[
https://issues.apache.org/jira/browse/CALCITE-1095?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15492409#comment-15492409
]
Jesus Camacho Rodriguez commented on CALCITE-1095:
--------------------------------------------------
[~julianhyde], I took a quick look and patch LGTM.
I just have a thought around hardcoding the precedence, maybe for a follow-up
if you think it makes sense. We could store the functions in a static data
structure so the precedence is inferred automatically depending on their
position there. We would be able to see clearly the precedence just by looking
at this structure, and we would not need to use magical numbers (_24, 26, 28,
etc_) in the function constructor, which I guess might get a bit tricky in case
new operators are added e.g. either in Calcite itself or by systems that use it
as a library.
> NOT precedence
> --------------
>
> Key: CALCITE-1095
> URL: https://issues.apache.org/jira/browse/CALCITE-1095
> Project: Calcite
> Issue Type: Bug
> Reporter: Julian Hyde
> Assignee: Julian Hyde
> Fix For: 1.9.0
>
>
> The precedence of the NOT operator is currently too high; it should be lower
> than '='. {{WHERE NOT deptno = 10}} works on other databases.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
