I am wanting to turn off the UpdateCheck functionality for all members (except their primary keys). Now I was following the example below as guidance, however my MetaDataMembers of the table are still set to Always.
http://www.the-lazy-coder.com/2013/04/set-updatecheck-to-never.html
The above code snippet just gets you to change the attribute, however it seems to never get picked up, as I can debug the code when it is running and I see all the properties being set, so I am presuming that the attributes changing does not change the underlying object.
Now if I were to change approach and just get the MetaDataMembers directly from the RowType I notice they have the UpdateCheck property, however only a getter. So is there a way to (via reflection if needed) overwrite this property once it is set? Even after looking at decompiled source it is an abstract class and I cannot find any implementations to use for reference.
I am using SQLMetal to generate the Context files, so there is no designer tinkering available, and although some people will say that I should run some text editing macros to parse and change the attributes, it all sounds too long winded when I should just be able to go into the object in memory and tell it to ignore whatever it has been told previously.
SO! Is there a way to override the property in the entities? I have tried running the original code in that link in both constructor, after the objects created and just before I am about to do an update, however none of the changes seem to stick or at least propagate to where it matters, and there is hardly any material on how to do any of this progmatically.
After searching around the internet I found no nice way to do it, and although there is the link I mentioned originally it doesn't work as it works on the attributes which are partly right but in the case above they are working on the attributes which are not in memory and are just the decorations, anyway the code below seems to work but is not nice:
public static void SetUpdateCheckStatus(this IDataContext dataContext, UpdateCheck updateCheckStatus)
{
var tables = dataContext.Mapping.GetTables();
foreach (var table in tables)
{
var dataMembers = table.RowType.DataMembers;
foreach (var dataMember in dataMembers)
{
if (!dataMember.IsPrimaryKey)
{
var dataMemberType = dataMember.GetType();
if (dataMemberType.Name == "AttributedMetaDataMember")
{
var underlyingAttributeField = dataMember.GetType().GetField("attrColumn", BindingFlags.Instance | BindingFlags.NonPublic);
if (underlyingAttributeField != null)
{
var underlyingAttribute = underlyingAttributeField.GetValue(dataMember) as ColumnAttribute;
if (underlyingAttribute != null)
{ underlyingAttribute.UpdateCheck = updateCheckStatus; }
}
}
else
{
var underlyingField = dataMember.Type.GetField("updateCheck", BindingFlags.Instance | BindingFlags.NonPublic);
if (underlyingField != null)
{ underlyingField.SetValue(dataMember, updateCheckStatus); }
}
}
}
}
}
The IDataContext is just a wrapper we put around a DataContext for mocking purposes, so feel free to change that to just DataContext. It is written extremely defensively as this way pulls back lots of members which do not have all the desired data so it has to filter them out and only work on the ones which do.
Related
I have an issue with removing old elements from my lists. I tried using the methods clear() and removeAllElements() and removeAll() wherever I could but that does not seem to clear them.
To help you understand the situation a little bit better:
d1 is an ArrayList that contains all available devices in our program.
availList2 and availList3 are using the DefaultListModel.
We wanted to make it so that when the user loads the products from the proper text file, if he did that a second time the products already listed in our gui would be overwritten with the ones in the original text file. However we ended up having duplicates of the products, even though we used the clear() method in both the d1 (ArrayList) and the JList.
Any useful tips or possible causes would be appreciated. Thank you very much in advance.
if(ev.getSource() == load_availables) {
int returnVal = chooser.showOpenDialog(mainApp.this);
if(returnVal == JFileChooser.APPROVE_OPTION) {
d1.returnDevices().removeAll(d1.returnDevices());
availList2.clear();
availList3.clear();
//availList2.removeAllElements();
//availList3.removeAllElements();
File file = chooser.getSelectedFile();
read.ReadDevices(file);
for(int i = 0; i < read.Size(); i++) {
d1.add_AvailableDevices(read.get(i));
}
}
}
If the list is not cleared then I would suggest you don't have the proper reference to the DefaultListModel that is being used by the JList when you invoke the clear() method.
Start by the reading the section from the Swing tutorial on How to Use Lists.
Download the ListDemo code and play with it. Change the "Fire" button to use the clear() method on the DefaultListModel to prove to yourself that is all you need to do.
Once you see the code working then you figure out how your code is different from the ListDemo working version.
I have a method on which I'd like to run a JUnit test. I'm mocking the cq5 page using JMockit.
My test method looks like this
#Mocked
Page page;
#Mocked
PageManager pageManager;
Tag testTag = pageManager.createTag("someID","someTitle","someDescription");//i've left out the try catch for brevety
System.out.println(testTag.getTitle()); // always null here
public void testSomeMethod() {
new Expectations() {
// variables declared here are mocked by default
{
page.getProperties();
propertyMap.put("cq:tags", testTag);
returns(new ValueMapDecorator(propertyMap));
}
};
String propertyValue = methodToBeTested(page);
Assert.assertEquals(propertyValue, "someTitle");
}
And the actual method to be tested does this :-
public static String getTopic(Page page) {
String topic = null;
Tag[] tags = page.getTags();
System.out.println(tags.size()); // returns 0 when I run the test.
for (int i = 0; i < tags.length; i++) {
Tag tag = tags[i];
topic = tag.getTitle();
}
}
return topic;
}
This always returns null when I run the test; however the method to be tested works correctly in the real scenario.
I suspect I'm not setting/mocking PageManager correctly, and consequently, my testTag is null
How do I mock this correctly to get the output I'm looking for?
You're getting to this testing from the wrong side. The way mocks (usually - I've never worked with jmockit specifically) work is, you create a blank object that acts as an impostor. This impostor is not a true PageManager - it only acts as one, and introduces himself as one whenever asked. When someone asks that impostor to do something (like calling it's method), the impostor does not know what to do, so it does nothing and returns null. However, you can tell the impostor how to behave in certain situations. Namely, you can tell it what to do when a method is called.
In your case, you don't need to create actual tags to test that method - you only need to mock a page object that, when asked for it's tags, will return an array containing a mocked tag which, in turn, when asked for it's title, will respond with the title you actually want to use in your test.
I don't know jmockit, so I cannot provide any code snippet. This, however, is a general question not strictly connected to CQ5/AEM
You may not be able to find any 'setter' methods for all objects you are trying to mock and this is anyways not the correct approach to mock.
The best way as mentioned by is to use mocked pages. You can use the Expectations class (mockit.Expectations) to mock the values to be returned by certain methods in the object.
See this example of mocking a 'SlingHttpServletRequest' object in a MockedClass class.
#Test
public void testMethod(#Mocked final SlingHttpServletRequest request){
String indicator ;
new Expectations() {
{
request.getParameter("archive");
returns("true");
}
};
indicator = OriginalClass.originalMethod(request);
Assert.assertEquals(indicator, "true");
}
In a similar way, you can mock other objects and their desired values.
I have answered the same question here: https://forums.adobe.com/thread/2536290
I ran into the same issue. in order to resolve Tags, they must exists under /content/cq:tags/your/tag or /etc/tags (legacy).
The Page#getTags implementation makes a call to TagManager#getTags which in turn tries to resolve the actual tag resource in the repo. Since you are testing in an AEM context, you have to load these tags in the appropriate location for the MockTagManager to resolve them.
What this means is that you need to load your tags into the AEM test context just like you've loaded your resources (via json).
Take a look at the aem-mock TagManager impl here: wcm-io-testing/MockTagManager.java at develop · wcm-io/wcm-io-testing · GitHub start with the resolve method and debug your way to figure out where you need to add those tags.
Right, I am trying to make a game and I would like a bit of code that whenever I place makes something applies to all objects. For example I have: A background, Lamp, Player and Text. I would like it so I don't have to make the Lamp, Background and Text a symbol but have one bit of code that refers to them whenever I type it, so I don't have to list them all individually.
If you are looking to affect a specific, known set of display objects, you could add them into an array when they get created, and then use that array as your collection of objects to do whatever you need to.
Code here is procedural, not OOP, but should give you the gist. If you paste this into an FLA file and create the three objects (instances names) you'll have a decent sample of what this approach may be able to help you work with.
This can become more advanced, creates asa its own AS3 class, and be made smarted to handle whatever visual display changes that might be needed. But you only need ot make it as complicated as suits your needs.
var myDisplayObjs_arr:Array = [];
function addObj(obj:DisplayObject):void{
myDisplayObjs_arr.push(obj);
}
function affectObjs(config:Object):void{
// config object that includes things like alpha, colorTransform, whatever
for (var i:int = 0; i<myDisplayObjs_arr.length; i++){
var dispObj:DisplayObject = myDisplayObjs_arr[i];
if (config.alpha) {
dispObj.alpha = config.alpha;
}
if (config.scaleX) {
dispObj.scaleX = config.scaleX;
}
}
}
addObj(lamp);
addObj(background);
addObj(header_txt);
// Call affectObjs(), passing it an object of some basic changes
affectObjs({alpha:.5, scaleX:2});
I cannot seem to find any simple examples of this.
I have a WPF UI that I wish to display a view as a child control within another view. The MvxWpfView inherits from UserControl so it should be possible, however I cannot seem to work out how to do the binding.
I get a BindingExpression path error, as it cannot find ChildView property in my ParentViewModel.
So how do I bind a view to control content?
Firstly it's possible that you just need to add the BViewModel you want displayed on AView as a property on ViewModelA
E.g.
public class AViewModel: MvxViewModel
{
public BViewModel ChildViewModel
{
get;set;//With appropriate property changed notifiers etc.
}
}
Then inside AView you just add a BView, and you can set the datacontext of BView as follows:
<UserControl DataContext="{Binding ChildViewModel}"/>
However, if you want something more flexible (and you want the presentation handled differently for different platforms) then you will need to use a Custom Presenter
Inside your setup.cs you override CreateViewPresenter:
protected override IMvxWpfViewPresenter CreateViewPresenter(Frame rootFrame)
{
return new CustomPresenter(contentControl);
}
Now create the class CustomPresenter you need to inherit from an existing presenter. You can choose between the one it's probably using already SimpleWpfPresenter or you might want to go back a bit more to basics and use the abstract implementation
The job of the presenter is to take the viewmodel you have asked it to present, and display it "somehow". Normally that mean identify a matching view, and bind the two together.
In your case what you want to do is take an existing view, and bind a part of it to the second view mode.
This shows how I have done this in WinRT - but the idea is very similar!
public override void Show(MvxViewModelRequest request)
{
if (request.ViewModelType == typeof (AddRoomViewModel))
{
var loader = Mvx.Resolve<IMvxViewModelLoader>();
var vm = loader.LoadViewModel(request, new MvxBundle());
if (_rootFrame.SourcePageType == typeof (HomeView))
{
HomeView view = _rootFrame.Content as HomeView;
view.ShowAddRoom(vm);
}
}
else
{
base.Show(request);
}
}
So what I'm doing is I'm saying if you want me to present ViewModel AddRoom, and I have a reference to the HomeView then I'm going to just pass the ViewModel straight to the view.
Inside HomeView I simply set the data context, and do any view logic I may need to do (such as making something visible now)
internal void ShowAddRoom(Cirrious.MvvmCross.ViewModels.IMvxViewModel vm)
{
AddRoomView.DataContext = vm;
}
Hopefully that makes sense! It's well worth putting a breakpoint in the show method of the presenters so you get a feel how they work - they are really simple when you get your head around them, and very powerful.
Can someone help me to find out why I'm getting the error message "Access to undefined property: removeChild(goBack)" on the following snipped?
BTW, this is for flash CS4
function nameOfFunction() {
var goBack:backButton_mc = new backButton_mc();
goBack.x = 10;
goBack.y = 700;
goBack.back_text.text = myXML.*[buildingName].NAME;
goBack.name = "backBtn";
goBack.buttonMode = true;
addChild(goBack);
goBack.addEventListener(MouseEvent.CLICK, anotherFunction);
}
function anotherFunction(e:MouseEvent):void {
removeChild(goBack);
}
You are wrong with the scope. (surprise :-D)
The variable goBack is just defined inside of "nameOfFunction", when you try to access this from a another function like "anotherFunction" it will not exists anymore (even if it is on the display list)
There are different possibilities to solve this problem:
function anotherFunction(e:MouseEvent):void {
removeChild(e.currentTarget);
}
Or the best way would be: promote goBack as a class member of the class holding both functions. (Or if you don't use classes make goBack "global".)
Hippo is correct, but I feel it is important to explain a little more.
You created a local variable, i.e. var someVariable:DataType; within a function. This means that that variable will only be available to objects in the scope (inside) of the function (local to), and it will only last for the lifetime of the function. Soon as that function has ran the code is gone until ran again. It looks like you are probable programming directly inside the flash IDE on the time-line, which is fine, but, if you were using a document class, you could merely declare you variable in the Class scope just above the constructor function, and then set the value in the same function that your using now. This way, the reference to the variable doesn't exist within the function, it is merely set from within. This will allow that variable to be accessed from anywhere in the same class even if set to private.
This may help:
//Frame 1, Actions layer
//Slap goBack right onto the root / stage
var goBack:MovieClip;
/*
I noticed you had this data-typed differently,
i prefer to type to an interface, not an implementation.
Since your class is a movieclip in the library it extends
MovieClip and therefor IS A MovieClip, but ok either way.
*/
function nameOfFunction():void
{
goBack = new backButton_mc();
goBack.x = 10;
goBack.y = 700;
goBack.back_text.text = myXML.*[buildingName].NAME;
goBack.name = "backBtn";
goBack.buttonMode = true;
addChild(goBack);
goBack.addEventListener(MouseEvent.CLICK, anotherFunction);
}
function anotherFunction(e:MouseEvent):void
{
removeChild(goBack);
}
Scope is very important and after a while very easy to tackle. Stick with it, experiment, read up on conventions and standards that can help your development and get to loving the DocumentClass becuase even though it may be daunting to some at first, once you learn it and get used to it, it so hard to go back to programming in the flash IDE on the timeline, where I believe only display objects and audio have any place being.