Hi Tetsuo, I stand corrected:For #getObjectClass() the model had to be bound to something revealing its type.
We cannot hold a reference to the type because this could lead to all sorts of class loading problems (see Wicket's IClassResolver). Furthermore it would unnecessarily increase the size of LazyModel in most cases (if everything is properly typed).
I improved LazyModel to behave like PropertyModel, i.e. return the result object's type as fallback or just return null if everything fails.
Sven On 06/14/2013 02:24 AM, tetsuo wrote:
I made the changes to make it work, added a test case, and sent a pull request on github. Now it works when you create an unbound model from a class, and then bind it to a plain model. On Thu, Jun 13, 2013 at 6:39 PM, tetsuo <ronald.tet...@gmail.com> wrote:Well, you were the one who said that if I created the unbound model (from(A.class)) then I could bind it to a plain IModel, without the extra type information :) I just assumed that it would keep the 'A.class' information passed in the beginning and use it at runtime. It seems that the Evaluation holds the initial 'from' type, but it isn't passed down to LazyModel when it is created (it only receives target and stack), so it has to try to discover the type by itself, and fails. I'll try to do some experiments here... On Thu, Jun 13, 2013 at 6:21 PM, Sven Meier <s...@meiers.net> wrote:Well, here again LazyModel needs the type of the bound target at runtime. Without any runtime information about the model's object type, LazyModel cannot derive the type of the evaluation result. Sven On 06/13/2013 10:55 PM, tetsuo wrote:this test also passes here, but assertEquals(B.class, ((IObjectClassAwareModel<B>) model.bind(a)).getObjectClass(**)); assertEquals(B.class, ((IObjectClassAwareModel<B>) model.bind(Model.of(a))).**getObjectClass()); While the first assert passes, but the second one doesn't. The exception is thrown not when getting the object, but the object class (the AbstractTextComponent class calls this before rendering and while converting input). On Thu, Jun 13, 2013 at 5:40 PM, Sven Meier <s...@meiers.net> wrote: Strange, works fine here:@Test public void bindToModelAndGet() { LazyModel<B> model = model(from(A.class).getB()); final A a = new A(); a.b = new B(); assertEquals(a.b, model.bind(Model.of(a)).****getObject()); } Sven On 06/13/2013 10:23 PM, tetsuo wrote: Thanks for the response!If I use an object instance, it works, but if I do as your third example (create a model from the class, then bind to a non-IObjectClassAwareModel-****model), it doesn't: public class TestPage extends WebPage { private String text; public TestPage(final PageParameters parameters) { super(parameters); add(new Form<Void>("form") .add(new TextField<String>("text", model(from(TestPage.class ).getText()).bind(Model.of(****this))))); } public String getText() { return text; } public void setText(String text) { this.text = text; } } If I change 'Model.of(this)' to 'this' or an IObjectClassAwareModel implementation, it works: class TestPageModel extends AbstractReadOnlyModel<****TestPage> implements IObjectClassAwareModel<****TestPage> { public TestPage getObject() { return TestPage.this; } public Class<TestPage> getObjectClass() { return TestPage.class; } } Is this a bug? I could create some wrapper models to make it work, but if this is a bug, I'd prefer to wait for a corrected version. Thanks again On Thu, Jun 13, 2013 at 4:52 PM, Sven Meier <s...@meiers.net> wrote: Hi,LazyModel needs to know the type of the model object to return an appropriate proxy: model(from(a).getB()); // works model(from(aModel).getB()); // aModel must be an IObjectClassAwareModel model(from(A.class).getB()).******bind(aModel); // works even if aModel does not reveal its object class Sven On 06/13/2013 09:35 PM, tetsuo wrote: wait, wait, do you actually do something likenew TextField<String>("name", new IModel<String>(){ public String getObject() { return entity.getName(); } public void setObject(String value) { entity.setName(value); } public void detach(){} }); for every single field in your system, or you use LazyModel? Well, I've been trying to use LazyModel, but with it I have to pass the type of every field (the last arg of TextField's constructor) because if I don't, this exception is thrown: Caused by: java.lang.******IllegalArgumentException: T is not a class or parameterizedType at org.wicketstuff.lazymodel.******reflect.Generics.getClass(** Generics.java:78) at org.wicketstuff.lazymodel.******LazyModel.getObjectType(** LazyModel.java:139) at org.wicketstuff.lazymodel.******LazyModel.getObjectClass(** LazyModel.java:124) at org.apache.wicket.markup.html.******form.** AbstractTextComponent.**** getModelType( AbstractTextComponent.java:******167) at org.apache.wicket.markup.html.******form.** AbstractTextComponent.**** resolveType( AbstractTextComponent.java:******152) at org.apache.wicket.markup.html.******form.** AbstractTextComponent.**** onBeforeRender( AbstractTextComponent.java:******142) Any thoughts? I guess I'll just go back to CompoundPropertyModel... (sigh) (and no, I don't spend that much time debugging property models. I usually don't rename properties that often, and when I have to do some refactoring, usually the structure changes, and I have to revise all pages anyway...) On Fri, May 31, 2013 at 10:49 AM, Martin Grigorov < mgrigo...@apache.org wrote:On Fri, May 31, 2013 at 4:24 PM, tetsuo <ronald.tet...@gmail.com> wrote: Black magic, or code generation? hard choice... :)I think I'll try the black magic, let's see how it goes :)I personally prefer writing the boilerplate of custom Model per field. It is a boring work but it gives me:* type safety * the fastest read/write access possible * easier debugging (who knows - maybe I've spent less time to write such models than you've spent to debug your issues with property model after refactoring) On Thu, May 30, 2013 at 8:16 PM, Igor Vaynberg <igor.vaynb...@gmail.com wrote:On Thu, May 30, 2013 at 12:37 PM, tetsuo <ronald.tet...@gmail.com wrote:-1000!This will be horrible! Even with the current API, most generics Ihavetodeclare in my code don't add anything to type safety. For example:while i am also not a fan of having component generified i do believe the example below is a bit contrived. first, i hope most people do not use PropertyModels because they are not compile-time-safe. there are plenty of project that implement compile-time-safe models, personally i prefer https://github.com/42Lines/******metagen<https://github.com/42Lines/****metagen> <https://github.com/**42Lines/**metagen<https://github.com/42Lines/**metagen> <https://github.com/**42Lines/**metagen<https://github.com/**42Lines/metagen> <https://github.com/**42Lines/metagen<https://github.com/42Lines/metagen>tousing proxy-based solutions. further, i hope even less people use compound property models. they are even more unsafe then property models and make your code even more fragile. i would hate to refactor code that uses CPMs. add(new Form<Person>("form", new CompoundPropertyModel<Person>( **** newPropertyModel<Person>(this, "person"))) .add(new TextField<String>("name")) .add(new TextField<Integer>("age")) .add(new TextField<Double>("salary")) .add(new Button("save", new PropertyModel<Person>(this,"******person")){public void onSubmit() { repository.save((Person)******getForm().****getDefaultModelObject()); }});In my experience, this kind of code is fairly common in Wicket applications. Every form component must be declared with a type, but nonehas *any* kind of type safety gain. but how often do you declare a form component without adding any validators to it? the generic type of component also makes sure you add the correct validator. for example it wont let you add a validator that expects strings to a component that produces integers. also, not sure why you are replicating the model in Button. first, the Button uses its model to fill its label; secondly, in real code the model would be in a final var or field that things like onsubmit can access easily. -igor - The property model uses reflection, so its type can't be verified by thecompiler (this.person could be anything, not just a Person).- Generics will guarantee that the form model will be of type Person, butsince it's all declared inline, and the real model isn't verifiable, it just adds lots of verbosity without any real gain.- Most form components use the implicit model, that also usesreflection,and also can't verify the actual type of the underlying property, at compilation time. Even in runtime, *the type information is lostdue erasure*, so it can't use it to do any additional verification.*- Worse, you can even declare the "name" TextField as <Integer> or <Double> (while maintaining the 'text' attribute as String), and sincethere is no type information at runtime, it doesn't matter. Itwon't eventhrow an exception (it will just work normally).* In this case, the type declaration is simply a lie. Just pain, no gain. In my code, I sometimes just add a@SuppressWarnings("rawtypes") to the class, and remove all useless generic type declarations. If everything will be required to declare them, I will have do itmorefrequently.That said, repeater components benefit greatly from generics. Sodocustommodels, validators, and converters. Or the rare cases that weexplicitlydeclare the form component model. But forcing everything to be generic-typed will just make Wicket extremely verbose to use,with verylittle benefit.On Thu, May 30, 2013 at 4:00 AM, Martin Grigorov < mgrigo...@apache.orgwrote:Hi,I just pushed some initial work for [1] and [2] inbranch generified-component-4930. So far it doesn't look nice. The added generics break somehow setMetaData/getMetaData methods - youcansee compilation errors in Component and Page classes. I think it iscaused by the anonymous instance of MetaDataKey ( new MetaDataKey<T>(type) {} ).Also the visit*** methods do not compile at the moment, but even ifwe finda way to fix their signature I think writing a visitor will becomequite cumbersome.At the moment we have IVisitorand org.apache.wicket.util.******iterator.**** AbstractHierarchyIterator which dothesame job. The Iterator API is supposed to be simpler to write forthe users. Maybe we can drop IVisitor ... ?!I'd like to ask for help with this task. It is supposed to be thebiggest API break for Wicket 7.0. My current feeling is that the end result won't be very pleasant for the user-land code. For example the application code will have to do something like:WebMarkupContainer<Void> wmc = new WebMarkupContainer<>("id") It is not that much but we have to decide whether we want it. But first let's try to fix the compilation problems. 1. https://issues.apache.org/******jira/browse/WICKET-4930<https://issues.apache.org/****jira/browse/WICKET-4930> <https:**//issues.apache.org/**jira/**browse/WICKET-4930<https://issues.apache.org/**jira/browse/WICKET-4930> <https:**//issues.apache.org/**jira/**browse/WICKET-4930<http://issues.apache.org/jira/**browse/WICKET-4930> <http**s://issues.apache.org/jira/**browse/WICKET-4930<https://issues.apache.org/jira/browse/WICKET-4930>(Addgenerics too.a.w.Component)2.https://cwiki.apache.org/******confluence/display/WICKET/**<https://cwiki.apache.org/****confluence/display/WICKET/**> <h**ttps://cwiki.apache.org/****confluence/display/WICKET/**<https://cwiki.apache.org/**confluence/display/WICKET/**> Wicket+7.0+Roadmap#Wicket7.******0Roadmap-Genericsfororg.**apache.wicket.Component<https:****//cwiki.apache.org/**confluence/** <http://cwiki.apache.org/confluence/**> display/WICKET/Wicket+7.0+****Roadmap#Wicket7.0Roadmap-** Genericsfororg.apache.wicket.****Component<https://cwiki.** apache.org/confluence/display/**WICKET/Wicket+7.0+Roadmap#** Wicket7.0Roadmap-**Genericsfororg.apache.wicket.**Component<https://cwiki.apache.org/confluence/display/WICKET/Wicket+7.0+Roadmap#Wicket7.0Roadmap-Genericsfororg.apache.wicket.Component>
