn. Coerce an empty string back to the attr's default (null) so the
// import is idempotent. A real alt survives verbatim (`|| undefined` keeps
// the truthy value; the default fills the empty case). `title` is coerced
// the same way for the whole class, even though `marked` does not
// currently emit `title=""` — defence in depth against any path that does.
// NOTE: this DIVERGES from editor-ext's literal image `alt` parseHTML
// (`getAttribute("alt")`, which returns "" verbatim), but CONVERGES on
// editor-ext's real STORED shape: an editor image inserted without alt
// renders with no `alt` attribute and re-parses as absent, never "".
alt: {
...parent.alt,
parseHTML: (el: HTMLElement) => el.getAttribute("alt") || null,
},
title: {
...parent.title,
parseHTML: (el: HTMLElement) => el.getAttribute("title") || null,
},
};
},
}).configure({ inline: false }),
@@ -1,443 +0,0 @@
/**
* Reusable round-trip-STABILITY matrix helper (fixtures-first).
*
* A single stored node authored WITHOUT a given string attribute (attr
* absent / undefined) must not gain a phantom EMPTY-STRING value after a
* markdown round-trip — the "empty-string-vs-absent" churn class. This helper,
* given a node spec, drives a matrix of attribute combinations through the REAL
* converter (`convertProseMirrorToMarkdown` -> `markdownToProseMirror`) and
* asserts byte-stability on two contours:
*
* 1. RAW round-trip: for the node under test, every attribute the round-trip
* materializes must equal what the INPUT authored — an authored attr keeps
* its value, an ABSENT attr may only reappear at its SCHEMA DEFAULT. If an
* absent attr comes back as a NON-default value (e.g. `alt: ""` where the
* default is `null`), that is an instability and is reported precisely as
* `type.attr: absent -> "<got>"`. This is the contour git-sync / stored
* JSON diffs on, so masking it only in `canonicalize` would leave the noise.
*
* 2. CANONICAL round-trip: `canonicalizeContent(original)` must deep-equal
* `canonicalizeContent(roundtrip)` (a second, semantic contour).
*
* The ONLY normalization the helper treats as allowed (not an instability) is
* the DOCUMENTED numeric width/height/size/aspectRatio -> string coercion the
* converter performs on purpose (a stored numeric `640` re-parses via
* `getAttribute` as the string `"640"`). It is encoded here as an explicit
* per-spec `numericStringAttrs` set applied to BOTH contours, NOT a silent skip.
*
* The helper is node-type agnostic: image and the whole media family share the
* `align !== "center"` predicate + `<!--name {…}-->` comment machinery, so one
* matrix guards the shared class.
*/
import { getSchema } from "@tiptap/core";
import {
convertProseMirrorToMarkdown,
markdownToProseMirror,
canonicalizeContent,
docmostExtensions,
} from "../src/lib/index.js";
import { firstDivergence } from "./roundtrip-helpers.js";
/** One attribute's two probe values. */
export interface AttrMatrixEntry {
/** Attribute name on the node. */
attr: string;
/**
* The "default" pick. `undefined` means the attribute is OMITTED entirely
* (the absent case — the one that can materialize an empty string on import).
* A concrete value is authored verbatim.
*/
default: unknown;
/** A representative NON-default value to exercise (must survive verbatim). */
nonDefault: unknown;
/**
* Marks the attr as a member of the EMPTY-STRING class the fix targets: a
* string attr whose schema default is `null`/absent and whose parseHTML
* coerces `"" -> default` (image/drawio `alt`+`title`, video `alt` via
* aria-label, pdf/attachment `name`, attachment `mime`). Set true to also
* drive the THIRD-STATE convergence case (see runConvergenceCase) for this
* attr. Attrs whose default is NOT null (e.g. embed `provider`, default "")
* or that are not `""`-coerced (control attrs) are left unset.
*/
emptyStringClass?: boolean;
}
/** A node type + the attribute matrix to sweep for it. */
export interface NodeStabilitySpec {
/** Node type (e.g. "image", "video"). */
type: string;
/** Attributes always present on the node (e.g. `{ src: "/i.png" }`). */
baseAttrs?: Record<string, unknown>;
/** Attributes to sweep at default and non-default. */
attrMatrix: AttrMatrixEntry[];
/**
* Attributes whose numeric -> string coercion on round-trip is DOCUMENTED and
* intentional; compared modulo `String(x)` on both sides. Defaults to the
* converter's known sizing set.
*/
numericStringAttrs?: string[];
}
/** A single unstable finding, legible enough to tie a gate-lock to. */
export interface Instability {
type: string;
attr: string;
/** What the input authored: the literal value, or the ABSENT sentinel. */
authored: unknown | typeof ABSENT;
/** What the round-trip produced. */
got: unknown;
/** What a stable round-trip should have produced (authored value or default). */
expected: unknown;
}
/** One matrix cell's result. */
export interface ComboResult {
label: string;
authored: Record<string, unknown>;
/** RAW-contour instabilities on the node under test. */
raw: Instability[];
/** CANONICAL-contour divergence (path + values) or null when equal. */
canonical: { path: string; a: unknown; b: unknown } | null;
/** True when the node type failed to round-trip at all (structural loss). */
missing: boolean;
md: string;
}
/** Whole-matrix report for one node spec. */
export interface MatrixReport {
type: string;
combos: ComboResult[];
}
/** Sentinel marking an attribute the input did NOT author. */
export const ABSENT = Symbol("ABSENT");
const DEFAULT_NUMERIC_STRING_ATTRS = [
"width",
"height",
"size",
"aspectRatio",
];
// The ProseMirror schema the converter targets — its attribute `default`s are
// the authoritative "what an absent attr should re-materialize as" oracle.
const schema = getSchema(docmostExtensions);
/** Read the schema default for every attribute of a node type. */
function schemaDefaults(type: string): Record<string, unknown> {
const specAttrs = (schema.nodes[type]?.spec?.attrs ?? {}) as Record<
string,
{ default: unknown }
>;
const out: Record<string, unknown> = {};
for (const [k, v] of Object.entries(specAttrs)) out[k] = v.default;
return out;
}
/** Find the first node of a given type anywhere in a PM doc tree. */
function findFirst(node: any, type: string): any {
if (node && node.type === type) return node;
for (const child of node?.content ?? []) {
const hit = findFirst(child, type);
if (hit) return hit;
}
return null;
}
/** Coerce a scalar for the documented numeric->string comparison. */
const numStr = (x: unknown): unknown => (x == null ? x : String(x));
/**
* Enumerate the cartesian product of the matrix: every attribute independently
* at its default (index 0) or non-default (index 1) pick. The all-default
* corner is included (the baseline). Small by construction (2^N over a handful
* of at-risk string attrs).
*/
function enumerateCombos(matrix: AttrMatrixEntry[]): number[][] {
let combos: number[][] = [[]];
for (let i = 0; i < matrix.length; i++) {
const next: number[][] = [];
for (const c of combos) {
next.push([...c, 0]);
next.push([...c, 1]);
}
combos = next;
}
return combos;
}
/** Build the authored attrs for one combo pick vector. */
function authoredAttrs(
spec: NodeStabilitySpec,
picks: number[],
): Record<string, unknown> {
const attrs: Record<string, unknown> = { ...(spec.baseAttrs ?? {}) };
spec.attrMatrix.forEach((entry, i) => {
if (picks[i] === 1) {
attrs[entry.attr] = entry.nonDefault;
} else if (entry.default !== undefined) {
attrs[entry.attr] = entry.default;
}
// default === undefined -> OMIT the attr entirely (the absent case).
});
return attrs;
}
/** Human-readable label for a combo (which attrs are at non-default). */
function comboLabel(spec: NodeStabilitySpec, picks: number[]): string {
const on = spec.attrMatrix
.filter((_, i) => picks[i] === 1)
.map((e) => e.attr);
return on.length === 0 ? "<all-default>" : on.join("+");
}
/**
* Run the full stability matrix for one node spec and return a structured
* report (does NOT throw — the caller asserts, so a failure can print the whole
* report). Every combo runs the real export->import pipeline once.
*/
export async function runStabilityMatrix(
spec: NodeStabilitySpec,
): Promise<MatrixReport> {
const numericStringAttrs = new Set(
spec.numericStringAttrs ?? DEFAULT_NUMERIC_STRING_ATTRS,
);
const defaults = schemaDefaults(spec.type);
const combos: ComboResult[] = [];
for (const picks of enumerateCombos(spec.attrMatrix)) {
const authored = authoredAttrs(spec, picks);
const doc = { type: "doc", content: [{ type: spec.type, attrs: authored }] };
const md = convertProseMirrorToMarkdown(doc);
const rt = await markdownToProseMirror(md);
const node = findFirst(rt, spec.type);
const result: ComboResult = {
label: comboLabel(spec, picks),
authored,
raw: [],
canonical: null,
missing: node == null,
md,
};
if (node != null) {
// RAW contour: every materialized attr must equal the authored value, or
// (for an absent attr) the schema default — modulo the documented numeric
// string coercion.
const rtAttrs = (node.attrs ?? {}) as Record<string, unknown>;
for (const key of Object.keys(rtAttrs)) {
const authoredHas = Object.prototype.hasOwnProperty.call(authored, key);
const expected = authoredHas ? authored[key] : defaults[key];
let got = rtAttrs[key];
let exp = expected;
if (numericStringAttrs.has(key)) {
got = numStr(got);
exp = numStr(exp);
}
if (firstDivergence(got, exp) !== null) {
result.raw.push({
type: spec.type,
attr: key,
authored: authoredHas ? authored[key] : ABSENT,
got: rtAttrs[key],
expected,
});
}
}
// CANONICAL contour: canonical forms deep-equal, modulo the same numeric
// string coercion (applied to both trees so a documented coercion is not
// counted as a divergence).
const ca = normalizeNumeric(canonicalizeContent(doc), numericStringAttrs);
const cb = normalizeNumeric(canonicalizeContent(rt), numericStringAttrs);
result.canonical = firstDivergence(ca, cb);
}
combos.push(result);
}
return { type: spec.type, combos };
}
/**
* Deep-copy a canonical tree, coercing the documented numeric->string attrs to
* their string form so an intentional `640 -> "640"` coercion is not reported
* as a canonical divergence. Only touches the listed attribute keys.
*/
function normalizeNumeric(node: any, attrs: Set<string>): any {
if (Array.isArray(node)) return node.map((n) => normalizeNumeric(n, attrs));
if (node === null || typeof node !== "object") return node;
const out: Record<string, unknown> = {};
for (const key of Object.keys(node)) {
if (key === "attrs" && node.attrs && typeof node.attrs === "object") {
const a: Record<string, unknown> = {};
for (const [k, v] of Object.entries(node.attrs)) {
a[k] = attrs.has(k) ? numStr(v) : v;
}
out.attrs = a;
} else {
out[key] = normalizeNumeric(node[key], attrs);
}
}
return out;
}
/** Flatten a report to just its unstable combos (for a terse assertion). */
export function unstableCombos(report: MatrixReport): ComboResult[] {
return report.combos.filter(
(c) => c.missing || c.raw.length > 0 || c.canonical !== null,
);
}
// ---------------------------------------------------------------------------
// THIRD STATE: an EXPLICITLY-STORED empty string on a string attr.
//
// The matrix above sweeps TWO states per string attr: absent/default and a
// non-default value — and asserts FIRST-pass byte-stability for both. There is
// a third, degenerate state the matrix does NOT cover: the attr stored as a
// LITERAL `""`. This is DISTINCT from "the node never had the attr": a user
// types an alt in the editor, then deletes it, and Tiptap's
// `updateAttributes({ alt: "" })` persists a literal `alt: ""` in the stored
// JSON. There is no absent-vs-"" distinction in the DOM once serialized, so the
// fix's `getAttribute("alt") || null` coercion canonicalizes BOTH to the
// default (`null`).
//
// Consequence — and this is CORRECT, not a bug: a doc carrying an explicit `""`
// converges to the default on the FIRST round-trip (a ONE-TIME diff: `"" ->
// null`), then is byte-stable from the SECOND round-trip on (idempotent). So
// this state must be pinned with a DIFFERENT contract than the matrix's:
// - do NOT assert first-pass byte-stability (the first pass legitimately
// changes `""` -> default), and
// - DO assert the first pass converges to the default AND the second pass is
// idempotent (rt2 deep-equals rt1).
//
// A future sync/QA pass diffing stored pages will see this one-time `"" -> null`
// normalization exactly once per affected node; it is the converter canon, not
// corruption, and must not be flagged as data loss.
// ---------------------------------------------------------------------------
/** Result of the third-state ("explicit empty string") convergence probe. */
export interface ConvergenceResult {
type: string;
attr: string;
/** The schema default the attr must converge to on pass 1 (null / absent). */
expectedDefault: unknown;
/** rt1's materialized value for the attr — must equal `expectedDefault`. */
firstPassValue: unknown;
/** True when the node round-tripped AND rt1 converged the attr to default. */
convergedToDefault: boolean;
/** rt1-vs-rt2 divergence; MUST be null (idempotent from pass 2 on). */
secondPassDivergence: { path: string; a: unknown; b: unknown } | null;
/** True when the node type failed to round-trip at all (structural loss). */
missing: boolean;
}
/** Round-trip a full PM doc through the real converter once. */
async function roundtripDoc(doc: any): Promise<any> {
return markdownToProseMirror(convertProseMirrorToMarkdown(doc));
}
/**
* Third-state convergence probe for one string attr of the empty-string class.
*
* (a) builds a doc with the attr EXPLICITLY set to `""` (baseAttrs + `""`),
* (b) rt1 = roundtrip(doc); asserts rt1's attr equals the schema default — the
* documented ONE-TIME `"" -> default` normalization (NOT byte-stable vs the
* `""` input, so first-pass stability is deliberately NOT asserted here),
* (c) rt2 = roundtrip(rt1); asserts rt2 deep-equals rt1 — idempotent from the
* second round-trip on.
*
* Returns a structured result (does NOT throw) so the caller can assert and
* print. Reusable across the whole node family: drive it for every attr flagged
* `emptyStringClass` on every spec (see convergenceCasesFor / the test driver).
*/
export async function runConvergenceCase(
spec: NodeStabilitySpec,
attr: string,
): Promise<ConvergenceResult> {
const expectedDefault = schemaDefaults(spec.type)[attr];
// (a) The degenerate third state: attr persisted as a LITERAL "".
const authored = { ...(spec.baseAttrs ?? {}), [attr]: "" };
const doc = { type: "doc", content: [{ type: spec.type, attrs: authored }] };
// (b) First round-trip: "" must normalize to the default (a one-time diff).
const rt1 = await roundtripDoc(doc);
const node1 = findFirst(rt1, spec.type);
const firstPassValue = node1?.attrs?.[attr];
const convergedToDefault =
node1 != null && firstDivergence(firstPassValue, expectedDefault) === null;
// (c) Second round-trip: must be byte-stable (rt2 deep-equals rt1). We compare
// the WHOLE docs — both are converter OUTPUTS already in the same materialized
// form (numeric attrs are strings on both sides), so no numeric normalization
// is needed here, unlike the raw/canonical contours above.
const rt2 = node1 != null ? await roundtripDoc(rt1) : rt1;
const secondPassDivergence =
node1 != null ? firstDivergence(rt1, rt2) : null;
return {
type: spec.type,
attr,
expectedDefault,
firstPassValue,
convergedToDefault,
secondPassDivergence,
missing: node1 == null,
};
}
/** The attrs of a spec flagged as members of the empty-string class. */
export function convergenceCasesFor(spec: NodeStabilitySpec): string[] {
return spec.attrMatrix
.filter((e) => e.emptyStringClass)
.map((e) => e.attr);
}
/** True when a convergence result honours the "converges once, then stable" contract. */
export function convergenceOk(r: ConvergenceResult): boolean {
return !r.missing && r.convergedToDefault && r.secondPassDivergence === null;
}
/** Render a convergence result as a legible one-liner for a failed assertion. */
export function formatConvergence(r: ConvergenceResult): string {
if (r.missing) return `${r.type}.${r.attr}: DID-NOT-ROUND-TRIP`;
const parts: string[] = [];
if (!r.convergedToDefault) {
parts.push(
`pass1 did NOT converge: got ${JSON.stringify(r.firstPassValue)} (expected default ${JSON.stringify(r.expectedDefault)})`,
);
}
if (r.secondPassDivergence) {
parts.push(
`pass2 NOT idempotent @ ${r.secondPassDivergence.path}: ${JSON.stringify(r.secondPassDivergence.a)} vs ${JSON.stringify(r.secondPassDivergence.b)}`,
);
}
const status = parts.length === 0 ? "converges-once-then-stable" : parts.join("; ");
return `${r.type}.${r.attr}: ${status}`;
}
/** Render a report as a legible multi-line string for a failed assertion. */
export function formatReport(report: MatrixReport): string {
const lines: string[] = [`node "${report.type}":`];
for (const c of report.combos) {
const flags: string[] = [];
if (c.missing) flags.push("DID-NOT-ROUND-TRIP");
for (const i of c.raw) {
const authored =
i.authored === ABSENT ? "absent" : JSON.stringify(i.authored);
flags.push(
`RAW ${i.type}.${i.attr}: ${authored} -> ${JSON.stringify(i.got)} (expected ${JSON.stringify(i.expected)})`,
);
}
if (c.canonical) {
flags.push(
`CANON @ ${c.canonical.path}: ${JSON.stringify(c.canonical.a)} vs ${JSON.stringify(c.canonical.b)}`,
);
}
const status = flags.length === 0 ? "stable" : flags.join("; ");
lines.push(` [${c.label}] ${status}`);
}
return lines.join("\n");
}
@@ -1,164 +0,0 @@
import { describe, expect, it } from "vitest";
import {
runStabilityMatrix,
unstableCombos,
formatReport,
runConvergenceCase,
convergenceCasesFor,
convergenceOk,
formatConvergence,
type NodeStabilitySpec,
} from "./roundtrip-stability.helper.js";
// ---------------------------------------------------------------------------
// Round-trip STABILITY matrix for image + the media family.
//
// Guards the "empty-string-vs-absent" churn class (GS-EDIT-REVERT family): a
// stored node authored WITHOUT a string attr (alt/title/caption/aria-label/...)
// must not gain a phantom `attr: ""` after `markdownToProseMirror(convert…)`.
// Each spec sweeps the at-risk string attrs at DEFAULT (absent) and at a real
// NON-default value; the helper asserts both the RAW round-trip (attrs equal the
// input's, modulo the documented numeric width/height/size/aspectRatio -> string
// coercion) and the CANONICAL round-trip (canonical forms deep-equal).
//
// The image + media family share the `align !== "center"` predicate and the
// `<!--name {…}-->` comment machinery, so one matrix guards the shared class.
// align is NOT part of this class (it round-trips correctly) and is not swept.
// ---------------------------------------------------------------------------
const SPECS: NodeStabilitySpec[] = [
{
// Image carries the most at-risk string attrs. `alt` is the one marked
// materializes as `<img alt="">` on `![](src)` import (the real bug); title
// and caption are covered as the same class. attachmentId is a string attr
// that must stay absent when unset (control).
type: "image",
baseAttrs: { src: "/i.png" },
attrMatrix: [
{ attr: "alt", default: undefined, nonDefault: "a real alt text", emptyStringClass: true },
{ attr: "title", default: undefined, nonDefault: "a real title", emptyStringClass: true },
{ attr: "caption", default: undefined, nonDefault: "a real caption" },
{ attr: "attachmentId", default: undefined, nonDefault: "att-42" },
],
},
{
// Video's `alt` rides the `aria-label` attribute (media aria-label at risk).
type: "video",
baseAttrs: { src: "/v.mp4" },
attrMatrix: [
{ attr: "alt", default: undefined, nonDefault: "a clip", emptyStringClass: true },
{ attr: "attachmentId", default: undefined, nonDefault: "att-1" },
],
},
{
// Audio carries no alt/title; attachmentId is its only optional string attr.
type: "audio",
baseAttrs: { src: "/a.mp3" },
attrMatrix: [
{ attr: "attachmentId", default: undefined, nonDefault: "att-2" },
],
},
{
// pdf: link-form media. `name` (filename) is its at-risk string attr.
type: "pdf",
baseAttrs: { src: "/d.pdf" },
attrMatrix: [
{ attr: "name", default: undefined, nonDefault: "report.pdf", emptyStringClass: true },
{ attr: "attachmentId", default: undefined, nonDefault: "att-3" },
],
},
{
// attachment: link-form media (file card). `name` + `mime` string attrs.
type: "attachment",
baseAttrs: { url: "/f.zip" },
attrMatrix: [
{ attr: "name", default: undefined, nonDefault: "bundle.zip", emptyStringClass: true },
{ attr: "mime", default: undefined, nonDefault: "application/zip", emptyStringClass: true },
{ attr: "attachmentId", default: undefined, nonDefault: "att-4" },
],
},
{
// embed: link-form media. `provider` is its at-risk string attr (schema
// default ""). embed's numeric width/height defaults (800/600) are a SEPARATE,
// documented limitation OUTSIDE the empty-string class: they are not in
// canonicalize's KNOWN_DEFAULTS, so an ABSENT width/height re-imports as the
// 800/600 default and diverges canonically (see the note in canonicalize.ts).
// That is canonicalize-owned and out of scope here, so we author the
// dimensions at their defaults (as real editor embeds carry them) to keep this
// guard focused on the empty-string/provider class.
// provider's schema default is "" (NOT null), so a re-imported "" is the
// correct value, not a phantom — it is outside the null-default empty-string
// class. We author it at its "" default (the default pick) so the sweep still
// asserts a non-default provider ("youtube") round-trips, without tripping the
// canonicalize KNOWN_DEFAULTS gap for embed's non-null defaults.
type: "embed",
baseAttrs: { src: "https://example.com/x", width: 800, height: 600 },
attrMatrix: [
{ attr: "provider", default: "", nonDefault: "youtube" },
],
},
{
// drawio: image-form diagram. `title` + `alt` string attrs (data-title/-alt).
type: "drawio",
baseAttrs: { src: "blob:drawio" },
attrMatrix: [
{ attr: "title", default: undefined, nonDefault: "flow chart", emptyStringClass: true },
{ attr: "alt", default: undefined, nonDefault: "an alt", emptyStringClass: true },
{ attr: "attachmentId", default: undefined, nonDefault: "att-5" },
],
},
{
// excalidraw: image-form diagram, same shared diagramAttributes set.
type: "excalidraw",
baseAttrs: { src: "blob:excalidraw" },
attrMatrix: [
{ attr: "title", default: undefined, nonDefault: "sketch", emptyStringClass: true },
{ attr: "alt", default: undefined, nonDefault: "an alt", emptyStringClass: true },
{ attr: "attachmentId", default: undefined, nonDefault: "att-6" },
],
},
];
describe("round-trip stability matrix (image + media family)", () => {
for (const spec of SPECS) {
it(`${spec.type}: no attr materializes an empty-string / phantom value`, async () => {
const report = await runStabilityMatrix(spec);
const unstable = unstableCombos(report);
// On failure, print the WHOLE matrix so which (attr, value) combos are
// unstable is legible.
expect(unstable, `\n${formatReport(report)}\n`).toEqual([]);
});
}
});
// ---------------------------------------------------------------------------
// THIRD STATE: an attr EXPLICITLY stored as a literal "" (GS-EDIT-REVERT: a user
// typed alt/title/name/... then deleted it, so Tiptap persisted `attr: ""` — a
// value DISTINCT from "attr was never set"). Unlike the absent case above, this
// state is NOT first-pass byte-stable: the fix's `"" -> default` coercion is a
// deliberate ONE-TIME normalization on the FIRST sync round-trip, stable
// thereafter. We therefore assert a DIFFERENT contract — "converges to default
// on pass 1, then idempotent from pass 2 on" — for every empty-string-class attr
// across the whole node family (image/video/pdf/attachment/drawio/excalidraw).
//
// IMPORTANT for a future sync/QA pass: the pass-1 `"" -> null` diff is the
// converter canon, not corruption. It appears at most once per affected node and
// must NOT be flagged as "the converter is losing/corrupting page data".
// ---------------------------------------------------------------------------
describe("round-trip third state: explicit empty string converges once, then idempotent", () => {
for (const spec of SPECS) {
for (const attr of convergenceCasesFor(spec)) {
it(`${spec.type}.${attr}: "" normalizes to default on pass 1, byte-stable from pass 2`, async () => {
const r = await runConvergenceCase(spec, attr);
// Pass 1 must converge "" -> the schema default (the one-time diff) and
// pass 2 (roundtrip of pass-1 output) must be byte-stable. formatConvergence
// prints exactly which half failed.
expect(convergenceOk(r), `\n${formatConvergence(r)}\n`).toBe(true);
// Spell the contract out explicitly so the intent is legible in the test:
expect(r.convergedToDefault, `\n${formatConvergence(r)}\n`).toBe(true);
expect(r.firstPassValue).toEqual(r.expectedDefault);
expect(r.secondPassDivergence, `\n${formatConvergence(r)}\n`).toBeNull();
});
}
}
});
+1 -4
View File
@@ -269,9 +269,6 @@ importers:
'@atlaskit/pragmatic-drag-and-drop-live-region':
specifier: 1.3.4
version: 1.3.4
'@braintree/sanitize-url':
specifier: 7.1.2
version: 7.1.2
'@casl/react':
specifier: 5.0.1
version: 5.0.1(@casl/ability@6.8.0)(react@18.3.1)
@@ -16156,7 +16153,7 @@ snapshots:
obug: 2.1.1
std-env: 4.1.0
tinyrainbow: 3.1.0
vitest: 4.1.6(@opentelemetry/api@1.9.0)(@types/node@22.19.1)(@vitest/coverage-v8@4.1.6)(happy-dom@20.8.9)(jsdom@25.0.0)(vite@8.0.5(@types/node@22.19.1)(esbuild@0.28.0)(jiti@2.4.2)(less@4.2.0)(sugarss@5.0.1(postcss@8.5.14))(terser@5.39.0)(tsx@4.21.0)(yaml@2.8.3))
vitest: 4.1.6(@opentelemetry/api@1.9.0)(@types/node@25.5.0)(@vitest/coverage-v8@4.1.6)(happy-dom@20.8.9)(jsdom@27.4.0(@noble/hashes@2.0.1))(vite@8.0.5(@types/node@25.5.0)(esbuild@0.28.0)(jiti@2.4.2)(less@4.2.0)(sugarss@5.0.1(postcss@8.5.14))(terser@5.39.0)(tsx@4.21.0)(yaml@2.8.3))
'@vitest/expect@4.1.6':
dependencies: