ListChangeListener, JavaFX - filter

I have a question related to filters on a ObservableList. The code that i have works fine, but i think is too slow. This is beacause i load like 40,000 orders at the beginning of the app, after that the app keeps receiving orders, but for now i only have the problem in the initial load. My main problem is that the copy of my original collection of orders is considerably slower, why? because the code that i have in the changeListener i think that could be better.. but i haven't found a solution yet. Well here's an example of my code.
public MainController()
{
filteredData.addAll(Repository.masterObservableList);
Repository.masterObservableList.addListener(new ListChangeListener<OrderVo>()
{
#Override
public void onChanged(ListChangeListener.Change<? extends OrderVo> change)
{
filteredData.clear();
for (OrderVo o : Repository.masterObservableList)
filteredData.add(o);
}
});
}
}
I'll explain the code a little bit. the "Repository" is a singleton, the masterObservableList is a ObservableList and as the name says.. is the "master" or the original. The filteredData is also a ObservableList but is only declared in the controller of my fxml (MainController) and works as the copy of the master collection. Every time that my master collection recieve a change (update or new order) the filteredData should apply that change.. but im doing a for each iteration and this is problem, because it works but, too slow. Why im saying that is slow? Because in the beginning i was using the master collection as the data provider of a TableView that shows the orders and it worked fast and clean. After that i wanted to add filters to the table and thats when i began to do a research and found the filtered data and another methods.. i keeped this method (is a larger code, but the main problem is here).. and it works.. but the time that took to load the orders in the beginning is like 1:30 mins more than before.. So guys, if you have any idea! of how to keep the filteredData updated without making a for each in the changeListener i will be very happy and grateful. Thanks for reading!

Your algorithm is forcefully resetting all elements of filteredData everytime masterObservableList changes.
This is too exhaustive, as the listener can be triggered upon each item added or removed.
Maybe it's better to only add or remove the elements that have changed?
Repository.masterObservableList.addListener(new ListChangeListener<OrderVo>()
{
#Override
public void onChanged(ListChangeListener.Change<? extends OrderVo> change)
{
while(change.next()) {
if(change.wasAdded()) {
List added = change.getAddedSubList();
//add those elements to your data
}
if(change.wasRemoved()) {
List removed = change.getRemoved();
//remove those elements from your data
}
}
}
});
}

Related

What determines which event will run first?

I've created a test plugin as a science project to try and determine if 2 of the same event are called at the same time, which will be executed first.
public class TestPlugin extends JavaPlugin implements Listener {
public void onEnable() {
Bukkit.getPluginManager().registerEvents(this, this);
}
#EventHandler(priority = EventPriority.HIGHEST)
public void event1(PlayerInteractEvent e) {
System.out.println("event 1");
}
#EventHandler(priority = EventPriority.HIGHEST)
public void event2(PlayerInteractEvent e) {
System.out.println("event 2");
}
}
the output that the plugin produced are
[17:01:51 INFO]: event 2
[17:01:51 INFO]: event 1
if event1 is listed first in the class file, why is it that event2 is fired first?
This is very hard to determine. Since registerEvents leads to the function createRegisteredListeners of https://github.com/Bukkit/Bukkit/blob/master/src/main/java/org/bukkit/plugin/java/JavaPluginLoader.java
In there they use a HashSet to store the Methods of your Listener-class. So it is the pseudo random hash they give each method to store it that determines which event gets registered first.
I'm not sure but I guess that each and every time you register the events it's kind of random which gets registered first. This is why you should use different eventPriorities in order to determine which is called first.
If you really want to get deeper you have to tear apart the function createRegisteredListeners of the github-link I posted earlier in this reply. But I guess it woll never be totally certain because of the HashSet used in line 235 of the JavaPluginLoader.java:
methods = new HashSet<Method>(publicMethods.length, Float.MAX_VALUE);
As Methods are added to the set it never is certain in which position in the RAM they end up.
I hope I could help you with this post even though it doesn't really answer your original question.

Determining if all items in a collection meet established critera

This seems to be a problem that comes up a lot. I've been coming up with the same solution nearly every time but was curious if people have a better method of accomplishing this.
I have one class that is a list of instances of another class. The state of the parent class is dependent upon state of ALL the children
As an example. Say I have the following classes
class Box
{
int _objectId= <insert_unique_id>;
int _itemCount = 0;
public void AddItem()
{
_itemCount = Max(_itemCount + 1, 5);
}
public int GetId()
{
return _objectId;
}
public bool IsFull()
{
return _itemCount == 5
}
}
class BiggerBox
{
Map<int, Box> _boxes;
public void AddToBox(int id)
{
_boxes[id].AddItem();
}
public bool IsFull()
{
foreach(var box in _boxes.Values)
if(!box.IsFull())
return false;
return true;
}
}
All additions to a "Box" are done via the BiggerBox.AddToBox call. What I would like to do is be able to determine box.IsFull() without iterating over every single item every time we add an element.
Typically i accomplish this by keeping a SET or a separate collection of what items are full.
Curious, has anyone come up to an ingenious solution to this or is the simple answer that there is no other way?
There are two things you need to do in order to accomplish what you want:
Be able to control every entrypoint to your collection
React to changes to the objects in the collection
For instance, if the objects in the collection are mutable (meaning, they can change after being added to your collection) you need your main object to react to that change.
As you say, you could create a separate set of the objects that are full, but if the objects can change afterwards, when they change you either need to take them out of that set, or add them to it.
This means that in order for you to optimize this, you need some way to observe the changes to the underlying objects, for instance if they implement INotifyPropertyChanged or similar.
If the objects cannot change after being added to your main object, or you don't really care if they do, you just need to control every entrypoint, meaning that you basically need to add the necessary checks to your AddItem method.
For your particular types I would implement an event on the Box class so that when it is full, it fires the event. Your BiggerBox class would then hook into this event in order to observe when an underlying box becomes full.
You can upkeep the number of complete (or non-complete) boxes in BiggerBox class, and update it in all the functions.
E.g., in AddToBox it could be:
bool wasFull = _boxes[id].IsFull;
_boxes[id].AddItem();
if (!wasFull && _boxes[id].IsFull) // After this update box has become full.
completeBoxes += 1;
It is also possible to implement this upkeep procedure in other hypothetical functions (like RemoveFromBox, AddBox, RemoveBox, etc.)

displaying images from an ArrayCollection in ListItemRenderer in flex

I am facing the following problem,
I have an object called "data". It has three properties, one of it being itemRendererData. The "itemRendererData" is an ArrayCollection of objects having many properties one of which is the property "imageURL" (datatype:String).
I am working in flex. I have defined the view and the item renderer properly. The view has the data. I am supposed to get the images from the url specified by imageURL property.
In the itemRenderer, I have declared, source
source = {data.itemRendererData.imageURL}
But the images are not being displayed.
Use a the FlexEvent.DATA_CHANGE handler rather than binding, which is actually the proper way to handle this and gives you far more control.
public function CustomItemRenderer() {
this.addEventListener(FlexEvent.DATA_CHANGE, this.dataChangeHandler);
this.addEventListener(FlexEvent.CREATION_COMPLETE, this.creationCompleteHandler);
}
private function creationCompleteHandler(e:FlexEvent) {
if (this.data) {
this.image.source = this.data.itemRendererData.imageURL;
}
}
private function dataChangeHandler(e:FlexEvent) {
if (this.data && this.initialized) {
this.image.source = this.data.itemRendererData.imageURL;
}
}
You will notice that I have a handler for FlexEvent.CREATION_COMPLETE as well. This is because the data is actually set before the components are created. So the first time a renderer is loaded, this.image is null and this.image.source will error out.
If that doesn't work, you also need to make sure that the Image/BitmapImage is not a direct child of the renderer. I never did figure out why this was, but adding it as a child of Group fixed that issue where the image was being set but not rendering. Again, I have no idea why this was and I tested for a few hours trying to figure it out.
As an added tip, avoid MXML-based ItemRenderers in mobile applications. They are noticeably slower than pure-AS3 renderers.

Thread safe caching

I am trying to analyze what problem i might be having with unsafe threading in my code.
In my mvc3 webapplication i try to the following:
// Caching code
public static class CacheExtensions
{
public static T GetOrStore<T>(this Cache cache, string key, Func<T> generator)
{
var result = cache[key];
if(result == null)
{
result = generator();
lock(sync) {
cache[key] = result;
}
}
return (T)result;
}
}
Using the caching like this:
// Using the cached stuff
public class SectionViewData
{
public IEnumerable<Product> Products {get;set;}
public IEnumerable<SomethingElse> SomethingElse {get;set;}
}
private void Testing()
{
var cachedSection = HttpContext.Current.Cache.GetOrStore("Some Key", 0 => GetSectionViewData());
// Threading problem?
foreach(var product in cachedSection.Products)
{
DosomestuffwithProduct...
}
}
private SectionViewData GetSectionViewData()
{
SectionViewData viewData = new SectionViewData();
viewData.Products = CreateProductList();
viewData.SomethingElse = CreateSomethingElse();
return viewData;
}
Could i run inte problem with the IEnumerable? I dont have much experience with threading problems. The cachedSection would not get touched if some other thread adds a new value to cache right? To me this would work!
Should i cache Products and SomethingElse indivually? Would that be better than caching the whole SectionViewData??
Threading is hard;
In your GetOrStore method, the get/generator sequence is entirely unsynchronized, so any nymber of threads can get null from the cache and run the generator function at the same time. This may - or may not - be a problem.
Your lock statement only locks the setter of cache[string], which is already thread safe and doesn't need to be "extra locked".
The variation of double-checked locking in the cache is suspect, I'd try to get rid of it. Since the thread that never enters the lock() section can get result without a memory barrier, result may not be entirely constructed by the time the thread gets it.
Enumerating the cached IEnumrators is safe as long as nothing modifies them at the same time. If GetSectionViewData() returns an object with immutable (as in non changing) collections, you're safe.
Your code is missing parts like how would Products be populated? Only in GetSectionViewData?
If so, then I don't see a major problem with your code.
There is however a chance that two threads generate the same data(CachedSection) for the same key, it shouldn't create a threading problem except that you are doing the work twice, so if this was an expensive operation I would change the code so it only generates it once per key. If it is not expensive, it works fine as is.
IEnumerable for Products is not touched (assuming you create it separately per thread, but the enumerator on the cache is modified for each insert operation, hence it is not thread safe. So if you are using this I would be careful about that.

ViewModels and IsolatedStorageSettings

Im working on a MVVM Windows phone app that displays weather info.
When the app loads up it opens MainPage.xaml. It makes a call the the service to get weather info and binds that data to the UI. Both Fahrenheit and Celcius info are returned but only one is displayed.
On the setting page, the user can select to view the temp in either Fahrenheit or Celcius.
The user can change this setting at any time and its stored in IsolatedStorageSettings.
The issue Im having is this:
when the user navigates to the Settings page and changes their preference for either Fahrenheit or Celcius, this change is not reflected on the main page.
This issue started me thinking about this in a broader context. I can see this being an issue in ANY MVVM app where the display depends on some setting in IsolatedStorage. Any time any setting in the IsoStore is updated, how does the ViewModels know this? When I navigate back in the NavigationStack from the settings page back to MainPage how can I force a rebind of the page?
The data in my model hasnt changed, only the data that I want to display has changed.
Am I missing something simple here?
Thanks in advance.
Alex
Probably you have code like this:
public double DisplayTemperature
{
get { return (IsCelsium) ? Celsium : Fahrenheit; }
}
And IsCelsium is:
public double IsCelsium
{
get { return (bool)settings["IsCelsium"]; }
set { settings["IsCelsium"] = value; }
}
So you need to add NotifyPropertyChanged event to notify UI to get new values from DisplayTemperature property:
public double IsCelsium
{
get { return (bool)settings["IsCelsium"]; }
set
{
settings["IsCelsium"] = value;
NotifyPropertyChanged("DisplayTemperature");
}
}
Take a look at Caliburn Micro. You could implement something similar or use CM itself. When using CM I don't even think about this stuff, CM makes it so simple.
When your ViewModel inherits from Screen there are life-cycle events that fire that you can override. For example, OnInitialize fires the very first time the ViewModel is Activated and OnActivate fires every time the VM is activated. There's also OnViewAttached and OnViewLoaded.
These methods are the perfect place to put logic to populate or re-populate data.
CM also has some special built in features for allowing one to easily tombstone a single property or an entire object graph into Iso or phone state.
ok, so Ive come up with a solution. Before I get to it, let me provide some background. The app that Im working on uses both MVVM Light and WP7Contrib. That being the case, I am using Funq for DI and the MVVMLight Toolkit. After I posted my initial question, I gave the question a bit more thought. I remembered a video that I watched a while back from MIX2011 called Deep Dive MVVM with Laurent Bugnion
http://channel9.msdn.com/Events/MIX/MIX11/OPN03
In it, he talks about just this problem (view models not living at the same time) on Windows Phone. The part in question starts around the 19 minute mark.
Anyway, after I remembered that and realized that the ViewModel locator is exposed in App.xaml, this became a trivial problem to solve. When the user changes the Fahrenheit/Celcius option on the setting page, I simply get a reference to the MainViewModel via the ViewModelLocator and reset the collection that is bound to the UI thus causing the bindings to update.
public bool AddOrUpdateValue(string Key, Object value)
{
bool valueChanged = false;
// If the key exists
if (settings.Contains(Key))
{
// If the value has changed
if (settings[Key] != value)
{
// Store the new value
settings[Key] = value;
valueChanged = true;
}
}
// Otherwise create the key.
else
{
settings.Add(Key, value);
valueChanged = true;
}
return valueChanged;
}
public bool ImperialSetting
{
get
{
return GetValueOrDefault<bool>(ImperialSettingKeyName, ImperialSettingDefault);
}
set
{
if (AddOrUpdateValue(ImperialSettingKeyName, value))
{
Save();
RaisePropertyChanged("ImperialSettingText");
var vml = new ViewModelLocator();
vml.MainViewModel.Cities = (App.Current as App).Cities;
}
}
}
It was a mistake on my part not to realize that I could get access to the viewModel via the ViewModelLocator. Hopefully this post saves someone else the time I burned on this issue.

Resources