Duplicate informal warnings about missing `crossing=*`

[skyper]

What steps will reproduce the problem?

1. Have a footway and a residential road crossing with a shared node tagged highway=crossing
2. Run Validator

What is the expected result?

One informal warning about missing crossing=* from mapcss

What happens instead?

Two informal warnings:

• Incomplete pedestrian crossing tagging. Required tags are highway=crossing|traffic_signals and crossing=*. from java test "Highways"
• highway=crossing without crossing from mapcss

Please provide any additional information below. Attach a screenshot if possible.

The java test should not warn about highway=crossing with missing crossing=* as this is covered in mapcss.

Relative:URL: ^/trunk
Repository:UUID: 0c6e7542-c601-0410-84e7-c038aed88b3b
Last:Changed Date: 2021-05-17 21:27:21 +0200 (Mon, 17 May 2021)
Revision:17903
Build-Date:2021-05-18 01:31:02
URL:https://josm.openstreetmap.de/svn/trunk

[GerdP]

Hmm, the java test is quite complex and doesn't warn when the footway ends on the residential road.
Not sure if that makes sense. I expected a message for a possibly obsolete highway=crossing tag for this case, but it is not produced.

[GerdP]

Possible patch to just get rid of the duplicate informational message:

src/org/openstreetmap/josm/data/validation/tests/Highways.java

@@ -211 +211 @@
         pedestrianWays = 0;
         cyclistWays = 0;
         carsWays = 0;
-
+        if (n.hasTag("highway","crossing") && !n.hasKey("crossing"))
+            return; // see #20905 handled by mapcss test
         for (Way w : n.getParentWays()) {
             String highway = w.get(HIGHWAY);
             if (highway != null) {

[skyper]

Replying to GerdP:

Hmm, the java test is quite complex and doesn't warn when the footway ends on the residential road.

No warning with only a single way from one side is on purpose, see #9141.

Not sure if that makes sense. I expected a message for a possibly obsolete highway=crossing tag for this case, but it is not produced.

#9245?

Recently, highway crossings tagging has been discussed on the German forum, but without a clear conclusion when to use crossing=unmarked and/or crossing=no and when to not tag the node at all.

[GerdP]

No warning with only a single way from one side is on purpose, see #9141.

Yes, I think's it's correct to not warn, but the mapcss test show's a warning as it doesn't check the existence of ways.

I'd rather remove the mapcss test here.

[skyper]

The mapcss tag is about a single missing subkey and it is useful even in cases with a single footway only from one side.

I prefer your patch in silencing the java test if highway=crossing is present.

[GerdP]

OK, will commit that patch next month.

[Don-vip]

Milestone renamed

[GerdP]

In 18273/josm:

fix #20905: Duplicate informal warnings about missing crossing=*

• get rid of the duplicate informational message for node with highway=crossing and no crossing=* tag