| postgres adapter [message #5720] |
Tue, 29 November 2005 09:53  |
Eclipse User |
|
|
|
Hi all
I'm willing to write an adapter for posgres db, I've got the sample
provider plugin from cvs, but it seems quite large, could anyone provide
me some light on what are the core extensions to implement?
thanks in advance!
|
|
|
| Re: postgres adapter [message #5739 is a reply to message #5720] |
Tue, 29 November 2005 13:40  |
| Eclipse User |
|
|
|
Hey Miguel,
The sample plug-in provided is a generic example that is not DB specific.
It simply stores a property containing a folder location and displays the
contents of that folder in the DSE when connected. (plug-in id is
org.eclipse.datatools.connectivity.sample.cp)
That said, there is a forthcoming implementation for Derby that should
serve as a complete DB specific example. This should be available in the
M2 release at the end of this month.
The following provides an overview of what is required for developing a
connection profile:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DTPConnectivityFrameworkDevelopersQSGuide.htm
The following provides specific documentation for the connectionProfile
extension point:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/CMF_I101905.html
The following provides specific documentation for the driver definition
framework:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DMF_I101205.html
Documentation for customizing the SQL model can be found at:
http://www.eclipse.org/datatools/project_modelbase/modelbase _doc/DTP%20Modelbase%20White%20paper.htm
Ideally, you should be able to use most of the classes in the
org.eclipse.datatools.connectivity.db.generic plug-in as your
implementation for your profile. You may want to create wizard and
property pages that are specific to postgres, but this is not necessary.
You should create a driver definition that is specific to postgres;
specifying appropriate default values, dbdefinition values, and required
JAR files; the JAR files can be specified using the format,
[<plug-id>]/<plug-in relative path to JAR>
This will allow the default JAR list to use the JARs distributed with the
plug-in. You can also just list the JAR names (when the user edits the
definition and specifies the actual JAR file, they are given the option to
update the other JARs listed using the same root folder). You should
sublcass the base wizard implementation to add a filter so only your
driver definitions are displayed to the user in your wizard (note, this
cannot currently be done programmatically, but should be available by M2).
You will need to create a dbdefinition instance and extension point for
postgres. You may need to create a specialized SQL model and catalog
loader for postgres (this will probably be the biggest task).
Look for the Derby connection profile in the M2 release.
Please let me know if you need anything else.
Rob
|
|
|
| Re: postgres adapter [message #5757 is a reply to message #5739] |
Thu, 01 December 2005 15:30 |
| Eclipse User |
|
|
|
thanks a lot!
I'll keep this arround until M2
Rob Cernich wrote:
> Hey Miguel,
>
> The sample plug-in provided is a generic example that is not DB
> specific. It simply stores a property containing a folder location and
> displays the contents of that folder in the DSE when connected.
> (plug-in id is org.eclipse.datatools.connectivity.sample.cp)
>
> That said, there is a forthcoming implementation for Derby that should
> serve as a complete DB specific example. This should be available in
> the M2 release at the end of this month.
>
> The following provides an overview of what is required for developing a
> connection profile:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DTPConnectivityFrameworkDevelopersQSGuide.htm
>
>
> The following provides specific documentation for the connectionProfile
> extension point:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/CMF_I101905.html
>
>
> The following provides specific documentation for the driver definition
> framework:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DMF_I101205.html
>
>
> Documentation for customizing the SQL model can be found at:
> http://www.eclipse.org/datatools/project_modelbase/modelbase _doc/DTP%20Modelbase%20White%20paper.htm
>
>
> Ideally, you should be able to use most of the classes in the
> org.eclipse.datatools.connectivity.db.generic plug-in as your
> implementation for your profile. You may want to create wizard and
> property pages that are specific to postgres, but this is not
> necessary. You should create a driver definition that is specific to
> postgres; specifying appropriate default values, dbdefinition values,
> and required JAR files; the JAR files can be specified using the format,
> [<plug-id>]/<plug-in relative path to JAR>
> This will allow the default JAR list to use the JARs distributed with
> the plug-in. You can also just list the JAR names (when the user edits
> the definition and specifies the actual JAR file, they are given the
> option to update the other JARs listed using the same root folder). You
> should sublcass the base wizard implementation to add a filter so only
> your driver definitions are displayed to the user in your wizard (note,
> this cannot currently be done programmatically, but should be available
> by M2). You will need to create a dbdefinition instance and extension
> point for postgres. You may need to create a specialized SQL model and
> catalog loader for postgres (this will probably be the biggest task).
>
> Look for the Derby connection profile in the M2 release.
>
> Please let me know if you need anything else.
>
> Rob
>
>
|
|
|
| Re: postgres adapter [message #568476 is a reply to message #5720] |
Tue, 29 November 2005 13:40 |
| Eclipse User |
|
|
|
Hey Miguel,
The sample plug-in provided is a generic example that is not DB specific.
It simply stores a property containing a folder location and displays the
contents of that folder in the DSE when connected. (plug-in id is
org.eclipse.datatools.connectivity.sample.cp)
That said, there is a forthcoming implementation for Derby that should
serve as a complete DB specific example. This should be available in the
M2 release at the end of this month.
The following provides an overview of what is required for developing a
connection profile:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DTPConnectivityFrameworkDevelopersQSGuide.htm
The following provides specific documentation for the connectionProfile
extension point:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/CMF_I101905.html
The following provides specific documentation for the driver definition
framework:
http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DMF_I101205.html
Documentation for customizing the SQL model can be found at:
http://www.eclipse.org/datatools/project_modelbase/modelbase _doc/DTP%20Modelbase%20White%20paper.htm
Ideally, you should be able to use most of the classes in the
org.eclipse.datatools.connectivity.db.generic plug-in as your
implementation for your profile. You may want to create wizard and
property pages that are specific to postgres, but this is not necessary.
You should create a driver definition that is specific to postgres;
specifying appropriate default values, dbdefinition values, and required
JAR files; the JAR files can be specified using the format,
[<plug-id>]/<plug-in relative path to JAR>
This will allow the default JAR list to use the JARs distributed with the
plug-in. You can also just list the JAR names (when the user edits the
definition and specifies the actual JAR file, they are given the option to
update the other JARs listed using the same root folder). You should
sublcass the base wizard implementation to add a filter so only your
driver definitions are displayed to the user in your wizard (note, this
cannot currently be done programmatically, but should be available by M2).
You will need to create a dbdefinition instance and extension point for
postgres. You may need to create a specialized SQL model and catalog
loader for postgres (this will probably be the biggest task).
Look for the Derby connection profile in the M2 release.
Please let me know if you need anything else.
Rob
|
|
|
| Re: postgres adapter [message #568498 is a reply to message #5739] |
Thu, 01 December 2005 15:30 |
| Eclipse User |
|
|
|
thanks a lot!
I'll keep this arround until M2
Rob Cernich wrote:
> Hey Miguel,
>
> The sample plug-in provided is a generic example that is not DB
> specific. It simply stores a property containing a folder location and
> displays the contents of that folder in the DSE when connected.
> (plug-in id is org.eclipse.datatools.connectivity.sample.cp)
>
> That said, there is a forthcoming implementation for Derby that should
> serve as a complete DB specific example. This should be available in
> the M2 release at the end of this month.
>
> The following provides an overview of what is required for developing a
> connection profile:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DTPConnectivityFrameworkDevelopersQSGuide.htm
>
>
> The following provides specific documentation for the connectionProfile
> extension point:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/CMF_I101905.html
>
>
> The following provides specific documentation for the driver definition
> framework:
> http://www.eclipse.org/datatools/project_connectivity/connec tivity_doc/DMF_I101205.html
>
>
> Documentation for customizing the SQL model can be found at:
> http://www.eclipse.org/datatools/project_modelbase/modelbase _doc/DTP%20Modelbase%20White%20paper.htm
>
>
> Ideally, you should be able to use most of the classes in the
> org.eclipse.datatools.connectivity.db.generic plug-in as your
> implementation for your profile. You may want to create wizard and
> property pages that are specific to postgres, but this is not
> necessary. You should create a driver definition that is specific to
> postgres; specifying appropriate default values, dbdefinition values,
> and required JAR files; the JAR files can be specified using the format,
> [<plug-id>]/<plug-in relative path to JAR>
> This will allow the default JAR list to use the JARs distributed with
> the plug-in. You can also just list the JAR names (when the user edits
> the definition and specifies the actual JAR file, they are given the
> option to update the other JARs listed using the same root folder). You
> should sublcass the base wizard implementation to add a filter so only
> your driver definitions are displayed to the user in your wizard (note,
> this cannot currently be done programmatically, but should be available
> by M2). You will need to create a dbdefinition instance and extension
> point for postgres. You may need to create a specialized SQL model and
> catalog loader for postgres (this will probably be the biggest task).
>
> Look for the Derby connection profile in the M2 release.
>
> Please let me know if you need anything else.
>
> Rob
>
>
|
|
|
Powered by
FUDForum. Page generated in 0.04991 seconds
Home