Since the properties adapter now depends on whenLoaded status of the ydoc, we may need to explicitly mark doc as emit if the workspace is created without using a provider.
A React wrapper for blocksuite editor from title/meta/doc/edgless fragments.
This PR only **clones** the `AffineEditorContainer`'s existing behavior and make it easier for extension in affine later.
fix TOV-315
### Some core issues:
A customized version of `createComponent` from `@lit/react`. The [existing and solutions in the community](https://github.com/lit/lit/issues/4435) does not work well in our case.
Alternatively in this PR the approach we have is to create web component instances in React lifecycle and then append them to DOM. However this make it hard to wrap the exported Lit component's using React and therefore we will have an additional wrapper tag for the wrapped web component.
To mitigate the migration issue on using React instead of Lit listed on last day, we now use [a proxy to mimic the wrapped React component](https://github.com/toeverything/AFFiNE/pull/5606/files#diff-5b7f0ae7b52a08739d50e78e9ec803c26ff3d3e5437581c692add0de12d3ede5R142-R183) into an `AffineEditorContainer` instance.
fix TOV-280
experimental features ui
- enabled in the workspace settings for a cloud workspace; only show for workspace owner + early access
- a disclaimer prompt will be shown before going to the next feature setting page
- for now only show the ai poc switch, which controls the display of the ai tab in editor's sidepanel
Use new `transformer` to import onboarding json templates.
The json files are generated via this gist
https://gist.github.com/pengx17/ef92c305ac23123803a1a6a20e31f822
Not using the all-in-one `ZipTransformer` to import onboarding via a zip file.
1. The main concerns is that we still need to serve the blob resources via CDN to reduce user's blob usage. Otherwise the user will get the onboarding images being uploaded to cloud server every time he creates a new workspace. In this PR we extracted parts of the code from `ZipTransformer` in blocksuite and mute some code for uploading blobs.
2. it maybe not necessary to use zip for loading snapshots.
This PR is a short term solution. whether or not to tune the transformer api design may need further discussions.
fix TOV-264
Before this change, when the user gets to an empty collection page & hide the sidebar, there is no sidebar toggle any longer.
Also added windows app control on windows.
Rewrite sidebar panel using a customized react-resizable-panels version that supports sidebar pixel sizing (not using flex percentages).
Now the left & right sidebar using the same `ResizePanel` impl.
fix https://github.com/toeverything/AFFiNE/issues/5271
fix TOV-163
fix TOV-146
fix TOV-168
fix TOV-109
fix TOV-165
Add color-scheme to html so that the web will use system scheme when scripts has not being loaded;
Use `(prefers-color-scheme: dark)` so that the css vars do not defer load based on `data-theme`, which is set by next-themes (which is deferred).
- moved `appSettingAtom` to infra since we now have different packages that depends on it. There is no better place to fit in for now
- use atomEffect to sync setting changes to updater related configs to Electron side
- refactored how Electron reacts to updater config changes.
Refactor AFFiNE layout to support new right sidebar.
The new layout:
**Highlights:**
- new sidebar UI/UX
- favoring top-down UI components that are composed by basic building blocks in each route, instead of creating universal component like `WorkspaceHeader` that renders every possible cases (which I think is really hard to maintain)
- remove plugin based solution
**Pros/cons for current plugin-based solution:**
The current solution is somewhat a Dependency Injection (DI) approach, where the layout is defined at the top and UI items can be injected using Jotai atom slots.
This approach works well if we want a fully configurable system with everything being handled by plugins. It provides flexibility for custom extensions.
However, this solution is more suitable for single-page applications where the UI is completely controlled by configuration. It becomes challenging to achieve an optimized and visually appealing UI that remains under our control. An example of such a scenario would be a customizable dashboard like Grafana.
Another drawback of the existing solution is that we need to use Jotai and hooks to access context values, resulting in an unclear data flow within the component hierarchy.
**Alternatively, our approach in this PR** provides layout building blocks such as headers and sidebars, which can then be composed in individual route components. The good is that we have cleaner biz component instead of vague all-in-one layout component (like `<WorkspaceHeader />`).
**Issues of the implementation in this PR:**
Some UI layouts that that seems to be defined at the root layout are now defined in individual route component instead.
New 3-col layout component like the right sidebar still needs some abstraction and they are right now just for the detail editor only.
`react-resizable-panels` will throw some errors sometime when showing history modal dialog.
I haven't checked the root cause, but upgrade it to the latest will get rid of the error.
At the moment on desktop the user's local blob data will be lost after enable cloud.
This is because blob data is only synced from old idb to new idb, but not sync into sqlitedb.
This pr will simply clone the db file for desktop app. It should also speed up the time when enabling cloud for a large local workspace.
Due to restrictions on how Electron package works, the `node_modules` should not be hoisted and not to use s/h-links at all. This is why we need to have two separate installs for electron and non-electron packages in the build.
Tested via the following script
```bash
#!/bin/bash
echo "step 1: clean up"
find . -name "node_modules" -prune -exec rm -rf '{}' +
# git clean -dfX
build_type=canary
echo "step 2: install web dependencies"
# firstly, build web static
PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1 SENTRYCLI_SKIP_DOWNLOAD=1 PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1 HUSKY=1 yarn
echo "step 3: generate assets"
BUILD_TYPE="$build_type" yarn workspace @affine/electron generate-assets
# cleanup node_modules
find . -name "node_modules" -prune -exec rm -rf '{}' +
echo "step 4: install electron dependencies"
# install electron deps
yarn config set nmHoistingLimits workspaces
yarn config set enableScripts false
yarn config set nmMode classic
PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1 HUSKY=0 yarn workspaces focus @affine/electron @affine/monorepo
echo "step 5: build native"
# build native
yarn workspace @affine/native build
yarn workspace @affine/storage build
echo "step 6: build electron"
# build electron
yarn workspace @affine/electron build
echo "step 7: package electron"
# package
SKIP_GENERATE_ASSETS=1 BUILD_TYPE="$build_type" HOIST_NODE_MODULES=1 yarn workspace @affine/electron package
```
I suspect HMR does not working properly on dev because we have multiple entries.
One relative issue: https://github.com/webpack/webpack-dev-server/issues/2792/
I think we do not need multiple entries for polyfills & plugins after all. They could be in the same chunk, and could be later optimized through splitChunks option.
`ses.ts` is changed to `ses-lockdown.ts` because `ses.ts` does not pass circular dependency check by madge. I haven't looked through the real root cause though. See https://github.com/pahen/madge/issues/355
It seems there are some cases that [this upstream PR](https://github.com/toeverything/blocksuite/pull/4747) will cause data loss.
Because of some historical reasons, the page id could be different with its doc id.
It might be caused by subdoc migration in the following (not 100% sure if all white screen issue is caused by it) https://github.com/toeverything/AFFiNE/blob/0714c12703ff8840acc180633e5489c755ddfba3/packages/common/infra/src/blocksuite/index.ts#L538-L540
In version 0.10, page id in spaces no longer has prefix "space:"
The data flow for fetching a doc's updates is:
- page id in `meta.pages` -> find `${page-id}` in `doc.spaces` -> `doc` -> `doc.guid`
if `doc` is not found in `doc.spaces`, a new doc will be created and its `doc.guid` is the same with its pageId
- because of guid logic change, the doc that previously prefixed with `space:` will not be found in `doc.spaces`
- when fetching the rows of this doc using the doc id === page id,
it will return EMPTY since there is no updates associated with the page id
The provided fix in the PR will patch the `spaces` field of the root doc so that after 0.10 the page doc can still be found in the `spaces` map. It shall apply to both of the idb & sqlite datasources.
Special thanks to @lawvs 's db file for investigation!
Peng Xiao
