[
https://issues.apache.org/jira/browse/CALCITE-2906?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16901094#comment-16901094
]
Xurenhe commented on CALCITE-2906:
----------------------------------
[~walterddr]
Sorry, I am free until today.
I have reviewed the issue of CALCITE-2721.
I guess the method of “Expression2b” should be refactored to support this case.
{code:java}
//代码占位符
tbl.foo(0).col.bar => `TBL`.`FOO`(0).`COL`.`BAR`
{code}
Please point out my mistake if my understanding is wrong.
> Record DOT extension does not distinguish between simple and compound
> identifiers
> ---------------------------------------------------------------------------------
>
> Key: CALCITE-2906
> URL: https://issues.apache.org/jira/browse/CALCITE-2906
> Project: Calcite
> Issue Type: Bug
> Components: core
> Reporter: Rong Rong
> Priority: Major
>
> Currently the DOT extensions only supports Simple Identifiers:
> For example:
> * *{{tbl.foo(0).col.bar}}* is parsed as: *{{((`TBL`.`FOO`(0).`COL`).`BAR`)}}*
> However, this is not the case in regular Expression field, for example:
> * *{{tbl.col.bar}}* is parsed as: *{{`TBL`.`COL`.`BAR`}}*
> Parser should support both compound identifier, and simple identifier chain
> after a record type SqlNode.
> Similarly, parsing should also be consistent for other types of expression,
> for example
> * *{{tbl.arr[0].foo.bar}}* is currently parsed as:
> *{{((`TBL`.`ARR`[0].`FOO`).`BAR`)}}*
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
