> For the oracle.sql types, it is a configuration parameter to use those types
> and was a design decision. [...] Using double mapped
> to BINARY_DOUBLE (IEEE-754) solved this as there should be no translation
> between Java's Double and BINARY_DOUBLE (they are both IEEE-754).

I see.

> Oracle
> also had some time zone issues with Date objects going in so a decision was
> made to stick with the oracle.sql types and handle the type mapping in our
> code.

I'm not sure if this is the same problem, but I've recently had
"trouble" with this as well. It turned out to be some subtlety of the
JDBC specification, which doesn't map well to Oracle's non-standard
implementation of DATE/TIMESTAMP. But after all, it was my
misunderstanding of JDBC. Here's some more insight:
http://stackoverflow.com/questions/9202857/timezones-in-sql-date-vs-java-sql-date

> Here is the mapping of your sample package.

I see, so they've had their own thoughts for method name
disambiguation. I'll think of something to properly disambiguate these
cases. Maybe, I'll add some methods to the GeneratorStrategy as
documented here:
http://www.jooq.org/manual/META/AdvancedConfiguration/

It might be useful for PL/SQL power users, to be able to influence
this section of the generated code more specifically

Thanks for your help!
Lukas

Reply via email to