Baymine opened a new pull request, #65911:
URL: https://github.com/apache/doris/pull/65911
### What problem does this PR solve?
Issue Number: close #65910
Problem Summary: Hive column types with explicit precision/length
annotations (e.g. `tinyint(3,0)`, `smallint(5,0)`, `int(10,0)`, `bigint(19,0)`,
`timestamp(19)`) and the Hive `datetime` type were not recognized by
`HiveMetaStoreClientHelper.hiveTypeToDorisType`. They fell through to the
`Type.UNSUPPORTED` fallback, which made the corresponding Hive columns
unreadable from Doris Hive/HMS external catalogs.
This change:
- Adds a `case "datetime"` arm to the exact-match switch, mapping it to
`datetimev2(timeScale)` (same as `timestamp`).
- Adds `startsWith` arms for `bigint`, `int`, `tinyint`, `smallint` and
`timestamp` so the precision-annotated forms (e.g. `int(10,0)`) resolve to the
same Doris type as their plain form. The `timestamp with local time zone` arm
is kept ahead of the new `timestamp` arm so it is matched first.
End-to-end: Hive columns declared as `tinyint(3,0)`, `int(10,0)`,
`bigint(19,0)`, `timestamp(19)` or `datetime` now resolve to the correct Doris
type instead of `UNSUPPORTED`.
### Release note
Hive columns of types `datetime`, `tinyint(n,m)`, `smallint(n,m)`,
`int(n,m)`, `bigint(n,m)` and `timestamp(n)` are now correctly mapped to the
corresponding Doris types when reading Hive/HMS external tables.
### Check List (For Author)
- Test
- [x] Unit Test — added `HiveTypeToDorisTypeTest` (`Tests run: 8,
Failures: 0, Errors: 0, Skipped: 0`).
- Behavior changed:
- [x] Yes — Hive types that previously resolved to `UNSUPPORTED` now
resolve to a concrete Doris type, so previously unreadable Hive columns become
readable.
- Does this need documentation?
- [x] No.
--
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]
