refs https://github.com/TryGhost/Product/issues/4209
- bumped Koenig packages
- `koenig-lexical` added nested editor TK support
- all packages dropped Node 16 support
- switched to using `isTKEnabled` prop and `<TKCountPlugin>`
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
fixes PROD-61
This adds a new default plan setting. It defaults to yearly, which is
the current default selected interval in Portal.
Behind the new portal improvements feature flag, the default plan can be
changed. It will also change automatically if the available intervals
are changed.
This PR also wires up passing the new setting to the Portal preview.
Refs https://github.com/TryGhost/Product/issues/4218
- Moving the TK confirmation step into a small modal as a more
lightweight solution
---------
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
refs https://ghost.slack.com/archives/CTH5NDJMS/p1701688836406919
Deleting members with email disabled, results in deleting all members
due to a broken NQL filter.
The filter `(email_disabled:1)` results in selecting all members because
of the surrounding brackets, which cause a `yg` filter to be generated
by NQL which is not supported by code that handles the Mongo filters.
This is a quick fix to reduce damage, this will need a proper fix in NQL
/ lower level.
no issue
- bumped `@tryghost/koenig-lexical` to version that no longer uses negative lookbehind in a regex which wasn't supported in Safari until version 16.4
no issue
- bumped `@tryghost/koenig-lexical` to version that no longer uses negative lookbehind in a regex which wasn't supported in Safari until version 16.4
refs TryGhost/Product#4175
- Added error handling to Sentry's beforeSend function in both Admin and
Core, so if there is any error in beforeSend, we will still send the
unmodified event to Sentry
- This is in response to an incident yesterday wherein the beforeSend
function threw an error due to an unexpected missing value in the
exception. The event sent to Sentry was the error in the beforeSend
function, and the original error never reached Sentry.
- If the original event had reached Sentry, even if unmodified by the
logic in beforeSend, we could have been alerted to the issue sooner and
more easily identified all affected sites.
- Also added defensive logic to protect for certain values in the
exception passed to beforeSend not existing and added unit tests for the
beforeSend function in admin and core
no refs
This reverts commit 329488139a.
- updates for mobile result in the locator queries finding two elements
instead of one, causing tests to fail
- reverting for now until we take a look closer with mobile, as it seems
CI is running at a mobile size
refs https://github.com/TryGhost/Product/issues/4184
- set up property on the editor controller for tracking number of TKs, action for updating it, and reset mechanism to ensure we go back to 0 when switching post
- uses random number for now pending `<TkPlugin>` being updated to expose the TK count
- passed TK count data to the publish flow modal so it can show a reminder step before the publish options step when there are still TKs in the post content
- added `onCountChange` prop to `<TkPlugin>` ready for the count feature to be implemented
refs https://github.com/TryGhost/Ghost/pull/19057
- the flag and toggle UI had been added but we were missing the final part of wiring up the feature in Admin and passthrough to the editor
fixes GRO-34
fixes GRO-33
This is a revision of a previous commit, that broke the browser tests
because changes in the data generator (requiring bookshelf had side
effects).
This adds a new way to run all tests with enforced numeric ObjectIDs.
These numeric ids cause issues if they are used withing NQL filters. So
they surface tiny bugs in our codebase.
You can run tests using this option via:
NUMERIC_IDS=1 yarn test:e2e
Removed some defensive logic that could be explained by this discovered
issue.
refs https://github.com/TryGhost/Product/issues/4152
---
<!-- Leave the line below if you'd like GitHub Copilot to generate a
summary from your commit -->
<!--
copilot:summary
-->
### <samp>🤖[[deprecated]](https://githubnext.com/copilot-for-prs-sunset)
Generated by Copilot at a28462f</samp>
This pull request adds a new admin-x app called `admin-x-demo`, which
demonstrates how to use the shared packages `admin-x-framework` and
`admin-x-design-system` to create a simple app that renders a button and
a modal. It also improves the development workflow, the vite
integration, the dependency management, and the type checking for the
admin-x apps and packages. It modifies some files in the
`admin-x-framework` and `admin-x-design-system` packages to make the
modals prop optional, to introduce a new type for the props from the
Ember app, to fix the z-index of the modal backdrop, and to use
consistent file extensions and module syntax.
fixes GRO-34
fixes GRO-33
This also adds a new way to run all tests with enforced numeric ObjectIDs.
These numeric ids cause issues if they are used withing NQL filters. So they
surface tiny bugs in our codebase.
You can run tests using this option via:
NUMERIC_IDS=1 yarn test:e2e
Also removed some defensive logic that could be explained by unquoted ids.
no issue
- The standard error message from Ember Data includes post/page ids in
the error message
- This causes Sentry to treat each instance as a unique issue and
results in many duplicate issues for the same error
- This change should mask the ids and allow Sentry to group the errors
correctly
no issue
- Fixed an issue where the ajax_url tag was exceeding the allowed length
of tags in Sentry
- Fixed the mechanism for deleting the ajax tags when the error is not
an ajax error
- Removed the isAjaxError tag, since we can use the other ajax tags to
filter for ajax errors now
refs TryGhost/Product#4160
- updated header to be ctrl+option/alt+1-5 for header to avoid conflict
with os behaviour
- updated strike to be ctrl+option/alt+u to avoid view source browser
behaviour
No ref
- In order to make it easier to distinguish between Admin and site in
browser tabs, we've updated the meta title to include `Ghost Admin` in
the title.
fixes GRO-25
Updated @tryghost/nql to 0.12.0 and other packages that depend on it
1. SQLite: when a filter string contains /.
When we use a NQL contain/starts/endsWith filter that contains a slash,
underlyingly the whole filter will get converted to a MongoDB query, in
which we just use a regexp to represent the filter. In here we will
escape the slash: \/ as expected in a regexp. Later when we convert this
MongoDB query back to knex/SQL, we use a SQL LIKE query. Currently we
don't remove the escaping here for a normal slash. MySQL seems to ignore
this (kinda incorrect). SQLite doesn't like it, and this breaks queries
on SQLite that use slashes. The solution here is simple: remove the
backslash escaping when converting the regexp to LIKE, just like we do
with other special regexp characters.
2. We don't escape % and _, which have a special meaning in LIKE queries
Usage of % and _ is now as expected and doesn't have the special SQL
meaning anymore.
refs
5f7c7a82dc
- The goal of the RewriteFrames plugin was to remove version identifiers
from the stack traces sent to Sentry, to improve the stack trace
grouping. That sort of worked, but it ended up breaking the sourcemaps
so the stack traces were ugly and hard to read within Sentry.
- This change removes the RewriteFrames plugin to fix the stack traces
again. Instead, planning to use Sentry's stack trace rules built into
the UI to fix the grouping.
fixes https://github.com/TryGhost/Product/issues/3738https://www.notion.so/ghost/Member-Session-Invalidation-13254316f2244c34bcbc65c101eb5cc4
- Adds the transient_id column to the members table. This defaults to
email, to keep it backwards compatible (not logging out all existing
sessions)
- Instead of using the email in the cookies, we now use the transient_id
- Updating the transient_id means invalidating all sessions of a member
- Adds an endpoint to the admin api to log out a member from all devices
- Added the `all` body property to the DELETE session endpoint in the
members API. Setting it to true will sign a member out from all devices.
- Adds a UI button in Admin to sign a member out from all devices
- Portal 'sign out of all devices' will not be added for now
Related changes (added because these areas were affected by the code
changes):
- Adds a serializer to member events / activity feed endpoints - all
member fields were returned here, so the transient_id would also be
returned - which is not needed and bloats the API response size
(`transient_id` is not a secret because the cookies are signed)
- Removed `loadMemberSession` from public settings browse (not used
anymore + bad pattern)
Performance tests on site with 50.000 members (on Macbook M1 Pro):
- Migrate: 6s (adding column 4s, setting to email is 1s, dropping
nullable: 1s)
- Rollback: 2s
closes https://github.com/TryGhost/Product/issues/4133
- we were creating a new JSDOM instance every time we rendered a card which lowered performance because JSDOM instantiation is heavy
- updated Koenig packages to remove the need for passing in an external `createDocument` option method as they now re-use the renderer's internal single instance of JSDOM
no issue
- Currently our stack traces in Production include the admin build
version in the paths, e.g. `/admin/1633/assets` instead of
`admin/assets`
- This confuses the error grouping logic in Sentry, resulting in many
duplicate issues being created every time we release a new version of
admin
- Ultimately, this makes it really difficult to determine if a 'New'
issue in Sentry is actually new, or if it's just the first time we've
seen it in this release.
- This commit adds the `RewriteFrames` integration to the Admin Sentry
client, which will strip the build version from the paths in the stack
traces, and allow Sentry to group issues correctly.
- With this, hopefully we will have far fewer 'New' issues created, so
we can again start alerting on the 'New' condition in Sentry.
fixes https://github.com/TryGhost/Product/issues/4108
- Updates filters behind a new alpha feature flag so you can also filter
on members who have email disabled (because the email had a permanent
bounce, they reported spam or the email address is invalid)
- When returning members, we now also use the email_disabled flag to set
email_suppression.suppressed correctly (in case they are out of sync,
which should normally never happen).
closes https://github.com/TryGhost/Ghost/issues/18949
- added check for `isComposing` in Enter key handler to avoid the move-to-post-body behaviour when the Enter keypress is used to end IME composition
---------
Co-authored-by: t8m8 <tomomasa.matsunaga@gmail.com>
no issue
- The fingerprinting on chunk files was happening twice (once by ember
and once by webpack), resulting in the .js file and the .map file not
matching
- This change prevents ember from fingerprinting the chunk.*.map files,
so the resulting .map and .js files will have the same basename
- No real functional difference here, just a bit easier to find the
corresponding .map file for a given .js file
no issue
- ember-cli-terser 4.0.2 apparently has a regression that breaks the
sourcemap generation for the admin ember app
- this reverts the package to 4.0.1, which fixes the sourcemaps and
should generate much more readable stack traces in Sentry
- Validating the sourcemaps locally succeeded, but will need to test
this on staging to confirm everything is working properly in CI and with
the CDN.
no issue
- we don't use `<i>` elements anywhere in our own code and this styling was causing odd in-editor previews of HTML cards when their content contained `<i>`
refs https://github.com/TryGhost/Product/issues/4055
---
<!-- Leave the line below if you'd like GitHub Copilot to generate a
summary from your commit -->
<!--
copilot:summary
-->
### <samp>🤖 Generated by Copilot at 3b24693</samp>
Removed unused components related to announcement settings, custom theme
settings, and file upload. These components were part of a deprecated
feature or a legacy UI.
no issue
- Somewhere along the process of building the admin app, the sourcemaps
are getting corrupted
- This commit is to test the theory that babel is the source of the
corruption, because it isn't generating sourcemaps, so we are missing a
step in the process
no issue
- The sourcemaps currently generated by the admin build do not include
the `sourcesContent` key
- This commit is to experiment with this change, in particular to see if
it improves the stacktraces available in Sentry from errors in the admin
app
- Currently the stacktraces in admin show the minified code — hopefully
this change will make the stacktraces show the original source code to
make it easier to debug issues from within Sentry
- Might thoroughly break the sourcemaps, in which case I will revert this
no issues
- The `path` tags added to Sentry events from admin aren't super useful, since they often include e.g. a
post id in the path, which isn't super useful for filtering/grouping
- We already have a `route` tag which can be used to filter to a
particular route, which renders the `path` tags unnecessary
no issue
- Follow up to
a382cd8a91
and
fb34e285fc
- The `sourcemap-postprocess` addon was ineffective because the assets
were being copied to `/ghost/core/core/built` before the addon was run,
so the changes weren't being applied in staging/production
- This change makes the required modifications to the sourcemaps before
copying them
refs https://github.com/TryGhost/Ghost/pull/18816
When opening a new post in the editor and immediately leaving we would hang because we attempted to save a revision rather than just throwing the blank post away.
- added additional `hadDirtyAttributes` check to the forced-revision save when leaving
- added additional destroyed post checks to the save task to avoid saving a deleted post when leaving the editor
- removed unnecessary waits for save tasks that prevented leaving the editor when the title on a new post had been filled in but not saved (matches previous editor behaviour)
no issue
- Follow up to https://github.com/TryGhost/Ghost/pull/18825
- Adding the sourceRoot key didn't seem to work, so I'm just removing
the `assets/` prefix from all the sources to hopefully correct the issue
no issues
- Added isAjaxError tag to all errors sent from admin with a true/false
value
- Removed all Ajax context and tags if the error is not an Ajax error
- Updated names of Ajax tags to use camelcase instead of dot notation
closes https://github.com/TryGhost/Koenig/pull/1038
- updated `<KoenigLexicalEditorInput>` to load `<EmojiPickerPlugin>` and compose it into the editor unless used with `<KoenigLexicalEditorInput @emojiPicker={{false}} />`
- bumped Koenig packages so `EmojiPickerPlugin` is available to import from `koenig-lexical`
no issue
- Currently AJAX errors that surface in Admin get logged to Sentry with
super generic messages like "Error"
- This commit adds more context to the information that's sent to sentry
and should make it easier to identify the error at a glance
no issue
- The sourcemaps generated for the admin app use relative links to the
source code files e.g. `assets/ghost.js`
- Since the sourcemaps themselves are hosted at `/assets` already, this
was leading to issues with sourcemaps in Sentry and in the browser
looking for the sources at `/assets/assets/ghost.js`
- This commit adds a `sourceRoot` key to the sourcemaps, which should
allow Sentry and the Browser to find the source code files at
`../assets/ghost.js` instead of `assets/ghost.js`
- We may need to iterate on this — not 100% sure if this is the best way
to do this without trying it in staging. If the `../` doesn't work in
all environments, we can try including the CDN url directly
closes https://github.com/TryGhost/Product/issues/4059
- modified `saveTask` so when it has the `leavingEditor` option it doesn't attempt to set a new post status
- save when leaving the editor is never a publish/unpublish-related event, rather it's a convenience autosave and a forced revision creation so modifying the post status should never happen for those saves
- updated the `willTransition` handling to avoid repeated saves
- sets a property on the controller whenever we attempt a save+transition retry on leaving the editor then skips any further attempted saves on transitions
- although it would be preferable we can't use `try/catch` on these saves because our save task always catches and doesn't re-throw, adjusting that would be a much larger change than we want to make for this fix
no refs
As part of the reduce Sentry noise work we decided to temporarily
prevent the submission of "handled" errors to Sentry. This is a
temporary measure to reduce noise while we work on improving the quality
of the error reporting / alerting
no issue
- All events sent from Admin should now be tagged with the current route
(e.g. `lexical-editor.edit`) and the path (e.g.
`editor/post/65388d98734d6ecc7bbafb87?test`)
- This will allow us to filter events by route and path in Sentry and
improve our alerting logic
no issue
- This change enables the Debug integration in Sentry when the app is
running in development mode (and Sentry is configured with a valid DSN).
- Whenever an event is sent to Sentry, the event & hint are passed to
the beforeSend() function will be logged to the browser console so you
can quickly see a) when events are sent to Sentry and b) exactly what
data & tags are sent with integration
no issue
- Unfortunately the RewriteFrames plugin was not successful in fixing
the duplicated `/assets/assets` in the Sentry stacktraces for admin
- This commit removes the RewriteFrames plugin and reverts the changes
so there will at least be some kind of source code reaching Sentry
no issue
- For whatever reason, our Sentry stack traces are adding an extra
`/assets` to the filenames in stacktraces from admin
- Attempted to fix this
[here](3e9ee16ffb)
but it didn't work because at this point the duplication hasn't occurred
yet
- This change strips the `/assets` segment from the filenames, so when
Sentry adds it back in, it doesn't duplicate anything. Would be good to
get to the bottom of why this is happening, but this should hopefully
improve our stacktraces in Sentry in the meantime.
no issue
- Our stack traces in Sentry are pointing to the wrong location for source code files. We attempted to fix this [here](https://github.com/TryGhost/Ghost/pull/18773) but this didn't work as expected in staging for some reason
- This commit just adds a few console logs to help with debugging in staging so we can figure out what's going wrong
fixes https://github.com/TryGhost/DevOps/issues/94
- we've seen in production that some filepaths contain a duplicate
assets folder, but we're not sure why
- instead of spending too much time trying to figure that out, we can
just fix that in preprocessing by using the RewriteFrames integration
in Sentry
- this should remove the duplicate folder from the stacktrace frame, if it exists
refs TryGhost/Product#4064
- In Sentry we were seeing the error "Discarded unknown attribute 'user.role'" for most events sent from admin
- This small change removes this error in Sentry and restores the user.role tag for error events in admin
no issue
- updates `@tryghost/koenig-lexical` with indent improvements
- prevents indents on paragraphs/headings etc which are not supported when rendering for front-end display
- improves indent behaviour for lists so <kbd>Tab</kbd> can be pressed anywhere in a list item to indent rather than only at the beginning of the list
closes https://github.com/TryGhost/Product/issues/4037
- bumps `@tryghost/kg-html-to-lexical` that includes better node normalization to handle `<br>` in top-level text nodes and after a nested block-level element
- also includes fix for button hrefs having `about:blank` prefixed to hash URLs
