jorisvandenbossche commented on PR #41823:
URL: https://github.com/apache/arrow/pull/41823#issuecomment-2139565365
To give a concrete example, to try to clarify:
* GDAL reads data from a file that contains a column with WKB values and
returns Arrow data with a column of binary type with the "ogc.wkb" annotation
* Should GDAL instead also use the "arrow.other" type?
* The ADBC postgresql driver reads data from a postgres table that has a
column with the polygon type, and returns Arrow data with a column of binary
type with the "arrow.other" annotation (and type_name "polygon" in the metadata)
* Why would the driver not return a custom extension type?
In both cases, a producer of Arrow data is returning data with a data type
that is not natively supported. So from an Arrow implementation (like Arrow
C++) point of view, both cases seem equivalent. But so why would we treat the
one case differently (assuming we add explicit support for "arrow.other"
extension type) than the other?
Maybe what I am after is some better guidelines for when data producers
should use the "arrow.other" type vs a custom extension type name.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
