berkaysynnada commented on issue #5105: URL: https://github.com/apache/arrow-rs/issues/5105#issuecomment-1824424257
I think we have 2 datatypes carrying risk now; decimals and intervals. Since decimals have a total order and intervals don't (I mean they cannot be ordered without some assumptions), we should follow different paths: 1) Decimals: They can continue as other ordinary types: https://github.com/apache/arrow-rs/blob/200e8c80084442d9579e00967e407cd83191565d/arrow-ord/src/cmp.rs#L207-L221 We would just need another struct that implements ArrayOrd with the capability of decimal comparison. 2) Interval: They need their own `apply()` function like `apply_partial()` (alternative to: https://github.com/apache/arrow-rs/blob/200e8c80084442d9579e00967e407cd83191565d/arrow-ord/src/cmp.rs#L289-L297). The logic would be specialized to intervals, but would use the same vectorized operations, possibly with some duplications. We cannot use ArrayOrd for intervals since there exist values of Interval A and Interval B, we can get both A<B=false and B<A=false. What do you think of such support? -- 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]
