On Mon 07 Nov 2011 10:32:33 AM CST, Łukasz Antoniak wrote:
>
> The whole idea of supporting multiline statements is that users would
> like to execute the scripts they already have (and for example execute
> in SQL*Plus for production) without applying extra modifications. The
> "continuation" symbol/operator would be useless for them.
Yes I get the idea behind import.sql :)
My point is less about allowing multi-line support and more about the
*how*.
> I think that introducing an interface like ImportSqlCommandExtracter,
> and providing two implementations (single line - the default for
> compatibility; and multiline - new) that can be switched by setting
> appropriate configuration parameter, would be a good idea. Do you
> agree?
Yes, that is why I suggested it :)
> I hope I didn't make you very angry with my suggestion. Just wanted to
> help/contribute and move forward the issue that six people voted for
> and exists unresolved since version 3.2. Max agreed that forcing ';'
> ending can be done in 3.3 version ("if we break every existing
> hibernate 3.2.x users existing import.sql it is not something we can
> put in. Then it would wait for 3.3.x." - JIRA HHH-2403). I will keep
> in mind that backwards compatibility is the priority.
Not angry at all. Not sure where you got that anyway tbh. If
backwards compatibility can be easily maintained then I just do not see
the reason to not. Whether we *could* break back-compat based on our
versioning schemes is a secondary question to my mind. *Should* is far
more important primary consideration. I just dont see this as a time
that we have to break compatibility.
--
st...@hibernate.org
http://hibernate.org
_______________________________________________
hibernate-dev mailing list
hibernate-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
