[
https://issues.apache.org/jira/browse/CALCITE-1790?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16008454#comment-16008454
]
Julian Hyde commented on CALCITE-1790:
--------------------------------------
Yeah, re-ordering expressions has been at the back of my mind. When we generate
code we rely on ordering, e.g. "x instance of Integer && x.intValue() = 10" but
we've never made any commitments regarding SQL and I don't think the standard
makes any either. I can see that ordering would matter when there are
exceptions (you cite divide-by-zero) and also when there are UDFs with
side-effects.
My policy has been "don't ask, don't tell", and to endeavor to make
simplifications such as this one if not order-preserving then at least stable.
For example, if you have to put the cases into a map, use a LinkedHashMap that
will preserve order rather than SortedMap that will re-order or a HashMap which
will create chaos.
> Simplify CASE P1 THEN <boolean> P@ THEN <booleans> ... ELSE TRUE/FALSE
> -----------------------------------------------------------------------
>
> Key: CALCITE-1790
> URL: https://issues.apache.org/jira/browse/CALCITE-1790
> Project: Calcite
> Issue Type: Improvement
> Components: core
> Reporter: Remus Rusanu
> Assignee: Remus Rusanu
> Priority: Minor
>
> In HIVE-14431 [~jcamachorodriguez] proposed a simplification for CASE when
> all branches are not nullable boolean expression into an alternative
> AND/OR/NOT based expression. This allows for more aggressive reductions and
> split/push-down on the whole. Meantime the simplifier code migrated to
> Calcite so I'm reviving this here.
> The proposed simplification is:
> {code}
> CASE
> WHEN p1 THEN ex1
> WHEN p2 THEN ex2
> ...
> WHEN pn THEN exn
> ELSE TRUE/FALSE
> END
> {code}
> to be transformed into:
> {code}
> (p1 AND ex1)
> OR (not(p1) AND p2 AND x2)
> ...
> OR (not(p1) AND not(p2) ... AND not(pn-1) AND Pn AND exn)
> [OR (not(p1) AND not(p2) ... AND not(pn))]
> {code}
> The last OR is depending on the ELSE branch being TRUE/FALSE.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
