Hi Zhe Hu, I remember there was a comparison-type configuration between char and varchar. Can you check if 'shouldConvertRaggedUnionTypesToVarying' has affected it?
Zhe Hu <[email protected]> 于2024年12月23日周一 08:48写道： > > Thank you, Mihai! > I’ll take a look at how the implicit cast works here. > > > Best regards > ZheHu > > ---- Replied Message ---- > | From | Mihai Budiu<[email protected]> | > | Date | 12/23/2024 03:32 | > | To | [email protected]<[email protected]> | > | Subject | Re: String comparison with DEFAULT conformance | > I don't know if the SQL standard mandates a particular result for this query. > It depends on how type inference works. If both sides are cast to char(4), > then the result should be "true". If both sides are cast to VARCHAR, then the > result should be "false". > > Whether this is a bug or not depends really on what MySQL and Oracle do about > this query. > > Mihai > > ________________________________ > From: Zhe Hu <[email protected]> > Sent: Friday, December 20, 2024 11:51 PM > To: [email protected] <[email protected]> > Subject: String comparison with DEFAULT conformance > > Hi, community! > I’ve encountered one confusing problem when I run the following test in > JdbcTest with Calcite's DEFAULT conformance. It seems that expression > “cast('Eric' as char(4))='Eric ‘“ > returns true. > However, when I use MYSQL or ORACLE conformance, the above comparison returns > false. > > > Is this intentional or a bug? > > > Best regards, > ZheHu > > >
