## Problem
Controls in the new table sheet's Columns section had inconsistent
styling, alignment and focus states. Some icon-only actions were faint
or lacked useful accessible names and tooltips, while related controls
used different dropdown treatments.
<img width="1428" height="250" alt="image"
src="https://github.com/user-attachments/assets/9d1ad80d-2de0-4736-8ddb-0260e10ddacc"
/>
## Solution
- Align the column headers and fields on the same grid, and keep the
foreign key action inside the name field in every state.
- Give the drag, foreign key, suggestions, extra options and remove
actions consistent button styling, sizing, hover and focus states, with
accessible names and tooltips.
- Move masking into Extra options and include it in the options count.
Remove the separators around the former inline controls.
- Use the design system's ComboboxTrigger for column types and align the
Default value select's radius with it. Update the suggestions dropdown
icon.
- Tidy related field labels and update the affected table editor tests.
The underlying column values and database behaviour are unchanged.
| Original before | Original after |
| --- | --- |
| <img width="1498" height="392" alt="CleanShot 2026-09-24 at 13 33
40@2x"
src="https://github.com/user-attachments/assets/56559e72-f289-44bc-999f-c1f3158c8f65"
/> | <img width="1496" height="416" alt="CleanShot 2026-09-24 at 13 32
42@2x"
src="https://github.com/user-attachments/assets/d05340fc-a460-4b5c-aaff-41127f692592"
/> |
## Review instructions
- Open **Table Editor → New table** and inspect the Columns section.
Check that Name, Type and Default value headers align with their fields.
- Tab through the row actions. Check their focus states, accessible
names and tooltips, including the foreign key control with and without a
relation.
- Drag a column to reorder it, open Extra options and toggle masking.
Check the options count and that the other settings still work.
- Compare the closed Type and Default value dropdowns, including an enum
default value.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Improvements**
* Updated the column editor layout and controls for naming,
sensitive-data settings, extra options, foreign-key editing, and column
removal, with clearer tooltips and more consistent interactions.
* Added accessible help tooltips for the Name and Default Value fields.
* Updated column-type and suggested-value controls, including clearer
guidance for choosing a suggested value.
* Sensitive-data settings are now available in the extra-options menu.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Danny White <3104761+dnywh@users.noreply.github.com>
Inserting a row through the table editor failed on any table with a
`GENERATED ALWAYS AS (...) STORED` column — the row editor sent an
explicit value for the generated column (e.g. `false` for booleans,
since the bool `Select` never hits the empty-string default heuristic
from #46826), which Postgres rejects with `428C9: cannot insert a
non-DEFAULT value into column`.
**Changed:**
- `RowField` now carries `isGenerated` (from pg-meta's `is_generated`,
previously unused by Studio)
- Generated columns are hidden from the row editor form (they're always
computed by the database, so there's nothing to input) but stay in
`rowFields` state so primary-key identifier logic is unaffected
- `generateRowObjectFromFields` skips generated fields, so they're
omitted from both insert and update payloads
- `validateFields` skips generated fields — an error on a hidden field
would be unfixable
**Added:**
- e2e test covering inserting a row into a table with a generated
boolean column
- unit tests for generated-column omission in insert/update payloads and
validation
## To test
1. Create a table with a generated column:
```sql
create table t (
id bigint generated by default as identity primary key,
base_price int,
discounted_price int,
is_discounted boolean generated always as (
base_price is distinct from discounted_price
) stored
);
```
2. Table Editor → `t` → Insert row — `is_discounted` should not appear
in the form
3. Fill the other fields and save — the insert should succeed and the
grid should show the computed value
4. Edit an existing row and save — should still work (generated column
untouched)
5. Sanity-check a normal table with identity/default columns — clearing
a default field on insert should still fall back to the default (#46826
behavior)
Addresses
[FE-3909](https://linear.app/supabase/issue/FE-3909/studio-insert-form-fails-on-generated-boolean-columns)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for generated columns in the table editor.
* Generated columns are automatically computed and excluded from insert
and update forms.
* Generated values now appear correctly in the table after saving a row.
* **Bug Fixes**
* Prevented validation errors for non-editable generated fields.
* Ensured generated columns are excluded from submitted row data.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: Alaister Young <10985857+alaister@users.noreply.github.com>
## What kind of change does this PR introduce?
A11y cleanup follow-up to #47984 /
[DEPR-626](https://linear.app/supabase/issue/DEPR-626).
## What is the current behavior?
Studio had 82 ratcheted `supabase/require-explicit-tabindex` violations
(raw `<button>` / `role="button"` without explicit `tabIndex`).
## What is the new behavior?
- Explicit `tabIndex={0}` (or disabled → `-1`) on those Studio call
sites across nav, `components/ui`, Database, Storage, and the remainder
- Ratchet baseline cleared (**82 → 0**) and the rule **removed from the
Studio ratchet** (debt is gone; ratchet is temporary)
- Rule remains a shared **`warn`** for now — promoting to `error` (and
sweeping www/docs/design-system) is a follow-up
- Also fixed the learn/ui-library call sites that surfaced while
experimenting with error promotion
- Small follow-ups where making controls focusable exposed gaps:
accessible names, disabled/focus consistency, focus-ring polish on
To-test surfaces, home section `KeyboardSensor`, and an E2E locator
tightened after `aria-label="Remove column"`
Prefer migrating to `Button` from `ui` in future touch-ups; this PR
takes the minimal path so Studio debt can stay at zero.
## Additional context
Batches landed together so baseline conflicts stayed simple while
chipping away:
- Hotspots / nav (FirstLevelNav, Marketplace, AttachmentUpload, Column,
Tabs, …)
- `components/ui` shared
- Database + Storage
- Remainder
**Out of scope / intentional deferrals**
- Promoting `supabase/require-explicit-tabindex` to a lint **error**
(follow-up after www/docs/design-system sweeps)
- Tabs/Radio roving, tooltips, context menus, in-menu items
- Full keyboard-accessible tab-close UX (close stays hover +
`tabIndex={-1}`; context menu still closes tabs)
- Data API docs links (`/project/<ref>/api` redirect)
**Reviewer notes**
- Rule only flags raw `<button>` / `role="button"` without a `tabIndex`
prop. `Button` from `ui` already bakes this in
- `tabIndex={-1}` is intentional for disabled controls, in-menu /
roving-focus children, and hover-only tab close
- For dnd-kit grips, put `tabIndex` **after** `{...attributes}` so it
isn’t overwritten (TS2783)
### To test
Use **Safari** with macOS Keyboard navigation **off** (System Settings →
Keyboard). Chrome once for a sanity pass. For each surface below: Tab
until the control is focused, then activate with Enter/Space where
relevant.
1. **API Docs side panel** (Table Editor → open a table → **API docs**)
- Floating API Docs panel — **not** `/project/<ref>/api` (that redirects
to Data API docs; language ToggleGroup uses arrow keys; links are out of
scope)
- Left nav buttons — Tab through several and activate one; active
highlight / navigation still works
2. **Integrations → Marketplace**
- Enable **Integrations layout** feature preview first (avatar menu →
Feature previews)
- `/org/<slug>/integrations` or project integrations marketplace
- “Clear all”, grid/list toggles — Tab + activate
3. **Table Editor → create a table → Columns**
- Drag handles only appear while **creating** (not when editing an
existing table)
- Tab to grip / remove (X) / sensitive-data eye if shown
4. **Project Home** — section drag handles
- Tab to a grip (visible focus ring)
- Optional: Space to pick up, arrows to move, Space/Esc to drop
(KeyboardSensor added)
- Mouse dnd still works
5. **Storage → Policies** — expand/collapse bucket list chevron
(design-system focus ring, no stuck grey open bg)
6. **Support form** (Help → Support) — attachment remove (×) and
add-attachment control when visible
Disabled controls should be **skipped** by Tab.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Accessibility Improvements**
* Improved keyboard navigation throughout Studio by explicitly managing
focus (`tabIndex`) across many interactive controls (menus, tabs,
tables, charts, dialogs, navigation, and form actions).
* Disabled or non-interactive controls are now removed from the tab
order (or made unfocusable), while available actions remain reachable.
* Ensured `type="button"` on relevant controls to prevent unintended
submissions, and refined keyboard focus behavior for various toggles and
copy/remove actions.
* **Chores**
* Updated the ESLint rule baseline configuration to match the new focus
behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Context
Migrates the remaining API requests to the pg-meta endpoint to use the
query endpoint directly with the SQL from the pg-meta package. This
touches the following:
- policies
- publications
- triggers
- views
- materialized views
- types
## To test
Just need to verify that we're still fetching the data correctly on
these pages
- Database policies
- Database publications
- Database triggers
- Database tables (views + materialized views)
- Database types
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved and stabilized loading of database metadata (views, triggers,
RLS policies, publications, materialized views, and enum types),
including more reliable schema-scoped filtering.
* Updated policy loading behavior and related UI queries to consistently
use schema arrays, improving cache correctness and consistency.
* **Tests**
* Updated end-to-end test synchronization to wait for the correct
metadata responses using more specific request identifiers.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
- closes https://github.com/supabase/supabase/issues/47606
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Copying cell values now preserves `false` and `0` instead of treating
them like empty values.
* Clipboard copy behavior now only returns blank for truly empty inputs,
helping keep table data accurate when copied.
* **Tests**
* Added end-to-end coverage for copying table cells with `false`, `0`,
and `true` values.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## TL;DR
fixes protected schema empty tables still exposing CSV import actions in
table editor...
## ref:
- closes https://github.com/supabase/supabase/issues/41358
- supersedes: https://github.com/supabase/supabase/pull/41362
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Updated the table empty state so CSV import actions are shown only
when the table is allowed to accept imports, and are hidden for
protected cases (including foreign-table scenarios).
* **Tests**
* Added an end-to-end test confirming that empty tables in protected
schemas do not expose the “Import data from CSV” button or the
drag-and-drop CSV hint.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Backports a batch of e2e test stabilization fixes — each commit is
scoped to a single failure class and only touches `e2e/studio/` files.
**Changed:**
- **`_global.setup` — playwright-locks cleanup was dead code**: the
lock-cleanup block was at the bottom of `Global Setup`, but every branch
above it returns early — so it never ran. Tests that use
`withFileOnceSetup` (cron-jobs) would see a stale `setup.done.json`
marker from the previous run and silently skip their setup, leaving e.g.
`pg_cron` uninstalled and all 11 cron-jobs specs failing. Moved the
cleanup to before any early return.
- **filter-bar — Home key**: macOS Chromium doesn't honor a standalone
`Home` keypress inside text inputs (macOS routes "go to line start" via
`Cmd+ArrowLeft` / `Fn+ArrowLeft`). Tests that expected the cursor to
jump to position 0 silently kept the previous selection. Replaced with
`el.setSelectionRange(0, 0)` so the assertion runs against a known
cursor position on every OS.
- **filter-bar — date filters**: tests inserted rows with `CURRENT_DATE`
/ `NOW()` (postgres session TZ = UTC) and asserted with JS-local dates
from `getDateValue()`. Near midnight the two diverged and the filter
returned 0 rows. Switched the inserts to explicit `getDateValue()`
strings so insert and assert use the same calendar day.
- **queue-table-operations — `networkidle`**: Studio holds long-poll /
SSE connections (PostHog, realtime), so `page.reload({ waitUntil:
'networkidle' })` never resolves and timed out. Replaced with a targeted
`waitForTableToLoad` API waiter.
- **sql-editor — RLS smoke test**: a hard-coded table name
(`pw_rls_smoke_test`) collided across 3 parallel workers running against
the same db. Suffixed with `test.info().parallelIndex`.
- **table-editor — FK spec timeout**: `waitForApiResponseWithTimeout`
for `query?key=table-update` returns `null` on timeout (silent), then
the panel-close assertion fails. Bumped 15s → 30s to absorb
parallel-load latency.
- **table-editor / storage-helpers — URL encoding & redirect race**:
post-action URL assertions were over-specific (`%20` vs `+` encoding)
and the bucket-delete redirect could race other history updates. Relaxed
the regex to accept both encodings; asserting the row removal directly
is a more stable signal than the redirect URL.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Improved end-to-end determinism with explicit dates/timestamps and
stable cursor positioning
* Prevented parallel-test collisions by using unique identifiers for
resources
* Made page reloads and API waits more robust for long-lived connections
and increased timeouts
* Strengthened assertions to rely on stable UI signals instead of
transient navigation/network state
* Ensured test setup reliably cleans up temporary locks before any setup
steps run
<!-- review_stack_entry_start -->
[](https://app.coderabbit.ai/change-stack/supabase/supabase/pull/46039?utm_source=github_walkthrough&utm_medium=github&utm_campaign=change_stack)
<!-- review_stack_entry_end -->
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Alaister Young <10985857+alaister@users.noreply.github.com>
Mark provenance of SQL via the branded types SafeSqlFragment and
UntrustedSqlFragment. Only SafeSqlFragment should be executed;
UntrustedSqlFragments require some kind of implicit user approval (show
on screen + user has to click something) before they are promoted to
SafeSqlFragment.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Editor and RLS tester show loading states for inferred/generated SQL
and include a dedicated user SQL editor for safer edits.
* **Refactor**
* Platform-wide SQL handling tightened: snippets and AI-generated SQL
are treated as untrusted/display-only until promoted, improving safety
and consistency.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Context
Resolves FE-3126
Just cleaning up the table editor header with a bit of refactors
(pre-req to investigating collapsing filter bar and table editor header
actions into a single row)
## Non-visual changes involved
- Break down components within `GridHeaderActions` into smaller ones
- `IndexAdvisorPopover`
- `SecurityDefinerViewPopover`
- `RealtimeToggle`
- Deprecate use of `useUrlState` in `GridHeaderActions` to use
`useQueryState` instead
- Improve types for `TwoOptionToggle`
## Visual changes involved
- Collapse realtime button toggle into a button icon, with no text (just
tooltip)
- Adjust layout of buttons a little
### Before
<img width="796" height="118" alt="image"
src="https://github.com/user-attachments/assets/436bca94-4d91-471a-a184-487c6f78dc04"
/>
### After
<img width="731" height="132" alt="image"
src="https://github.com/user-attachments/assets/5fd30982-a1fc-4f92-a590-146d1e69d52a"
/>
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Index Advisor popover with recommendations.
* Realtime toggle to manage realtime table publication.
* Security Definer view popover with optional autofix.
* Insert menu for adding rows/columns and CSV import.
* **Bug Fixes**
* Adjusted filter bar input sizing for improved readability.
* **Refactor**
* Header layout updated and insert/import actions moved into dedicated
components.
* **Tests**
* Updated end-to-end selectors for the Insert row menu item.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## TL;DR
- extends https://github.com/supabase/supabase/pull/45076
- closes https://github.com/supabase/supabase/issues/45113
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Sequence values now properly sync after bulk row inserts, ensuring
correct auto-increment behavior for subsequent inserts.
* **Tests**
* Added end-to-end coverage for CSV import, including a "Paste text"
import path that verifies custom-owned sequences are synchronized before
the next insert.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Gildas Garcia <1122076+djhi@users.noreply.github.com>
Co-authored-by: Ali Waseem <waseema393@gmail.com>
## TL;DR
fixes csv imports with explicit id values leaving auto generated primary
keys out of sync
by resolving the primary key sequence correctly after import
## ex:
| Before | After |
| --- | --- |
| <img width="378" height="201" alt="Before: stale sequence after CSV
import"
src="https://github.com/user-attachments/assets/34c827f5-c69b-4c05-aa67-9fb6fd65a040"
/> | <img width="435" height="196" alt="After: sequence synced after CSV
import"
src="https://github.com/user-attachments/assets/bb84f286-1b3e-44a5-965b-5f2faa3f7622"
/> |
## ref:
- closes https://github.com/supabase/supabase/issues/45073
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **Bug Fixes**
* Fixed identity sequence synchronization during CSV imports. After
spreadsheet data is inserted, the system now properly updates sequence
values, ensuring subsequent rows receive correct identifiers.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
- Makes the new table filter bar the permanent default by hardcoding
`useIsTableFilterBarEnabled` to return `true`
- Removes the feature preview toggle (opt-out) from the preview modal
- Cleans up E2E tests: removes old filter UI test, removes
`enableFilterBar` helper, fixes race condition in column-drop test
- Old filter code paths are left in place for a follow-up cleanup PR
Closes FE-2819
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Removed the table filter bar preview and its opt-in/local-storage
preview key.
* Cleared preview content from the feature preview modal.
* **Tests**
* Removed the UI filtering e2e test and associated preview opt-in
helper.
* Updated filter-bar e2e flows to adjust navigation/wait behavior (added
explicit waits, removed redundant reloads).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
## Problem
Since the drag & drop refactor, users cannot double click to select the
column name in the columns headers of the table editor. Besides, the
advisor button cannot be clicked either.
## Solution
- Add a delay constraint on the drag & drop operation, allowing click
events to trigger
- Add e2e tests to prevent future reggressions
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Double-click column headers to copy column names to clipboard
* Index Advisor warnings now accessible in the Table Editor
* **Improvements**
* Improved drag start timing and tolerance for smoother column dragging
* Better screen reader announcement for index suggestion actions
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## TL;DR
the table editor definition panel was showing incomplete SQL for views
with `WITH (security_invoker = true)`
ignoring the reloption and making it easy to accidentally strip it when
recreating the view
## prob
When viewing a security invoker view in the Table Editor, the Definition
panel only showed `CREATE VIEW ... AS ...`
without the `WITH (security_invoker = true)` clause
which caused two issues:
1. the displayed SQL was incomplete and didn't match the actual view
definition
2. users copying the SQL to recreate the view would unintentionally lose
the security_invoker setting
## ex:
| Before | After |
|--------|-------|
| `create view public.exposed_api as`<br>`select id, secret from
public.rls_protected_table;` | `create view public.exposed_api with
(security_invoker = true) as`<br>`select id, secret from
public.rls_protected_table;` |
## ref:
- closes https://github.com/supabase/supabase/issues/44934
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* View definitions now show the full CREATE statement (including
materialized views and WITH (...) options) and preserve security options
like security_invoker when viewed or opened in the SQL editor.
* **Tests**
* Added end-to-end test verifying security option preservation in view
definitions and when opening them in the SQL editor.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Problem
We currently have 3 different libraries for drag & drop, two of which
are not actively maintained anymore.
## Solution
Migrate all usage of the two unmaintained libraries to DndKit.
This PR focuses on using DndKit instead of `react-dnd` for column
reordering in the table editor
## Screencast
https://github.com/user-attachments/assets/54fb36f4-5671-42e2-9698-2ae928a69f55
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Column drag-and-drop rebuilt with a live drag preview and improved
pointer/keyboard handling
* **Improvements**
* More reliable column reordering with refined move/freeze/unfreeze
behavior and stable index recomputation
* Standardized column type/format display and simplified grid rendering
* **Accessibility**
* Added accessible labeling for column actions controls
* **Tests**
* End-to-end locator updated to target the column actions button
deterministically
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Context
Resolves https://github.com/supabase/supabase/issues/43548
There's currently an issue with the Table Editor where if you have, for
example, a nullable `text` column with a default value, inserting a new
row and selecting "Set to NULL" doesn't do anything, and saving will
insert the row with the default value
<img width="700" height="258" alt="image"
src="https://github.com/user-attachments/assets/6a284ebb-c346-40a6-9a30-793118844084"
/>
This stems from a legacy logic in the Table Editor whereby we treat
`null` values as "no input" - which is incorrect as `null` values are
also valid values. So the PR here changes a few things to resolve this
properly:
## Changes involved
Main fix:
- `undefined` will be the "no input" value instead, and it'll be the
default value when generating the row object for inserting a new row
- `NULL` or even empty string like `''` will be treated as they are
(valid inputs)
Secondary adjustments:
- (Queue operations) Queueing an insert with no value but default value
is NULL, will show the placeholder as `DEFAULT` instead of `NULL` for
better accuracy in representation
<img width="892" height="96" alt="image"
src="https://github.com/user-attachments/assets/02cf86bf-c17b-4e25-9a8f-17960b1d2575"
/>
- Added a `Set to Default` CTA here, but will only show up if adding a
new row or updating a queued insert row operation, which will set the
value of the input field back to `undefined` for PG to handle it as the
default value
<img width="734" height="208" alt="image"
src="https://github.com/user-attachments/assets/23887c0c-533e-4494-acbe-61309ff5d7c5"
/>
## To test
Verify within the Table Editor (along with queue operation feature
preview)
- For inserting a new row, setting value to NULL and setting value to
Default works
- For updating a row, setting value to NULL works
## I have read the
[CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md)
file.
YES/NO
## What kind of change does this PR introduce?
fixes: #37454
The context menu on the FK row is a lie, its just exposing the cell
underneath. Ideally this should have its own context menu, right now we
only support copy operations.
---------
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
## I have read the
[CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md)
file.
YES
## What kind of change does this PR introduce?
Since queue operations is a feature users can opt-out of, we need to
make it cleaner to toggle between queuing vs straight edits. To do this,
refactor all the operations into a single hook and reference it in
places where we mutate the rows.
## Testing
- Test edit cells, rows, and deletes for non queue operations
- Test edit cells, rows, and deletes for queue operations, also double
check modifying the same rows that are not yet added
---------
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
## Summary
- fix the Studio SQL row formatter to emit raw boolean and numeric
literals instead of quoted strings
- serialize text array members as SQL string literals inside
`ARRAY[...]`
- escape fallback string formats outside `text`/`varchar`, and add
regression coverage for that path
Closes#44024
## Test plan
- [x] Ran a direct `tsx` smoke against `formatTableRowsToSQL()` for the
`storage.buckets` case and confirmed it now emits `true`, `false`,
`10485760`, and `ARRAY['image/*']`
- [x] Ran focused formatter smokes for JSON escaping, text arrays, and
fallback string formats like `citext`
- [x] `pnpm --filter studio test -- TableEntity.utils.test.ts`
Note: the targeted Vitest run is still blocked in this environment
before the test executes (`localStorage.getItem is not a function`).
---------
Co-authored-by: Alaister Young <alaister@users.noreply.github.com>
## Problem
Some tests rely on hard coded timeouts. That makes them
- brittle if the timeout is not long enough
- take longer than necessary if the timeout is too long
## Solution
- Rely on playwright `expect` retries when possible
- Rely on UI updates when possible
Fixes#42089
Supersedes #43352 and
https://github.com/supabase/postgres-meta/pull/1049
## Problem
When creating a table in the Dashboard, if a column-specific error
occurs (e.g., invalid enum value or broken foreign key), the Dashboard
displays an error toast, but the side panel remains open. In the
background, the table is actually created (partially), leading to a
_42P07: relation already exists_ error if the user tries to click _Save_
again.
## Solution
We actually had nested transactions which is not supported by Postgres.
- [x] Allow generating SQL without transactions from `@supabase/pg-meta`
- [x] Fix the table creation
- [x] Fix some accessibility issues
- [x] Add tests
## How to test
1. Click _New table_ in the Table editor
2. Add a column with an _int8_ type and set its default value to
`bazinga`
3. Click _Save_
You should see an error.
4. Fix the default value by setting it to `10`
5. Click _Save_
Results:
- You should see a success message about the table creation
- The table should have the column with the correct default value
## Problem
e2e tests are still flaky and not as fast as they could be
## Solution
- [x] Reset supabase instance: this makes the database visualiser tests
faster as there are less tables to screenshot
- [x] Improve the single file setup utilities so that they never block
local tests by cleaning them up before starting
- [x] Disable animations while running the tests (less time waiting for
animations to complete
- [x] Add utility functions that help reproducing flaky tests locally
## Context
Addresses an error when drag and dropping a CSV in the Table Editor when
the table is empty, specifically on Safari leading to a `undefined is
not an object (evaluating 'n.type')` error
For context, Safari obscures the item data from the drag event and only
exposes it on the drop event
([ref](https://github.com/mdn/browser-compat-data/issues/24898)) hence
why this issue
## Changes involved
Am opting to change the UX instead of fixing this browser specific
behaviour
- Ignore file validity when dragging, let the on drop handle validate
the type and reject unsupported files instead
- This is better either way as currently our Table Editor shows "CSV
files only", when we actually also support TSV files
- Also made a slight adjustment to the drag over UI - border around the
grid empty area, rather than the whole grid
- Before:
<img width="1151" height="911" alt="image"
src="https://github.com/user-attachments/assets/049199aa-86bc-4e88-8444-5a7ee236c8d0"
/>
- After:
<img width="1150" height="912" alt="image"
src="https://github.com/user-attachments/assets/698ca8f3-a326-4a82-b1b4-4448aa8d777a"
/>
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Refactor**
* Simplified the empty-table display and CSV import interface for a more
consistent user experience.
* Import button and drag-and-drop hints now remain visible when the
table is empty, removing prior conditional hiding during drag.
* Streamlined empty-state styling and messaging for a cleaner, more
predictable appearance.
* **Tests**
* Updated end-to-end test expectations to reflect the persistent
empty-state message during drag-over.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Cleanup / chore — removing a feature flag that has been at 100% rollout
with no issues.
## What is the current behavior?
The `dataApiExposedBadge` feature flag is checked at runtime via
ConfigCat, even though it has been set to 100% for a while with no
issues.
## What is the new behavior?
The feature flag check is removed and the gated behavior is now
unconditionally enabled. A reminder has been set to delete the flag from
ConfigCat in a month.
## Summary by CodeRabbit
**Refactor**
* Simplified API access status indicators in the studio editor, ensuring
security-related tooltips now display consistently based on actual
access conditions.
* Updated component interfaces and consolidated internal hook
dependencies for improved code maintainability and organization.
## I have read the
[CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md)
file.
YES
## What kind of change does this PR introduce?
Chore: test drag and drop on table editor
follow up for https://github.com/supabase/supabase/issues/42655
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Added comprehensive end-to-end test for CSV drag-and-drop import
functionality that validates the complete import workflow: drag-drop
detection on empty tables, import dialog display with accurate row
counts, and successful data import to the grid.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
* attempt to make table editor tests less flaky
* updated race conditions for table editor
* updated playwright config with optimized launch options
* reduce workers
* updated timeouts for tests
* updated config to have a global timeout of 20 seconds
* updated to be visible
Exporting all rows (in CSV, SQL, or JSON format) currently uses offset pagination, which can cause performance problems if the table is large. There is also a correctness problem if the table is being actively updated as the export happens, because the relative row offsets could shift between queries.
Now that composite filters are available in postgres-meta, we can change to using cursor pagination on the primary key (or any non-null unique keys) wherever possible. Where this is not possible, the user will be shown a confirmation dialog explaining the possible performance impact.
---------
Co-authored-by: Ali Waseem <waseema393@gmail.com>
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
* updated because of zero falsey value not copying
* test: add E2E test for copying first and second row values
Adds regression test to verify that users can copy cell values from
the first row (index 0) and second row. This prevents the bug where
falsy index checks would prevent copying from row 0.
Addresses FE-2178
Table editor tests were being run serially because they would otherwise
stomp all over each other in setup. But this made them really slow. This
PR changes them to run in parallel, and makes sure the setup step runs
only once by using a file lock.
Also cleans up miscellaneous other sources of flakiness in tests.
Co-authored-by: Ali Waseem <waseema393@gmail.com>
* updated selector code to be more efficient
* updated styling to show a cleaner table loader
* updated format assignment
* disable selector when loading
* updated request to fetch tables
* updated key rendering
* a new hope
* run tests in ci against cli mode
* summary
* try vercel action to run e2e against studio self hosted preview
* believe
* debug
* gh pages artifact
* test
* rm pages step
* fix automation bypass missing
* continue on error
* only install necessary deps for CI
* fix bypass
* remove
* fail job if test fails
* disable customer query if is_platform false
* vercel check
* fix var name, make comment update instead
* check bypass on runtime
* add env var
* fix tests going to project ref instead of default
* fix
* better dates in comment
* Update E2E test workflow to include flaky test detection and improve summary output
* fix
* fix dumb mistake
* chore: add final e2e tests for table-editor and sql-editor
* chore: update tests to run in staging
* chore: minor updates
* chore: fix PR feedback
---------
Co-authored-by: Jordi Enric <37541088+jordienr@users.noreply.github.com>
* chore: add e2e test for sql editor snippets
* trigger build
* chore: fix failing test and add rls
---------
Co-authored-by: Terry Sutton <saltcod@gmail.com>
