YiwenWu commented on PR #3669:
URL: https://github.com/apache/calcite/pull/3669#issuecomment-1926097975
> > > Just a question, how to ensure that the semantics of these functions
are consistent with those in SPARK?
> >
> >
> > @macroguo-ghy I executed the Function Sqls in the spark engine. And
checked the function Definition/OperandType/ReturnType is consistent with the
Spark definition.
>
> Some functions have subtle differences for corner cases of the arguments.
And Calcite tests often do _not_ test the corner cases. I haven't checked Spark
in particular, but we are writing a lot of tests for existing functions and we
find lots of such cases.
Strongly agree, there may be many subtle differences in the execution of
functions of each engine. This problem also troubles us very much, For example,
precision issues and null value issues。
In my opinion, the current function definition cannot guarantee the
correctness of execution, mainly limiting the input types and return types of
Function.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
