I am guessing this was a result of the recent optimization:
Reprex:
import sedona.db
sd = sedona.db.connect()
url = "https://github.com/geoarrow/geoarrow-data/releases/download/v0.2.0/ns-water_water-point.parquet"
sd.read_parquet(url).to_view("pts")
sd.sql("""SELECT COUNT(*) AS n, ST_Envelope_Agg(geometry) FROM pts GROUP BY "FEAT_CODE" """).show()
#> SedonaError: could not cast array of type Binary to arrow_array::array::byte_view_array::GenericByteViewArray<arrow_array::types::BinaryViewType>.
I am guessing this was a result of the recent optimization:
Reprex: