bugfix/17086-mapline-in-non-topojson

[kamil-musialowski]

Fixed #17086, mapline didn't work with lon/lat coordinates on non-TopoJSON maps

[highsoft-bot]

File size comparison

Sizes for compiled+gzipped (bold) and compiled files.

	master	candidate	difference
highmaps.js	125.4 kB 381.3 kB	125.4 kB 381.5 kB	39 B 244 B

[highsoft-bot]

Visual test results - Differences found

Found 1 diffing sample(s). Please review the differences.

[TorsteinHonsi]

Hi, this is how geometries work with the legacy GeoJSON files. Not because they are GeoJSON, but because they have the crs specifying the projection and the hc-transform attribute that defines how geometries should be translated. So basically, all geometries in the legacy files are pre-projected.

The Point geometry is an exception because I made it work identically to the lon and lat point options. So if there is a bug, that is it. But if no-one is reporting it, I don't think we should waste more time on supporting the legacy maps.

In the OP demo, you can see that the fix only works with LineString, but not MultiLineString, Polygon etc.

Please revert the added code in this PR. But the question is if we should add some heads-up in the docs under the geometry option?

[KacperMadej]

@TorsteinHonsi

But the question is if we should add some heads-up in the docs under the geometry option?

That would be very nice. The current description mentions only geoJSON (by the way, for consistent should it be GeoJSON?). Should it also contain info about the geometry of 'Point' type working differently? or is that just a small inconsistency waiting to be addressed when reported?

[TorsteinHonsi]

That would be very nice. The current description mentions only geoJSON (by the way, for consistent should it be GeoJSON?).

Yes, should be GeoJSON. The description is technically correct. TopoJSON is missing because it unpacks into GeoJSON, but should probably be mentioned nonetheless.

The destinction in the behaviour is not between GeoJSON and TopoJSON per se, but between sources that are pre-projected (like our legacy GeoJSON files are), and those that are not (like our TopoJSON files are). So I guess this should be described.

Should it also contain info about the geometry of 'Point' type working differently? or is that just a small inconsistency waiting to be addressed when reported?

I think the description of the mappoint.data.geometry is correct, so no need to elaborate there for the time being.

[kamil-musialowski]

Moved to #18148