[
https://issues.apache.org/jira/browse/CALCITE-1659?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15886437#comment-15886437
]
Remus Rusanu edited comment on CALCITE-1659 at 2/27/17 8:06 PM:
----------------------------------------------------------------
bq. Seems you figured out how to run Quidem tests. How could the documentation
be improved? Should we add more documentation to the {{QuidemTest}} class?
In the end the page at
https://calcite.apache.org/docs/howto.html#running-integration-tests was the
one that got me on the right track. Via trial-and-error I ended up modifying
misc.iq and running the following, because it was quick and yielded the result
and diff I needed:
{noformat}
>mvn -f core/pom.xml verify -Pti -Dtest=Quidem
{noformat}
although I'm not sure the {{-Dtest=...}} does what I think it does, because
when the .iq test begin to pass then many more tests run, so I believe I was
just lucky in the order. Also, the actual .iq expected result and plan I
copy-pasted them from the mvn diff output, I did not locate any output file in
core/target to copy from.
As for what to improve, I think the calcite.apache.org HOWTO is the best
resource to get the ball rolling. Had the {{QuidemTest}} class contain detailed
explanation exactly step by step what to do, I would still probably not stumble
upon it, and not know where to start.
was (Author: rusanu):
bq. Seems you figured out how to run Quidem tests. How could the documentation
be improved? Should we add more documentation to the {{QuidemTest}} class?
In the end the page at
https://calcite.apache.org/docs/howto.html#running-integration-tests was the
one that got me on the right track. Via trial-and-error I ended up modifying
misc.iq and running the following, because it was quick and yielded the result
and diff I needed:
{{noformat}}
>mvn -f core/pom.xml verify -Pti -Dtest=Quidem
{{noformat}}
although I'm not sure the {{-Dtest=...}} does what I think it does, because
when the .iq test begin to pass then many more tests run, so I believe I was
just lucky in the order. Also, the actual .iq expected result and plan I
copy-pasted them from the mvn diff output, I did not locate any output file in
core/target to copy from.
As for what to improve, I think the calcite.apache.org HOWTO is the best
resource to get the ball rolling. Had the {{QuidemTest}} class contain detailed
explanation exactly step by step what to do, I would still probably not stumble
upon it, and not know where to start.
> Simplifying CAST(characterLiteral as TIMESTAMP) should round the sub-second
> fraction
> ------------------------------------------------------------------------------------
>
> Key: CALCITE-1659
> URL: https://issues.apache.org/jira/browse/CALCITE-1659
> Project: Calcite
> Issue Type: Bug
> Reporter: Remus Rusanu
> Assignee: Julian Hyde
> Fix For: 1.12.0
>
>
> Timestamps are losing sub-second parts after Calcite literal constant
> reduction.
> {noformat}
> select cast("1970-12-31 15:59:58.174" as TIMESTAMP) from src limit 1;
> {noformat}
> yields {{1970-12-31 15:59:58}} if CBO is enabled in Hive.
> {{ZonelessTimestamp.toString()}} contains a truncation 0-19 (ie. removes
> sub-second part from string representation).
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
