Github user liancheng commented on the pull request:
https://github.com/apache/spark/pull/5499#issuecomment-93123075
Good catch! The change LGTM.
`NullableColumnBuilderSuite` uses a mock builder because the real builders
also mix in `CompressibleColumnBuilder`, while `NullableColumnBuilderSuite` is
used to test nulls.
This wasn't detected before because we were using a column type ID in the
columnar buffer, and column accessors are constructed according to the type ID.
However, while adding column type for fixed-decimal, decimals with different
precisions and scales are actually different types, a single type ID is not
enough to represent complete type information. Thus column type ID is
invalidated and now column accessors are built according to the data types in
the schema.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
