[pdal] Overlaying uint64 ids into las file

[Stephen Knox]

I have another question.

I am reading a LAS 1.2 file (laz compressed), trying to overlay it with a
geopackage with 64bit integer ids, and then save it out to a LAZ again.

Full pipeline is:

[
    "/data/input.LAZ",
    {
      "type":"filters.range",
      "limits":"Classification[6:6]"
    },
    {
      "type":"filters.ferry",
      "dimensions":"=> BldId"
     },
    {
      "type":"filters.overlay",
      "dimension":"BldId",
      "datasource":"/data/all_tiles_extract.gpkg",
      "column":"bld_id", (gdal reads as Integer64)
      "layer":"layer_name"
    },
    {
      "type":"filters.range",
      "limits":"BldId[1:18446744073709551615]"
    },
     {
        "type":"writers.las",
        "filename":"/data/blds_filter_then_overlaylaz",
        "extra_dims":"all",
        "minor_version":4
    }
]

My resulting BldId dimension has a max value of 2147483647,, which is the
signed 32 bit integer max. The schema is:

      {
        "name": "BldId",
        "size": 8,
        "type": "double" (or unsigned)
      }

Things I have tried:
- Converting the 1.2 LAZ to 1.4 before making the conversion
- Changing the code so that default double is an unsigned int (as above)
- Using the name of an "known" uint64 field, e.g. ShotNumber
- reading the file with --readers.las.extra_dims=BldId=uint64
- writing as a 1.2 LAS and using the --readers.las.use_eb_vlr=true
- copying an existing field e.g.
"type":"filters.ferry",
"dimensions":"Z=> BldId"

None of these seem to work

Can anyone help me to map 64 bit integer ids into a LAS file?

Thanks

Stephen
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pdal/attachments/20201130/96c62794/attachment.html>