[
https://issues.apache.org/jira/browse/HIVE-21955?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16878391#comment-16878391
]
Zihao Ye commented on HIVE-21955:
---------------------------------
An example:
select * from t where ((rfr_page = 'search' and uid is not null) or (rfr_page =
'index' and uid is not null));
Filter text is:
Filter text = (((rfr_page = 'search') and uid is not null) or ((rfr_page =
'index') and uid is not null))
However the sarg is:
sarg: leaf-0 = (EQUALS rfr_page search), leaf-1 = (EQUALS rfr_page index),
leaf-2 = (IS_NULL uid), expr = (and (or leaf-0 leaf-1) (or (not leaf-2) leaf-1)
(or leaf-0 (not leaf-1)) (or (not leaf-2) (not leaf-1)))
> SearchArgumentImpl generates wrong ExpressionTree in some cases which might
> result in loss of data
> ---------------------------------------------------------------------------------------------------
>
> Key: HIVE-21955
> URL: https://issues.apache.org/jira/browse/HIVE-21955
> Project: Hive
> Issue Type: Bug
> Components: Hive, ORC
> Reporter: Zihao Ye
> Assignee: Zihao Ye
> Priority: Critical
>
> ExpressionBuilder applies `pushDownNot`, `foldMaybe`, `flatten`,
> `convertToCNF`, `flatten` and `buildLeafList` in order to form a
> non-normalized expression into a CNF expression with the unique leaves.
> After an expression is converted to CNF, there might be more than one
> non-leaf node which are exactly the same object in the expression tree. If
> this happens, those non-leaf node will be visited more than once in
> `buildLeafList` function. As a result, a wrong ExpressionTree is generated.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
