implement unselectable features

[warm-coolguy]

Summary

• For the GFI, a method has been added to mark features as unselectable. They are not styled differently, but just ignored. If visible/displayed in another way, styling has to remove them there, too.
• For the extendedMasterportalapiMarkers in the core, the feature has been doubled; since it is orthogonal to GFI, two configurations are required to get the feature running in this mode, one on the markers, one in the GFI. However, styling has been resolved here, as it is internal to this feature.

Instructions for local reproduction and review

Use the snowbox to check the behaviour. Use cases have been added.

• GFI (pure): Switch the "Compensation area" layer visible. The feature left of the Flugzeugwerft (left of Finkenwerder) should not be clickable, but the one to its right should be.
• GFI + extendedMasterportalapiMarkers: Switch the "Reports (MML)" layer on. Arbitrarily, features with uneven ID have been deemed unselectable. Check whether the behaviour seems plausible/correct to you.

You may also test GFI on the GeoJSON layer by removing the extendedMasterportalapiMarkers and enableClustering related code. In that case, all pins are visible, but only pins with even number display a GFI window. As in the first case, styling will have to be added later on the layer itself.

Also, please confirm that nothing in the Meldemichel broke.

[dopenguin]

If I configure featureList, non-selectable features are still selectable. Thus, they also change their styling if bindWithCoreHoverSelect is used.
I would have expected that features that not selectable in the map are not being shown in the featureList. Would you agree?

Edit: Meldemichel seems to be working fine as far as I can tell.

🏓 @warm-coolguy

[warm-coolguy]

If I configure featureList, non-selectable features are still selectable. Thus, they also change their styling if bindWithCoreHoverSelect is used. I would have expected that features that not selectable in the map are not being shown in the featureList. Would you agree?

🏓 @warm-coolguy

Yes, I agree. As discussed before, I'm shoveling the PR over to you and added you as assignee. Should any question about my implementation so far arise, call me anytime.

🏓 @dopenguin

[dopenguin]

If I configure featureList, non-selectable features are still selectable. Thus, they also change their styling if bindWithCoreHoverSelect is used. I would have expected that features that not selectable in the map are not being shown in the featureList. Would you agree?
🏓 @warm-coolguy

Yes, I agree. As discussed before, I'm shoveling the PR over to you and added you as assignee. Should any question about my implementation so far arise, call me anytime.

🏓 @dopenguin

I've fixed that behaviour in c38065c.
If you want to reproduce it, I've attached a patch with the relevant changes to @polar/client-snowbox.

🏓 @warm-coolguy

[warm-coolguy]

If I configure featureList, non-selectable features are still selectable. Thus, they also change their styling if bindWithCoreHoverSelect is used. I would have expected that features that not selectable in the map are not being shown in the featureList. Would you agree?
🏓 @warm-coolguy

Yes, I agree. As discussed before, I'm shoveling the PR over to you and added you as assignee. Should any question about my implementation so far arise, call me anytime.
🏓 @dopenguin

I've fixed that behaviour in c38065c. If you want to reproduce it, I've attached a patch with the relevant changes to @polar/client-snowbox.

🏓 @warm-coolguy

🚀 @dopenguin Checked and just works. Nice.

[dopenguin]

If I configure featureList, non-selectable features are still selectable. Thus, they also change their styling if bindWithCoreHoverSelect is used. I would have expected that features that not selectable in the map are not being shown in the featureList. Would you agree?
🏓 @warm-coolguy

Yes, I agree. As discussed before, I'm shoveling the PR over to you and added you as assignee. Should any question about my implementation so far arise, call me anytime.
🏓 @dopenguin

I've fixed that behaviour in c38065c. If you want to reproduce it, I've attached a patch with the relevant changes to @polar/client-snowbox.
🏓 @warm-coolguy

🚀 @dopenguin Checked and just works. Nice.

🚀 @warm-coolguy weeeee