* only send permissions updates for groups that have been modified in the graph

* Support partial permission graph updates on the BE (#44797)

* handles keeping track of advanced permissions changes to send relevant group data when updating the graph, merges response

* tests and optimizations

* respect group-ids set when fetching connection impersonations, sandboxes, and admin group perms

* fix typo

* add e2e test

* fix BE test

* impls pr feedback, reduces code needed since advanced permissions state is always reset on save

* last fixes

---------

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

* recent views sometimes don't show special colls

colls with a non-null namespace are special: they can't interact with
normal collections, dashboards, cards, etc at all. (Currently the only
use for these we have is snippets.) At some point we could add a
`namespace` argument to this endpoint to support looking at recent
collections in a different namespace, but that'd be a bigger refactor
without a whole lot to gain. For now we can just exclude these
completely, which is what we want for current usages of the recents
endpoint anyway.

We won't exclude non-null types across the board, because sometimes
we'll want to see these. But we will be particular about when to display
them.

We'll never display the `trash` collection, but will display the
instance analytics collection when we're not trying to make selections.

* Hack an e2e test around an existing bug

So, the entity picker that pops up when you're selecting a collection to
put a Snippet into shows you "Recent" collections - which previously
included snippet collections.

We should either make the Recents API support a namespace parameter for
collections so it could actually support this case, or the entity picker
shouldn't show a recents tab here (realistically, how many collections
are people going to have for snippets that they would need this tab
anyway).

* Calculate metadata without running queries [JDBC]

* Appease Kondo

* Test fixes 

* Overhauled Card metadata calculation code

* Test fixes 

* Kondo fix 

* Ok maybe I fixed stuff

* Work around https://github.com/aws/amazon-redshift-jdbc-driver/issues/118

* Maybe fixx the Cypress tests

* Experimental possible e2e test fix 

* Test fixes 

* Do things a little differently.

* Try re-enabling the field refs

* question marks

* remove all question mark notes except about json

Fix - Invalid join clause does not disappear after changing query source to use a different database (#45025)

* Reset JoinDraft state on database change (fixes metabase#42385)

* Add repro for #42385

* Use admin user

* add default font family

* use metabase provider internal component for tests

* Add repro for #34350

* Reduce diff

* make ad-hoc changelog generation actually useful

* add github action

* remove test code

* fix whitespace

* fix formatting

Add a repro for 'Permission error before finishing referencing a question in the native query' (#44995)

* Create new release note sections

* add explanatory text

* return data about the proper billing status

* fix copy

* fix the tests

* support sorting on funnel bar charts

* fix funnel bar row display setting

* spec

* mark a defsetting as being :sensitive?

* add test

* Fix SDK components loaders

* Review fixes

* Add reproduction for #38989

* Use saveQuestion instead of clicking trough the UI

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

* Use openOrdersTable instead of manually clicking through the ui

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

* Create question programmatically in test

---------

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

* Add a test table with many schemas

* Add a repro for #28106

* Format code

* Add issue tag

Repro #27521 - Query fails when selecting two fields from nested query, when fields are identical except for their join-alias (#44909)

* Add repro for #27521

* Improve assertions

* Add comments

* Use saveQuestion helper

* Use openOrdersTable helper

Add a reproduction for "Cannot find the new column in pick columns after alter schema and synced" (#44882)

* Add a reproduction for metabase#41765

* Navigate manually to the admin page

* Wrap non-latin characters in a span specifying working font

Fixes: #38753

CSSBox seems to have a bug where font fallback doesn't work properly. This is noticeable when a font does not contain
glyphs that are present in the string being rendered. For example, Lato does not have many international characters,
so the rendered version of tables (that show up in Slack messages) will not render properly, making the card
unreadable.

Since this appears to be a downstream bug, I've opted to work around this limitation by wrapping any non-latin
characters in a <span> that specifies the font family to be sans-serif, which should contain the glyphs to properly
render.

This leaves Lato in place for other characters.

For now, I figured it's worth trying this solution over using Noto for 2 reasons:
- we can keep Lato, which has been the decided font style for the app for some time (this keeps things consistent
where possible)
- the Noto font containing all glyphs is quite a large font (>20mb, I think) and it would be nice to avoid bundling
that if we can.

* stub installed fonts test

* typo

* Do wrapping, but now per-string, and in Clojure data not html string

I've decided that a reasonable solution is to still wrap strings containing non-lato characters. But it's not done
with str/replace to the html string but rather in a postwalk over the Hiccup data prior to rendering.

Seems like a decent compromise of issues without patching CSSBox or fixing upstream (may be good to do, but will take
longer to get a fix in).

* add test checking that glyphs render correctly

* Add a test that directly checks the wrapping fn

* Change the string to keep the linter quiet

* Change how we check if string can be rendered to faster method, new Lato Font

With Sanya's help, the way I check if a given string is renderable with Lato is now faster.

Also use the full Lato font, not just the 'latin' lato so we can cover more chars

* change lato so that it loads the fn even in a fresh test run

* improve cartesian charts performance

* drop unused import

Add a repro for 'Filter values in dropdown are not getting populated in SQL question based dashboard' (#44927)

* prevent pin maps from filtering out rows with null values

- since pin maps are just displaying where the data is by latitude and longitude, we do not need to filter the pin just because one of the rows is null.
- the metric column selection does not make sense for pins as they just represent that data through coordinates (no aggregations or values displayed), so we should not be filtering out any data just because it is null

* add unit test