noexit=yes is displayed over road class in object descriptions

[matkoniecz]

See https://www.openstreetmap.org/way/714073113

(yes, noexit=yes on way is a bad tagging scheme - but accepted enough to make mass removing it as approach that is not viable)

iD using its presets displays it correctly - see https://www.openstreetmap.org/edit?editor=id&way=714073113#map=19/52.61365/13.43482

This seems distinct from building=yes displaying bad labels for multipolygon buildings as highway=living_street is defined without catchall.

[Helium314]

I suspect the difference comes from SC determining feature by tags only.
iD probably doesn't show No Exit because noexit=yes is not valid on a way, but SC doesn't check geometry type, see

StreetComplete/app/src/main/java/de/westnordost/streetcomplete/util/NameAndLocationLabel.kt

Lines 75 to 76 in 7a976ed

	// not for geometry because at this point we cannot tell apart points and vertices
	// .forGeometry(element?.geometryType)

(now with the cache, such a geometry check might be more feasible)

[matkoniecz]

Also, even if points and vertices or way and areas cannot be distinguished we can still distinguish points/vertices from ways/areas.

[westnordost]

Let's do the solution @matkoniecz mentioned, it doesn't introduce any more new calls to the (hopefully) cached data but also solves this issue.