[
https://issues.apache.org/jira/browse/HIVE-24940?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated HIVE-24940:
----------------------------------
Labels: pull-request-available (was: )
> PartitionPruner may reject partitionfilter expressions evaluating to unknown
> if the filter is safe
> --------------------------------------------------------------------------------------------------
>
> Key: HIVE-24940
> URL: https://issues.apache.org/jira/browse/HIVE-24940
> Project: Hive
> Issue Type: Improvement
> Reporter: Zoltan Haindrich
> Assignee: Zoltan Haindrich
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> {code}
> CREATE TABLE t1pstring (col1 string) PARTITIONED BY (p1 string);
> INSERT INTO t1pstring PARTITION (p1) VALUES
> ("2020","2020"),("2021","2021"),("2021_backup","2021_backup"),('8888','8888'),('88889','88889'),('8888_a','8888_a'),('a8888','a8888');
> explain extended SELECT count(*) FROM t1pstring WHERE p1=8888;
> [...]
> | Truncated Path -> Alias: |
> | /t1pstring/p1=2021_backup [t1pstring] |
> | /t1pstring/p1=8888 [t1pstring] |
> | /t1pstring/p1=8888_a [t1pstring] |
> | /t1pstring/p1=a8888 [t1pstring] |
> [...]
> {code}
> note:
> * for all values which are interpretable as integers - the equals is
> evaluated and the result is false
> * for {{NULL}} values and for non-integer values ({{a8888}}) the comparision
> results a {{NULL}} which is retained
> * {{NULL}} -s are retained because there is a preprocessing step which
> removes functions which are dependant of non-partition columns as well - and
> replaces them with a {{NULL}}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
