Timm0 commented on PR #28979: URL: https://github.com/apache/flink/pull/28979#issuecomment-6267203909
> I'm looking at it again and I'm now thinking we should remove it. Looking at how all of this gets mapped to SQL statements: > > `IS_NOT_NULL` -> `IS NOT NULL` this is apparently built into the grammar `IS_NOT_NAN` -> `IS_NOT_NAN` aka a simple function > > the grammatical carve out for `IS NOT NULL` feels like evidence enough to special case `isNotNull` in the code. Special casing a boolean function flip (`IS_NOT_NAN(x)` vs `NOT IS_NAN(x)`) feels unnecessary. > > The only provider I'm seeing that special cases that is oracle and they do it grammatically. Every other one appears to just negate the `IS_NAN` function. > > Is everyone on board with me correcting this? My thought is remove the dedicated `isNotNan` functions from java all the way up to python and support those being handled though idiomatic negations `not(isNan(..))`. If everyone agrees, do we want to open a new ticket + PR for this? Or do we want to handle it through an addendum on this PR? First of all, thank you for having another look at it! I'm fine with both directions here. If we want to remove it I think we can handle this by re-using the same ticket as @dianfu already suggested. -- 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]
