[FEATURE] Adjustment to ShapefileReader

[jbellander]

Is your feature request related to a problem? Please describe.
I'm using the shape file linked below. But the reader as it is now does not accept the file as ShapefileReader currently is implemented.

https://geodata.naturvardsverket.se/nedladdning/friluftsliv/Anordningar_shp.zip

Just as FYI that shape originates from Sweden's Environmental Protection Agencys open data.

Describe the solution you'd like
ShapefileReader:56 changed from transformToLatLon = CRS.findMathTransform(src, dest); to transformToLatLon = CRS.findMathTransform(src, dest, true); solves the issue. However not everyone will want to have it this so and option is probably needed.

Describe alternatives you've considered
n/a

Additional context
n/a

[msbarry]

Hello @jbellander ! I tried opening that shapefile in planetiler and I get an error Bad reference system ... Bursa wolf parameters required - Are you passing in an explicit projection string?

Looks like the full coordinate reference system specification is:

PROJCS["SWEREF99_TM", 
  GEOGCS["GCS_SWEREF99", 
    DATUM["D_SWEREF99", 
      SPHEROID["GRS_1980", 6378137.0, 298.257222101]], 
    PRIMEM["Greenwich", 0.0], 
    UNIT["degree", 0.017453292519943295], 
    AXIS["Longitude", EAST], 
    AXIS["Latitude", NORTH]], 
  PROJECTION["Transverse_Mercator"], 
  PARAMETER["central_meridian", 15.0], 
  PARAMETER["latitude_of_origin", 0.0], 
  PARAMETER["scale_factor", 0.9996], 
  PARAMETER["false_easting", 500000.0], 
  PARAMETER["false_northing", 0.0], 
  UNIT["m", 1.0], 
  AXIS["x", EAST], 
  AXIS["y", NORTH]]

[jbellander]

I should be honest to say that I'm on very deep water when it comes this topic. I got the same error as you. But setting lenient to true makes it work. I understand based on the article below, is that it will not make it 100% correct but good enough for display. (Which is good enough for me).

https://docs.geotools.org/stable/userguide/faq.html#q-bursa-wolf-parameters-required

Edit: Missed the link..

[msbarry]

Ahh I was getting mixed up with CRS.decode(projection, longitudeFirst);

I'd like to move the sources to accept individual arguments like source.shp?lenient=true&projection=EPSG:3857 but that's not quite supported yet.

What do you think in the meantime what if we log a warning in this case and retry with lenient=true? So at least it will accept the input but let you know there might be issues?

Do you have a minimal shapefile that reproduces this case that I could include in the repo as a test case?

[jbellander]

I understand. But doing it as a fallback with a warning would be a simple solution that would work fine for my purposes. Nice!