# Introduction
Introduced `EachTesting` pattern for the builder unit tests.
As always any suggestions are more than welcomed !
Still need to:
- [x] implem basic tests for field
- [x] create `get-flat-index-field-metadata.mock.ts`
- [x] Implement basic tests for index and index-fields
- [ ] Implem standard edges cases tests TDD style
## Misc
- was https://github.com/twentyhq/twenty/pull/13132 closed due to mess
to rebase on main
resolve#12345
The issue was caused by the delete running after the update, which led
to both the old and new options being deleted when they shared the same
value.
---------
Co-authored-by: Marie Stoppa <marie.stoppa@essec.edu>
_(AI generated)_
### Summary
This PR fixes a validation bug in the GraphQL API that prevented
relation fields from being programmatically deactivated. The validation
was incorrectly triggering a "name cannot be changed" error even when
the update payload did not include a name, making it impossible to
disable the field.
Issue #13200
### Problem
- The [updateOneField] mutation failed when trying to set `isActive:
false` on a `RELATION` field.
- The root cause was a validation check in
[FieldMetadataValidationService] that compared the incoming `name` with
the existing one. If the input `name` was `undefined`, the check
`undefined !== existingName` would incorrectly fail.
- This created a catch-22 where a field could not be deleted (because it
had to be deactivated first) and could not be deactivated (due to this
validation error).
### Solution
- The validation logic in [field-metadata-validation.service.ts] has
been updated to only check for a name change if a new name is
**explicitly provided** in the input
(`isDefined(fieldMetadataInput.name)`).
- This change correctly enforces the rule that relation field names
cannot be changed, while allowing other properties like `isActive` to be
updated without issue.
### How to Test
1. Create a custom field of type `RELATION`.
2. Using the GraphQL API, call the [updateOneField] mutation with the
field's ID and the payload `{ "isActive": false }`.
3. Verify that the mutation succeeds and the field is now inactive.
4. Call the [deleteOneField] mutation to delete the field.
5. Verify that the deletion is successful.
### Additional Changes
_to be deleted if not necessary_
- Added a new integration test
[successful-field-metadata-relation-update.integration-spec.ts] to cover
this specific use case and prevent future regressions. The existing test
for failing updates remains untouched and continues to pass.
Instead of initializing model at start time we do it at run time to be
able to swap model provider more easily.
Also introduce a third driver for openai-compatible providers, which
among other allows for local models with Ollama
Step operand will more or less be the same as view filter operand.
This PR:
- moves `ViewFilterOperand` to twenty-shared
- use it as step operand
- check what operand should be available based on the selected field
type in filter action
- rewrite the function that evaluates filters so it uses
ViewFilterOperand instead
ViewFilterOperand may be renamed in a future PR.
In the BE we throw custom errors with precise error codes (e.g.
"LABEL_ALREADY_EXISTS") before catching them in filters and rethrowing
BaseGraphQLErrors (standard errors such as NotFoundError, UserInputError
etc.).
In the FE we were grouping sentries based on the error codes but we were
actually grouping by very broad codes such as "NOT_FOUND" or
"BAD_USER_INPUT", extracted from the BaseGraphQLErrors.
To fix that, we update the BaseGraphQLError constructor api to allow to
pass on the CustomError directly and retrieve from it the original code
and store it in existing property `subCode` that we will use in the FE
to send errors to sentry.
This new api also eases usage of `userFriendlyMessage` that is passed on
to the api response and therefore to the FE when CustomError is passed
on directly to the BaseGraphQLError constructor.
# Introduction
- Added `INDEX` action generation
- Refactored the naming, mainly `WorkspaceMigrationV2ObjectInput` ->
`FlattenObjectMetadata` and same for field. The transpilation will be
done above, agnostically of the workspace migration
Still need to:
- Create testing toolbox and follow each testing pattern for each
actions and make a complex one ( remove static current tests )
- Handle standard and custom edges cases
Notes:
`workspace-migration-v2/types` and `workspace-migration-v2/utils` could
be located outside of this folder, my hunch is that we will move them
once we work on flatten tranpilers
Fixes https://github.com/twentyhq/twenty/issues/12867
Issue:
when you have a variable `toto` which is: `Record<string, MyType>` and
you do toto['xxx'], this will be typed as `MyType` instead of `MyType |
undefined`
Solutions:
- activate `noUncheckedIndexedAccess` check in tsconfig, this is the
preferred solution but will take time to get there (this raises 600+
errors)
- use a Map: cf https://github.com/twentyhq/twenty/pull/13125/files
- set the type to Partial<Record<string, MyType>>. Drawback is that when
you do Object.values(toto), you'll get `Array<MyType | undefined>`.
Hence why we have to filter these behind
<img width="1512" alt="image"
src="https://github.com/user-attachments/assets/d0a0bfed-c441-4e53-84c2-2da98ccbcf50"
/>
- update publishOneServerlessFunction so it does not break the workflow
if failing
- update upgrade documentation to update `docker-compose.yml` when
mutated
Fixes https://github.com/twentyhq/twenty/issues/13058
Large PR, sorry for that. Don't hesitate to reach me to have full
context (env. 500lines for integration and unit tests)
- Add connect logic in Workspace Entity Manager
- Update QueryDeepPartialEntity type to enable dev to use connect
- Add integration test on createOne / createMany
- Add unit test to cover main utils
- Remove feature flag on connect
closes https://github.com/twentyhq/core-team-issues/issues/1148
closes https://github.com/twentyhq/core-team-issues/issues/1147
I rebuilt the advanced filters used in views and workflow search for a
specific filter step.
Components structure remains the same, using `stepFilterGroups` and
`stepFilters`. But those filters are directly sent to backend.
Also re-using the same kind of states we use for advanced filters to
share the current filters used. And a context to share what's coming
from workflow props (function to update step settings and readonly)
⚠️ this PR only focusses on the content of the step. There is still a
lot to do on the filter icon behavior in the workflow
https://github.com/user-attachments/assets/8a6a76f0-11fa-444a-82b9-71fc96b18af4
In this PR
- introduction of fieldPermission entity
- addition of upsertFieldPermission in role resolver
- computing of permissions taking fieldPermission into account. In order
to limit what is stored in Redis we only store fields restrictions. For
instance for objectMetadata with id XXX with a restriction on field with
id YYY we store:
`"XXX":{"canRead":true,"canUpdate":false,"canSoftDelete":false,"canDestroy":false,"restrictedFields":{"YYY":{"canRead":false,"canUpdate":null}}}`
---------
Co-authored-by: Charles Bochet <charlesBochet@users.noreply.github.com>
## Expected behavior
Described behavior regarding: (update | create) x (custom | standard) x
(icon, label, name, isSynced)
**Custom:**
- Field RELATION create: name, label, isSynced, icon should be editable
- Field RELATION update: name should not, icon label, isSynced should
- For other fields, icon, label, name, isSynced should be editable at
field creation | update
To simplify: Field RELATION name should not be editable at update
**Standards**
- Field: create does not makes sense
- Field: name should not, icon label, isSynced should (this will end up
in overrides)
To simplify, no Field RELATION edge case, name should not be editable at
update
**Note:** the FE logic is quite different as the UI is hiding some
details behind the syncWithLabel. See my comments and TODO there
## What I've tested:
(update | create) x (custom | standard) x (icon, label, name, isSynced,
description)
Updates yarn to the latest version 4.9.2 (from 4.4.0).
Also removes the explicit `enableHardenedMode` from yarnrc as it
significantly slows down installation.
This is already enabled automatically for pull requests on Github, thus
preventing lockfile poisoning where it's relevant.
See <https://yarnpkg.com/features/security#hardened-mode>:
> in most cases you won't even have to think about it - the hardened
mode is enabled by default when Yarn detects it runs in a pull request
from a public GitHub repository.
It can additionally be enabled explicitly for specific CI jobs by using
an environment variable, if desired:
> The hardened mode can be set (or disabled) [...] by defining
`YARN_ENABLE_HARDENED_MODE=1|0` in your environment variables
If this is the case, yarn still recommends **not** enabling it
everywhere:
> **DANGER**
>
> The hardened mode makes installs significantly slower as Yarn has to
query the registry to make sure the information contained in the
lockfile are accurate. If your CI pipeline runs multiple jobs, we
recommend disabling the hardened mode in all but one of them so as to
limit the performance impact.
---------
Co-authored-by: prastoin <paul@twenty.com>
In certain scenarios, the front directory may not be writable.
When this is the case, writing the patched `index.html` should be
skipped just like when the file does not exist.
For brevity, I have replaced the `existsSync` check with a try-catch
that catches any errors occuring during read
or write of the index file.
The alternative would be `fs.accessSync` with `W_OK`, but that would
still throw an error if the file is not writable so I think it is
reasonable to skip it altogether and go straight for the read and write
attempts.
A specific scenario where the front directory is immutable is NixOS,
where the directory may be located in the read-only nix store.
# Introduction
In this PR we've mainly refactor the typing to be extending existing
entities.
Also handling relations through the field abstraction layer rather than
a dedicated one. We reverted midway
We then still need to:
- Handle indexing
- Uniqueness
- Add strong coverage and avoid static inline snapshoting as right now +
building a coherent testing set
- Deprecate the `standardId` in favor of a `uniqueIdentifier` on each
`objectMetadata` and `fieldMetadata`
- Rename types `input` to `flattened`
- Handle custom or non custom edit edge cases. ( e.g cannot delete a
standard field or object )
## Notes
Right I preferred including too many information ( whole object and
field input ) in the action context, we might wanna evict redundant
information in the future when implementing the runners
This PR is purely technical, it does produces any functional change to
the user
- add Lock mecanism to run steps concurrently
- update `workflow-executor.workspace-service.ts` to handle multi branch
workflow execution
- stop passing `context` through steps, it causes race condition issue
- refactor a little bit
- simplify `workflow-run.workspace-service.ts` to prepare `output` and
`context` removal
- move workflowRun status computing from `run-workflow.job.ts` to
`workflow-executor.workspace-service.ts`
## NOTA BENE
When a code step depends of 2 parents like in this config (see image
below)
If the form is submitted before the "Code - 2s" step succeed, the branch
merge "Form" step is launched twice.
- once because form is submission Succeed resumes the workflow in an
asynchronous job
- the second time is when the asynchronous job is launched when "Code -
2s" is succeeded
- the merge "Form" step makes the workflow waiting for response to
trigger the resume in another job
- during that time, the first resume job is launched, running the merge
"Form" step again
This issue only occurs with branch workflows. It will be solved by
checking if the currentStepToExecute is already in a SUCCESS state or
not
<img width="505" alt="image"
src="https://github.com/user-attachments/assets/b73839a1-16fe-45e1-a0d9-3efa26ab4f8b"
/>
# Introduction
In this PR we've initialized the `workspace-migration-v2` folder.
Focusing on the builder in the first place.
From now it contains:
- Basic temporary types ( `fieldMetadataEntity` and
`ObjectMetadataEntity` )
- Object actions builder ( create, delete, update )
- Fields actions builder ( create, delete ) ( update coming in a
following PR )
We will still have to handle specific conditions such as:
- Index creation
- Uniqueness addition removal
- Relation
We also need to determine when we want to compute and transpile the
object no field `uniqueIdentifier`
We're aiming to merge this first in order to avoid accumulating code in
this PR
---------
Co-authored-by: prastoin <paul@twenty.com>
This PR is raised to close the issue #13044
But there are some doubts that needs to be approved.
If the fields are not custom then we were saving the changes in
**standardOverrides** obj in which only three fields are
overridableFields **label, icon & description** can be updated for a
field.
You can see this in _before-update-one-field.hook.ts_ on line 85
```ts
const overridableFields = ['label', 'icon', 'description'];
```
If the field to be updated are from these three we are putting this in a
**standardOverrides** obj and passing it
However in our _field-metadata.service.ts_ file. We have **updateOne**
function inside it we have wrote a condition if **isCustom** is false
then the purpose was to build the updatableFields from the
**standardOverrides** obj that we got in **fieldMetadataInput** but
there was an error in it. As you can see below
```ts
const updatableFieldInput =
existingFieldMetadata.isCustom === false
? this.buildUpdatableStandardFieldInput(
fieldMetadataInput,
existingFieldMetadata,
)
: fieldMetadataInput;
```
However, the issue was that we were placing the entire
**standardOverrides** object inside **updatableFieldInput** again —
instead of merging its individual fields (label, icon, description)
directly into the update payload.
This PR fixes that by correctly applying the overrides into the
top-level object.
Please refer to the file changes for the full context.
But the thing i don't know.
[ ] - Is this the correct expected flow??
[ ]- Will this change break anything elsewhere?
I still have doubts on these two. Let me know if I missed something.
---------
Co-authored-by: Jagss24 <btwitsjagannat12@gmail.com>
Co-authored-by: Charles Bochet <charles@twenty.com>
In this PR, I'm fixing a bug introduced in recent performance work on
the cache.
Bug context: https://github.com/twentyhq/twenty/issues/12865
Related PR opened by a contributor:
https://github.com/twentyhq/twenty/pull/13003
## Root cause
We cache all objectMetadataItems at graphql level : see
`useCachedMetadata` hook:
- instead of going through the regular resolvers, we direlcty load data
from the cache. However this data must be localized regarding labels and
descriptions
In a precedent refactoring, we introduced the notion of locale in the
cache key. However, the user locale was not properly taken into account
as we did not have the information in this hook.
## Fix
1. **Introduce locale in userWorkspace entity**. The locale is stored on
workspaceMember in each postgres workspaceSchema (workspace_xxx) which
is the alter ego of userWorkspace in postgres core schema. Note that we
can't store it in user as a user can be part of multiple workspaces (the
locale already there must be seen as a default for this user), and we
cannot rely on workspaceMember as we would need to query the
workspaceSchema in the authentication layer which we want to avoid for
performance reasons.
2. During request hydration from token (containing the userWorkspaceId),
we fetch the userWorkspace and store it in the Request (this impact both
AuthContext and Request interface)
3. Leverage userWorkspace.locale in the useCachedMetadata hook
## Additional notes
There is no need to change the way we store and retrieve the
object-metadata-maps object itself which is different from the graphql
layer cache. object-metadadata-maps are not localized
Context :
- Phones import is a bit complex if not all subfields are provided.
- Phones subfield validation are absent or different from BE validation.
Solution :
- Normalize callingCode and countryCode validation (BE/FE)
- Ease phone import if only phoneNumber is provided
Currently, when a server query or mutation from the front-end fails, the
error message defined server-side is displayed in a snackbar in the
front-end.
These error messages usually contain technical details that don't belong
to the user interface, such as "ObjectMetadataCollection not found" or
"invalid ENUM value for ...".
**BE**
In addition to the original error message that is still needed (for the
request response, debugging, sentry monitoring etc.), we add a
`displayedErrorMessage` that will be used in the snackbars. It's only
relevant to add it for the messages that will reach the FE (ie. not in
jobs or in rest api for instance) and if it can help the user sort out /
fix things (ie. we do add displayedErrorMessage for "Cannot create
multiple draft versions for the same workflow" or "Cannot delete
[field], please update the label identifier field first", but not
"Object metadata does not exist"), even if in practice in the FE users
should not be able to perform an action that will not work (ie should
not be able to save creation of multiple draft versions of the same
workflows).
**FE**
To ease the usage we replaced enqueueSnackBar with enqueueErrorSnackBar
and enqueueSuccessSnackBar with an api that only requires to pass on the
error.
If no displayedErrorMessage is specified then the default error message
is `An error occured.`
