Prevent inaccessible links from being displayed in LinkFormatter

republishing after discussion that this can be handled in public.

MR created (PS: Nightwatch tests seem unstable, I had to re-run).

Please check #4 and #13, I feel like this issue extends beyond link formatters so leaving to needs work...
Here is a static patch from MR for composer.

Rebased and moving to review, to get feedback on the approach, thanks.

Left some comments on MR.

Thanks, though I wasn't looking for a code review yet (tests and upgrade path still todo) but rather opinions on the overall approach / scope as it may extend way beyond link formatters.
See #13.

I am marking the following as duplicates:

- ✨ Only display internal links to content the user has access to view Closed: duplicate -- older, but has no code
- ✨ Link field > Access to internal links is not checked on display. Active -- more recent, has a MR
- #2968609: Link fields do not check access for entity refs → -- older, has a MR, but the MR has not been updated in longer

Code from those issues with MRs should be compared with what is here, in case there are ideas that should be integrated into the MR here.

Participants in all those issues should have credit added here.

✨ Link field > Access to internal links is not checked on display. Active 's MR is pretty basic.

#2968609: Link fields do not check access for entity refs → 's MR has a lot of stuff that isn't here.

1. More work on the URL, e.g.:

      if ($url->isRouted() && preg_match('/^entity\.(\w+)\.canonical$/', $url->getRouteName(), $matches)) {
        // Check access to the canonical entity route.
        $link_entity_type = $matches[1];
        if (!empty($url->getRouteParameters()[$link_entity_type])) 

Should we be doing that here?

2. More stuff on caching that I've not looked at in depth

3. A function test (though it could really be a kernel test)

Also, outputting broken internal links really looks like a bug to me -- one of the duplicate issues filed it as such too.

I've tested the current MR with field data in all forms:

- a path alias to a node: /mynode
- a system path to a node: /node/1
- using the lookup to store the entity

In all cases, the item is skipped when the user doesn't have access, so that's working well!

I don't know about the caching side of things though -- see my comment above.