| SQL Editor using an Oracle Connection [message #24920] |
Fri, 05 January 2007 04:00  |
Eclipse User |
|
|
|
Hi,
I'm using dtp 1.0 with an Oracle 10 database. I'm able to explore the
schemas and other objects in the database in the Datasource Explorer and
I can run several statements.
But follwing features seam not to work:
- Stored Procedures cannot be opened or edited: The editor says 'Failed
to get source' (not entries in the eclipse log)
- In the sql editor tables, columns or other database objects are never
proposed by the content assist
How can I avoid these problems?
Kind regards,
Stefan
|
|
|
|
| Re: SQL Editor using an Oracle Connection [message #25269 is a reply to message #25190] |
Wed, 10 January 2007 21:53  |
| Eclipse User |
|
|
|
Originally posted by: hcao.sybase.com
Hi Stefan,
Please see the inline comments below:
Max (Hui) Cao
Sybase Inc.
"Stefan Holzknecht" <s.holzknecht@intershop.de> wrote in message
news:eo27hv$t9j$1@utils.eclipse.org...
> Stefan Holzknecht wrote:
>> Hi,
>> I'm using dtp 1.0 with an Oracle 10 database. I'm able to explore the
>> schemas and other objects in the database in the Datasource Explorer and
>> I can run several statements.
>> But follwing features seam not to work:
>> - Stored Procedures cannot be opened or edited: The editor says 'Failed
>> to get source' (not entries in the eclipse log)
The generic catalog loader can't load the stored procedure source for you
since it's dependent on the JDBC api. You'll have to create your own catalog
loader for Oracle.
>> - In the sql editor tables, columns or other database objects are never
>> proposed by the content assist
The generic sql parser does not help much in this case. A fully-fledged
content assist feature needs the support of a customized parser. But I agree
there's something to improve, like the ModelUtil.getSchemas() method you
mentioned below.
>> How can I avoid these problems?
>> Kind regards,
>> Stefan
> I forget to say that other features cannot be used either. Eg. I'm unable
> to specify schema filters based on a selection when editing properties of
> a database. I remember a M? build with which I got a list of all schemas
> to choose from here.
>
The schema filter function is hosted in the connectivity project and I don't
know the details.
> Probably the wrong behaviour is connected to the fact that the method
> org.eclipse.datatools.sqltools.sql.util.ModelUtil.getSchemas (Database,
> String) is called with a catalog name null (while debugging the content
> assist invokes this in SQLDBProposalsService.loadDBProposals(). The field
> SQLDBProposalsRequest._database is set to null.). The database name is set
> as default catalog name. It seams to me that the empty string is a better
> solution here (at least the Database object contained a proper filled
> catalog with empty name).
> This information probably helps in order to provide a better working sql
> editor.
This is a good suggestion, I'v created BZ170153 for it. Thanks!
> Stefan
|
|
|
| Re: SQL Editor using an Oracle Connection [message #583852 is a reply to message #24920] |
Wed, 10 January 2007 03:22 |
| Eclipse User |
|
|
|
Stefan Holzknecht wrote:
> Hi,
> I'm using dtp 1.0 with an Oracle 10 database. I'm able to explore the
> schemas and other objects in the database in the Datasource Explorer and
> I can run several statements.
> But follwing features seam not to work:
> - Stored Procedures cannot be opened or edited: The editor says 'Failed
> to get source' (not entries in the eclipse log)
> - In the sql editor tables, columns or other database objects are never
> proposed by the content assist
> How can I avoid these problems?
> Kind regards,
> Stefan
I forget to say that other features cannot be used either. Eg. I'm
unable to specify schema filters based on a selection when editing
properties of a database. I remember a M? build with which I got a list
of all schemas to choose from here.
Probably the wrong behaviour is connected to the fact that the method
org.eclipse.datatools.sqltools.sql.util.ModelUtil.getSchemas (Database,
String) is called with a catalog name null (while debugging the content
assist invokes this in SQLDBProposalsService.loadDBProposals(). The
field SQLDBProposalsRequest._database is set to null.). The database
name is set as default catalog name. It seams to me that the empty
string is a better solution here (at least the Database object contained
a proper filled catalog with empty name).
This information probably helps in order to provide a better working sql
editor.
Stefan
|
|
|
| Re: SQL Editor using an Oracle Connection [message #583882 is a reply to message #25190] |
Wed, 10 January 2007 21:53 |
| Eclipse User |
|
|
|
Hi Stefan,
Please see the inline comments below:
Max (Hui) Cao
Sybase Inc.
"Stefan Holzknecht" <s.holzknecht@intershop.de> wrote in message
news:eo27hv$t9j$1@utils.eclipse.org...
> Stefan Holzknecht wrote:
>> Hi,
>> I'm using dtp 1.0 with an Oracle 10 database. I'm able to explore the
>> schemas and other objects in the database in the Datasource Explorer and
>> I can run several statements.
>> But follwing features seam not to work:
>> - Stored Procedures cannot be opened or edited: The editor says 'Failed
>> to get source' (not entries in the eclipse log)
The generic catalog loader can't load the stored procedure source for you
since it's dependent on the JDBC api. You'll have to create your own catalog
loader for Oracle.
>> - In the sql editor tables, columns or other database objects are never
>> proposed by the content assist
The generic sql parser does not help much in this case. A fully-fledged
content assist feature needs the support of a customized parser. But I agree
there's something to improve, like the ModelUtil.getSchemas() method you
mentioned below.
>> How can I avoid these problems?
>> Kind regards,
>> Stefan
> I forget to say that other features cannot be used either. Eg. I'm unable
> to specify schema filters based on a selection when editing properties of
> a database. I remember a M? build with which I got a list of all schemas
> to choose from here.
>
The schema filter function is hosted in the connectivity project and I don't
know the details.
> Probably the wrong behaviour is connected to the fact that the method
> org.eclipse.datatools.sqltools.sql.util.ModelUtil.getSchemas (Database,
> String) is called with a catalog name null (while debugging the content
> assist invokes this in SQLDBProposalsService.loadDBProposals(). The field
> SQLDBProposalsRequest._database is set to null.). The database name is set
> as default catalog name. It seams to me that the empty string is a better
> solution here (at least the Database object contained a proper filled
> catalog with empty name).
> This information probably helps in order to provide a better working sql
> editor.
This is a good suggestion, I'v created BZ170153 for it. Thanks!
> Stefan
|
|
|
Powered by
FUDForum. Page generated in 0.25289 seconds
Home