Continuation of https://github.com/payloadcms/payload/pull/6245.
This PR allows you to pass `blocksAsJSON: true` to SQL adapters and the
adapter instead of aligning with the SQL preferred relation approach for
blocks will just use a simple JSON column, which can improve performance
with a large amount of blocks.
To try these changes you can install `3.43.0-internal.c5bbc84`.
### What?
Adds `constructorOptions` property to the upload config to allow any of
[these options](https://sharp.pixelplumbing.com/api-constructor/) to be
passed to the Sharp library.
### Why?
Users should be able to extend the Sharp library config as needed, to
define useful properties like `limitInputPixels` etc.
### How?
Creates new config option `constructorOptions` which passes any
compatible options directly to the Sharp library.
#### Reported by client.
Fixes https://github.com/payloadcms/payload/issues/12776
- Adds a new `jobs.config.enabled` property to the sanitized config,
which can be used to easily check if the jobs system is enabled (i.e.,
if the payload-jobs collection was added during sanitization). This is
then checked before Payload sets up job autoruns.
- Fixes some type issues that occurred due to still deep-requiring the
jobs config - we forgot to omit it from the `DeepRequired(Config)` call.
The deep-required jobs config was then incorrectly merged with the
sanitized jobs config, resulting in a SanitizedConfig where all jobs
config properties were marked as required, even though they may be
undefined.
The custom save button showing on export enabled collections in the edit
view. Instead it was meant to only appear in the export collection. This
makes it so it only appears in the export drawer.
Type declaration extending `custom.['plugin-import-export']` was
incorrectly named `toCSVFunction` instead of `toCSV`. This changes the
type to match the correct property name `toCSV`.
By default, calling `payload.jobs.run()` will incorrectly run all jobs
from all queues. The `npx payload jobs:run` bin script behaves the same
way.
The `payload-jobs/run` endpoint runs jobs from the `default` queue,
which is the correct behavior.
This PR does the following:
- Change `payload.jobs.run()` to only runs jobs from the `default` queue
by default
- Change `npx payload jobs:run` bin script to only runs jobs from the
`default` queue by default
- Add new allQueues / --all-queues arguments/queryparams/flags for the
local API, rest API and bin script to allow you to run all jobs from all
queues
- Clarify the docs
Introduces the ability to customize the order of both default and custom
tabs. This way you can make custom tabs appear before default ones, or
change the order of tabs as you see fit.
To do this, use the new `tab.order` property in your edit view's config:
```ts
import type { CollectionConfig } from 'payload'
export const MyCollectionConfig: CollectionConfig = {
// ...
admin: {
components: {
views: {
edit: {
myCustomView: {
path: '/my-custom-view',
Component: '/path/to/component',
tab: {
href: '/my-custom-view',
order: 100, // This will put this tab in the first position
},
}
}
}
}
}
}
```
---------
Co-authored-by: Jacob Fletcher <jacobsfletch@gmail.com>
Simplifies the rendering logic around document tabs in the following
ways:
- Merges default tabs with custom tabs in a more predictable way, now
there is only a single array of tabs to iterate over, no more concept of
"custom" tabs vs "default" tabs, there's now just "tabs"
- Deduplicates rendering conditions for all tabs, now any changes to
default tabs would also apply to custom tabs with half the code
- Removes unnecessary `getCustomViews` function, this is a relic of the
past
- Removes unnecessary `getViewConfig` function, this is a relic of the
past
- Removes unused `references`, `relationships`, and `version` key
placeholders, these are relics of the past
- Prevents tab conditions from running twice unnecessarily
- Other misc. cleanup like unnecessarily casting the tab conditions
result to a boolean, etc.
### What?
Updates the redirects plugin types to make collections a required type
### Why?
Currently not including the collections object when importing the plugin
causes an error to occur when going to the page in the UI, also it
cannot generate types. Likely due to it unable to make a reference to a
collection.
### How?
Makes collections required
Fixes#12709
---------
Co-authored-by: Patrick Roelofs <patrick.roelofs@iquality.nl>
Co-authored-by: Germán Jabloñski <43938777+GermanJablo@users.noreply.github.com>
### What?
Fixes inconsistent `pill` sizes across the Admin Panel.
### How?
Pills without a specified size default to **medium**. In the folders
[PR](https://github.com/payloadcms/payload/pull/10030), additional
padding was to the medium size. As a result, any pills without an
explicit size now appear larger than intended.
This PR fixes that by updating any pills that should be small to
explicitly set `size="small"`.
Fixes#12752
Simplifies document routing in the following ways:
- Removes duplicative code blocks that made it easy to make changes to
collections but not globals
- Consolidates `CustomView`, `DefaultView`, and `ErrorView` into just
`View`
- Removes unnecessary `overrideDocPermissions` arg
- Standardizes the 404 logic when the doc lacks read access
- Fixes styling issue where `UnauthorizedView` is rendered without
margins, e.g. when you lack permission to read versions but navigate to
`/versions`
Filters URLs to avoid issues with SSRF
Had to use `undici` instead of native `fetch` because it was the only
viable alternative that supported both overriding agent/dispatch and
also implemented `credentials: include`.
[More info
here.](https://blog.doyensec.com/2023/03/16/ssrf-remediation-bypass.html)
---------
Co-authored-by: Elliot DeNolf <denolfe@gmail.com>
Customizing the `path` property on default document views is currently
not supported, but the types suggest that it is. You can only provide a
path to custom views. This PR ensures that `path` cannot be set on
default views as expected.
For example:
```ts
import type { CollectionConfig } from 'payload'
export const MyCollectionConfig: CollectionConfig = {
// ...
admin: {
components: {
views: {
edit: {
default: {
path: '/' // THIS IS NOT ALLOWED!
},
myCustomView: {
path: '/edit', // THIS IS ALLOWED!
Component: '/collections/CustomViews3/MyEditView.js#MyEditView',
},
},
},
},
},
}
```
For background context, this was deeply explored in #12701. This is not
planned, however, due to [performance and maintainability
concerns](https://github.com/payloadcms/payload/pull/12701#issuecomment-2963926925),
plus [there are alternatives to achieve
this](https://github.com/payloadcms/payload/pull/12772).
This PR also fixes and improves various jsdocs, and fixes a typo found
in the docs.
When saving a global with versioning enabled as draft, and then
publishing it, the following error may appear: `[16:50:35] ERROR: Could
not find createdAt or updatedAt in latestVersion.version`
This is due to an incorrect check to appease typescript strict mode. We
shouldn't throw if `version.updatedAt` doesn't exist - the purpose of
this logic is to add that property if it doesn't exist
Currently, we globally enable both DOM and Node.js types. While this
mostly works, it can cause conflicts - particularly with `fetch`. For
example, TypeScript may incorrectly allow browser-only properties (like
`cache`) and reject valid Node.js ones like `dispatcher`.
This PR disables DOM types for server-only packages like payload,
ensuring Node-specific typings are applied. This caught a few instances
of incorrect fetch usage that were previously masked by overlapping DOM
types.
This is not a perfect solution - packages that contain both server and
client code (like richtext-lexical or next) will still suffer from this
issue. However, it's an improvement in cases where we can cleanly
separate server and client types, like for the `payload` package which
is server-only.
## Use-case
This change enables https://github.com/payloadcms/payload/pull/12622 to
explore using node-native fetch + `dispatcher`, instead of `node-fetch`
+ `agent`.
Currently, it will incorrectly report that `dispatcher` is not a valid
property for node-native fetch
### What?
The azure storage adapter returns a 500 internal server error when a
file is not found.
It's expected that it will return 404 when a file is not found.
### Why?
There is no checking if the blockBlobClient exists before it's used, so
it throws a RestError when used and the blob does not exist.
### How?
Check if exception thrown is of type RestError and have a 404 error from
the Azure API and return a 404 in that case.
An alternative way would be to call the exists() method on the
blockBlobClient, but that will be one more API call for blobs that does
exist. So I chose to check the exception instead.
Also added integration tests for azure storage in the same manner as s3,
as it was missing for azure storage.
### What?
The results when querying orderable collections can be incorrect due to
how the underlying database handles sorting when capitalized letters are
introduced.
### Why?
The original fractional indexing logic uses base 62 characters to
maximize the amount of data per character. This optimization saves a few
characters of text in the database but fails to return accurate results
when mixing uppercase and lowercase characters.
### How?
Instead we can use base 36 values instead (0-9,a-z) so that all
databases handle the sort consistently without needing to introduce
collation or other alternate solutions.
Fixes#12397
### What
This PR updates the `afterChange` hook for collections and globals to
include the `data` argument.
While the `doc` argument provides the saved version of the document,
having access to the original `data` allows for additional context—such
as detecting omitted fields, raw client input, or conditional logic
based on user-supplied data.
### Changes
- Adds the `data` argument to the `afterChange` hook args.
- Applies to both `collection` and `global` hooks.
### Example
```
afterChange: [
({ context, data, doc, operation, previousDoc, req }) => {
if (data?.customFlag) {
// Perform logic based on raw input
}
},
],
```
<!--
Thank you for the PR! Please go through the checklist below and make
sure you've completed all the steps.
Please review the
[CONTRIBUTING.md](https://github.com/payloadcms/payload/blob/main/CONTRIBUTING.md)
document in this repository if you haven't already.
The following items will ensure that your PR is handled as smoothly as
possible:
- PR Title must follow conventional commits format. For example, `feat:
my new feature`, `fix(plugin-seo): my fix`.
- Minimal description explained as if explained to someone not
immediately familiar with the code.
- Provide before/after screenshots or code diffs if applicable.
- Link any related issues/discussions from GitHub or Discord.
- Add review comments if necessary to explain to the reviewer the logic
behind a change
### What?
### Why?
### How?
Fixes #
-->
### What?
In a similar vein to #11734, #11733, #10327 - this PR returns a 404 in
the response when a file is not found while using the `storage-gcs`
adapter. Currently a 500 is returned.
### Why?
To return the correct error level in the response when a file is not
found when using `storage-gcs`.
### How?
The GCS nodejs library exposes the `ApiError` as a general error - these
changes check that the caught error is an instance of this class and if
the provided code is a `404`.
Previously, every time the drawer re-rendered a new entry animation may
be triggered. This PR fixes this by setting the open state to
`modalState[slug]?.isOpen` instead of `false`.
Additionally, I was able to simplify this component while maintaining
functionality. Got rid of one `useEffect` and one `useState` call. The
remaining useEffect also runs less often (previously, it ran every time
`modalState` changed => it re-ran if _any_ modal opened or closed, not
just the current one)
When making an export of a collection, and no fields are selected then
you get am empty CSV. The intended behavior is that all data is exported
by default.
This fixes the issue that from the admin UI, when the fields selector
the resulting CSV has no columns.
Fixes https://github.com/payloadcms/payload/issues/12628
When using sqlite, the error from the db is a bit different than
Postgres.
This PR allows us to extract the fieldName when using sqlite for the
unique constraint error.
### What?
This PR updates the import names for the Bengali (Bangladesh) and
Bengali (India) translations to follow the camelCase convention used for
other hyphenated locales:
- `bnBD → bnBd`
- `bnIN → bnIn`
This aligns with the existing pattern already used for translations like
`zhTw.`
Locale keys in the `translations` map (`'bn-BD'`, `'bn-IN'`) remain
unchanged.
Important: An intentional effort is being made during migration to not
modify runtime behavior. This implies that there will be several
assertions, non-null assertions, and @ts-expect-error. This philosophy
applies only to migrating old code to TypeScript strict, not to writing
new code. For a more detailed justification for this reasoning, see
#11840 (comment).
In this PR, instead of following the approach of migrating a subset of
files, I'm migrating all files by disabling specific rules. The first
commits are named after the rule being disabled.
With this PR, the migration of the payload package is complete 🚀
Adds support for read replicas
https://orm.drizzle.team/docs/read-replicas that can be used to offload
read-heavy traffic.
To use (both `db-postgres` and `db-vercel-postgres` are supported):
```ts
import { postgresAdapter } from '@payloadcms/db-postgres'
database: postgresAdapter({
pool: {
connectionString: process.env.POSTGRES_URL,
},
readReplicas: [process.env.POSTGRES_REPLICA_URL],
})
```
---------
Co-authored-by: Dan Ribbens <dan.ribbens@gmail.com>
Previously, if you enabled presigned URL downloads for a collection, all
the files would use them. However, it might be possible that you want to
use presigned URLs only for specific files (like videos), this PR allows
you to pass `shouldUseSignedURL` to control that behavior dynamically.
```ts
s3Storage({
collections: {
media: {
signedDownloads: {
shouldUseSignedURL: ({ collection, filename, req }) => {
return req.headers.get('X-Disable-Signed-URL') !== 'false'
},
},
},
},
})
```
### What?
In a project that has `jobs` configured and the import/export plugin
will error on start:
`payload-jobs validation failed: taskSlug: createCollectionExport is not
a valid enum value for path taskSlug.`
### Why?
The plugin was not properly extending the jobs configuration.
### How?
Properly extend existing config.jobs to add the `createCollectionExport`
task.
Fixes #
This makes it possible to add custom logic into how we map the document
data into the CSV data on a field-by-field basis.
- Allow custom data transformation to be added to
`custom.['plugin-import-export'].toCSV inside the field config
- Add type declaration to FieldCustom to improve types
- Export with `depth: 1`
Example:
```ts
{
name: 'customRelationship',
type: 'relationship',
relationTo: 'users',
custom: {
'plugin-import-export': {
toCSV: ({ value, columnName, row, siblingDoc, doc }) => {
row[`${columnName}_id`] = value.id
row[`${columnName}_email`] = value.email
},
},
},
},
```
Fixes https://github.com/payloadcms/payload/issues/12639
Currently, if an `importmap.js` file is not found — it throws an error.
This change allows the file to be created if the directory can be found.
If you specify your own `importmap` location in the config, it will
attempt to create when missing and throw an error if it cannot.
Fixes#12723
## Problem
With two or more state groups, a falsy state entry would run
`dom.style.cssText = ''`, wiping all inline styles - including valid
styles just set by earlier groups. This caused earlier groups to lack
css styling.
## Solution
**1. Collect first, apply once**
During the stateMap.forEach loop, gather each active group's styles into
a shared `mergedStyles` object.
Inactive groups still clear their own `data-*` attribute but leave
styles untouched.
**2. Single reset + single write**
After the loop, perform one cssText reset, instead of one cssText in
each iteration. Then apply each state group's css all at once, after the
blank reset
Result: every state group can coexist without overwriting styles from
the others
