Hi,
committed in rev 1503763 on STANBOL-1097
cheers,
Enrico
On 16 July 2013 12:53, Reto Bachmann-Gmür <r...@wymiwyg.com> wrote:
> Hi enrico
> >
> > BTW, if I create a new class, say JobResultData extending ResultData and
> > explicitly declaring the method, then it works.
>
> So a bit more code is required but I think this is the best approach.
> Can you commit your change?
> >
> > FYI the key method should be
> > freemarker.ext.beans.BeansWrapper.discoverAccessibleMethods (if I
> followed
> > correctly the method calls...).
>
> I got there too ;) It seems it would be easy to fix
> (discoverAccessibleMethods could get a bit shorter). The question is
> if's worth the effort of trying to get this into freemarker. On the
> other hand just having our own ObjectWrapper isn't so easy because as
> all the relevant methods are private so we cannot just provide a
> subclass.
>
> >
> > So the model should be a Java Bean, and I don't think that anonymous
> > classes are it.
> >
> > So we could change the object wrapper to support our inline methods, or
> ask
> > the freemarker ml howto.
> >
> > But at this point I am not sure that it is a recommended way of providing
> > the pojo to fm.
>
> Well the recommended way in Stanbol is to use RdfViewable. This also
> ensures that the resource can be dereferenced as RDF as well. And for
> porting the existing code we have to use public value objects (as with
> JobResultData).
>
> Cheers,
> Reto
>
> >
> > Cheers,
> > Enrico
> >
> >
> > On 16 July 2013 11:04, Reto Bachmann-Gmür <r...@wymiwyg.com> wrote:
> >
> >> Hi Enrico
> >>
> >> Thanks for analyzing this further. It' really weird that freemarker
> >> doesn't get all the methods of the anonymous inner class. I guess an
> >> optimization attempt going wrong. Debugging freemarker code is not a
> >> nice thing to do ...
> >>
> >> Cheers,
> >> Reto
> >>
> >> On Tue, Jul 16, 2013 at 11:44 AM, Enrico Daga <enricod...@gmail.com>
> >> wrote:
> >> > (some) more insight.
> >> >
> >> > Another thing I tried is the following in the fm template:
> >> >
> >> > ${it.class.getMethod("getJobInfo").getName()}
> >> > ${it.jobInfo!"NULL"}
> >> > ${it.getJobInfo()}
> >> >
> >> > While the first line works, displaying the method name, the second
> >> displays
> >> > NULL and the third throws the exception. Another note is that the
> method
> >> > getJobInfo() is never called by the template, while if I override a
> >> method
> >> > defined in a supertype (for example getUriInfo) it is executed
> regularly.
> >> > So I guess there is some problem in the way the template engine
> evaluate
> >> > the expression looking for the actual method of an inline instance.
> >> >
> >> > I have setup a sling logger for freemarker and found that the method
> >> > getJobInfo is not included in the introspection info for the actual
> >> object
> >> > (big log [1]).
> >> >
> >> > So this pattern won't work at least with the actual configuration of
> the
> >> > template engine.
> >> >
> >> > Enrico
> >> >
> >> >
> >> >
> >> > [1] Follows the log:
> >> > -------------
> >> >
> >> > 16.07.2013 10:32:40.573 *DEBUG* [1106793701@qtp-2017787826-51]
> >> > freemarker.beans Key "jobInfo" was not found on instance of
> >> > org.apache.stanbol.commons.jobs.web.resources.JobsResource$2.
> >> Introspection
> >> > information for the class is: {getRootUrl=public java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRootUrl(),
> >> > getRequestUri=public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRequestUri(),
> >> > mainMenuItems=java.beans.PropertyDescriptor@6398d702, hashCode=public
> >> > native int java.lang.Object.hashCode(),
> >> > registeredLinkResources=java.beans.PropertyDescriptor@bf736045,
> >> > java.lang.Object@11496fc2={public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getPublicBaseUri()=[Ljava.lang.Class;@2c69c4e9
> >> ,
> >> > public boolean
> >> > java.lang.Object.equals(java.lang.Object)=[Ljava.lang.Class;@53feb3c6
> ,
> >> > public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getNavigationLinks()=[Ljava.lang.Class;@76ccb73e
> >> ,
> >> > public java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getStaticResourcesRootUrl()=[Ljava.lang.Class;@7a128460
> >> ,
> >> > public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRegisteredLinkResources()=[Ljava.lang.Class;@55b0f93c
> >> ,
> >> > public final native java.lang.Class
> >> > java.lang.Object.getClass()=[Ljava.lang.Class;@55bafe17, public
> >> > java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getStaticRootUrl()=[Ljava.lang.Class;@677fc3f9
> >> ,
> >> > public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getConsoleBaseUri()=[Ljava.lang.Class;@751b4584
> >> ,
> >> > public native int
> java.lang.Object.hashCode()=[Ljava.lang.Class;@5171192c
> >> ,
> >> > public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRegisteredScriptResources()=[Ljava.lang.Class;@582bc179
> >> ,
> >> > public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getMainMenuItems()=[Ljava.lang.Class;@20ab1c28
> >> ,
> >> > public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRequestUri()=[Ljava.lang.Class;@41c596d9
> >> ,
> >> > public java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRootUrl()=[Ljava.lang.Class;@1332a24b
> >> ,
> >> > public java.lang.String
> >> > java.lang.Object.toString()=[Ljava.lang.Class;@613cf3c5},
> >> > requestUri=java.beans.PropertyDescriptor@a9193b22,
> >> > staticRootUrl=java.beans.PropertyDescriptor@ac5c5fb,
> >> > getNavigationLinks=public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getNavigationLinks(),
> >> > getPublicBaseUri=public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getPublicBaseUri(),
> >> > getRegisteredScriptResources=public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRegisteredScriptResources(),
> >> > registeredScriptResources=java.beans.PropertyDescriptor@1b975b12,
> >> > getClass=public final native java.lang.Class
> java.lang.Object.getClass(),
> >> > getMainMenuItems=public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getMainMenuItems(),
> >> > equals=public boolean java.lang.Object.equals(java.lang.Object),
> >> > class=java.beans.PropertyDescriptor@88f57024,
> >> > staticResourcesRootUrl=java.beans.PropertyDescriptor@dd57bb70,
> >> > consoleBaseUri=java.beans.PropertyDescriptor@5cd93387,
> >> > getConsoleBaseUri=public java.net.URI
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getConsoleBaseUri(),
> >> > getStaticRootUrl=public java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getStaticRootUrl(),
> >> > getStaticResourcesRootUrl=public java.lang.String
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getStaticResourcesRootUrl(),
> >> > rootUrl=java.beans.PropertyDescriptor@575a742d,
> >> > getRegisteredLinkResources=public java.util.List
> >> >
> >>
> org.apache.stanbol.commons.web.base.resource.TemplateLayoutConfiguration.getRegisteredLinkResources(),
> >> > publicBaseUri=java.beans.PropertyDescriptor@14a41d71,
> >> > navigationLinks=java.beans.PropertyDescriptor@9297e601,
> toString=public
> >> > java.lang.String java.lang.Object.toString()}
> >> >
> >> >
> >> >
> >> > On 14 July 2013 23:20, Enrico Daga <enricod...@gmail.com> wrote:
> >> >
> >> >> Hi Reto,
> >> >> Yes 1.2 is around and works well.
> >> >>
> >> >> I had a look at JobsResource and I found that freemarker variables do
> >> not
> >> >> work at all. I can't say why at the moment, but replacing "this"
> with an
> >> >> inline subclass of ResultData deals to an InvalidReferenceException
> for
> >> >> anything I try to reference from it even if the "it" var is not null
> >> and is
> >> >> the actual pojo.
> >> >> Invoking /jobs/test starts a fake background job, you can easily see
> the
> >> >> problem by following the url to ping the status (/jobs/{jobId}).
> >> >> Needs more insight.
> >> >>
> >> >>
> >> >> On 14 July 2013 13:05, Reto Bachmann-Gmür <r...@wymiwyg.com> wrote:
> >> >>
> >> >>> Hi Enrico
> >> >>>
> >> >>>
> >> >>> Had quite some problems with this things. On hand OSGi stuff:
> >> >>>
> >> >>>
> >> >>>
> >>
> http://mail-archives.apache.org/mod_mbox/felix-users/201306.mbox/%3CCALvhUEWRe+XKW=2M0e7DwNC8bm=mrakpqng5uq+ax5mocws...@mail.gmail.com%3E
> >> >>>
> >> >>> On the other hands jersey dependency on stuff that wasn't yet in
> >> central.
> >> >>>
> >> >>> If the final version's out and things are working, great!
> >> >>>
> >> >>> Reto
> >> >>>
> >> >>>
> >> >>> On Sun, Jul 14, 2013 at 1:13 PM, Enrico Daga <enricod...@gmail.com>
> >> >>> wrote:
> >> >>> > Hi Reto,
> >> >>> > I am trying to run the mini launcher as test but jersey did not
> start
> >> >>> > because javax.annotation was 1.1.9-b02 (saw on felix console). The
> >> weird
> >> >>> > thing is that the deployed bundle was 1.2-b04 (as configured in
> >> >>> bundlelist
> >> >>> > stanbolcommons and present in .../startup/1/). I changed the
> bundle
> >> >>> version
> >> >>> > to 1.2 and now it starts. Don't know why 1.2-b04 was seen as
> >> 1.1.9-b02,
> >> >>> but
> >> >>> > is there a special reason why we would need 1.2-b04?
> >> >>> >
> >> >>> > Enrico
> >> >>> >
> >> >>> >
> >> >>> > On 12 July 2013 19:15, Enrico Daga <enricod...@gmail.com> wrote:
> >> >>> >
> >> >>> >> Hi Reto, yes, now works. Thank you.
> >> >>> >>
> >> >>> >>
> >> >>> >> On 12 July 2013 13:27, Reto Bachmann-Gmür <r...@wymiwyg.com>
> wrote:
> >> >>> >>
> >> >>> >>> Hi Enrico
> >> >>> >>>
> >> >>> >>> Thanks a lot for looking into this.
> >> >>> >>>
> >> >>> >>> The aim of the ng branch is to be more portable. Most notably
> this
> >> is
> >> >>> >>> independence of jersey and of the ServletContext hacks.
> >> >>> >>>
> >> >>> >>> I'm sorry I removed the jersey specific test only locally. I
> >> committed
> >> >>> >>> the removal and hope that it now compiles for you too.
> >> >>> >>>
> >> >>> >>> Cheers,
> >> >>> >>> Reto
> >> >>> >>>
> >> >>> >>> On Fri, Jul 12, 2013 at 2:13 PM, Enrico Daga <
> enricod...@gmail.com
> >> >
> >> >>> >>> wrote:
> >> >>> >>> > Hi Reto,
> >> >>> >>> > sorry for this late reply, for some reasons I skipped this
> thread
> >> >>> >>> > entirely...
> >> >>> >>> > For what concerns the Commons Jobs API it is only used by
> >> >>> reasoners. It
> >> >>> >>> is
> >> >>> >>> > a bit explained in a README file even if probably a
> documentation
> >> >>> >>> targeted
> >> >>> >>> > to developers would help. I could add it to the web site
> >> >>> documentation.
> >> >>> >>> > About the branch commons-ng, if I understood the aim is to
> clear
> >> the
> >> >>> >>> JAX-RS
> >> >>> >>> > implementation in order to be jersey independent, right?
> >> >>> >>> > I downloded the branch to have a look to your examples but had
> >> the
> >> >>> >>> > following:
> >> >>> >>> >
> >> >>> >>> > [ERROR] Failed to execute goal
> >> >>> >>> >
> org.apache.maven.plugins:maven-compiler-plugin:2.3.1:testCompile
> >> >>> >>> > (default-testCompile) on project
> >> >>> org.apache.stanbol.entityhub.jersey:
> >> >>> >>> > Compilation failure: Compilation failure:
> >> >>> >>> > [ERROR]
> >> >>> >>> >
> >> >>> >>>
> >> >>>
> >>
> /Users/enricox/Development/Stanbol-commons-ng/entityhub/jersey/src/test/java/org/apache/stanbol/entityhub/jersey/parsers/RepresetnationReaderTest.java:[27,40]
> >> >>> >>> > package com.sun.jersey.api.representation does not exist
> >> >>> >>> >
> >> >>> >>> > My idea was to have a look, learn howto and migrate the
> >> reasoners.
> >> >>> I'll
> >> >>> >>> add
> >> >>> >>> > the documentation of Jobs, if you think it is useful (I like
> that
> >> >>> small
> >> >>> >>> > thing btw, I would like to see it reused)
> >> >>> >>> >
> >> >>> >>> > Cheers,
> >> >>> >>> > Enrico
> >> >>> >>> >
> >> >>> >>> >
> >> >>> >>> >
> >> >>> >>> > On 10 July 2013 15:11, Rupert Westenthaler <
> >> >>> >>> rupert.westentha...@gmail.com>wrote:
> >> >>> >>> >
> >> >>> >>> >> Hi Reto
> >> >>> >>> >>
> >> >>> >>> >> Generally I find this API very interesting and I was
> considering
> >> >>> this
> >> >>> >>> >> as useful for the implementation of the Stanbol Enhancer
> "/task"
> >> >>> >>> >> endpoint as suggested by David in [1]. For the current
> services
> >> >>> >>> >> provided by the Enhancer and Entityhub this is not
> applicable as
> >> >>> there
> >> >>> >>> >> are no async services where users would like to interact with
> >> >>> running
> >> >>> >>> >> jobs.
> >> >>> >>> >>
> >> >>> >>> >> best
> >> >>> >>> >> Rupert
> >> >>> >>> >>
> >> >>> >>> >> [1] http://stanbol.markmail.org/thread/npwf7olw2jemvu37
> >> >>> >>> >>
> >> >>> >>> >> On Wed, Jul 10, 2013 at 3:56 PM, Reto Bachmann-Gmür <
> >> >>> r...@apache.org>
> >> >>> >>> >> wrote:
> >> >>> >>> >> > After not getting any answer I used grep myself ;)
> >> >>> >>> >> >
> >> >>> >>> >> > It seems commons.job is currently only used in the
> >> reasoner.web.
> >> >>> >>> >> >
> >> >>> >>> >> > Are there plans to use this elsewhere? Or to add some
> >> >>> documentation
> >> >>> >>> >> > for it to be more likely to be adopted?
> >> >>> >>> >> >
> >> >>> >>> >> > Cheers,
> >> >>> >>> >> > Reto
> >> >>> >>> >> >
> >> >>> >>> >> > On Wed, Jun 26, 2013 at 10:14 PM, Reto Bachmann-Gmür <
> >> >>> >>> r...@apache.org>
> >> >>> >>> >> wrote:
> >> >>> >>> >> >> Did anybody have a chance to try this out? Where is
> Commons
> >> Job
> >> >>> >>> >> documented?
> >> >>> >>> >> >>
> >> >>> >>> >> >> Cheers,
> >> >>> >>> >> >> Reto
> >> >>> >>> >> >>
> >> >>> >>> >> >>
> >> >>> >>> >> >> On Fri, Jun 7, 2013 at 11:56 PM, Reto Bachmann-Gmür <
> >> >>> >>> r...@apache.org>
> >> >>> >>> >> wrote:
> >> >>> >>> >> >>>
> >> >>> >>> >> >>> At least I think it runs again.I'm not sure how to test
> it.
> >> >>> Would
> >> >>> >>> be
> >> >>> >>> >> great
> >> >>> >>> >> >>> if somebody else could check it out.
> >> >>> >>> >> >>>
> >> >>> >>> >> >>> Cheers,
> >> >>> >>> >> >>> Reto
> >> >>> >>> >> >>
> >> >>> >>> >> >>
> >> >>> >>> >>
> >> >>> >>> >>
> >> >>> >>> >>
> >> >>> >>> >> --
> >> >>> >>> >> | Rupert Westenthaler
> rupert.westentha...@gmail.com
> >> >>> >>> >> | Bodenlehenstraße 11
> >> ++43-699-11108907
> >> >>> >>> >> | A-5500 Bischofshofen
> >> >>> >>> >>
> >> >>> >>> >
> >> >>> >>> >
> >> >>> >>> >
> >> >>> >>> > --
> >> >>> >>> > Enrico Daga
> >> >>> >>> >
> >> >>> >>> > --
> >> >>> >>> > http://www.enridaga.net
> >> >>> >>> > skype: enri-pan
> >> >>> >>>
> >> >>> >>
> >> >>> >>
> >> >>> >>
> >> >>> >> --
> >> >>> >> Enrico Daga
> >> >>> >>
> >> >>> >> --
> >> >>> >> http://www.enridaga.net
> >> >>> >> skype: enri-pan
> >> >>> >>
> >> >>> >>
> >> >>> >>
> >> >>> >
> >> >>> >
> >> >>> > --
> >> >>> > Enrico Daga
> >> >>> >
> >> >>> > --
> >> >>> > http://www.enridaga.net
> >> >>> > skype: enri-pan
> >> >>>
> >> >>
> >> >>
> >> >>
> >> >> --
> >> >> Enrico Daga
> >> >>
> >> >> --
> >> >> http://www.enridaga.net
> >> >> skype: enri-pan
> >> >>
> >> >>
> >> >>
> >> >
> >> >
> >> > --
> >> > Enrico Daga
> >> >
> >> > --
> >> > http://www.enridaga.net
> >> > skype: enri-pan
> >>
> >
> >
> >
> > --
> > Enrico Daga
> >
> > --
> > http://www.enridaga.net
> > skype: enri-pan
>
--
Enrico Daga
--
http://www.enridaga.net
skype: enri-pan
