Fixes https://github.com/payloadcms/payload/issues/13433. Testing
release: `3.54.0-internal.90cf7d5`
Previously, when calling `getPayload`, you would always use the same,
cached payload instance within a single process, regardless of the
arguments passed to the `getPayload` function. This resulted in the
following issues - both are fixed by this PR:
- If, in your frontend you're calling `getPayload` without `cron: true`,
and you're hosting the Payload Admin Panel in the same process, crons
will not be enabled even if you visit the admin panel which calls
`getPayload` with `cron: true`. This will break jobs autorun depending
on which page you visit first - admin panel or frontend
- Within the same process, you are unable to use `getPayload` twice for
different instances of payload with different Payload Configs.
On postgres, you can get around this by manually calling new
`BasePayload()` which skips the cache. This did not work on mongoose
though, as mongoose was caching the models on a global singleton (this
PR addresses this).
In order to bust the cache for different Payload Config, this PR
introduces a new, optional `key` property to `getPayload`.
## Mongoose - disable using global singleton
This PR refactors the Payload Mongoose adapter to stop relying on the
global mongoose singleton. Instead, each adapter instance now creates
and manages its own scoped Connection object.
### Motivation
Previously, calling `getPayload()` more than once in the same process
would throw `Cannot overwrite model` errors because models were compiled
into the global singleton. This prevented running multiple Payload
instances side-by-side, even when pointing at different databases.
### Changes
- Replace usage of `mongoose.connect()` / `mongoose.model()` with
instance-scoped `createConnection()` and `connection.model()`.
- Ensure models, globals, and versions are compiled per connection, not
globally.
- Added proper `close()` handling on `this.connection` instead of
`mongoose.disconnect()`.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211114366468745
### What?
Brackets (`[ ]`) in option values end up in GraphQL enum names via
`formatName`, causing schema generation to fail. This PR adds a single
rule to `formatName`:
- replace `[` and `]` with `_`
### Why?
Using `_` (instead of removing the brackets) is safer and more
consistent:
- Avoid collisions: removal can merge distinct strings (`"A[B]"` →
`"AB"`). `_` keeps them distinct (`"A_B"`).
- **Consistency**: `formatName` already maps punctuation to `_` (`. - /
+ , ( ) '`). Brackets follow the same rule.
Readability: `mb-[150px]` → `mb__150px_` is clearer than `mb150px`.
Digits/units safety: removal can jam characters (`w-[2/3]` → `w23`); `_`
avoids that (`w_2_3_`).
### How?
Update formatName to include a bracket replacement step:
```
.replace(/\[|\]/g, '_')
```
No other call sites or value semantics change; only names containing
brackets are affected.
Fixes#13466
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211141396953194
Fixes#13574.
When editing an autosave-enabled document within a document drawer, any
changes made will while the autosave is processing are ultimately
discarded from local form state. This makes is difficult or even
impossible to edit fields.
This is because we server-render, then replace, the entire document on
every save. This includes form state, which is stale because it was
rendered while new changes were still being made. We don't need to
re-render the entire view on every save, though, only on create. We
don't do this on the top-level edit view, for example. Instead, we only
need to replace form state.
This change is also a performance improvement because we are no longer
rendering all components unnecessarily, especially on every autosave
interval.
Before:
https://github.com/user-attachments/assets/e9c221bf-4800-4153-af55-8b82e93b3c26
After:
https://github.com/user-attachments/assets/d77ef2f3-b98b-41d6-ba6c-b502b9bb99cc
Note: ignore the flashing autosave status and doc controls. This is
horrible and we're actively fixing it, but is outside the scope of this
PR.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211139422639700
### What?
Ensure the Rich Text field is read-only when viewing a locked document.
### Why?
When opening a locked doc in read-only mode, the Rich Text field could
still appear editable. This is inconsistent with other fields and allows
users to try typing into a locked document.
### How?
- Pass `readOnly={isTrashedDoc || isLocked}` into `buildFormState` in
the edit view renderer.
### What?
Prevents the Auth component from rendering an empty `.auth-fields`
wrapper.
### Why?
When `disableLocalStrategy` is true and `enableFields` is false, but
`useAPIKey` is true while
read access to API key fields is denied, the component still rendered
the parent wrapper with a
background—showing a blank box.
### How?
Introduce `hasVisibleContent`:
- `showAuthBlock = enableFields`
- `showAPIKeyBlock = useAPIKey && canReadApiKey`
- `showVerifyBlock = verify && isEditing`
If none are true, return `null`. (`disableLocalStrategy` is already
accounted for via `enableFields`.)
Fixes#12089
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211117270523574
Continuation of https://github.com/payloadcms/payload/pull/13501.
When merging server form state with `acceptValues: true`, like on submit
(not autosave), rows are not deeply merged causing custom row
components, like row labels, to disappear. This is because we never
attach components to the form state response unless it has re-rendered
server-side, so unless we merge these rows with the current state, we
lose them.
Instead of allowing `acceptValues` to override all local changes to
rows, we need to flag any newly added rows with `addedByServer` so they
can bypass the merge strategy. Existing rows would continue to be merged
as expected, and new rows are simply appended to the end.
Discovered here:
https://discord.com/channels/967097582721572934/967097582721572937/1408367321797365840
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211115023863814
Previously, when manually setting `createdAt` or `updatedAt` in a
`payload.db.*` or `payload.*` operation, the value may have been
ignored. In some cases it was _impossible_ to change the `updatedAt`
value, even when using direct db adapter calls. On top of that, this
behavior sometimes differed between db adapters. For example, mongodb
did accept `updatedAt` when calling `payload.db.updateVersion` -
postgres ignored it.
This PR changes this behavior to consistently respect `createdAt` and
`updatedAt` values for `payload.db.*` operations.
For `payload.*` operations, this also works with the following
exception:
- update operations do no respect `updatedAt`, as updates are commonly
performed by spreading the old data, e.g. `payload.update({ data:
{...oldData} })` - in these cases, we usually still want the `updatedAt`
to be updated. If you need to get around this, you can use the
`payload.db.updateOne` operation instead.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210919646303994
Fixes#10526
### What?
`updateVersion` throws error at deleting before version of draft.
### Why?
When triggers `db.updateVersion` with non-mongodb environment,
`saveVersion` in `payload/src/versions/saveVersions.ts` doesn't remove
`id` in `versionData`.
### How?
Add `delete versionData.id` for non-mongodb environments.
---------
Co-authored-by: German Jablonski <43938777+GermanJablo@users.noreply.github.com>
### What?
Make the document `id` available to server-rendered admin components
that expect it—specifically `EditMenuItems` and
`BeforeDocumentControls`—so `props.id` matches the official docs.
### Why?
The docs show examples using `props.id`, but the runtime `serverProps`
and TS types didn’t include it. This led to `undefined` at render time.
### How?
- Add id to ServerProps and set it in renderDocumentSlots from
req.routeParams.id.
Fixes#13420
This PR fixes some incorrect field paths handling (=> should not pass
path and schema paths that contain index paths down to sub-fields
outside of the indexPath property) when building the version fields.
Fixes https://github.com/payloadcms/payload/issues/12376
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210932060696925
No need to re-fetch doc permissions during autosave. This will save us
from making two additional client-side requests on every autosave
interval, on top of the two existing requests needed to autosave and
refresh form state.
This _does_ mean that the UI will not fully reflect permissions again
until you fully save, or until you navigating back, but that has always
been the behavior anyway (until #13416). Maybe we can find another
solution for this in the future, or otherwise consider this to be
expected behavior.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211094073049052
If you have a beforeChange hook that manipulates arrays or blocks by
_adding rows_, the result of that hook will not be reflected in the UI
after save or autosave as you might expect.
For example, this hook that ensures at least one array row is populated:
```ts
{
type: 'array',
hooks: {
beforeChange: [
({ value }) =>
!value?.length
? [
// this is an added/computed row if attempt to save with no rows
]
: value,
],
},
// ...
}
```
When you save without any rows, this hook will have automatically
computed a row for you and saved it to the database. Form state will not
reflect this fact, however, until you refresh or navigate back.
This is for two reasons:
1. When merging server form state, we receive the new fields, but do not
receive the new rows. This is because the `acceptValues` flag only
applies to the `value` property of fields, but should also apply to the
`rows` property on `array` and `blocks` fields too.
2. When creating new form state on the server, the newly added rows are
not being flagged with `addedByServer`, and so never make it into form
state when it is merged in on the client. To do this we need to send the
previous form state to the server and set `renderAllFields` to false in
order receive this property as expected. Fixed by #13524.
Before:
https://github.com/user-attachments/assets/3ab07ef5-3afd-456f-a9a8-737909b75016
After:
https://github.com/user-attachments/assets/27ad1d83-9313-45a9-b44a-db1e64452a99
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211094073049042
### What?
The PR #12763 added significant improvements for third-party databases
that are compatible with the MongoDB API. While the original PR was
focused on Firestore, other databases like DocumentDB also benefit from
these compatibility features.
In particular, the aggregate JOIN strategy does not work on AWS
DocumentDB and thus needs to be disabled. The current PR aims to provide
this as a sensible default in the `compatibilityOptions` that are
provided by Payload out-of-the-box.
As a bonus, it also fixes a small typo from `compat(a)bility` to
`compat(i)bility`.
### Why?
Because our Payload instance, which is backed by AWS DocumentDB, crashes
upon trying to access any `join` field.
### How?
By adding the existing `useJoinAggregations` with value `false` to the
compatiblity layer. Individual developers can still choose to override
it in their own local config as needed.
Completes https://github.com/payloadcms/payload/pull/13513
That PR fixed it for the `admin` suite, but I had also encountered the
same issue in `live-preview`.
When searching for instances, I found others with the same pattern
within packages, which I took the opportunity to fix in case the same
error occurs in the future.
Adds various helpers to make it easier and more standard to manage array
fields within e2e tests. Retrofits existing tests to ensure consistent
interactions across the board, and also organizes existing blocks and
relationship field helpers in the same way.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211094073049046
### What?
Update `SelectFieldBaseClientProps` type so `value` accepts `string[]`
for `hasMany` selects
### Why?
Multi-selects currently error with “Type 'string[]' is not assignable to
type 'string'”.
### How?
- Change `value?: string` to `value?: string | string[]`
- Also adds additional multi select custom component to `admin` test
suite for testing
---------
Co-authored-by: Jarrod Flesch <jarrodmflesch@gmail.com>
### What?
This PR makes `filterFields` recurse into **fields with subfields**
(e.g., tabs, row, group, collapsible, array) so nested fields with
`admin.disableListColumn: true` (or hidden/disabled fields) are properly
excluded.
### Why?
Nested fields with `admin.disableListColumn: true` were still appearing
in the list view.
Example: a text field inside a `row` or `group` continued to show as a
column despite being marked `disableListColumn`.
### How?
- Call `filterFields` recursively for `tab.fields` and for any field
exposing a `fields` array.
Fixes#13496
### What?
This PR applies `mergeFieldStyles` to the `BlocksField` component,
ensuring that custom admin styles such as `width` are correctly
respected when Blocks fields are placed inside row layouts.
### Why?
Previously, Blocks fields did not inherit or apply their `admin.width`
(or other merged field styles). For example, when placing two Blocks
fields side by side inside a row with `width: '50%'`, the widths were
ignored, causing layout issues.
### How?
- Imported and used `mergeFieldStyles` within `BlocksField`.
- Applied the merged styles to the root `<div>` via the `style` prop,
consistent with how other field components (like `TextField`) handle
styles.
Fixes#13498
### What?
When exporting, if no `sort` parameter is set but a `groupBy` parameter
is present in the list-view query, the export will treat `groupBy` as
the SortBy field and default to ascending order.
Additionally, the SortOrder field in the export UI is now hidden when no
sort is present, reducing visual noise and preventing irrelevant order
selection.
### Why?
Previously, exports ignored `groupBy` entirely when no sort was set,
leading to unsorted output even if the list view was grouped. Also,
SortOrder was always shown, even when no sort field was selected, which
could be confusing. These changes ensure exports reflect the list view’s
grouping and keep the UI focused.
### How?
- Check for `groupBy` in the query only when `sort` is unset.
- If found, set SortBy to `groupBy` and SortOrder to ascending.
- Hide the SortOrder field when `sort` is not set.
- Leave sorting unset if neither `sort` nor `groupBy` are present.
Currently, if you don't have delete access to the document, the UI
doesn't allow you to replace the file, which isn't expected. This is
also a UI only restriction, and the API allows you do this fine.
This PR makes so the "remove file" button renders even if you don't have
delete access, while still ensures you have update access.
---------
Co-authored-by: Paul Popus <paul@payloadcms.com>
### What?
This PR adds a dedicated `sortOrder` select field (Ascending /
Descending) to the import-export plugin, alongside updates to the
existing `SortBy` component. The new field and component logic keep the
sort field (`sort`) in sync with the selected sort direction.
### Why?
Previously, descending sorting did not work. While the `SortBy` field
could read the list view’s `query.sort` param, if the value contained a
leading dash (e.g. `-title`), it would not be handled correctly. Only
ascending sorts such as `sort=title` worked, and the preview table would
not reflect a descending order.
### How?
- Added a new `sortOrder` select field to the export options schema.
- Implemented a `SortOrder` custom component using ReactSelect:
- On new exports, reads `query.sort` from the list view and sets both
`sortOrder` and `sort` (combined value with or without a leading dash).
- Handles external changes to `sort` that include a leading dash.
- Updated `SortBy`:
- No longer writes to `sort` during initial hydration—`SortOrder` owns
initial value setting.
- On user field changes, writes the combined value using the current
`sortOrder`.
### What:
This PR adds `limit` and `page` fields to the export options, allowing
users to control the number of documents exported and the page from
which to start the export. It also enforces that limit must be a
positive multiple of 100.
### Why:
This feature is needed to provide pagination support for large exports,
enabling users to export manageable chunks of data rather than the
entire dataset at once. Enforcing multiples-of-100 for `limit` ensures
consistent chunking behavior and prevents unexpected export issues.
### How:
- The `limit` field determines the maximum number of documents to export
and **must be a positive multiple of 100**.
- The `page` field defines the starting page of the export and is
displayed only when a `limit` is specified.
- If `limit` is cleared, the `page` resets to 1 to maintain consistency.
- Export logic was adjusted to respect the `limit` and `page` values
when fetching documents.
---------
Co-authored-by: Patrik Kozak <35232443+PatrikKozak@users.noreply.github.com>
### What?
Fixes an issue where CSV exports and the preview table displayed all
fields of documents in hasMany monomorphic relationships instead of only
their IDs.
### Why?
This caused cluttered output and inconsistent CSV formats, since only
IDs should be exported for hasMany monomorphic relationships.
### How?
Added explicit `toCSV` handling for all relationship types in
`getCustomFieldFunctions`, updated `flattenObject` to delegate to these
handlers, and adjusted `getFlattenedFieldKeys` to generate the correct
headers.
Follow-up to #13416. Supersedes #13434.
When autosave is triggered and the user continues to modify fields,
their changes are overridden by the server's value, i.e. the value at
the time the form state request was made. This makes it almost
impossible to edit fields when using a small autosave interval and/or a
slow network.
This is because autosave is now merged into form state, which by default
uses `acceptValues: true`. This does exactly what it sounds like,
accepts all the values from the server—which may be stale if underlying
changes have been made. We ignore these values for onChange events,
because the user is actively making changes. But during form
submissions, we can accept them because the form is disabled while
processing anyway.
This pattern allows us to render "computed values" from the server, i.e.
a field with an `beforeChange` hook that modifies its value.
Autosave, on the other hand, happens in the background _while the form
is still active_. This means changes may have been made since sending
the request. We still need to accept computed values from the server,
but we need to avoid doing this if the user has active changes since the
time of the request.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1211027929253429
### What?
Adds e2e tests that verify group-by functionality within the trash view
of a collection.
### Why?
To ensure group-by behaves correctly when viewing soft-deleted
documents, including:
- Clearing the group-by selection via the reset button.
- Navigating from grouped rows to the trashed document's edit view.
### How?
- Added `should properly clear group-by in trash view` to test the reset
button behavior.
- Added `should properly navigate to trashed doc edit view from group-by
in trash view` to confirm correct linking and routing.
By default, `payload.jobs.run` only runs jobs from the `default` queue
(since https://github.com/payloadcms/payload/pull/12799). It exposes an
`allQueues` property to run jobs from all queues.
For handling schedules (`payload.jobs.handleSchedules` and
`config.jobs.autoRun`), this behaves differently - jobs are run from all
queues by default, and no `allQueues` property exists.
This PR adds an `allQueues` property to scheduling, as well as changes
the default behavior to only handle schedules for the `default` queue.
That way, the behavior of running and scheduling jobs matches.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210982048221260
Allows user to override more of the tenant field config. Now you can
override most of the field config with:
### At the root level
```ts
/**
* Field configuration for the field added to all tenant enabled collections
*/
tenantField?: RootTenantFieldConfigOverrides
```
### At the collection level
Setting collection level overrides will replace the root level overrides
shown above.
```ts
collections: {
[key in CollectionSlug]?: {
// ... rest of the types
/**
* Overrides for the tenant field, will override the entire tenantField configuration
*/
tenantFieldOverrides?: CollectionTenantFieldConfigOverrides
}
}
```
Fixes#10515. Needed for #12956.
Hooks run within autosave are not reflected in form state.
Similar to #10268, but for autosave events.
For example, if you are using a computed value, like this:
```ts
[
// ...
{
name: 'title',
type: 'text',
},
{
name: 'computedTitle',
type: 'text',
hooks: {
beforeChange: [({ data }) => data?.title],
},
},
]
```
In the example above, when an autosave event is triggered after changing
the `title` field, we expect the `computedTitle` field to match. But
although this takes place on the database level, the UI does not reflect
this change unless you refresh the page or navigate back and forth.
Here's an example:
Before:
https://github.com/user-attachments/assets/c8c68a78-9957-45a8-a710-84d954d15bcc
After:
https://github.com/user-attachments/assets/16cb87a5-83ca-4891-b01f-f5c4b0a34362
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210561273449855
## What
Before this PR, an internal link in the Lexical editor could reference a
document from a different tenant than the active one.
Reproduction:
1. `pnpm dev plugin-multi-tenant`
2. Log in with `dev@payloadcms.com` and password `test`
3. Go to `http://localhost:3000/admin/collections/food-items` and switch
between the `Blue Dog` and `Steel Cat` tenants to see which food items
each tenant has.
4. Go to http://localhost:3000/admin/collections/food-items/create, and
in the new richtext field enter an internal link
5. In the relationship select menu, you will see the 6 food items at
once (3 of each of those tenants). In the relationship select menu, you
would previously see all 6 food items at once (3 from each of those
tenants). Now, you'll only see the 3 from the active tenant.
The new test verifies that this is fixed.
## How
`baseListFilter` is used, but now it's called `baseFilter` for obvious
reasons: it doesn't just filter the List View. Having two different
properties where the same function was supposed to be placed wasn't
feasible. `baseListFilter` is still supported for backwards
compatibility. It's used as a fallback if `baseFilter` isn't defined,
and it's documented as deprecated.
`baseFilter` is injected into `filterOptions` of the internal link field
in the Lexical Editor.
When grouping by a checkbox field, boolean values are not translated,
causing labels to render incorrectly, and falsey values to render
without a heading.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210979856538211
When grouping by a date field and its value is null, the list view
crashes.
---
- To see the specific tasks where the Asana app for GitHub is being
used, see below:
- https://app.asana.com/0/0/1210979856538208
### What?
Fix the folder view for upload documents only using
`formatFolderOrDocumentItem()` function and only if the upload is an
image, even when there's a `thumbnailURL` available.
### Why?
Folder view for upload collections (especially those with sharp resizing
disabled) renders different thumbnails between the folder view and list
view. With sharp resizing disabled and an `adminThumbnail` fn provided,
the list view will correctly render optimised images, while the folder
view renders full source images - resulting in a huge discrepancy in
loaded image sizes.
### How?
We're passing the `value.thumbnailURL` **before** the
`formatFolderOrDocumentItem()` call rather than passing it directly as a
function parameter to cover cases where non-image uploads have a
`thumbnailURL` defined.
Fixes#13246
### What?
Allows document to successfully be saved when `fallback to default
locale` checked without throwing an error.
### Why?
The `fallback to default locale` checkbox allows users to successfully
save a document in the admin panel while using fallback data for
required fields, this has been broken since the release of `v3`.
Without the checkbox override, the user would be prevented from saving
the document in the UI because the field is required and will throw an
error.
The logic of using fallback data is not affected by this checkbox - it
is purely to allow saving the document in the UI.
### How?
The `fallback` checkbox used to have an `onChange` function that
replaces the field value with null, allowing it to get processed through
the standard localization logic and get replaced by fallback data.
However, this `onChange` was removed at some point and the field was
passing the actual checkbox value `true`/`false` which then breaks the
form and prevent it from saving.
This fallback checkbox is only displayed when `fallback: true` is set in
the localization config.
This PR also updated the checkbox to only be displayed when `required:
true` - when it's the field is not `required` this checkbox serves no
purpose.
Also adds tests to `localization/e2e`.
Fixes#11245
---------
Co-authored-by: Jarrod Flesch <jarrodmflesch@gmail.com>
Extension of https://github.com/payloadcms/payload/pull/13213
This PR correctly filters tenants, users and documents based on the
users assigned tenants if any are set. If a user is assigned tenants
then list results should only show documents with those tenants (when
selector is not set). Previously you could construct access results that
allows them to see them, but in the confines of the admin panel they
should not see them. If you wanted a user to be able to see a "public"
tenant while inside the admin panel they either need to be added to the
tenant or have no tenants at all.
Note that this is for filtering only, access control still controls what
documents a user has _access_ to a document. The filters are and always
have been a way to filter out results in the list view.
