[
https://issues.apache.org/jira/browse/CALCITE-4740?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17400748#comment-17400748
]
yanjing.wang commented on CALCITE-4740:
---------------------------------------
Thanks [~julianhyde], I added the dialect conformance check. I think the
converted SqlNode is correct, So I put the stripping HAVING clause logic
together with SELECT_LIST.
{quote}don't generate references to aliases that start "expr$"?
{quote}
Your suggestion is good and more simpler, but maybe cause the converted SqlNode
stripping logic scatter and *"expr$"* appears in multiple places. Do you agree?
> JDBC adapter generates incorrect HAVING clause in BigQuery dialect
> ------------------------------------------------------------------
>
> Key: CALCITE-4740
> URL: https://issues.apache.org/jira/browse/CALCITE-4740
> Project: Calcite
> Issue Type: Bug
> Components: core
> Affects Versions: 1.27.0
> Environment: jdk8
> macos
> Reporter: yanjing.wang
> Priority: Minor
>
> I'm converting
> {code:java}
> String query = select
> "product_id" , sum("shelf_width")
> from "foodmart"."product"
> where "product_id" > 0
> group by "product_id" having sum("shelf_width") > 1{code}
> with big query dialect, but results
> {code:java}
> SELECT product_id, SUM(shelf_width)
> FROM foodmart.product
> WHERE product_id > 0
> GROUP BY product_id
> HAVING `EXPR$1` > 1
> {code}
> Maybe we need replace having clause aliases when stripping trivial aliases.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
