springframework-developer — Spring Framework developers list

Re: [Springframework-developer] Commons-Validator for Spring

[Seth L. <se...@eh...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


| (2) Improve the handling of Errors objects to allow message resources
to be
| specified as arguments that would not be resolved to their corresponding
| messages until reaching the view. I really like this solution, but I have
| not thought it through completely. One more note: since the errorArgs
| parameter is an array of Objects (not an array of Strings), maybe there
| would be a way to use a "Message" object that would be resolved at a later
| time, while preserving the current functionality of Strings that are used
| directly as arguments. For all I know, there is already a way to do
this (it
| seems logical enough), and I just don't know about it. This J2EE stuff
is so
| hard to keep up with...

Daniel,

I just posted a simple patch w/ unit test for the functionality we've
discussed.  Please review and see if it meets your needs.

http://opensource.atlassian.com/projects/spring/secure/ViewIssue.jspa?key=SPR-56


Seth
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFAQ9Ua5EIB1scRes8RAjA6AJ4vfkVAqr7FSl0h/du+Yuob+pSzJwCfaZfO
QBboOgIetd7TU2y+/pNUZ+Q=
=fPxm
-----END PGP SIGNATURE-----

RE: [Springframework-developer] JMS brainstorming

[Mark P. <mar...@co...>]

Hi,

I've added Andre's code into the sandbox. =20

I'd like to get some input on the following.... Can we count on using
the JMS 1.1 API or should we be compatible with the JMS 1.0.2 API? The
reason is that JMS 1.1 introduced "domain unification", meaning that
there the classes are no longer split into "Queue" and "Topic" class
hierarchies.  The code can now refer to a "ConnectionFactory" instead of
it being a "QueueConnectionFactory" or a "TopicConnectionFactory".  This
applies up and down the class hierarchy since the split is rather high
level.  The end result is that using the JMS 1.1 api makes the client
code much easier to use since there is less obvious duplication.  Other
advantages can be read in the spec...such as being able to use  queues
and topics in the same transaction (probably since they now can share
the same Session object).

The templates now are written using the 1.0.2 style classes, which I
haven't used in a long time....but I'd like to hear what people are
using considering how much it will affect the shape of the code...  I'd
vote for dropping 1.0.2 support.

Cheers,
Mark



-----Original Message-----
From: spr...@li...
[mailto:spr...@li...] On Behalf
Of Andre Biryukov
Sent: Thursday, February 26, 2004 11:46 AM
To: rod...@in...;
spr...@li...
Subject: Re: [Springframework-developer] JMS brainstorming


Hi Rod, Hi All,

I've been working on the JMS support for a few weeks now and I think I'm

ready to go public with it to find out your opinion. So far, I got Mark
to=20
volunteer to look at the code, make suggestions and possibly contribute
some=20
code. If anybody is interested to look at it as well, just drop me a
mail,=20
I'll be happy to forward what I have so far to you.

I'm not only interested in the feedback about the code, but would like
to=20
hear opinions on what we should include and what not. The API for
messaging=20
is huge, well, big. The challenge is to try and put as much
functionality as=20
possible, but at the same token, try not to make it look complex.

Here's the list of things I've implemented:

1. Templates for publish/subscribe and point-to-point messaging. I did=20
include support for both durable and non-durable messaging.
2. The consumers will be able to not only create messages, but also will

have some control over how they will be delivered. I.e. modify default=20
delivery mode, priority, time to live.
3. Message listeners will be instrumental in automatically registering
beans=20
as subscribers to topics and queues.

I left out support for the transacted sessions. Sessions created are all
set=20
to "no-transaction" and "auto-ACK". Is this something we would like to
see=20
in the framework or should we just keep it simple for now ? The more=20
suggestions I get now, the better, so please open the flood gate. :)

Thanks,

Andre.

>From: "Rod Johnson" <rod...@in...>
>To: <spr...@li...>
>CC: "Andre Biryukov" <and...@ho...>
>Subject: Re: [Springframework-developer] JMS brainstorming
>Date: Thu, 26 Feb 2004 06:35:33 -0000
>
>Rong
>
>JMS support is in the works, and should make 1.1.
>
>Andre Biryukov is currently working on this, and should be joining the=20
>Spring team soon.
>
>Andre, if you want to share your work in progress with the Spring dev=20
>list (or with Rong directly), now would be a good time :-)
>
>Regards,
>Rod
>
>----- Original Message -----
>From: "Ou, Rong" <Ro...@sa...>
>To: <spr...@li...>
>Sent: Tuesday, February 24, 2004 9:44 PM
>Subject: [Springframework-developer] JMS brainstorming
>
>
>Hi all,
>
>We currently have a project that does a lot of JMS messaging, and we=20
>desperately need an API to simplify the interface layer to JMS. I heard

>rumors that JMS support will be added to Spring 1.1, but we just can't=20
>wait that long. So I am thinking about doing something in the interim.=20
>Who knows, if it comes out good, maybe I can contribute the code back.
>
>In Rod's book there is only a couple of pages that briefly touches upon
>JMS.
>I looked at the interface21 code, the JMS portion is also pretty=20
>simplistic.
>I understand Spring has evolved a lot since the interface21 days. So to
all
>you experienced Spring'ers, if you were to add JMS support today, what=20
>would
>you do? I am still fairly new to Spring, so any suggestions and help
are
>welcome.
>
>Thanks,
>
>Rong
>
>
>-------------------------------------------------------
>SF.Net is sponsored by: Speed Start Your Linux Apps Now.
>Build and deploy apps & Web services for Linux with
>a free DVD software kit from IBM. Click Now!=20
>http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
>_______________________________________________
>Springframework-developer mailing list=20
>Spr...@li...
>https://lists.sourceforge.net/lists/listinfo/springframework-developer
>
>

_________________________________________________________________
Stay informed on Election 2004 and the race to Super Tuesday.=20
http://special.msn.com/msn/election2004.armx



-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=3D1356&alloc_id=3D3438&op=3Dclick
_______________________________________________
Springframework-developer mailing list
Spr...@li...
https://lists.sourceforge.net/lists/listinfo/springframework-developer

[Springframework-developer] Website and docs up-to-date

[Alef A. <al...@jt...>]

Thomas, all,

=20

While updating the documentation I took the liberty to also update the
website when J=FCrgen announced RC2, so don=92t worry about that =
anymore=85

=20

Alef

=20

Re: [Springframework-developer] Commons-Validator for Spring

[Seth L. <se...@eh...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


| (2) Improve the handling of Errors objects to allow message resources
to be
| specified as arguments that would not be resolved to their corresponding
| messages until reaching the view. I really like this solution, but I have
| not thought it through completely. One more note: since the errorArgs
| parameter is an array of Objects (not an array of Strings), maybe there
| would be a way to use a "Message" object that would be resolved at a later
| time, while preserving the current functionality of Strings that are used
| directly as arguments. For all I know, there is already a way to do
this (it
| seems logical enough), and I just don't know about it. This J2EE stuff
is so
| hard to keep up with...
|

Daniel,

Spring seems to have an interface that says "Hi, I'm a resolvable
Message".  This is similar to the Message object you propose above.

http://www.springframework.org/docs/api/org/springframework/context/MessageSourceResolvable.html

I'm now subclasses ResourceBundleMessageSource to search through the
message args, look for MessageSourceResolvable, and resolve it.

I'm specifically reimplementing the getMessageInternal method from
AbstractMessageSource.  That's a little ugly, but that seems to be the
point where the MessageFormat is used.

Hope that helps,
Seth
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFAQ81F5EIB1scRes8RAqRiAJ46Xm6+uoI/J7ci/3trE5JDcv+/sACghEdE
Gc6YPhbEalOzhAAkbSC/chE=
=1VaR
-----END PGP SIGNATURE-----

[Springframework-developer] Spring Framework 1.0 RC2 released

[<jue...@we...>]

Dear Spring community,
=20
I'm pleased to announce the release of Spring Framework 1.0 RC2. Besides =
some internal refactorings and clarifications, this release introduces =
the following new features and changes (among others):
=20
* scheduling support via Quartz and Timer
* support for the SqlMapClient API of iBATIS SQL Maps 2
* JdbcTemplate convenience methods, taking prepared statement arguments =
as Object array
* support for custom RMI socket factories and additional RMI invocation =
parameters
* PreferencesPlaceholderConfigurer, resolving placeholders via J2SE 1.4 =
Preferences
* optional "type" attribute for "constructor-arg" tag in XML bean =
definitions
=20
* revised BeanFactoryLocator implementations
* simplified AOP Advisor interface hierarchy
* revised DataFieldMaxValueIncrementer implementation hierarchy
* extended mapping configuration options in LocalSessionFactoryBean
* metadata support uses latest Commons Attributes snapshot
* VelocityView supports Velocity Tools 1.1
=20
As usual, see the changelog for details. Note that the Image Database =
sample app in "samples/imagedb" also illustrates scheduling support now.
=20
We plan to release Spring 1.0 final shortly, with an updated AOP =
Alliance version and completed docs. Please report any feedback on the =
RC2 release promptly.
=20
To see what's coming up next, have a look at the major milestones for =
Spring 1.1 that we've gathered in our JIRA =
(http://opensource.atlassian.com/projects/spring). Note that this list =
is still subject to change.
=20
Regards,
Juergen
=20

RE: [Springframework-developer] JavaMailSenderImpl

[Oliver H. <Oli...@ou...>]

Thanks Juergen and every one else who gave a suggestion, it's the little =
things like this that make Spring great to work with.=20


-----Original Message-----
From: j=FCrgen h=F6ller [werk3AT] [mailto:jue...@we...]=20
Sent: Monday, 1 March 2004 6:18 PM
To: spr...@li...
Subject: Re: [Springframework-developer] JavaMailSenderImpl


Or simply subclass JavaMailSenderImpl and invoke setSession with a =
custom Session. However, I see the original point: I'll add =
"javaMailProperties".
=20
Will still make it into RC2, BTW, as I'm currently preparing it :-)
=20
Juergen
=20

Re: [Springframework-developer] JavaMailSenderImpl

[Seth L. <se...@eh...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

jürgen höller [werk3AT] wrote:
| Or simply subclass JavaMailSenderImpl and invoke setSession with a
custom Session. However, I see the original point: I'll add
"javaMailProperties".
|
| Will still make it into RC2, BTW, as I'm currently preparing it :-)


I'd just like to say thanks for fielding requests for new features.
It's refreshing to see such an understanding and open group.

Seth
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFAQ6Dp5EIB1scRes8RAuVOAJ9aGoqBKtp+iLwHDXKuf3bFO7qKyQCggvh+
pZJIWthi+NXHMldKdaYj+Nk=
=Z7+y
-----END PGP SIGNATURE-----

[Springframework-developer] spring-rcp subproject structure and initial commits

[Keith D. <kd...@cs...>]

Hey yall -
=20
The spring-rcp sub project structure now exists in the CVS repository as =
a
new module.  You'll find that the structure very much mirrors that of =
the
spring module.  I made some initial commits last night including:
=20
org.springframework.rcp.image
    A image and icon loading/caching library using resource-bundle style
lookup.  It abstracts away hardcoded image paths in your code, supports
at-runtime resizable icons (for accessbility), and provides built in =
broken
image indicator support.  The design is very MessageSource like and also
builds off Spring's core.io.Resource classes for accessing image =
resources.
=20
org.springframework.rcp.util
  Misc utility classes (mostly static-utility) that aim to reduce repeat
coding.  Several of these likely make sense for inclusion in the core at
some point...
=20
I'll be continuing commiting additional packages this week including:
- The rules-based data validation stuff with source-level attributes
(currently I'm working on this with Seth Ladd and Sam Newman.)
- The action contribution framework (the UI management side of actions.)
Sam Newman has also been working with me on a general command framework =
that
integrates with the visual side of an action for rich-clients.
- Test suites and samples for all of the above.
=20
Obviously the code, since we're just getting started, is considered
experimental and may undergo significant changes. =20
Keep the feedback coming!  :-) I've gotten a lot of useful ideas and
suggestions so far!  I will post announcements as new packages are
committed.  Keith
=20

=20

=20

Re: [Springframework-developer] Commons-Validator for Spring

[Seth L. <se...@eh...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


| (2) Improve the handling of Errors objects to allow message resources
to be
| specified as arguments that would not be resolved to their corresponding
| messages until reaching the view. I really like this solution, but I have
| not thought it through completely. One more note: since the errorArgs
| parameter is an array of Objects (not an array of Strings), maybe there
| would be a way to use a "Message" object that would be resolved at a later
| time, while preserving the current functionality of Strings that are used
| directly as arguments. For all I know, there is already a way to do
this (it
| seems logical enough), and I just don't know about it. This J2EE stuff
is so
| hard to keep up with...

Daniel,

I asked the list this exact question last week!  We're running into the
same problems. :)

I like your option (2).  For a quick solution over here, we did this: if
the String in the errorArgs array is this: "${something.key}" then we
treated that as a message key.  This is close to your proposal to place
Message objects into the errorArgs array.

I hope that helps.
Seth
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFAQ4U95EIB1scRes8RAtfZAKCEp0T69XvVBXfxg9d7YUJIqS68PQCcD5cY
vWqs9ssqri9zLhw9i1zSIZ4=
=2ubC
-----END PGP SIGNATURE-----

Re: [Springframework-developer] Tiles section in reference docs

[Darren D. <da...@da...>]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Monday 01 March 2004 16:01, Alef Arendsen wrote:
> I happen to - finally - have some spare time here, I'll take of it...
>
> Alef

Thx Alef - I had originally said I'd do it but time once again defeated me.
=2D --=20

Darren Davison
Public Key: http://www.davison.uk.net/key.jsp
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAQ2NhKLMLAN01aw0RAvEfAJ9NQqERThxMWPw9fzdrpr4Ssbm2XACcCoxj
IzZ4fqalfmTIq+uPqgzxj7U=3D
=3D9JhL
=2D----END PGP SIGNATURE-----

RE: [Springframework-developer] Tiles section in reference docs

[Alef A. <al...@jt...>]

I happen to - finally - have some spare time here, I'll take of it...

Alef


> -----Original Message-----
> From: spr...@li...
> [mailto:spr...@li...] On =
Behalf
> Of j=FCrgen h=F6ller [werk3AT]
> Sent: Monday, March 01, 2004 4:25 PM
> To: spr...@li...
> Subject: [Springframework-developer] Tiles section in reference docs
>=20
> Alef, others,
>=20
> We need to merge the tiles.html document into the reference docs for =
1.0
> final. If anybody finds the time to do this, please go ahead promptly =
:-)
>=20
> Juergen
>=20
>=20
>=20
> -------------------------------------------------------
> SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> Build and deploy apps & Web services for Linux with
> a free DVD software kit from IBM. Click Now!
> http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
> _______________________________________________
> Springframework-developer mailing list
> Spr...@li...
> https://lists.sourceforge.net/lists/listinfo/springframework-developer

[Springframework-developer] OT: New book and release date

[Lars F. <lar...@gm...>]

Rod and Jürgen,

just pre-ordered your new book yesterday.

Amazon says it will be released in April - is this still realistic ?

I'm asking this question in this forum because other readers
might be interested in buying the book.

Thanks,
Lars

[Springframework-developer] Tiles section in reference docs

[<jue...@we...>]

Alef, others,
=20
We need to merge the tiles.html document into the reference docs for 1.0 =
final. If anybody finds the time to do this, please go ahead promptly =
:-)
=20
Juergen
=20

Re: [Springframework-developer] VelocityEngineFactory

[<jue...@we...>]

Good point - I've just made initVelocityEngine synchronized. =
initVelocityEngine forces (re)initialization of the engine, without the =
need to access it; it's typically invoked in startup code. =
getVelocityEngine returns the engine, lazily initializing it if not =
already initialized. I see value in both methods; I don't like code that =
forces initialization through calling a getter, without needing the =
returned instance.
=20
BTW, RC2 is gonna get uploaded to SourceForge any minute now...
=20
Juergen
=20

________________________________

Von: spr...@li... im Auftrag =
von Darren Davison
Gesendet: Mo 01.03.2004 14:16
An: spr...@li...
Betreff: [Springframework-developer] VelocityEngineFactory



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


VelocityEngineFactory.initVelocityEngine is a public unsynchronized =
method.=20
It's scope was widened in 1.0-m3.  It seems to have been made public so
that VelocityConfigurer can call it.=20

If it's ok to leave it public and unsynchronized,then it looks odd that =
the
VelocityEngineFactory.getVelocityEngine() method is synchronized.  It
returns the initialized VelocityEngine instance or calls
initVelocityEngine() before returning the instance.

Should VelocityConfigurer call VelocityEngineFactory.getVelocityEngine()
instead so that initVelocityEngine() can be protected?  Or am I missing
something else (as usual)?

- --

Darren Davison
Public Key: http://www.davison.uk.net/key.jsp
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAQze1KLMLAN01aw0RAlQRAJ4vY6cTubGcpHUXFXmmqUQNr5q7fQCgnWgM
wwz8DGKkMulwA90hH6ZtpvI=3D
=3DV4Ys
-----END PGP SIGNATURE-----


-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
_______________________________________________
Springframework-developer mailing list
Spr...@li...
https://lists.sourceforge.net/lists/listinfo/springframework-developer

[Springframework-developer] VelocityEngineFactory

[Darren D. <da...@da...>]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


VelocityEngineFactory.initVelocityEngine is a public unsynchronized method.=
 =20
It's scope was widened in 1.0-m3.  It seems to have been made public so=20
that VelocityConfigurer can call it. =20

If it's ok to leave it public and unsynchronized,then it looks odd that the=
=20
VelocityEngineFactory.getVelocityEngine() method is synchronized.  It=20
returns the initialized VelocityEngine instance or calls=20
initVelocityEngine() before returning the instance.

Should VelocityConfigurer call VelocityEngineFactory.getVelocityEngine()=20
instead so that initVelocityEngine() can be protected?  Or am I missing=20
something else (as usual)?

=2D --=20

Darren Davison
Public Key: http://www.davison.uk.net/key.jsp
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAQze1KLMLAN01aw0RAlQRAJ4vY6cTubGcpHUXFXmmqUQNr5q7fQCgnWgM
wwz8DGKkMulwA90hH6ZtpvI=3D
=3DV4Ys
=2D----END PGP SIGNATURE-----

Re: [Springframework-developer] Hibernate meta data

[Jozsa K. <dy...@on...>]

On Fri, Feb 27, 2004 at 02:56:56PM -0500, Keith Donald wrote:
> From what I remember from the Hibernate APIs, there is no way to get the
> Configuration from the Session Factory.  The Configuration does create
> SessionFactories, but the SessionFactory does not have a way to traverse
> back to its associated Configuration.

Yeah that was always a problem for me too, every time I wanted to use
SchemaExport. Currently, I gotta maintain the spring-based hibernate
configuration, and a separate one, to be able to use SchemaExport to flush
my database configuration during development.

dyn
-- 
.Digital.Yearning.for.Networked.Assassination.and.Xenocide

Re: [Springframework-developer] AbstractStatelessSessionBean

[<jue...@we...>]

Good point - just changed it. This is also consistent with how =
ContextLoaderListener and FrameworkServlet behave in a web container.
=20
Juergen
=20

________________________________

Von: spr...@li... im Auftrag =
von Luke Taylor
Gesendet: So 29.02.2004 22:01
An: spr...@li...
Betreff: [Springframework-developer] AbstractStatelessSessionBean



I've just been looking at this class after someone had a problem in the
user list and it throws a CreateException if it fails to load the bean
factory correctly. CreateException is supposed to be treated as an
application exception by an EJB container so is unlikely to be logged by
the container (it doesn't seem to be by JBoss at any rate).

Since ejbCreate is called on bean creation, rather than when the client
   calls create() there's a danger the exception could be silently
swallowed when the session bean pool is being populated and the client
won't see it. Even if the server does log it, it might not obviously
correspond to the client error.

I think it might be better to just throw the BeansException directly. It
will be logged by the container and the bean will be destroyed. In fact,
it might be better to encourage this behaviour from onEJBCreate by
removing CreateException from the throws clause altogether. From a quick
look at the spec, it seems that only a runtime exception will cause the
bean to be destroyed, and I can't think of a situation with an SSB where
one would want to throw an exception on creation yet have the bean
remain in the pool to be invoked by a client later. At the very least
the BeansException should probably be logged.

What does everyone think?

I really wonder why I ever thought EJBs might be a good thing...
probably because they seemed simpler than CORBA :).

Luke.



--
  Luke Taylor.                      Monkey Machine Ltd.
  PGP Key ID: 0x57E9523C            http://www.monkeymachine.ltd.uk




-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=3D1356&alloc_id=3D3438&op=3Dclick
_______________________________________________
Springframework-developer mailing list
Spr...@li...
https://lists.sourceforge.net/lists/listinfo/springframework-developer

Re: [Springframework-developer] JavaMailSenderImpl

[<jue...@we...>]

Or simply subclass JavaMailSenderImpl and invoke setSession with a =
custom Session. However, I see the original point: I'll add =
"javaMailProperties".
=20
Will still make it into RC2, BTW, as I'm currently preparing it :-)
=20
Juergen
=20

________________________________

Von: spr...@li... im Auftrag =
von Alef Arendsen
Gesendet: Mo 01.03.2004 07:50
An: spr...@li...
Betreff: RE: [Springframework-developer] JavaMailSenderImpl



Well, you could implement a (beans.factory.)FactoryBean
(http://monkeymachine.co.uk/spring/ref/beans.html#beans-factory-lifecycle=
-fa
ctorybean) here that retrieves your custom session:

public class OCSessionFactory implements FactoryBean {
  public Object getObject() {
    return OCSession.getInstance(new Properties());
  }
  public Class getObjectType() {
    return OCSession.class;
  }
}

And after that use this to set your own session:

<bean id=3D"mailSender"
    class=3D"org.springframework.mail.javamail.JavaMailSenderImpl">
  <property name=3D"session">
    <bean class=3D"ourcommunity.util.email.OCSessionFactory"/>
  </property>
</bean>

Hope this helps.

Alef

> -----Original Message-----
> From: spr...@li...
> [mailto:spr...@li...] On =
Behalf
> Of Oliver Hutchison
> Sent: Monday, March 01, 2004 4:35 AM
> To: spr...@li...
> Subject: RE: [Springframework-developer] JavaMailSenderImpl
>
> How would you go about doing this if you don't use JNDI? Is there a =
way
> in spring to invoke the Session.getInstance method? I'd like to be =
able
> to have all of the configuration for my MailSender in one place.
>
> At the moment I've simply extended JavaMailSenderImpl so now my config
> looks like this:
>
>     <bean id=3D"javaMailSender"
> class=3D"ourcommunity.util.email.OCJavaMailSenderImpl">
>         <property
> name=3D"host"><value>${oc.mail.server.addr}</value></property>
>         <property name=3D"javaMailProperties">
>             <props>
>                 <prop =
key=3D"mail.smtp.dsn.notify">SUCCESS,FAILURE</prop>
>                 <prop key=3D"mail.smtp.dsn.ret">HDRS</prop>
>             </props>
>         </property>
>     </bean>
>
> And OCJavaMailSenderImpl.java looks like this
>
> package ourcommunity.util.email;
>
> import java.util.Properties;
> import javax.mail.Session;
> import org.springframework.mail.javamail.JavaMailSenderImpl;
>
> public class OCJavaMailSenderImpl
> extends JavaMailSenderImpl
> {
>     public void setJavaMailProperties(Properties props)
>     {
>         setSession(Session.getInstance(props));
>     }
> }
>
> But it seems to me this would be very useful for others if it was just
> part of the standard Spring functionality.
>
> -----Original Message-----
> From: Darren Davison [mailto:da...@da...]
> Sent: Monday, 1 March 2004 2:01 PM
> To: spr...@li...
> Subject: Re: [Springframework-developer] JavaMailSenderImpl
>
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On Monday 01 March 2004 02:42, Oliver Hutchison wrote:
> > Hi,
> >
> > Except for extending JavaMailSenderImpl is there a way to specify
> > extra properties that are used to instantiate the Session?
>
> you can create your own javax.mail.Session object (populated with your
> additional properties) and simply pass it to the
> JavaMailSenderImpl.setSession(Session) method.  This is particularly
> handy
> if you obtain your session from JNDI.
>
>
>
>
> -------------------------------------------------------
> SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> Build and deploy apps & Web services for Linux with
> a free DVD software kit from IBM. Click Now!
> http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
> _______________________________________________
> Springframework-developer mailing list
> Spr...@li...
> https://lists.sourceforge.net/lists/listinfo/springframework-developer



-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
_______________________________________________
Springframework-developer mailing list
Spr...@li...
https://lists.sourceforge.net/lists/listinfo/springframework-developer

RE: [Springframework-developer] JavaMailSenderImpl

[Alef A. <al...@jt...>]

Well, you could implement a (beans.factory.)FactoryBean
(http://monkeymachine.co.uk/spring/ref/beans.html#beans-factory-lifecycle=
-fa
ctorybean) here that retrieves your custom session:

public class OCSessionFactory implements FactoryBean {
  public Object getObject() {
    return OCSession.getInstance(new Properties());
  }
  public Class getObjectType() {
    return OCSession.class;
  }
}

And after that use this to set your own session:

<bean id=3D"mailSender"
    class=3D"org.springframework.mail.javamail.JavaMailSenderImpl">
  <property name=3D"session">
    <bean class=3D"ourcommunity.util.email.OCSessionFactory"/>
  </property>
</bean>

Hope this helps.

Alef

> -----Original Message-----
> From: spr...@li...
> [mailto:spr...@li...] On =
Behalf
> Of Oliver Hutchison
> Sent: Monday, March 01, 2004 4:35 AM
> To: spr...@li...
> Subject: RE: [Springframework-developer] JavaMailSenderImpl
>=20
> How would you go about doing this if you don't use JNDI? Is there a =
way
> in spring to invoke the Session.getInstance method? I'd like to be =
able
> to have all of the configuration for my MailSender in one place.
>=20
> At the moment I've simply extended JavaMailSenderImpl so now my config
> looks like this:
>=20
>     <bean id=3D"javaMailSender"
> class=3D"ourcommunity.util.email.OCJavaMailSenderImpl">
>         <property
> name=3D"host"><value>${oc.mail.server.addr}</value></property>
>         <property name=3D"javaMailProperties">
>             <props>
>                 <prop =
key=3D"mail.smtp.dsn.notify">SUCCESS,FAILURE</prop>
>                 <prop key=3D"mail.smtp.dsn.ret">HDRS</prop>
>             </props>
>         </property>
>     </bean>
>=20
> And OCJavaMailSenderImpl.java looks like this
>=20
> package ourcommunity.util.email;
>=20
> import java.util.Properties;
> import javax.mail.Session;
> import org.springframework.mail.javamail.JavaMailSenderImpl;
>=20
> public class OCJavaMailSenderImpl
> extends JavaMailSenderImpl
> {
>     public void setJavaMailProperties(Properties props)
>     {
>         setSession(Session.getInstance(props));
>     }
> }
>=20
> But it seems to me this would be very useful for others if it was just
> part of the standard Spring functionality.
>=20
> -----Original Message-----
> From: Darren Davison [mailto:da...@da...]
> Sent: Monday, 1 March 2004 2:01 PM
> To: spr...@li...
> Subject: Re: [Springframework-developer] JavaMailSenderImpl
>=20
>=20
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>=20
> On Monday 01 March 2004 02:42, Oliver Hutchison wrote:
> > Hi,
> >
> > Except for extending JavaMailSenderImpl is there a way to specify
> > extra properties that are used to instantiate the Session?
>=20
> you can create your own javax.mail.Session object (populated with your
> additional properties) and simply pass it to the
> JavaMailSenderImpl.setSession(Session) method.  This is particularly
> handy
> if you obtain your session from JNDI.
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> Build and deploy apps & Web services for Linux with
> a free DVD software kit from IBM. Click Now!
> http://ads.osdn.com/?ad_id=1356&alloc_id438&op=3Dick
> _______________________________________________
> Springframework-developer mailing list
> Spr...@li...
> https://lists.sourceforge.net/lists/listinfo/springframework-developer

RE: [Springframework-developer] JavaMailSenderImpl

[Oliver H. <Oli...@ou...>]

How would you go about doing this if you don't use JNDI? Is there a way
in spring to invoke the Session.getInstance method? I'd like to be able
to have all of the configuration for my MailSender in one place.

At the moment I've simply extended JavaMailSenderImpl so now my config
looks like this:
  =20
    <bean id=3D"javaMailSender"
class=3D"ourcommunity.util.email.OCJavaMailSenderImpl">
        <property
name=3D"host"><value>${oc.mail.server.addr}</value></property>
        <property name=3D"javaMailProperties">
            <props>
                <prop =
key=3D"mail.smtp.dsn.notify">SUCCESS,FAILURE</prop>
                <prop key=3D"mail.smtp.dsn.ret">HDRS</prop>
            </props>
        </property>
    </bean>=20

And OCJavaMailSenderImpl.java looks like this

package ourcommunity.util.email;

import java.util.Properties;
import javax.mail.Session;
import org.springframework.mail.javamail.JavaMailSenderImpl;

public class OCJavaMailSenderImpl
extends JavaMailSenderImpl
{
    public void setJavaMailProperties(Properties props)
    {
        setSession(Session.getInstance(props));
    }
}

But it seems to me this would be very useful for others if it was just
part of the standard Spring functionality.

-----Original Message-----
From: Darren Davison [mailto:da...@da...]=20
Sent: Monday, 1 March 2004 2:01 PM
To: spr...@li...
Subject: Re: [Springframework-developer] JavaMailSenderImpl


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Monday 01 March 2004 02:42, Oliver Hutchison wrote:
> Hi,
>
> Except for extending JavaMailSenderImpl is there a way to specify=20
> extra properties that are used to instantiate the Session?

you can create your own javax.mail.Session object (populated with your=20
additional properties) and simply pass it to the=20
JavaMailSenderImpl.setSession(Session) method.  This is particularly
handy=20
if you obtain your session from JNDI.

Re: [Springframework-developer] JavaMailSenderImpl

[Darren D. <da...@da...>]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Monday 01 March 2004 02:42, Oliver Hutchison wrote:
> Hi,
>
> Except for extending JavaMailSenderImpl is there a way to specify extra
> properties that are used to instantiate the Session?

you can create your own javax.mail.Session object (populated with your=20
additional properties) and simply pass it to the=20
JavaMailSenderImpl.setSession(Session) method.  This is particularly handy=
=20
if you obtain your session from JNDI.

=2D --=20

Darren Davison
Public Key: http://www.davison.uk.net/key.jsp
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAQqdbKLMLAN01aw0RAgYFAJ4tU1yRQR7voAPHTybnyIORUlM7QwCfcY5Q
iV4fWRINX+oE+kmsfNcjS8o=3D
=3Dl5Ln
=2D----END PGP SIGNATURE-----

[Springframework-developer] JavaMailSenderImpl

[Oliver H. <Oli...@ou...>]

Hi,

Except for extending JavaMailSenderImpl is there a way to specify extra
properties that are used to instantiate the Session?

After a cursory glance at the code this doesn't seem possible but there
always seems to be another option in Spring :-)

If this is not possible from a users point of view this would seem to me
a very handy thing to add. Sun's SMTP implementation has quite a few
options that can only be activated through the Properties passed into
the Session.getInstance. I guess you'd just need to add a
javaMailProperties field.

Thanks,

Ollie

[Springframework-developer] Commons-Validator for Spring

[Daniel M. <mi...@pa...>]

Hello everyone,

I have recently been working on a commons-validator extension/addition to
the Spring framework. My current implementation consists of a
org.springframework.validation.commons package with about six classes. It
provides an implementation of the org.springframework.validation.Validator
interface that would encapsulate the functionality of the commons validator.
A user's project could simply add the necessary commons-validator config
files (validator-rules.xml, validation.xml), jars (commons-validator.jar),
and wire up a few beans and have completely xml configured validation.

By the way, my implementation for Spring will be much handier than the
Struts validator: it will allow easy extension of the validator. A user
could simply wire up a custom validator with my Spring-commons-validator
object. The commons-validator could easily be invoked by the custom
validator. For instance, this provides a logical way for users to implement
complex validation routines in a custom validator and leave the boring stuff
to the commons-validator (required, int, date, etc). To get back to the
point, the only way to extend the validator in Struts is to (1) either code
pluggable validators for the commons validator or (2) add a validate()
method to your ActionForm that does the complex stuff. I don't like the
first approach because it's cumbersome and one must understand the innards
of the commons-validator to do it. The second approach usually ends up
adding unnecessary complexity to a form. Furthermore, if you're using
DynaValidatorForms you have to extend the DynaValidatorForm class to add
this functionality, which (imo) puts validation logic in a not-so-logical
location.

Now this may all be sounding fine and dandy, and you're all saying "OK.
Let's have a look at the code so we can see if we'd like to use it." or even
better "Hey, give us your code...we really need that." But there's a
catch...

Background Info
The commons-validator uses message resources to identify arguments (such as
field names) for validation messages. For example, the base message would be
error.required="{0} is required." and arg0="formName.fieldName.label" (or
any other message resource identifier). Upon validation failure, the
validator resolves this message resource identifier to its corresponding
message in order to add it to the arguments array for the Errors object
(ActionErrors in Struts). The Struts validator adds a request object to the
commons-validator's map of resources, so request attributes may be used in
validation methods. Hence, the validator has a way to access the locale of
the current request as well as message resources to resolve message resource
identifiers. Currently, the Validator interface in Spring only specifies the
command object and the errors object as arguments, so I have no way of
getting these resources.

Ideally, the validator should not have to worry about resolving message
resources--it should only do validation. However, the commons-validator not
only does this, but also provides for locale-specific validation
configurations, so it seems necessary to have such information.

I have come up with two possible solutions for this problem.

(1) Add another method to the Spring Validator interface with something like
the following method signature:

public void validate(javax.servlet.http.HttpServletRequest request,
java.lang.Object obj, Errors errors)

I don't like this because it depends on a web context. One of the things I
was hoping for was the ability to use the commons-validator to validate
business objects in the middle tier, instead of only for the web. Of course,
the new method could simply provide the Locale and MessageResources, which
would possibly solve that problem.

(2) Improve the handling of Errors objects to allow message resources to be
specified as arguments that would not be resolved to their corresponding
messages until reaching the view. I really like this solution, but I have
not thought it through completely. One more note: since the errorArgs
parameter is an array of Objects (not an array of Strings), maybe there
would be a way to use a "Message" object that would be resolved at a later
time, while preserving the current functionality of Strings that are used
directly as arguments. For all I know, there is already a way to do this (it
seems logical enough), and I just don't know about it. This J2EE stuff is so
hard to keep up with...

Please provide response, and I will attempt to accommodate (or at least
consider) any wishes that may be expressed pertaining to additional features
or other adjustments. One area that I know should be supported, but I have
not worked on yet is the WizardController (multiple pages/stages of
validation per bean/form). I believe the commons-validator already supports
this, so it should only be a matter of adding another class or so to my
package to allow such functionality.

Thank you,
Daniel Miller

P.S. Sorry for being so long winded. I'll shut up now until I hear back from
you. :)

[Springframework-developer] AbstractStatelessSessionBean

[Luke T. <ne...@fr...>]

I've just been looking at this class after someone had a problem in the 
user list and it throws a CreateException if it fails to load the bean 
factory correctly. CreateException is supposed to be treated as an 
application exception by an EJB container so is unlikely to be logged by 
the container (it doesn't seem to be by JBoss at any rate).

Since ejbCreate is called on bean creation, rather than when the client 
   calls create() there's a danger the exception could be silently 
swallowed when the session bean pool is being populated and the client 
won't see it. Even if the server does log it, it might not obviously 
correspond to the client error.

I think it might be better to just throw the BeansException directly. It 
will be logged by the container and the bean will be destroyed. In fact, 
it might be better to encourage this behaviour from onEJBCreate by 
removing CreateException from the throws clause altogether. From a quick 
look at the spec, it seems that only a runtime exception will cause the 
bean to be destroyed, and I can't think of a situation with an SSB where 
one would want to throw an exception on creation yet have the bean 
remain in the pool to be invoked by a client later. At the very least 
the BeansException should probably be logged.

What does everyone think?

I really wonder why I ever thought EJBs might be a good thing... 
probably because they seemed simpler than CORBA :).

Luke.



-- 
  Luke Taylor.                      Monkey Machine Ltd.
  PGP Key ID: 0x57E9523C            http://www.monkeymachine.ltd.uk

Re: [Springframework-developer] Please add the license to the web site

[Keith D. <kd...@cs...>]

I think we should use version 2.0 since it makes it more clear the apache
license can be used by other projects not under the control of the ASF,
right?

One thing to consider is updating the source file header template to refl=
ect
any change/clarification in license. Keith

----- Original Message -----=20
From: <tri...@tr...>
To: <spr...@li...>
Sent: Sunday, February 29, 2004 10:29 AM
Subject: Re: [Springframework-developer] Please add the license to the we=
b
site


> A couple of questions that I think we should resolve before 1.0 final.
>
> Should we use the Apache License, Version 2.0 instead - any Apache
projects
> released after March 1st is going to use the 2.0 version.  Also, should=
n't
we
> have a Contributor License Agreement for each developer if we are going=
 to
> assign the copyrights to the Spring Framework Project?
>
> Thomas
>
>
> Quoting "j=FCrgen h=F6ller [werk3AT]" <jue...@we...>:
>
> > Everyone,
> > =3D20
> > I've recently adapted the text of our license a bit for 1.0 RC2.
> > =3D20
> > It clarifies that it applies to all files contained in the download n=
ow,
=3D
> > except the third-party libraries in the "lib" directory. I've also ad=
ded
=3D
> > Rod and myself as representatives of the "Spring Framework Project"; =
=3D
> > I've read somewhere that a copyright should in some way refer to lega=
l =3D
> > entities rather than just a legally undefined "project".
> > =3D20
> > If there are any objections or suggestions regarding the wording, ple=
ase
=3D
> > raise your hand!
> > =3D20
> > Juergen
> > =3D20
> >
> > ________________________________
> >
> > Von: spr...@li... im Auftrag=
 =3D
> > von Cameron Braid
> > Gesendet: So 29.02.2004 14:04
> > An: spr...@li...
> > Betreff: [Springframework-developer] Please add the license to the we=
b =3D
> > site
> >
> >
> > I know that you are using an apache license, though there is no menti=
on
=3D
> > if this readily availible on the web site.
> > =3D20
> > Thanks.
> > =3D20
> > Cameron
> >
> >
> > -------------------------------------------------------
> > SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> > Build and deploy apps & Web services for Linux with
> > a free DVD software kit from IBM. Click Now!
> > http://ads.osdn.com/?ad_id=3D1356&alloc_id=3D3438&op=3Dclick
> > _______________________________________________
> > Springframework-developer mailing list
> > Spr...@li...
> > https://lists.sourceforge.net/lists/listinfo/springframework-develope=
r
> >
>
>
>
>
>
> -------------------------------------------------------
> SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> Build and deploy apps & Web services for Linux with
> a free DVD software kit from IBM. Click Now!
> http://ads.osdn.com/?ad_id=3D1356&alloc_id=3D3438&op=3Dclick
> _______________________________________________
> Springframework-developer mailing list
> Spr...@li...
> https://lists.sourceforge.net/lists/listinfo/springframework-developer