* Add Unit Test Stress Test Flake Fix workflow

* Add "frontend" prefix to the workflow to disambiguate between backend and frontend

This is a follow-up after #38011.
See this comment: https://github.com/metabase/metabase/pull/38011#discussion_r1462885451

Resolves https://github.com/metabase/metabase/security/dependabot/130

* Revert "Fix trailing whitespace (#38057)"

This reverts commit 1de0326c.

* Revert "Add upload test results to Trunk (#37490)"

This reverts commit a36347a1.

* Fix flake in fingerprint test:

```
FAIL in metabase.sync.analyze.fingerprint-test/fingerprint-test (fingerprint_test.clj:287)

:sparksql
 Fingerprints should actually get saved with the correct values Text fingerprints
expected: {:global {:distinct-count 100, :nil% 0.0},
           :type {:type/Text {:average-length #<Fn@2ee5402f metabase.sync.analyze.fingerprint_test/fn__139755[fn/fn]>,
                              :percent-email 0.0,
                              :percent-json 0.0,
                              :percent-state 0.0,
                              :percent-url 0.0}}}
  actual: nil
```

* Typo fix

Fixes #37345.

Fixes https://github.com/metabase/metabase/issues/38009

### Description

This makes the feature flag loading code more robust by making the function re-entrant. To understand how this function could deadlock before, see the description in https://github.com/metabase/metabase/pull/38027.

It works by ensuring that the call to initialize the nonce comes from the same thread that initially locked on fetching the tokens.

There is a subtle behavior change here - was allow for a 10s timeout on each call individually now. Since it is unlikely that both endpoints will timeout consecutively, and the combined wait time is still fairly short, I think this is OK.

### How to verify

You can use the same steps as https://github.com/metabase/metabase/pull/38027, either of these PRs are sufficient to fix the bug.

Resolves #38010

* Formatting for time types

Currently, the frontend does not do any special formatting for time of day types. These are always formatted as 12 hour AM/PM times. This PR adds a new function to `frontend/src/metabase/lib/formatting/time.ts`, `formatTimeWithOptions`, that applies user-defined formatting to the time. Previously, `formatTime` was called, which takes no formatting options, so none can be applied.

Fixes #11398

* e2e tests for time of day formatting.

* Update frontend/src/metabase/lib/formatting/time.ts

Co-authored-by: Emmad Usmani <emmadusmani@berkeley.edu>

* Replacing formatTime with formatTimeWithOptions

This replaces uses of `formatTime` with `formatTimeWithOptions`, updates unit tests, removes references to `formatTime`, and incorporates some code suggestions.

* Unit tests

* Adding test id for field info popover.

* Refactoring `formatTimeWithOptions` to `formatTime`

* Moved `TimeOnlyOptions` interface to formatting types and made it an ancestor of `OptionsType`

---------

Co-authored-by: Emmad Usmani <emmadusmani@berkeley.edu>

Fixes #36856.

* fix auto y-axis split on native queries

* specs

* Fix wrong `target` prop type on `DashCardParameterMapper`

* Remove unneeded prop from the test setup

* Filter out audit tables in search results

* add test

* Use ASCII arrow characters instead of SVG icons

* Increase comparison text font size

* Fix salt in pulse unsub hash
* Workaround for read-only settings not supporting temp overrides yet

Ensure we read and update nested JSON in a transaction

- Introduce an "init" hook for application settings.
- Add guards against introducing race conditions again in other ways.
- Fix the usage of hygienic scope macros with read only settings.
- Lots of misc improvements to settings API and implementation.

This init hooks better ergonomics, and protects against race conditions. 
The previous pattern of calling setters inside getters almost caused a P0. 

See https://github.com/metabase/metabase/issues/37801.

* Migrate Question `isNative` and `isStructured` methods to MLv2

* Mark methods as deprecated

* Add the guard to the `isNativeDashCard` util

* Add additional guard to the `buildFieldFilterUiParameter` helper

* Update all `isNative` method instances to MLv2

* Fix the negation logic 



* Remove `isNative()` method altogether

* Simplify the `shouldResolveFkField` check

Co-authored-by: Kamil Mielnik <kamil@kamilmielnik.com>

* Remove explicit Boolean conversion

* Simplify conditional logic for `getTargetsForQuestion`

---------

Co-authored-by: Kamil Mielnik <kamil@kamilmielnik.com>

To prevent ambiguous dashboard URLs, labels should not be called 'tab'.

* Enabling dragging of long-named TabButtons

The primary change needed to support rearranging of tabs with super long names is the to shorten the names only when dragging. This was achieved by exposing the `isDragging` prop of the `useSortable` hook in the `RenameableTabButtonStyled` component. The label of this component is dynamically shortened if it is very long on a drag operation by the following logic and function:

```
// Update the label prop when dragging
label={isDragging ? dragLabel(label) : label}
```

The `dragLabel` function is as follows:
```
  const dragLabel = (s: string) => {
    if (s.length < 20) {
      return s;
    } else {
      return `${s.slice(0, 17)}...`;
    }
  };
```

The other change needed to support e2e testing for this operation was the addition of the `MouseSensor` hook to the DnDContext. See [this issue comment](https://github.com/clauderic/dnd-kit/issues/208#issuecomment-824469766) for explanation.

Once that change was in place it was possible to do a drag and drop rearrange test in the "should allow me to rearrange long tabs (#34970)" spec.

This test:
- Creates a dashboard with 3 tabs
- Ensures their order
- Gives the last tab a super long name that would not be rearrangeable before fix
- Drags that tab to the first position
- Saves the dashboard
- Asserts the new tab order is saved

Fixes #34970

* Fixing type from T to unknown to satisfy the type of the render function in TabButton.unit.spec.ts

* Fixing type from T to any to satisfy the type of the render function in TabButton.unit.spec.ts

* Fixing divergent tests

The tests to ensure DnD works required the addition of the `mouseSensor` in TabRow.tsx. This, in turn, broke the tests in DashboardTabs.unit.spec.tsx. Adding in the `activationConstraint: { distance: 10 }` prop to the mouseSensor fixed the DashboardTabs.unit.spec.tsx tests, but broke the DnD tests. Based on the docs [here](https://docs.dndkit.com/api-documentation/sensors/mouse) it looks like you need to move a "distance, in pixels, by which the mouse needs to be moved before a drag start event is emitted." So, the DnD test in tabs.cy.spec.js added an addtional mousemove trigger of 11 pixels to activate the mouseSensor. Yeah, it totally makes sense. 

* prettier

Part of #37173.

Co-authored-by: Noah Moss <32746338+noahmoss@users.noreply.github.com>

* Cleanup component

* Replace filters with MLv2 methods

* Cleanup unused functions

* Make filter work

* Make filter work

* Fix rebase issues

* Update ML methods usage

* pass params in the correct order

* pass params in the correct order

* Fix argumetns order

* Remove duplicated test and unskip correct one

* [Leaflet] Replace brush filter with MLv2 alternative (#37295)

* Update ML methods usage

* Fix argumetns order

* Unskip test

* Pass metadata

* Pass metadata

* add force render to auth settings, read only auth card

* adding footer prop to auth cards, force render differt auth cards

This reverts commit a5dd4444.

These are helpers for going from an aggregation or breakout
clause to the relevant `:metadata/column`. (Nil for top-level
aggregations like `:count`.)

Fixes #37120.