2:57 a.m.
Hi, guys,
who could tell me what a default data access should be used in case of
entity mapping through orm.xml (field/property)?
Is it written anywhere?
I ask because there is seems a bug in hibernate - it use "property"
access by default.
Dmitry Geraskov.
3:56 a.m.
Hi Dmitry,
for JPA 2 the official reference is the final version of the JSR 317 (see
also http://java.sun.com/xml/ns/persistence/orm_2_0.xsd)
The spec dedicates a whole paragraph about the default access type using
using annotations, but I could
not find anything regarding the default access type using orm.xml only. It
seems it is up to
the implementor to decide on the the default. Why would you consider
'property' access to be the wrong default?
In any case, you can enforce a specific access type in xml using one of
the many ways to explicitly set
the access type, for example <persistence-unit-defaults>
--Hardy
On Thu, 24 Jun 2010 09:57:46 +0200, Dmitry Geraskov <dgeraskov(a)exadel.com>
wrote:
Hi, guys,
who could tell me what a default data access should be used in case of
entity mapping through orm.xml (field/property)?
Is it written anywhere?
I ask because there is seems a bug in hibernate - it use "property"
access by default.
Dmitry Geraskov.
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
4:09 a.m.
Hi, Hardy,
but I could not find anything regarding the default access type using
orm.xml only
the same for me, I also wanted to find orm.dtd and look for default
values like I found for hibernate-mapping here:
http://www.koders.com/xml/fidD6EE54ED93539E461E947B87F881E8A50DF4DEC5.asp...
I decided that here could be bug because I thought that the default
value specified somewhere, and I noticed that jpt (Dali) plugins for
Eclipse show that default value is "field". Of course they are not for
Hibernate implementation, but I expected the same behaviour.
Thanks for the help, Hardy.
24/06/2010 11:56, Hardy Ferentschik wrote:
Hi Dmitry,
for JPA 2 the official reference is the final version of the JSR 317
(see also http://java.sun.com/xml/ns/persistence/orm_2_0.xsd)
The spec dedicates a whole paragraph about the default access type
using using annotations, but I could
not find anything regarding the default access type using orm.xml
only. It seems it is up to
the implementor to decide on the the default. Why would you consider
'property' access to be the wrong default?
In any case, you can enforce a specific access type in xml using one
of the many ways to explicitly set
the access type, for example <persistence-unit-defaults>
--Hardy
On Thu, 24 Jun 2010 09:57:46 +0200, Dmitry Geraskov
<dgeraskov(a)exadel.com> wrote:
> Hi, guys,
> who could tell me what a default data access should be used in case of
> entity mapping through orm.xml (field/property)?
> Is it written anywhere?
> I ask because there is seems a bug in hibernate - it use "property"
> access by default.
>
> Dmitry Geraskov.
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
4:06 a.m.
The spec dedicates a whole paragraph about the default access type
using
using annotations, but I could
not find anything regarding the default access type using orm.xml only. It
seems it is up to
the implementor to decide on the the default. Why would you consider
'property' access to be the wrong default?
The problem is that eclipse tooling tries and validate if the orm.xml actually maps to
something
"legal".
Currently the default for orm.xml seem to be field access first thus it fails validation
if there is no
matching field.
The Eclipse tooling claims to be following the JPA spec thus why we are asking what
Hibernate
is basing its choice on.
/max
9:43 a.m.
Give me the spec paragraph that says that field has priority over getter.
On 1 juil. 2010, at 11:06, Andersen Max <max.andersen(a)redhat.com> wrote:
> The spec dedicates a whole paragraph about the default access
type using
> using annotations, but I could
> not find anything regarding the default access type using orm.xml only. It
> seems it is up to
> the implementor to decide on the the default. Why would you consider
> 'property' access to be the wrong default?
The problem is that eclipse tooling tries and validate if the orm.xml actually maps to
something
"legal".
Currently the default for orm.xml seem to be field access first thus it fails validation
if there is no
matching field.
The Eclipse tooling claims to be following the JPA spec thus why we are asking what
Hibernate
is basing its choice on.
/max
_______________________________________________
hibernate-dev mailing list
hibernate-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/hibernate-dev
11:28 a.m.
Give me the spec paragraph that says that field has priority over
getter.
Where is the spec that says getter has priority over field ? :)
In a separate discussion with the Dali (eclipse jpa tooling) team it seems like
for orm.xml it is not at all specified and thus left up to interpretation of the vendor
and it seems the ref implementation (eclipselink) takes the opposite approach to
this than Hibernate and we get stuck in the middle.
The "fix" seems to be that the validation in eclipse will be customizable to
which default
it should validate against.
/max
On 1 juil. 2010, at 11:06, Andersen Max <max.andersen(a)redhat.com> wrote:
>> The spec dedicates a whole paragraph about the default access type using
>> using annotations, but I could
>> not find anything regarding the default access type using orm.xml only. It
>> seems it is up to
>> the implementor to decide on the the default. Why would you consider
>> 'property' access to be the wrong default?
>
> The problem is that eclipse tooling tries and validate if the orm.xml actually maps
to something
> "legal".
>
> Currently the default for orm.xml seem to be field access first thus it fails
validation if there is no
> matching field.
>
> The Eclipse tooling claims to be following the JPA spec thus why we are asking what
Hibernate
> is basing its choice on.
>
> /max
>
>
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
3:48 p.m.
On 1 juil. 2010, at 18:28, Andersen Max wrote:
> Give me the spec paragraph that says that field has priority over getter.
Where is the spec that says getter has priority over field ? :)
The spec does not, that's the point :)
In a separate discussion with the Dali (eclipse jpa tooling) team it
seems like
for orm.xml it is not at all specified and thus left up to interpretation of the vendor
and it seems the ref implementation (eclipselink) takes the opposite approach to
this than Hibernate and we get stuck in the middle.
+1 some kind of "dialect". I do believe getter is a better choice for Hibernate
and in general for various reasons, but I am also sympathetic to the other side.
12:56 a.m.
On Jul 1, 2010, at 22:48, Emmanuel Bernard wrote:
On 1 juil. 2010, at 18:28, Andersen Max wrote:
>
>> Give me the spec paragraph that says that field has priority over getter.
>
> Where is the spec that says getter has priority over field ? :)
The spec does not, that's the point :)
> In a separate discussion with the Dali (eclipse jpa tooling) team it seems like
> for orm.xml it is not at all specified and thus left up to interpretation of the
vendor
> and it seems the ref implementation (eclipselink) takes the opposite approach to
> this than Hibernate and we get stuck in the middle.
+1 some kind of "dialect". I do believe getter is a better choice for Hibernate
and in general for various reasons, but I am also sympathetic to the other side.
it just means orm.xml even if using only spec compliant elements are not actually
portable.
(I know there are other scenarios where portability breaks, but handling defaults at the
basic metadata specification
sounds almost too easy to miss...but I guess we don't live in a perfect world ;)
/max
5853
days inactive
5861
days old
7 comments
4 participants
participants (4)
-
Andersen Max -
Dmitry Geraskov -
Emmanuel Bernard -
Hardy Ferentschik