You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@calcite.apache.org by Kumar Ramaiyer <ku...@workday.com.INVALID> on 2023/05/11 21:12:58 UTC

Calcite Examples

Hello

Calcite is an amazing tool.  I am starting to get into it and would like to expose SQL interface for our SaaS application.  Goal is then to (1) connect to any BI tool that our customers may have instead of having to write to our proprietary API (2) Able to build a conversational AI application.

I went through all the videos and tutorials I could see (including the Boss 21) on youtube.  The Lucene based tutorial is pretty good till the point of customer operator and custom rules.  It looks fairly complicated and I am not sure how I can translate that to our domain.

I would like to see examples – if there are – for the following:


  1.  Pushing the filter rule to our API
  2.  Pushing the limit rule to our API
  3.  Pushing the group by to our API

I want to provide implementations for these.

Thanks for your help.

Regards
Kumar

Re: Calcite Examples

Posted by Benchao Li <li...@apache.org>.

Hi Kumar,

You can take a look at some internal rules in Calcite:
- push project to scan :
https://github.com/apache/calcite/blob/289d08205584efefa7b399318e7bb81c4f7f11aa/core/src/main/java/org/apache/calcite/rel/rules/ProjectTableScanRule.java
- push filter to scan :
https://github.com/apache/calcite/blob/289d08205584efefa7b399318e7bb81c4f7f11aa/core/src/main/java/org/apache/calcite/rel/rules/FilterTableScanRule.java

And also some downstream projects' implementation such as Flink:
-  push limit to scan :
https://github.com/apache/flink/blob/06688f345f6793a8964ec00002175f44cda13c33/flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/rules/logical/PushLimitIntoTableSourceScanRule.java
- push filter to scan :
https://github.com/apache/flink/blob/06688f345f6793a8964ec00002175f44cda13c33/flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/rules/logical/PushFilterIntoTableSourceScanRule.java
- push project to scan :
https://github.com/apache/flink/blob/06688f345f6793a8964ec00002175f44cda13c33/flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/rules/logical/PushProjectIntoTableSourceScanRule.java

Kumar Ramaiyer <ku...@workday.com.invalid> 于2023年5月17日周三 04:19写道：

> Hello
>
> Calcite is an amazing tool.  I am starting to get into it and would like
> to expose SQL interface for our SaaS application.  Goal is then to (1)
> connect to any BI tool that our customers may have instead of having to
> write to our proprietary API (2) Able to build a conversational AI
> application.
>
> I went through all the videos and tutorials I could see (including the
> Boss 21) on youtube.  The Lucene based tutorial is pretty good till the
> point of customer operator and custom rules.  It looks fairly complicated
> and I am not sure how I can translate that to our domain.
>
> I would like to see examples – if there are – for the following:
>
>
>   1.  Pushing the filter rule to our API
>   2.  Pushing the limit rule to our API
>   3.  Pushing the group by to our API
>
> I want to provide implementations for these.
>
> Thanks for your help.
>
> Regards
> Kumar
>


-- 

Best,
Benchao Li