[
https://issues.apache.org/jira/browse/CALCITE-2918?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16791602#comment-16791602
]
Stamatis Zampetakis commented on CALCITE-2918:
----------------------------------------------
Nice script [~krisden]!
I checked the commit by [~kgyrtkirk]. It is not that commit that caused the
problem but is the one which revealed it. With this commit x is distinct from y
is now translated to x=y IS NOT TRUE and ((x is not null) or (y is not null))
which was not the case before.
I think that the failure will still apear if we attempt to execute the
following query in Postgres:
{code:sql}
SELECT * FROM table WHERE x=y IS NOT TRUE
{code}
I haven't tried the query above but will do if I find some time today.
Given the above, I think we are not dealing with a regression but just a hidden
(existing) bug that we could also deal with after the release of 1.19. What do
you think?
> Integration tests against postgres are broken
> ---------------------------------------------
>
> Key: CALCITE-2918
> URL: https://issues.apache.org/jira/browse/CALCITE-2918
> Project: Calcite
> Issue Type: Bug
> Affects Versions: 1.18.0, 1.19.0
> Reporter: Kevin Risden
> Priority: Blocker
> Fix For: 1.19.0
>
> Attachments: test.sh
>
>
> As part of the release process, integration testing against postgres found a
> failure.
> {code:java}
> ./mvnw verify -Pit
> [INFO] -------------------------------------------------------
> [INFO] T E S T S
> [INFO] -------------------------------------------------------
> [INFO] Running org.apache.calcite.test.JdbcTest
> 2019-03-11 11:23:41,539 [main] INFO - open start - state modified
> 2019-03-11 11:23:41,555 [main] INFO - Checkpoint start
> 2019-03-11 11:23:41,555 [main] INFO - Checkpoint end - txts: 25
> [ERROR] Tests run: 290, Failures: 0, Errors: 1, Skipped: 21, Time elapsed:
> 65.154 s <<< FAILURE! - in org.apache.calcite.test.JdbcTest
> [ERROR] testIsNotDistinctInFilter(org.apache.calcite.test.JdbcTest) Time
> elapsed: 0.041 s <<< ERROR!
> java.sql.SQLException:
> Error while executing SQL "select *
> from "foodmart"."employee" as e1
> where e1."last_name" is distinct from e1."last_name"": While executing SQL
> [SELECT *
> FROM "foodmart"."employee"
> WHERE ("last_name" IS NOT NULL OR "last_name" IS NOT NULL) AND "last_name" =
> "last_name" IS NOT TRUE] on JDBC sub-schema
> at
> org.apache.calcite.test.JdbcTest.testIsNotDistinctInFilter(JdbcTest.java:1585)
> Caused by: java.lang.RuntimeException:
> While executing SQL [SELECT *
> FROM "foodmart"."employee"
> WHERE ("last_name" IS NOT NULL OR "last_name" IS NOT NULL) AND "last_name" =
> "last_name" IS NOT TRUE] on JDBC sub-schema
> at
> org.apache.calcite.test.JdbcTest.testIsNotDistinctInFilter(JdbcTest.java:1585)
> Caused by: org.postgresql.util.PSQLException:
> ERROR: argument of IS NOT TRUE must be type boolean, not type character
> varying
> Position: 114
> at
> org.apache.calcite.test.JdbcTest.testIsNotDistinctInFilter(JdbcTest.java:1585)
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
