[
https://issues.apache.org/jira/browse/HIVE-11504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14681993#comment-14681993
]
Sergio Peña commented on HIVE-11504:
------------------------------------
Thanks [~Ferd]. Here are some comments about the patch:
- LeafFilterFactory.java
Remove unused imports:
import org.apache.hadoop.hive.ql.io.sarg.PredicateLeaf;
import org.apache.parquet.schema.MessageType;
- ParquetFilterPredicateConverter.java
If toFilterPredicate() passes a null schema, then
buildFilterPredicateFromPredicateLeaf() will receive a null type, and the
builder will not be generated:
{{builder = leafFilterFactory.getLeafFilterBuilderByType(type);}}
So, there won't be any filter predicate.
The old toFilterPredicate() used to return a FilterPredicate even if the
schema was null. What did you decide to remove such function?
I think it was much cleaner before. Were there errors on some of the tests?
- parquet_predicate_pushdown.q
Tests passes on my computer, but not on Jenkins upstream. I think we should
re-arrange some
of the queries to allow Jenkins to pass the tests.
> Predicate pushing down doesn't work for float type for Parquet
> --------------------------------------------------------------
>
> Key: HIVE-11504
> URL: https://issues.apache.org/jira/browse/HIVE-11504
> Project: Hive
> Issue Type: Sub-task
> Reporter: Ferdinand Xu
> Assignee: Ferdinand Xu
> Attachments: HIVE-11504.1.patch, HIVE-11504.2.patch,
> HIVE-11504.2.patch, HIVE-11504.patch
>
>
> Predicate builder should use PrimitiveTypeName type in parquet side to
> construct predicate leaf instead of the type provided by PredicateLeaf.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
