laskoviymishka commented on issue #1090:
URL: https://github.com/apache/iceberg-go/issues/1090#issuecomment-4482190545
Quick answers, following the pattern this repo already uses for cross-client
binary fixtures — for example `puffin/testdata` and `table/dv/testdata/deletes`:
1. **Directory**
I’d put these under `table/testdata/geo/`, or a similar single location
next to the geo plumbing once it lands. One directory, flat layout, with a
`README.md` like the existing fixtures.
2. **Format and provenance**
The repo convention so far is to copy canonical binary blobs directly
from Apache Iceberg Java test resources and pin them byte-for-byte in
`testdata/`. The `README.md` then lists the upstream URL for each file.
That’s what makes them useful as cross-client pins: the bytes are exactly
what Java emits / reads, not whatever a local generator happens to produce
today.
So I’d use small Parquet files with WKB-encoded geometry / geography
columns, copied as-is from Java’s test corpus, and possibly from
`parquet-testing` where relevant. No in-repo generator script or
package-manager pinning needed — the source URL in the `README.md` is the pin.
3. **Iceberg tables vs. raw Parquet**
I’d start with raw Parquet files that exercise the geo encoding and
Parquet GeoStatistics shapes. Full Iceberg table fixtures can come later if a
sub-issue actually needs them. That keeps this PR small and easier to review.
Scope-wise, I’d keep this PR tightly focused on fixtures plus a small loader
test that reads each file and checks that it parses. That should let it merge
ahead of the feature work, and the rest of the chain can build on it.
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
