Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Eclipse Projects » Dali » DDL Generation observations
DDL Generation observations [message #433932] Tue, 25 April 2006 20:25 Go to next message
Kevin Sutter is currently offline Kevin SutterFriend
Messages: 104
Registered: July 2009
Senior Member
First off, now that I finally got DDL generation to actually kick off and
process, I was quite impressed with the result. I had read through the
gotchas already discovered, so I was ready for those. But, overall, it
generated the necessary ddl that matched my annotated source. Thanks.

But, of course, I have a few observations...

o Are Discriminator columns supported yet? I had a Discriminator Column
defined for one of my tables, and I didn't see that get generated in my
DDL. I modified my script before running it.

o Should Embedded classes work yet? One of my Entities referenced an
Embeddable class for a couple of its properties. But, the generated ddl
for this table didn't include these additional fields. They were
TIMESTAMP fields, if that makes any difference.

o The statement terminator doesn't seem to be working quite right. By
default, it used "GO" as the statement terminator. But, the generated
script didn't have the special statement to indicate the terminator:

-- <ScriptOptions statementTerminator="GO">

o Related to this, why wouldn't the semi-colon (;) be the default value
for the statement terminator? The word "GO" seems too common since it
found that in several locations of my generated DDL and, of course, died
all over the place when I first tried to run it... :-)

o The DDL generation seems to combine table definitions from several
files. (Again, this may be due to only having the option of running this
tool at the Project level.) Because I was experimenting with generating
Entities at the same time, I had multiple Entity class files that were
referencing the same table names. The DDL generator just combined all of
these definitions into a single table definition. This doesn't seem
right. I would have preferred a warning or an error that it couldn't
generate the ddl due to conflicting annotations.

That's it. Once I worked around these items, I was able to generate ddl
that actually dropped my old tables and re-created new ones based on the
annotations.

Thanks,
Kevin
Re: DDL Generation observations [message #433935 is a reply to message #433932] Tue, 25 April 2006 21:20 Go to previous messageGo to next message
Karen Butzke is currently offline Karen ButzkeFriend
Messages: 220
Registered: July 2009
Senior Member
Hi Kevin,

> o Are Discriminator columns supported yet? I had a Discriminator Column
> defined for one of my tables, and I didn't see that get generated in my
> DDL. I modified my script before running it.

I am currently adding support for Discriminator Column in our model and
UI. Once this is complete, we could take advantage of that in the DDL
gen. Could you open an enhancement request that is specific to
DiscriminatorColumn support in DDL gen?

> o Should Embedded classes work yet? One of my Entities referenced an
> Embeddable class for a couple of its properties. But, the generated ddl
> for this table didn't include these additional fields. They were
> TIMESTAMP fields, if that makes any difference.

This sounds like something that probably wasn't in our model when the ddl
gen was first written. Open an enhancement request for this as well.

> o The statement terminator doesn't seem to be working quite right. By
> default, it used "GO" as the statement terminator. But, the generated
> script didn't have the special statement to indicate the terminator:

> -- <ScriptOptions statementTerminator="GO">

> o Related to this, why wouldn't the semi-colon (;) be the default value
> for the statement terminator? The word "GO" seems too common since it
> found that in several locations of my generated DDL and, of course, died
> all over the place when I first tried to run it... :-)

I would report this to the WTP RDB component. We take our model and turn
it into an RDB model and then generate the ddl through through an RDB
wizard. After our .5 release we will be migrating to using DTP project
along with WTP, so you might want to report it to DTP as well as RDB.


> o The DDL generation seems to combine table definitions from several
> files. (Again, this may be due to only having the option of running this
> tool at the Project level.) Because I was experimenting with generating
> Entities at the same time, I had multiple Entity class files that were
> referencing the same table names. The DDL generator just combined all of
> these definitions into a single table definition. This doesn't seem
> right. I would have preferred a warning or an error that it couldn't
> generate the ddl due to conflicting annotations.

I agree, it does seem that we should warn you or let you choose whether
that is what you want to do. Looks like another enhancement request is
necessary :)

Thanks for all your feedback.

Karen
Re: DDL Generation observations [message #433942 is a reply to message #433935] Wed, 26 April 2006 14:06 Go to previous message
Kevin Sutter is currently offline Kevin SutterFriend
Messages: 104
Registered: July 2009
Senior Member
Thanks, Karen. I have opened the requested Enhancement and Bug reports.
Details of the reports are posted below...

Karen wrote:

> Hi Kevin,

>> o Are Discriminator columns supported yet? I had a Discriminator Column
>> defined for one of my tables, and I didn't see that get generated in my
>> DDL. I modified my script before running it.

> I am currently adding support for Discriminator Column in our model and
> UI. Once this is complete, we could take advantage of that in the DDL
> gen. Could you open an enhancement request that is specific to
> DiscriminatorColumn support in DDL gen?

Enhancement 138625 was opened.

>> o Should Embedded classes work yet? One of my Entities referenced an
>> Embeddable class for a couple of its properties. But, the generated ddl
>> for this table didn't include these additional fields. They were
>> TIMESTAMP fields, if that makes any difference.

> This sounds like something that probably wasn't in our model when the ddl
> gen was first written. Open an enhancement request for this as well.

Enhancement 138627 was opened.

>> o The statement terminator doesn't seem to be working quite right. By
>> default, it used "GO" as the statement terminator. But, the generated
>> script didn't have the special statement to indicate the terminator:

>> -- <ScriptOptions statementTerminator="GO">

>> o Related to this, why wouldn't the semi-colon (;) be the default value
>> for the statement terminator? The word "GO" seems too common since it
>> found that in several locations of my generated DDL and, of course, died
>> all over the place when I first tried to run it... :-)

> I would report this to the WTP RDB component. We take our model and turn
> it into an RDB model and then generate the ddl through through an RDB
> wizard. After our .5 release we will be migrating to using DTP project
> along with WTP, so you might want to report it to DTP as well as RDB.

Bug reports 138631 and 138633 were opened against rdp and dtp,
respectively.

>> o The DDL generation seems to combine table definitions from several
>> files. (Again, this may be due to only having the option of running this
>> tool at the Project level.) Because I was experimenting with generating
>> Entities at the same time, I had multiple Entity class files that were
>> referencing the same table names. The DDL generator just combined all of
>> these definitions into a single table definition. This doesn't seem
>> right. I would have preferred a warning or an error that it couldn't
>> generate the ddl due to conflicting annotations.

> I agree, it does seem that we should warn you or let you choose whether
> that is what you want to do. Looks like another enhancement request is
> necessary :)

Enhancement 138634 was opened.

> Thanks for all your feedback.

> Karen
Re: DDL Generation observations [message #584892 is a reply to message #433932] Tue, 25 April 2006 21:20 Go to previous message
Karen Butzke is currently offline Karen ButzkeFriend
Messages: 220
Registered: July 2009
Senior Member
Hi Kevin,

> o Are Discriminator columns supported yet? I had a Discriminator Column
> defined for one of my tables, and I didn't see that get generated in my
> DDL. I modified my script before running it.

I am currently adding support for Discriminator Column in our model and
UI. Once this is complete, we could take advantage of that in the DDL
gen. Could you open an enhancement request that is specific to
DiscriminatorColumn support in DDL gen?

> o Should Embedded classes work yet? One of my Entities referenced an
> Embeddable class for a couple of its properties. But, the generated ddl
> for this table didn't include these additional fields. They were
> TIMESTAMP fields, if that makes any difference.

This sounds like something that probably wasn't in our model when the ddl
gen was first written. Open an enhancement request for this as well.

> o The statement terminator doesn't seem to be working quite right. By
> default, it used "GO" as the statement terminator. But, the generated
> script didn't have the special statement to indicate the terminator:

> -- <ScriptOptions statementTerminator="GO">

> o Related to this, why wouldn't the semi-colon (;) be the default value
> for the statement terminator? The word "GO" seems too common since it
> found that in several locations of my generated DDL and, of course, died
> all over the place when I first tried to run it... :-)

I would report this to the WTP RDB component. We take our model and turn
it into an RDB model and then generate the ddl through through an RDB
wizard. After our .5 release we will be migrating to using DTP project
along with WTP, so you might want to report it to DTP as well as RDB.


> o The DDL generation seems to combine table definitions from several
> files. (Again, this may be due to only having the option of running this
> tool at the Project level.) Because I was experimenting with generating
> Entities at the same time, I had multiple Entity class files that were
> referencing the same table names. The DDL generator just combined all of
> these definitions into a single table definition. This doesn't seem
> right. I would have preferred a warning or an error that it couldn't
> generate the ddl due to conflicting annotations.

I agree, it does seem that we should warn you or let you choose whether
that is what you want to do. Looks like another enhancement request is
necessary :)

Thanks for all your feedback.

Karen
Re: DDL Generation observations [message #584968 is a reply to message #433935] Wed, 26 April 2006 14:06 Go to previous message
Kevin Sutter is currently offline Kevin SutterFriend
Messages: 104
Registered: July 2009
Senior Member
Thanks, Karen. I have opened the requested Enhancement and Bug reports.
Details of the reports are posted below...

Karen wrote:

> Hi Kevin,

>> o Are Discriminator columns supported yet? I had a Discriminator Column
>> defined for one of my tables, and I didn't see that get generated in my
>> DDL. I modified my script before running it.

> I am currently adding support for Discriminator Column in our model and
> UI. Once this is complete, we could take advantage of that in the DDL
> gen. Could you open an enhancement request that is specific to
> DiscriminatorColumn support in DDL gen?

Enhancement 138625 was opened.

>> o Should Embedded classes work yet? One of my Entities referenced an
>> Embeddable class for a couple of its properties. But, the generated ddl
>> for this table didn't include these additional fields. They were
>> TIMESTAMP fields, if that makes any difference.

> This sounds like something that probably wasn't in our model when the ddl
> gen was first written. Open an enhancement request for this as well.

Enhancement 138627 was opened.

>> o The statement terminator doesn't seem to be working quite right. By
>> default, it used "GO" as the statement terminator. But, the generated
>> script didn't have the special statement to indicate the terminator:

>> -- <ScriptOptions statementTerminator="GO">

>> o Related to this, why wouldn't the semi-colon (;) be the default value
>> for the statement terminator? The word "GO" seems too common since it
>> found that in several locations of my generated DDL and, of course, died
>> all over the place when I first tried to run it... :-)

> I would report this to the WTP RDB component. We take our model and turn
> it into an RDB model and then generate the ddl through through an RDB
> wizard. After our .5 release we will be migrating to using DTP project
> along with WTP, so you might want to report it to DTP as well as RDB.

Bug reports 138631 and 138633 were opened against rdp and dtp,
respectively.

>> o The DDL generation seems to combine table definitions from several
>> files. (Again, this may be due to only having the option of running this
>> tool at the Project level.) Because I was experimenting with generating
>> Entities at the same time, I had multiple Entity class files that were
>> referencing the same table names. The DDL generator just combined all of
>> these definitions into a single table definition. This doesn't seem
>> right. I would have preferred a warning or an error that it couldn't
>> generate the ddl due to conflicting annotations.

> I agree, it does seem that we should warn you or let you choose whether
> that is what you want to do. Looks like another enhancement request is
> necessary :)

Enhancement 138634 was opened.

> Thanks for all your feedback.

> Karen
Previous Topic:persistence.xml management
Next Topic:Refreshing of Persistence Properties
Goto Forum:
  


Current Time: Wed Nov 26 07:57:24 GMT 2014

Powered by FUDForum. Page generated in 0.03229 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.2.
Copyright ©2001-2010 FUDforum Bulletin Board Software