#17765 closed enhancement (fixed)
[Patch] Move code for error "Area style way is not closed" to UnclosedWays
Reported by: | GerdP | Owned by: | team |
---|---|---|---|
Priority: | normal | Milestone: | 19.05 |
Component: | Core validator | Version: | |
Keywords: | Cc: | Klumbumbus |
Description
Don't know why this message is produced by MultipolygonTest
instead of UnclosedWays
.
The message is produced for unclosed ways with e.g. just one tag highway=services.
The attached patch moves the corresponding code from MultipolygonTest
to UnclosedWays
and the message to the error group "Unclosed way" with new error code 1131 (old was 1609)
@Klumbumbus:
With r15094 we get two messages for such a way with highway=services. Is that intended?
Attachments (1)
Change History (11)
by , 6 years ago
Attachment: | 17765.patch added |
---|
comment:1 by , 6 years ago
comment:2 by , 6 years ago
Ah, probably the list in the new rule needs $
and ^
so that service doesn't match services?
comment:3 by , 6 years ago
Ah, now I get it. So yes and no. The missing $ and ^ are intended to match all the ..._link highways. However services should not match as this is used on areas. I'll fix the rule.
comment:5 by , 6 years ago
(I wanted to add asserts but they didn't work due to the :AreaStyle pseudo class)
comment:9 by , 6 years ago
Oops, I only looked at the file multipolygon.osm...
I'll add a new test for UnclosedWays.
For the example tag combination from #17623 (highway=trunk + junction=yes) on a unclosed way the "Area style way is not closed" warning doesn't show up, thats why I wasn't aware of that test while creating r15094.
The "Area style way is not closed" test and the second rule in r15094 have little different goals. (The first "assumes" the area is not closed, i.e. error in geometry while the second "assumes" that there is an error in the tags.) Not sure if and how they could be merged.