In the search modal, under the sub docs,
we now display the top parent of the doc, to give
more context to the user about the doc they are
looking for.
We refactorize the filters to get more flexibility
and avoid too much props drilling.
In the search endpoint, we were returning all the parents for a document
mathing the search. We want instead to return only the top parent the
user has access to.
In the content_retrieve action, if the document is not existing on the
object storage, we return an empty binary string. In the case the app is
ran as an ASGI application we still have a warning explaining it must
consume the iterator before sending it.
The tests are using the local object storage configured in the docker
compose stack. Every time a test using the DocumentFactory is executed,
an object is created in the object storage and will stay here but never
used. The data/media directory is growing and time to time we have to
delete it in order to free some gigabytes. In this commit we add a
global conftest responsible to delete the object created during the test
execution. The space will be really free when the object storage
container will be shutdown.
Changed:
- 💄(frontend) display emoji button on hover
Fixed:
- 🐛(backend) close thread DB connections to fix test teardown
OperationalError
- 🐛(frontend) fix crash when orphaned threads
- 🐛(backend) order trashbin response by most recently deleted
- 🐛(backend) stream document content with an async
iterator under ASGI
- 🐛(frontend) fix long titles in table of content
When docs is ran using an ASGI server, the django StreamingHttpResponse
expect to consume an async generator. If a sync generator is used, it
fully consumes it and then return all the content and we loose the
benefits of using a StreamingHttpResponse.
The app expect threads to always have
comments, but in some cases threads can end up without
comments, which causes crashes. This PR adds a
check to delete orphaned threads without comments
to avoid this issue.
When all the tests are ran, there is a PytestWarning log present to warn
us that there is remaining database connection open and can not be
closed. It appears that the tests about concurrence are responsible of
leaving connection open. We need to manually close them in these tests.
Added:
- ✨(backend) support creating subdoc from file
- ✨(frontend) comment side panel
- ✨(buildpack) add PaaS deployment support, tested with Scalingo
- 🔧(backend) allow configuring settings
OIDC_OP_USER_ENDPOINT_FORMAT
- ⚡️(helm) create a dedicated svc and deployment for
yprovider converter
- ✨(backend) allow to leave a document
- ✨(frontend) add the presenter mode
- 📈(backend) create a utils to capture event with posthog
- 🔧(backend) new setting DOCUMENT_ALL_ENDPOINT_ENABLED
Changed:
- ♻️(backend) allow global search in sub documents
- ✨(backend) add a breadcrumb in the search response
- ♻️(frontend) move doc action buttons to fix toolbar
- ♿️(frontend) add aria-hidden to decorative avatar SVGs
in share modal
- 🏗️(frontend) move comments to its own folder feature
Fixed:
- 🐛(docs) run migration 0027 without superuser role
- 🐛(backend) prevent admins/owners from overwriting other
users comments
- 🐛(y-provider) return empty output when converting empty
Yjs document
- 🐛(backend) use computed_link_reach in
handle_onboarding_document
- 🐛(frontend) fix toolbar blocknote hidden
- 🐛(frontend) fix application crashes when using
GTranslate and zoom
- 🐛(frontend) fix emoji pdf not matching
- 🐛(backend) fix UnorderedObjectListWarning for
DocumentAskForAccess viewset
In the tests there is a UnorderedObjectListWarning, we also find this same
warning in the logs in production. This warning appears because the
queryset used by the DocumentAskForAccess paginator is not ordered and
can lead to inconsistent result. Ordering the queryset fix this warning.
We want to allow or not the usage of the /documents/all/ endpoint. It has
been created to be used for an other purpose than the js client. For
those who don't use it this new settings allow to disable it. By default
it is set to True to keep the same behavior.
Add a Playwright spec that opens the presenter overlay from the
doc options menu, walks through a multi-slide document built with
dividers, and verifies that Escape closes the overlay.
Cover the three hooks that drive the presenter overlay:
useSlides for divider-based block segmentation,
useBrowserFullscreen for the Fullscreen API wrapper,
and usePresenterShortcuts for the keyboard navigation bindings.
Add a presenter overlay that turns the current document into a
slide deck. The editor's blocks are snapshot at open time and
split into slides on each divider; navigation is driven by
keyboard shortcuts and a floating bar with browser fullscreen
support. The overlay is wired to the doc header toolbox via a
new "Present" entry, lazy-loaded to keep the editor bundle lean.
cunningham-react 4.3.0 splits Modal props per variant. Switch
AlertModal and SideModal from the now-removed ModalProps alias to
the more precise ModalDefaultVariantProps type so the modal
wrappers keep type-checking against the new API.
We bumped the Blocknote editor to version 0.51.4,
which includes several bug fixes and improvements,
particularly 1 bug fix about emoji on the side bar.
The feature "doc-editor" start to be quiete big,
and the comments related code is quite a lot.
We move all the comments related code to a new
folder "doc-comments" to make the code more
organized and easier to maintain.
The header could hide the formatting bar of the
blocknote editor. This commit improve the z-index
of the header to ensure that the formatting bar
is always visible.
Error boundary page is used when an error occurs
in the application, this page can have states still
set, causing side effects depend the component that
throw the error. In this case, the left panel is
still loaded and try to load the tree, but the
provider is not set, causing a crash error inside
the error boundary page.
We don't need the left panel in the error boundary
page, so we can just not load it to avoid this issue.
The emoji algorithm to find the emoji was not totally
correct, it could choose a bad family of emojis.
The first part of the emoji "code" is
what should be used first to find the "family" of the
emoji, we can then try to find the exact match in
that specific family.
When an AI action is called from ai_translate, ai_transform and
ai_proxy, an event doc_ai_action is captured with a method property
giving the information from where it is called.
Anthony LC