So, I am working on a class called DMFWriteExportData and trying to get it run in Batch.
I am at a point where I need to figure out a way to get rid of fieldControl and the reason being it does not let me Run the class on the server and throws an error because it is not supposed to be running on server? (not sure)
Error: "The method Dialog Control.control cannot be called from the server; use methods on the Dialog Field class instead."
-
public Object dialog()
{
DialogRunbase dialog = new DialogRunbase("#DMF372", this);
FormStringControl control;
dialogExecution = dialog.addFieldValue(extendedTypeStr(dMFExecutionId), executionId);
control = dialogExecution.fieldControl();
control.mandatory(true);
control.displayLength(24);
control.registerOverrideMethod(methodstr(FormStringControl, lookup), methodstr(DMFWriteExecutionParameters, executionIdLookup), this);
control.registerOverrideMethod(methodstr(FormStringControl, modified), methodstr(DMFWriteExecutionParameters, executionIdModified), this);
dialogdescription=dialog.addFieldValue(extendedTypeStr(description),DMFExecution::find(executionId).Description);
dialogdescription.enabled(false);
return dialog;
}
I am wondering:
If it is actually true that this class cannot be set to server
when using control.registerOverrideMethod
If yes, what would be the ideal solution to overcome this situation,
is there any way I can create custom lookups? I see there is method
called registerOverrideMethod in the DialogField class.
Any help would be appreciated.
Thanks,
Khosla
The reason why you cannot (and should) run the code above in batch is because it uses dialog controls that only exist on the client side. You should never run this kind of code on server. Please check runon property of your class and set it to called from.
However, I assume you are using RunBaseBatch. If you are on AX 2012, you should use the SysOperation framework instead.
When using RunBaseBatch, all code is on the same class. This way, you are mixing client side code (main method, dialog method etc) with the code that should run on server (run method). For this reason you should set the "runon" property of the class to CalledFrom, not Server.
You can solve this by using SysOperation which applies the Model View Controller (MVC) pattern that neatly sepperates the two.
For an introduction to SysOperation, check my blog here:
AX2012: SysOperation introduction
Related
I have a servicereference with a method I need to use in a test.
The servicereference class is defined as:
public class MyServiceReference : Clientbase<IMyServiceReference>, IMyServiceReference
{
public MyServiceReference()
{
}
..... methods is then defined
}
From my testmethod I have tried both
private MyServiceReference myServiceReferenceFake = A.Fake<MyServiceReference>();
// And
private MyServiceReference myServiceReference = new MyServiceReference();
For both of these is crashes in the constructor with the message:
System.InvalidOperationException: Could not find default endpoint element that references contract.
All I need is to have a callto definition from a method in that class.
How can this be solved?
I've no experience with Clientbase, which I assume to be a System.ServiceModel.ClientBase<TChannel>,but I can make some general comments.
Since you tried first to fake a MyServiceReference, I'll assume that you're not testing that class, and you want to use it as a collaborator for the system under test. In that case, your best bet is to try faking IMyServiceReference. interfaces are very easy to fake, since they don't bring along any behaviour or baggage like faking a class does.
If you feel you really need to fake a MyServiceReference, then we have to contend with the fact that FakeItEasy will eventually call MyServiceReference(), which will call ClientBase<IMyServiceReference>(), whose documentation says
Initializes a new instance of the ClientBase<TChannel> class using the default target endpoint from the application configuration file.
Based on the error you reported, I assume that the application configuration file is not found or does not include the configuration required to create a MyServiceReference. The fact that you get the same error when you just try to instantiate a MyServiceReference directly strengthens my belief.
So I think your paths forward are either to try faking IMyServiceReference or to provide the configuration that ClientBase<IMyServiceReference> needs.
I have been playing with Revit API through pyRevit for quite a while (so, using Python) and I now also need to use C# for some parts of my code.
I am trying to access the VersionNumber property, provided by the Application class. Doing that using pyRevit was quite straightforward, as all I did was
app = __revit__.Application
That's it. I can do it anywhere, it works fine.
Now, I look at C# examples illustrating this (I'm a total C# beginner, by the way), and all examples seem to indicate the same way of proceeding:
UIApplication uiApp = commandData.Application;
Application app = uiApp.Application;
with commandData being an ExternalCommandData object passed as an argument in the Execute method of the class.
My question is: how can I access commandData if I'm not in the Execute method? i.e. if commandData isn't passed as an argument. Can I just instantiate an Application object, just as I did with Python? How?
Thanks a lot,
Arnaud.
Whether you are in a context of Revit's ExternalCommand or not you need access to the UIApplication. You are correct that ExternalCommand provides it via commandData. You can also obtain it from an idling event like so:
private static void OnIdling(object sender, IdlingEventArgs e)
{
var app = (UIApplication)sender;
}
It is also accessible from an ExternalEventHandler `Execute method like so:
public void Execute(UIApplication app)
{
}
How to clean SSJS (Server Side Javascript) in Domino server after someone used javascript prototype in a nsf?
Mark Roden discovered a huge weakness in XPages SSJS: (thanks to David Leedy for tell me about this and show me the article).
If you have the following SSJS code:
var dummyObj = {}
dummyObj.prototype.NAME = "Johann"
XPages SSJS doesn't care that you uses var (var means the variable must be local) and it makes dummyObj.NAME visible in the whole server with the value Johann. So if another nsf in the same server uses a var with the same name it inherits the whole prototype:
var dummyObj = {}
println(dummyObj.NAME) /*prints "Johann" */
This is a huge bug (one that makes unreliable XPages SSJS IMO). Even if you don't use prototype at all, if someone else in his application do something like this:
String.prototype.split = function(){ return "I broke this method" }
It will broke all applications in the same server that uses the innocent split().
So, the question is: if someone "by mistake" writes the following SSJS (XPages Server Side Javascript) in a NSF:
String.prototype.split = function(){ return "I broke this method" }
How can I fix String.prototype.split() to his original value?
As Mark Roden said, restarting HTTP task doesn't fix it.
///////////////////////////////////////////////////////////
Edit 1: Why I think this is a huge bug:
I'm a Javascript fan but IMHO #MarkyRoden has discovered a huge bug in SSJS. Shims and polyfills aren't really the main problem. Eval is known to be a bad practice but the prototype object is a fundamental element of basic Javascript. It's the standard and preferred way to add methods to Javascript classes, it's also needed for inheritance and all kind of OOP stuff. So you will need some kind of namespace at server level in order to avoid collisions. All this is really bad but the huge problem is that just a line of code in one application can broke all applications in a server. Yes, you can trust in your developers but one of them can write a bad line by mistake and also a Domino server can have hundreds of applications from different software vendors. Set the responsability in code reviews is not a reliable enought procedure. Maybe it's time to have a real javascript engine in SSJS, like V8, Spidermonkey, Chakra or Rhino. As a workaround, I'm thinking in something like Tommy Valand's idea with Rhino in SSJS.
Edit 2: It's even worse. You can do things like:
prototype.importPackage = null
or
prototype.Array = null
As you can see in #SvenHasselbach's article: http://hasselba.ch/blog/?p=1371
Edit 3: IBM: you told me I could use SSJS. COME ONE! PLEASE FIX THIS, it's AWFUL. Please let's officially report this issue to IBM.
You can reset the SSJS interpreter with the following Java code:
FacesContextExImpl fc = (FacesContextExImpl) FacesContextExImpl.getCurrentInstance();
UIViewRootEx2 uiRoot = (UIViewRootEx2) fc.getViewRoot();
JSContext jsContext = uiRoot.getJSInterpreter().getJSContext();
jsContext.getRegistry().init(jsContext);
This reinitializes the registry and all prototype functions.
EDIT: Changed the declaration of fc to the correct type.
EDIT 2:
Here is the SSJS version:
var uiRoot = facesContext.getViewRoot();
var jsContext = uiRoot.getJSInterpreter().getJSContext();
var reg = jsContext.getRegistry();
reg.init( jsContext );
Does I understand you correctly, that you want to clean up the SSJS interpreter to avoid a collision with your own prototype extension?
Just to clarify the answer above: This reinitializes the SSJS interpreter once. And only once.
You have to do this over and over again, because directly after reinitializing, another application on the server can overwrite the prototype functionality again. That's why this is not a real solution, it is an answer to your initial question.
It will have interessting consequences if another application will do the same while your code tries to use your extension...
try to do a Restart Task Http instead
tell http restart will not do a full restart of the http task
I'm having trouble synchronising my GEF editor with the EMF-based model. I think this is due to the fact that the model-internal EMF Adapter, or rather the methods it calls, aren't finished before the editor's Adapter's notifyChanged() is called and updates the model children. This leads to the editor view being out-of-sync with the model itself, or rather, the changes to the model not being represented in the view when they should be.
Consider this set up. A Command "CreateNodeCommand" adds a node to the underlying model:
#Override
public void execute() {
...
getNewNode().setGraph(getGraph());
...
}
The GraphEditPart has an internal class extending org.eclipse.emf.common.notify.Adapter. It's notifyChanged() method is indeed notified, as tested similar to below (incomplete code):
#Override
public void notifyChanged(Notification notification) {
switch (notification.getEventType()) {
case Notification.ADD:
System.err.println("ADD occurred!");
refreshChildren();
}
The problem is, that the (third-party) model itself also implements an Adapter, which in turn runs a number of methods on the new model element, such as adding an ID, etc.
It seems to me that the fact that the new element's figure doesn't show up in the editor directly after it's been created - but only after the next editing step, the figure for which then doesn't appear - suggests that the model adapter is still busy setting up the new element while refreshChildren() is already being called by the editor adapter.
This seems to call for synchronisation, but I'm unsure whether this can be achieved with built-in Java functionality for multithreading, or calls for an EMF-based approach.
Please share your knowledge about synchronising in EMF.
Many thanks in advance!
EDIT
On request, here is the source code for the getModelChildren() method:
#Override
protected List<EObject> getModelChildren() {
List<EObject> allModelObjects = new ArrayList<EObject>();
allModelObjects.addAll(((MyGraph) getModel()).getTokens());
allModelObjects.addAll(((MyGraph) getModel()).getNodes());
return allModelObjects;
}
Debugging the (3rd party) model, I found out that the Graph's enotify() fired the notification before the actual adding took place, hence my Adapterreceived the notification too early, i.e., before the node had been added.
The notification is now called after the add and everything works fine.
Thanks for all of your help!
Try extending EContentAdapter instead of AdapterImpl, and not forget to call
super.notifyChanged(Notification notification);
in it. It's an adapter, which will add itself to new elements of the model, and notify you then they are changed.
I have just started with Eclipse RCP.
I created Eclipse RCP View with TableViewer and WritableList to get data from other thread.
But I cannot see any changes. I need only to show content of List that other thread is managing.
public class View extends ViewPart {
private TableViewer viewer;
private WritableList input;
I also can get error,
org.eclipse.core.runtime.AssertionFailedException: assertion failed: Getter called outside realm of observable org.eclipse.core.databinding.observable.list.WritableList
I know what is UI Thread. I just don't know how to write. Please help with example.
UPDATE. Was not solved, because of lack of time, and missing good and focused tutorial.
I received this error message also with my code.
Databinding observables (WritableList, WritableValue...) inherit from ChangeManager, which provides ChangeManager#getRealm and the realm has Realm#exec. Within the runnable provided to exec, the operation runs in the correct thread.
This line caused the error (Getter called outside realm of observable):
WritableValue value = getEditor().getWritableValue();
System.out.println(((RcpEditorModel) value.getValue()).getNumber());
And this prevented the exception:
WritableValue value = getEditor().getWritableValue();
value.getRealm().exec(() -> System.out.println(((RcpEditorModel) value.getValue()).getNumber()));
The same will work with WritableList since it also inherits from ChangeManager.