Recogonize Oracle Packages

[Jack]

1) It would be nice if the ORM would take into account Oracle packages in lieu of just dumping everything into stored procedures.  Even if it were just grouping in the SchemaExplorer.  I can have the same named proc in different packages.

2) Will Functions be supported?  both inside and outside packages?

3) What about quering data via SProc?  I currently have ~200+ views but I have SProcs written to return the data as strongly typed ref cursors

4) What about functions that return multiple ref cursors?  For example to populate a large complex screen I'm selecting all the data in one pass.  All the master data records and all the detail data records.  In my BO or my UI I can then use LINQ/filters to handle the parent/child.  How would I model that in the ORM?

5) I'm a bit confused on what one can actually do with a SP that is imported...  I can get it into the Schema Explorer but can't do anything from there.

6) Is there an ETA on when I can use existing SProcs for CRUD?  In the simple case a ORM generated SPROC is fine but how do I tie in my extra logic?  Currently I might do a simple insert/update but then I do some math and update other tables.  How do I simply extend an autogenerated CUD proc?

7) I'm missing how to have a single entity vs. a collection.  Say for example I have Person and PersonList both which derive from my v_person view.  Currently I have a proc pkg_person.get_person(an_id in number, arec_data in out system_refcursor) and pkg_person.get_person_list(arec_data in out system_refcursor). 

What I'm hoping for is that in my BO I can do Person person =  MyDAL.Person.GetById(123) and/or Person person = MyDal.PersonList(x=>x.Id == 123).  PersonList was somehow populated earlier.

I think I'm missing where all the retrieval logic goes...  I'm currently have an existing library of retrieval packages/procedures - how do I plug them in so they are wrapped and the data transformation is all completed for me?  Is that not available?

Thanks

jack

I feel like the ORM has potential... my goal is to have this giant DAL that

[Jack]

Sorry I digressed a bit and went off topic and likely posted in the wrong forum.  I've put in a related ticket instead.  This could be wacked or moved elsewhere.  I'm happy to update it with answers from the ticket.

thanks

jack

[Damyan Bogoev]

Hello Jack,

Firstly I want to thank you for the valuable input.
1-3) We will consider your suggestions and investigate further options to implement them.
4)     I am afraid that functions that return multiple ref cursors are not supported yet.
5)     When the OpenAccess Data Model Wizard generates a domain model from a database, it creates methods in the context class for each stored procedure in the model. I suggest you reading the following help topic which explains how to use the imported stored procedures.
6)    This is actually on our to-do list. We plan to provide support for this feature in the future.
7)    It is not necessary to use stored procedure in order to retrieve data from the database. What you could do is to use the Telerik OpenAccess ORM visual designer to create a domain model. Later on you will be able to query the domain model using Linq.
Hope that helps.

Best wishes,
Damyan Bogoev
the Telerik team

[Henk]

Hi Damyan,

1. I noticed that in the latest q3 release there are still no distinction when different packages has the same stored procedures/functions names. Just put package_name.procedure/function including the parameters, like the package specification.
2. Also schema name with underscore like at_admin become atadmin. 
3. It would be nice when you can disable/enable extra columns If you import a table/view with 20 columns using the domain model wizard  and the system updates all the necessary references accordingly. Yes, you can delete a column but not disable it.If I delete a column and choose update from database then I cannot choose the deleted column to add it back into the model...
4. If i create a view with multiple tables and with outer joins you dont have perse an identity column based on the pk's of the joined tables. But you need an identity column, yes?

[Damyan Bogoev]

Hi,

1-2. Currently we are working on improving the schema reading in Oracle.
3. We will consider your suggestions in the future improvements of the wizard.
4. I am afraid that currently we do not provide support for this scenario. The schema reading tool does not read the primary key column for a specific view.
I am sorry for the inconvenience caused.

All the best,
Damyan Bogoev
the Telerik team