[
https://issues.apache.org/jira/browse/CALCITE-4860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17905770#comment-17905770
]
ZheHu commented on CALCITE-4860:
--------------------------------
For this case, I suppose we can enable "NULLS LAST/FIRST" syntax without
aggregation. Because we've already filed
[CALCITE-4868|https://issues.apache.org/jira/browse/CALCITE-4868] for such
scene.
For example, the following case should work normally.
{code:java}
select col1 from view order by col1 asc nulls first
{code}
And make CALCITE-4868 focus on
{code:java}
select col1 from view group by col1 order by col1 asc nulls first
{code}
Appreciate any comments.
> In Elasticsearch adapter, support NULLS FIRST and NULLS LAST query
> ------------------------------------------------------------------
>
> Key: CALCITE-4860
> URL: https://issues.apache.org/jira/browse/CALCITE-4860
> Project: Calcite
> Issue Type: Bug
> Components: elasticsearch-adapter
> Affects Versions: 1.27.0
> Reporter: xiong duan
> Assignee: ZheHu
> Priority: Major
> Labels: pull-request-available
> Time Spent: 5h 50m
> Remaining Estimate: 0h
>
> In Elasticsearch Adapter, The SQL:
> {code:java}
> @Test void integerCat3() {
> CalciteAssert.that()
> .with(newConnectionFactory())
> .query("select cat1 from view order by cat1 nulls first")
> .returns("cat1=a\n" +
> "cat1=b\n" +
> "cat1=null\n");
> }
> {code}
> NULLS FIRST don't handle.
> According
> [es_missing_values|https://www.elastic.co/guide/en/elasticsearch/reference/current/sort-search-results.html#_missing_values]
> . We can add another parameter in the script to implement this.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
