Home » Modeling » EMF "Technology" (Ecore Tools, EMFatic, etc) » [Teneo] Persisting an Ecore Model
| [Teneo] Persisting an Ecore Model [message #84695] |
Thu, 24 May 2007 02:19  |
Eclipse User |
|
|
|
Originally posted by: irbull.cs.uvic.ca
I have been trying to persist a bunch of data (each to its own database)
and the corresponding Ecore model to which each of the datasets conforms
(each Ecore model would be persisted to its own database). At some
point in the future I plan on loading one of the datasets, but before I
can load the dataset, I need to load the ecore model. So the algorithm
is basically:
determine what dataset I need
create a hbds with EcorePackage and XMLTypePackage as the ePackages
do query("from ecore_EObject as EObject where dtype='EPackage'");
create a hbds with the result of the query as the ePackage
query the new hbds on the instance data
However, when I loaded the ePACKAGE I received a number of errors
because it seems that the eTypes are not persisted. In order to get
this to work, I had to add the following lines when I persisted the
models, and again, when I load them back:
EcorePackage.eINSTANCE.getETypedElement().getEStructuralFeat ure(EcorePackage.ETYPED_ELEMENT__ETYPE).setVolatile(false);
EcorePackage.eINSTANCE.getEClass().getEStructuralFeature(Eco rePackage.ECLASS__INSTANCE_CLASS_NAME).setVolatile(false);
I am posting this here in case others have similar problems.
Martin, Ed, does this make sense to you guys?
cheers,
ian
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #84710 is a reply to message #84695] |
Thu, 24 May 2007 09:43  |
| Eclipse User |
|
|
|
Originally posted by: merks.ca.ibm.com
Ian,
I wouldn't expect that the volatile setting should affect
serialization. Making something volatile should only affect code
generation and has (should have) no direct implication that the feature
is derived or transient. I would expect that only the transient setting
should affect serialization. (Of course what I expect and what actually
is might not match.)
Ian Bull wrote:
> I have been trying to persist a bunch of data (each to its own
> database) and the corresponding Ecore model to which each of the
> datasets conforms (each Ecore model would be persisted to its own
> database). At some point in the future I plan on loading one of the
> datasets, but before I can load the dataset, I need to load the ecore
> model. So the algorithm is basically:
>
> determine what dataset I need
> create a hbds with EcorePackage and XMLTypePackage as the ePackages
> do query("from ecore_EObject as EObject where dtype='EPackage'");
> create a hbds with the result of the query as the ePackage
> query the new hbds on the instance data
>
> However, when I loaded the ePACKAGE I received a number of errors
> because it seems that the eTypes are not persisted. In order to get
> this to work, I had to add the following lines when I persisted the
> models, and again, when I load them back:
>
> EcorePackage.eINSTANCE.getETypedElement().getEStructuralFeat ure(EcorePackage.ETYPED_ELEMENT__ETYPE).setVolatile(false);
>
> EcorePackage.eINSTANCE.getEClass().getEStructuralFeature(Eco rePackage.ECLASS__INSTANCE_CLASS_NAME).setVolatile(false);
>
>
> I am posting this here in case others have similar problems.
>
> Martin, Ed, does this make sense to you guys?
>
> cheers,
> ian
|
|
| |
| Re: [Teneo] Persisting an Ecore Model [message #84741 is a reply to message #84725] |
Thu, 24 May 2007 09:58 |
| Eclipse User |
|
|
|
Originally posted by: merks.ca.ibm.com
This is a multi-part message in MIME format.
--------------010104050407010800020408
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Christian,
The core runtime doesn't use isVolatile for anything. Arguably it
should never have been a property in Ecore itself but rather should
have been a generator property, i.e., it should have been a property on
the GenFeature since it has no meaningful impact outside of generated code.
Christian W. Damus wrote:
> Hi, Ed,
>
> It used to be, that in an EStructuralFeature
>
> isVolatile && isTransient && !isChangeable
>
> indicated that a feature was derived (before the isDerived attribute was
> introduced). Does that interpretation still hold in some places? How does
> EMF deal with models created before isDerived came into being?
>
> Cheers,
>
> Christian
>
>
> Ed Merks wrote:
>
>
>> Ian,
>>
>> I wouldn't expect that the volatile setting should affect
>> serialization. Making something volatile should only affect code
>> generation and has (should have) no direct implication that the feature
>> is derived or transient. I would expect that only the transient setting
>> should affect serialization. (Of course what I expect and what actually
>> is might not match.)
>>
>>
>>
>
> <snip>
>
--------------010104050407010800020408
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Christian,<br>
<br>
The core runtime doesn't use isVolatile for anything. Arguably it
should never have been a property in Ecore itself but rather should
have been a generator property, i.e., it should have been a property on
the GenFeature since it has no meaningful impact outside of generated
code.<br>
<br>
<br>
Christian W. Damus wrote:
<blockquote cite="midf345fr$hm0$1@build.eclipse.org" type="cite">
<pre wrap="">Hi, Ed,
It used to be, that in an EStructuralFeature
isVolatile && isTransient && !isChangeable
indicated that a feature was derived (before the isDerived attribute was
introduced). Does that interpretation still hold in some places? How does
EMF deal with models created before isDerived came into being?
Cheers,
Christian
Ed Merks wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Ian,
I wouldn't expect that the volatile setting should affect
serialization. Making something volatile should only affect code
generation and has (should have) no direct implication that the feature
is derived or transient. I would expect that only the transient setting
should affect serialization. (Of course what I expect and what actually
is might not match.)
</pre>
</blockquote>
<pre wrap=""><!---->
<snip>
</pre>
</blockquote>
<br>
</body>
</html>
--------------010104050407010800020408--
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #84756 is a reply to message #84741] |
Thu, 24 May 2007 10:40 |
| Eclipse User |
|
|
|
Hi,
Thanks Ian for letting us know.
The reason that Teneo uses it is that for jpox it is required to have an actual member in the
generated code. For hibernate this is less of an issue.
Ian, can you enter a bugzilla for this on Teneo
https://bugs.eclipse.org/bugs/enter_bug.cgi?product=EMFT
Then I can let Teneo handle this better.
gr. Martin
Ed Merks wrote:
> Christian,
>
> The core runtime doesn't use isVolatile for anything. Arguably it
> should never have been a property in Ecore itself but rather should
> have been a generator property, i.e., it should have been a property on
> the GenFeature since it has no meaningful impact outside of generated code.
>
>
> Christian W. Damus wrote:
>> Hi, Ed,
>>
>> It used to be, that in an EStructuralFeature
>>
>> isVolatile && isTransient && !isChangeable
>>
>> indicated that a feature was derived (before the isDerived attribute was
>> introduced). Does that interpretation still hold in some places? How does
>> EMF deal with models created before isDerived came into being?
>>
>> Cheers,
>>
>> Christian
>>
>>
>> Ed Merks wrote:
>>
>>
>>> Ian,
>>>
>>> I wouldn't expect that the volatile setting should affect
>>> serialization. Making something volatile should only affect code
>>> generation and has (should have) no direct implication that the feature
>>> is derived or transient. I would expect that only the transient setting
>>> should affect serialization. (Of course what I expect and what actually
>>> is might not match.)
>>>
>>>
>>>
>>
>> <snip>
>>
>
--
With Regards, Martin Taal
Springsite/Elver.org
Office: Hardwareweg 4, 3821 BV Amersfoort
Postal: Nassaulaan 7, 3941 EC Doorn
The Netherlands
Tel: +31 (0)84 420 2397
Fax: +31 (0)84 225 9307
Mail: mtaal@springsite.com - mtaal@elver.org
Web: www.springsite.com - www.elver.org
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #84770 is a reply to message #84756] |
Thu, 24 May 2007 12:32 |
| Eclipse User |
|
|
|
Originally posted by: irbull.cs.uvic.ca
Done:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=188973
Thanks guys!
p.s. EMF / EMFT / Eclipse Modeling ROCKS!!!
Martin Taal wrote:
> Hi,
> Thanks Ian for letting us know.
> The reason that Teneo uses it is that for jpox it is required to have an
> actual member in the generated code. For hibernate this is less of an
> issue.
> Ian, can you enter a bugzilla for this on Teneo
> https://bugs.eclipse.org/bugs/enter_bug.cgi?product=EMFT
>
> Then I can let Teneo handle this better.
>
> gr. Martin
>
> Ed Merks wrote:
>> Christian,
>>
>> The core runtime doesn't use isVolatile for anything. Arguably it
>> should never have been a property in Ecore itself but rather should
>> have been a generator property, i.e., it should have been a property
>> on the GenFeature since it has no meaningful impact outside of
>> generated code.
>>
>>
>> Christian W. Damus wrote:
>>> Hi, Ed,
>>>
>>> It used to be, that in an EStructuralFeature
>>>
>>> isVolatile && isTransient && !isChangeable
>>>
>>> indicated that a feature was derived (before the isDerived attribute was
>>> introduced). Does that interpretation still hold in some places?
>>> How does
>>> EMF deal with models created before isDerived came into being?
>>>
>>> Cheers,
>>>
>>> Christian
>>>
>>>
>>> Ed Merks wrote:
>>>
>>>
>>>> Ian,
>>>>
>>>> I wouldn't expect that the volatile setting should affect
>>>> serialization. Making something volatile should only affect code
>>>> generation and has (should have) no direct implication that the feature
>>>> is derived or transient. I would expect that only the transient
>>>> setting
>>>> should affect serialization. (Of course what I expect and what
>>>> actually
>>>> is might not match.)
>>>>
>>>>
>>>>
>>>
>>> <snip>
>>>
>>
>
>
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #606777 is a reply to message #84695] |
Thu, 24 May 2007 09:43 |
| Eclipse User |
|
|
|
Ian,
I wouldn't expect that the volatile setting should affect
serialization. Making something volatile should only affect code
generation and has (should have) no direct implication that the feature
is derived or transient. I would expect that only the transient setting
should affect serialization. (Of course what I expect and what actually
is might not match.)
Ian Bull wrote:
> I have been trying to persist a bunch of data (each to its own
> database) and the corresponding Ecore model to which each of the
> datasets conforms (each Ecore model would be persisted to its own
> database). At some point in the future I plan on loading one of the
> datasets, but before I can load the dataset, I need to load the ecore
> model. So the algorithm is basically:
>
> determine what dataset I need
> create a hbds with EcorePackage and XMLTypePackage as the ePackages
> do query("from ecore_EObject as EObject where dtype='EPackage'");
> create a hbds with the result of the query as the ePackage
> query the new hbds on the instance data
>
> However, when I loaded the ePACKAGE I received a number of errors
> because it seems that the eTypes are not persisted. In order to get
> this to work, I had to add the following lines when I persisted the
> models, and again, when I load them back:
>
> EcorePackage.eINSTANCE.getETypedElement().getEStructuralFeat ure(EcorePackage.ETYPED_ELEMENT__ETYPE).setVolatile(false);
>
> EcorePackage.eINSTANCE.getEClass().getEStructuralFeature(Eco rePackage.ECLASS__INSTANCE_CLASS_NAME).setVolatile(false);
>
>
> I am posting this here in case others have similar problems.
>
> Martin, Ed, does this make sense to you guys?
>
> cheers,
> ian
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #606778 is a reply to message #84710] |
Thu, 24 May 2007 09:55 |
| Eclipse User |
|
|
|
Originally posted by: cdamus.ca.ibm.com
Hi, Ed,
It used to be, that in an EStructuralFeature
isVolatile && isTransient && !isChangeable
indicated that a feature was derived (before the isDerived attribute was
introduced). Does that interpretation still hold in some places? How does
EMF deal with models created before isDerived came into being?
Cheers,
Christian
Ed Merks wrote:
> Ian,
>
> I wouldn't expect that the volatile setting should affect
> serialization. Making something volatile should only affect code
> generation and has (should have) no direct implication that the feature
> is derived or transient. I would expect that only the transient setting
> should affect serialization. (Of course what I expect and what actually
> is might not match.)
>
>
<snip>
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #606779 is a reply to message #84725] |
Thu, 24 May 2007 09:58 |
| Eclipse User |
|
|
|
This is a multi-part message in MIME format.
--------------010104050407010800020408
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Christian,
The core runtime doesn't use isVolatile for anything. Arguably it
should never have been a property in Ecore itself but rather should
have been a generator property, i.e., it should have been a property on
the GenFeature since it has no meaningful impact outside of generated code.
Christian W. Damus wrote:
> Hi, Ed,
>
> It used to be, that in an EStructuralFeature
>
> isVolatile && isTransient && !isChangeable
>
> indicated that a feature was derived (before the isDerived attribute was
> introduced). Does that interpretation still hold in some places? How does
> EMF deal with models created before isDerived came into being?
>
> Cheers,
>
> Christian
>
>
> Ed Merks wrote:
>
>
>> Ian,
>>
>> I wouldn't expect that the volatile setting should affect
>> serialization. Making something volatile should only affect code
>> generation and has (should have) no direct implication that the feature
>> is derived or transient. I would expect that only the transient setting
>> should affect serialization. (Of course what I expect and what actually
>> is might not match.)
>>
>>
>>
>
> <snip>
>
--------------010104050407010800020408
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Christian,<br>
<br>
The core runtime doesn't use isVolatile for anything. Arguably it
should never have been a property in Ecore itself but rather should
have been a generator property, i.e., it should have been a property on
the GenFeature since it has no meaningful impact outside of generated
code.<br>
<br>
<br>
Christian W. Damus wrote:
<blockquote cite="midf345fr$hm0$1@build.eclipse.org" type="cite">
<pre wrap="">Hi, Ed,
It used to be, that in an EStructuralFeature
isVolatile && isTransient && !isChangeable
indicated that a feature was derived (before the isDerived attribute was
introduced). Does that interpretation still hold in some places? How does
EMF deal with models created before isDerived came into being?
Cheers,
Christian
Ed Merks wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Ian,
I wouldn't expect that the volatile setting should affect
serialization. Making something volatile should only affect code
generation and has (should have) no direct implication that the feature
is derived or transient. I would expect that only the transient setting
should affect serialization. (Of course what I expect and what actually
is might not match.)
</pre>
</blockquote>
<pre wrap=""><!---->
<snip>
</pre>
</blockquote>
<br>
</body>
</html>
--------------010104050407010800020408--
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #606780 is a reply to message #84741] |
Thu, 24 May 2007 10:40 |
| Eclipse User |
|
|
|
Hi,
Thanks Ian for letting us know.
The reason that Teneo uses it is that for jpox it is required to have an actual member in the
generated code. For hibernate this is less of an issue.
Ian, can you enter a bugzilla for this on Teneo
https://bugs.eclipse.org/bugs/enter_bug.cgi?product=EMFT
Then I can let Teneo handle this better.
gr. Martin
Ed Merks wrote:
> Christian,
>
> The core runtime doesn't use isVolatile for anything. Arguably it
> should never have been a property in Ecore itself but rather should
> have been a generator property, i.e., it should have been a property on
> the GenFeature since it has no meaningful impact outside of generated code.
>
>
> Christian W. Damus wrote:
>> Hi, Ed,
>>
>> It used to be, that in an EStructuralFeature
>>
>> isVolatile && isTransient && !isChangeable
>>
>> indicated that a feature was derived (before the isDerived attribute was
>> introduced). Does that interpretation still hold in some places? How does
>> EMF deal with models created before isDerived came into being?
>>
>> Cheers,
>>
>> Christian
>>
>>
>> Ed Merks wrote:
>>
>>
>>> Ian,
>>>
>>> I wouldn't expect that the volatile setting should affect
>>> serialization. Making something volatile should only affect code
>>> generation and has (should have) no direct implication that the feature
>>> is derived or transient. I would expect that only the transient setting
>>> should affect serialization. (Of course what I expect and what actually
>>> is might not match.)
>>>
>>>
>>>
>>
>> <snip>
>>
>
--
With Regards, Martin Taal
Springsite/Elver.org
Office: Hardwareweg 4, 3821 BV Amersfoort
Postal: Nassaulaan 7, 3941 EC Doorn
The Netherlands
Tel: +31 (0)84 420 2397
Fax: +31 (0)84 225 9307
Mail: mtaal@springsite.com - mtaal@elver.org
Web: www.springsite.com - www.elver.org
|
|
|
| Re: [Teneo] Persisting an Ecore Model [message #606781 is a reply to message #84756] |
Thu, 24 May 2007 12:32 |
| Eclipse User |
|
|
|
Done:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=188973
Thanks guys!
p.s. EMF / EMFT / Eclipse Modeling ROCKS!!!
Martin Taal wrote:
> Hi,
> Thanks Ian for letting us know.
> The reason that Teneo uses it is that for jpox it is required to have an
> actual member in the generated code. For hibernate this is less of an
> issue.
> Ian, can you enter a bugzilla for this on Teneo
> https://bugs.eclipse.org/bugs/enter_bug.cgi?product=EMFT
>
> Then I can let Teneo handle this better.
>
> gr. Martin
>
> Ed Merks wrote:
>> Christian,
>>
>> The core runtime doesn't use isVolatile for anything. Arguably it
>> should never have been a property in Ecore itself but rather should
>> have been a generator property, i.e., it should have been a property
>> on the GenFeature since it has no meaningful impact outside of
>> generated code.
>>
>>
>> Christian W. Damus wrote:
>>> Hi, Ed,
>>>
>>> It used to be, that in an EStructuralFeature
>>>
>>> isVolatile && isTransient && !isChangeable
>>>
>>> indicated that a feature was derived (before the isDerived attribute was
>>> introduced). Does that interpretation still hold in some places?
>>> How does
>>> EMF deal with models created before isDerived came into being?
>>>
>>> Cheers,
>>>
>>> Christian
>>>
>>>
>>> Ed Merks wrote:
>>>
>>>
>>>> Ian,
>>>>
>>>> I wouldn't expect that the volatile setting should affect
>>>> serialization. Making something volatile should only affect code
>>>> generation and has (should have) no direct implication that the feature
>>>> is derived or transient. I would expect that only the transient
>>>> setting
>>>> should affect serialization. (Of course what I expect and what
>>>> actually
>>>> is might not match.)
>>>>
>>>>
>>>>
>>>
>>> <snip>
>>>
>>
>
>
|
|
|
Goto Forum:
Current Time: Thu May 08 18:15:06 EDT 2025
Powered by FUDForum. Page generated in 0.05985 seconds |
Home