Hi

Depending the case DTO are not an option.

I agree in rest app i wouldnt it but if not possible (maybe through another
Bean) it would kill this module for half of the usages i see since i'd need
to add this layer.
Le 12 juil. 2013 06:55, "hantsy" <[email protected]> a écrit :

> No DTO please, data module for data access, why we care about DTO.
>
> A question about the data, the difference for EJB and none EJB environment.
>
> if possible in a EJB envoriment, proxy the Repository and add @Stateless
> and transaction declaration to Repository automatically at runtime.
>
> Regards
>
> Hantsy
> On 7/10/2013 23:23, Thomas Hug wrote:
> > I wouldn't label the feature with DTO but rather as some general result
> > transformation - might also be useful for e.g. native queries. Going back
> > to the API suggestion, from that perspective such an annotation should
> > probably also work on method level, so I'd keep the forEntity out there.
> >
> >
> > On Wed, Jul 10, 2013 at 4:22 PM, John D. Ament <[email protected]
> >wrote:
> >
> >> Personally, I don't like this idea.
> >>
> >> A DAO should do DAO stuff.
> >> A DTO should do DTO stuff.
> >>
> >> The transformation of your entities into some other POJO shouldn't be
> >> inside your DAO.
> >>
> >> Right now, I use google guava to do DTO work on entities going back and
> >> forth over a REST API.  Works well IMHO.
> >>
> >> John
> >>
> >>
> >> On Wed, Jul 10, 2013 at 9:21 AM, Romain Manni-Bucau
> >> <[email protected]>wrote:
> >>
> >>> globally my answer meant "if forEntity is sometimes mandatory,
> sometimes
> >>> not this is maybe not the right place"
> >>>
> >>> i thought to add it to mapper config
> >>>
> >>> *Romain Manni-Bucau*
> >>> *Twitter: @rmannibucau <https://twitter.com/rmannibucau>*
> >>> *Blog: **http://rmannibucau.wordpress.com/*<
> >>> http://rmannibucau.wordpress.com/>
> >>> *LinkedIn: **http://fr.linkedin.com/in/rmannibucau*
> >>> *Github: https://github.com/rmannibucau*
> >>>
> >>>
> >>>
> >>> 2013/7/10 Thomas Hug <[email protected]>
> >>>
> >>>> Making forEntity non-optional would then be redundant for the regular
> >>> cases
> >>>> using the base interface, so I wouldn't. But I see that it should be
> >>>> clearly documented then as things might get confusing...
> >>>>
> >>>>
> >>>> On Wed, Jul 10, 2013 at 3:02 PM, Romain Manni-Bucau
> >>>> <[email protected]>wrote:
> >>>>
> >>>>> do you mean you force forEntity = Person.class?
> >>>>>
> >>>>> looks ok for me since the only constraint is to add the dto types
> >>>> somewhere
> >>>>> :)
> >>>>>
> >>>>> *Romain Manni-Bucau*
> >>>>> *Twitter: @rmannibucau <https://twitter.com/rmannibucau>*
> >>>>> *Blog: **http://rmannibucau.wordpress.com/*<
> >>>>> http://rmannibucau.wordpress.com/>
> >>>>> *LinkedIn: **http://fr.linkedin.com/in/rmannibucau*
> >>>>> *Github: https://github.com/rmannibucau*
> >>>>>
> >>>>>
> >>>>>
> >>>>> 2013/7/10 Thomas Hug <[email protected]>
> >>>>>
> >>>>>> Hmm and I assumed DTOs are dead and buried :-)
> >>>>>>
> >>>>>> Packing this in the base interface feels kind of clunky to me -
> >> also
> >>>>>> considering that there are repositories without the need to extend
> >>> the
> >>>>> base
> >>>>>> interface. What about something like
> >>>>>>
> >>>>>> @Repository(forEntity = Person.class)
> >>>>>> @ResultMapper(entityMapper = MapperX.class, keyMapper =
> >>> MapperY.class)
> >>>>>> public interface PersonRepository extends
> >> EntityRepository<PersonDto,
> >>>>>> DtoPk> { ... }
> >>>>>>
> >>>>>> Having the Entity on @Repository takes precedence and the type
> >>>> parameters
> >>>>>> are in this case just for convenience.
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>> On Wed, Jul 10, 2013 at 2:35 PM, Romain Manni-Bucau
> >>>>>> <[email protected]>wrote:
> >>>>>>
> >>>>>>> +1
> >>>>>>>
> >>>>>>> just to complete this thread the main issue is not the
> >>> implementation
> >>>>> but
> >>>>>>> the exposed API:
> >>>>>>>
> >>>>>>> public interface EntityRepository<E, PK extends Serializable>
> >>>>>>>
> >>>>>>> would become
> >>>>>>>
> >>>>>>> public interface EntityDtoRepository<E, PK extends Serializable,
> >>> Dto,
> >>>>>>> DtoPk>
> >>>>>>>
> >>>>>>> *Romain Manni-Bucau*
> >>>>>>> *Twitter: @rmannibucau <https://twitter.com/rmannibucau>*
> >>>>>>> *Blog: **http://rmannibucau.wordpress.com/*<
> >>>>>>> http://rmannibucau.wordpress.com/>
> >>>>>>> *LinkedIn: **http://fr.linkedin.com/in/rmannibucau*
> >>>>>>> *Github: https://github.com/rmannibucau*
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> 2013/7/10 Jean-Louis MONTEIRO <[email protected]>
> >>>>>>>
> >>>>>>>> Hello guys,
> >>>>>>>>
> >>>>>>>> Just used DS Data module yesturday, and I was wondering if we
> >>> could
> >>>>>> add a
> >>>>>>>> feature allowing on-the-fly conversion to DTO.
> >>>>>>>> For example, we could use modelmapper (or similar to convert
> >> DAO
> >>>>> return
> >>>>>>>> values to DTO objects).
> >>>>>>>>
> >>>>>>>> Adding a mapper interface to delegate to would also allow
> >> people
> >>> to
> >>>>>> plug
> >>>>>>>> their own implementation in.
> >>>>>>>>
> >>>>>>>> WDYT?
> >>>>>>>>
> >>>>>>>> JLouis
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> 2013/7/1 Thomas Hug <[email protected]>
> >>>>>>>>
> >>>>>>>>> Hi John
> >>>>>>>>> Thnx for the message, missed that one. Looks like there's a
> >>>> default
> >>>>>>>> profile
> >>>>>>>>> needed (test-persistence.xml only part of the specific server
> >>>>>>> profiles).
> >>>>>>>>> Will check tonight.
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> On Mon, Jul 1, 2013 at 2:42 AM, John D. Ament <
> >>>>>> [email protected]
> >>>>>>>>>> wrote:
> >>>>>>>>>> Hi
> >>>>>>>>>>
> >>>>>>>>>> Whoever brought in the data module, can you double check
> >> your
> >>>>> tests
> >>>>>>> and
> >>>>>>>>>> license headers?
> >>>>>>>>>>
> >>>>>>>>>> I think it's just your tests, but it's failing during a rat
> >>>> check
> >>>>>>>>>>
> >>>>>>>>>>
> >>
> https://builds.apache.org/job/DeltaSpike%20RAT-Check/org.apache.deltaspike.modules$deltaspike-data-module-impl/558/testReport/org.apache.deltaspike.data.impl/QueryResultTest/org_apache_deltaspike_data_impl_QueryResultTest/
> >>>>>>>>>> John
> >>>>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> --
> >>>>>>>> Jean-Louis
> >>>>>>>>
>
>

Reply via email to