I am creating varied number of MvxSpinners programmatically. The number of the MvxSpinners generated cannot be predetermined. It is determined by the user input.
I have a List<Beneficiary>. Each MvxSpinner is meant to update each Beneficiary in the collection.
Since I cannot determine the number of MvxSpinner (which corresponds to the count of the Beneficiary in the collection) to be generated, I am forced to have one ICommand to handle all the HandleSelectedItem event of the MvxSpinners.
The Challenge
I am having difficulty determining the index of the List<Beneficiary> to update depending on the MvxSpinner the user clicked.
An Example
let
var BeneficiaryList=new List<Beneficiary>()
If there are 5 Beneficiary object in the collection, 5 MvxSpinner will be generated.
If the user selects a MVXSpinner which is meant to update index 2 of the collection, how do i determine the index of Beneficary to update?
What I have tried
private IList<Beneficiary> _beneficiaryList;
public IList<Beneficiary> BeneficiaryList
{
get { return _beneficiaryList; }
set { _beneficiaryList= value; RaisePropertyChanged(() => BeneficiaryList); }
}
public ICommand UpdateBeneficiary=> new MvxCommand<Beneficiary>(item =>
{
//item is of type Beneficiary
//But I do not know which index of BeneficiaryList to update
});
Your help will be deeply appreciated.
You probably need a List of ICommands too, one for each spinner. Something like this in your view model...
private IList<ICommand> _commands;
public IList<ICommand> Commands {
get {
if (_commands == null) {
_commands = BeneficiaryList.Select(x => new MvxCommand<Beneficiary>(item => {
...
}));
}
return _commands;
}
}
And set up your bindings like this (assuming you've got a list of spinners)
for (int i = 0; i < spinners.Count; i++) {
var spinner = spinners[i];
set.Bind (spinner).To(vm => vm.Commands[i]);
}
Well, it is interesting to answer my own question.
What I did was to give each Spinner a unique ID that corresponds to the index of the collection.
I created a custom Spinner called MvxSpinnerIndexer extending MvxSpinner (I really do not think it matters. You can just extend Spinner). MvxSpinnerIndexer retrieved the Id and the SelectedItem and then placed the two into a Dictionary
Here is the source for MvxSpinnerIndexer
public class MvxSpinnerIndexer : Spinner
{
public MvxSpinnerIndexer(Context context, IAttributeSet attrs)
: this(
context, attrs,
new MvxAdapter(context)
{
SimpleViewLayoutId = global::Android.Resource.Layout.SimpleDropDownItem1Line
})
{ }
public MvxSpinnerIndexer(Context context, IAttributeSet attrs, IMvxAdapter adapter)
: base(context, attrs)
{
var itemTemplateId = MvxAttributeHelpers.ReadListItemTemplateId(context, attrs);
var dropDownItemTemplateId = MvxAttributeHelpers.ReadDropDownListItemTemplateId(context, attrs);
adapter.ItemTemplateId = itemTemplateId;
adapter.DropDownItemTemplateId = dropDownItemTemplateId;
Adapter = adapter;
SetupHandleItemSelected();
}
public new IMvxAdapter Adapter
{
get { return base.Adapter as IMvxAdapter; }
set
{
var existing = Adapter;
if (existing == value)
return;
if (existing != null && value != null)
{
value.ItemsSource = existing.ItemsSource;
value.ItemTemplateId = existing.ItemTemplateId;
}
base.Adapter = value;
}
}
[MvxSetToNullAfterBinding]
public IEnumerable ItemsSource
{
get { return Adapter.ItemsSource; }
set { Adapter.ItemsSource = value; }
}
public int ItemTemplateId
{
get { return Adapter.ItemTemplateId; }
set { Adapter.ItemTemplateId = value; }
}
public int DropDownItemTemplateId
{
get { return Adapter.DropDownItemTemplateId; }
set { Adapter.DropDownItemTemplateId = value; }
}
public ICommand HandleItemSelected { get; set; }
public int ViewId { get; set; }
private void SetupHandleItemSelected()
{
ItemSelected += (sender, args) =>
{
//sender.
var control = (MvxSpinnerIndexer)sender;
var controlId = control.Id;
var position = args.Position;
HandleSelected(position, controlId);
};
}
protected virtual void HandleSelected(int position, int? controlId)
{
var item = Adapter.GetRawItem(position);
var content = new Dictionary<string, object> {{"Index", controlId}, {"SelectedItem", item}};
//var param = new ListItemWithIndexModel { Index = controlId, SelectedItem = item };
if (HandleItemSelected == null
|| item == null
|| !HandleItemSelected.CanExecute(content))
return;
HandleItemSelected.Execute(content);
}
}
In your ViewModel
public ICommand SpinnerSelected => new MvxCommand<Dictionary<string, object>>(item =>
{
var selectedItem = item["SelectedItem"] as ListItemModel;//Cast to the actual model
var index = item["Index"] as int?;//The index of the collection to update
});
I believe this will be useful to the community.
Related
I have a razor view in ASP.NET MVC looping over an array of objects from my model and generating corresponding html controls.
My html elements are properly bound, except my drop down lists who can't seem to select the value provided to them by the model.
My view: (in the hereby case, I'm simply displaying a list of countries)
#for (var i = 0; i < Model.answers.Count(); i++)
{
<div class="form-group">
...
#switch (Model.answers[i].Statement.QuestionType)
{
...
case ExternalEnums.QuestionTypeEnum.country:
#Html.DropDownListFor(Model => Model.answers[i].Value,
new SelectList(Model.Pays, "Value", "Text"))
break;
}
...
</div>
}
My view controller, generating the country list items and retrieving the existing model entries:
public class HomeIndexViewModel
{
private QuestionsModelContainer dbContext;
private AdmcommonEntities admCommonContext;
...
public List<Answer> answers { get; private set; }
private IEnumerable<SelectListItem> _countries;
public IEnumerable<SelectListItem> Pays
{
get
{
if (_countries == null)
SetCountries();
return _countries;
}
}
public HomeIndexViewModel()
{
Init(-1, null);
}
public HomeIndexViewModel(int page, string _pageWideError = null)
{
Init(page, _pageWideError);
}
private void Init(int page, string _pageWideError = null)
{
dbContext = new QuestionsModelContainer();
PageNum = page;
pageWideError = _pageWideError;
answers = GetAnswers();
...
}
private void SetCountries()
{
using (admCommonContext = new AdmcommonEntities())
{
var localEntities = admCommonContext.Pays.ToList();
var localList = new List<SelectListItem>();
localList.Add(new SelectListItem());
foreach (var item in localEntities)
{
var newItemList = new SelectListItem();
newItemList.Text = item.Libelle;
newItemList.Value = item.Libelle;
localList.Add(newItemList);
}
_countries = localList;
}
}
public List<Statement> GetStatements()
{
var statements = dbContext.StatementSet.Where(w => w.Page == PageNum).OrderBy(w => w.Order).ToList();
return statements;
}
public List<Answer> GetAnswers()
{
var statements = GetStatements();
var ExistingAnswers = new List<Answer>();
if (AdminPermissionManager.IsUserAuthenticated()) //Loading existing entries.
ExistingAnswers = Answer.GetExistingAnswers(statements, dbContext);
var answers = new List<Answer>();
foreach (var item in statements)
{
var answer = ExistingAnswers.Where(w => w.StatementId == item.Id).FirstOrDefault();
if (answer == null)
{
answer = new Answer();
answer.StatementId = item.Id;
answer.Statement = item;
}
answers.Add(answer);
}
return answers;
}
}
My model class, simply containing the value I'm trying to display:
[MetadataType(typeof(AnswerMetaData))]
public partial class Answer
{
...
public static List<Answer> GetExistingAnswers(List<int> statementIds, QuestionsModelContainer dbContext)
{
List<Answer> ExistingAnswers;
var usercode = AdminPermissionManager.GetUserCode();
ExistingAnswers = dbContext.AnswerSet.Where(w => statementIds.Contains(w.StatementId) && w.ChildCode == usercode).ToList();
return ExistingAnswers;
}
public static List<Answer> GetExistingAnswers(List<Statement> statements, QuestionsModelContainer dbContext)
{
var statementIds = statements.Select(w => w.Id).ToList();
return GetExistingAnswers(statementIds, dbContext);
}
}
public class AnswerMetaData
{
[InternalValidation]
public string Value { get; set; }
private class InternalValidationAttribute : ValidationAttribute
{
...
}
}
I'm sure there's something very obvious that I'm missing, but can't figure out what exactly :/...
You're nearly there actually, this part in the View:
#Html.DropDownListFor(
Model => Model.answers[i].Value,
new SelectList(Model.Pays, "Value", "Text")
)
You create a new selectlist - each time, but you already have a IEnumerable<SelectListItem> created, so you don't have to recreate that list. The only thing you might be missing (most likely) is the "Selected" item option.
If you already have a value selected (and it isn't the first one) it will not be selected dropdown option - also because you pass the value of the selected option as the "ID" of the field (not the actual value) - DropDownListFor is kinda weird in that regard.
So you want to change your #Html.DropDownListFor to something like this:
#Html.DropDownListFor(
Model => Model.answers[i].Name,
Pays(Model.answers[i].Value)
)
When that being done you should change your property "Pays" in the ViewModel to a method that accepts a value (idk what you're using, but let's assume it's string) - to something along the lines of this:
public IEnumerable<SelectListItem> Pays(string selectedValue)
{
if (_countries == null) SetCountries();
var value = new List<SelectListItem>();
foreach(var item in _countries)
{
item.Selected = (item.Value == selectedValue);
value.Add(item);
}
return value;
}
This above is a bit pseudocoded since I'm typing this from memory, but it should get you into the correct direction. Also remember to check with the inspect element in the browser if the dropdown HTML element really has the correct name attribute.
I'm writing a plugin for resharper which I want to use to navigate from a ConcreteCommand -> ConcreteCommandHandler where those types look like this
public class ConcreteCommand : ICommand
public class ConcreteCommandHandler : ICommandHandler<ConcreteCommand>
I've got as far as adding my navigation menu option when the cursor is on a ICommand instance/definition (currently only by checking if the name contains 'Command' and not 'CommandHandler'), and I think I have the code necessary to actually search for a type which inherits something, but my issue is that the only thing I actually have a type for is my ConcereteCommand and I need to create (or get a reference to) the generic type ICommandHandler<T> with T being the type the cursor is currently on.
So I have 2 things I still want to know:
How can I check if my IDeclaredElement is an implementation of a particular interface (ideally by specifying the full name in a string from config)?
How can I create a ITypeElement which is a generic type of a specific interface where I can set the generic type from my existing IDeclaredElements type, so I can then find classes which inherit this?
My existing code looks like this:
[ContextNavigationProvider]
public class CommandHandlerNavigationProvider : INavigateFromHereProvider
{
public IEnumerable<ContextNavigation> CreateWorkflow(IDataContext dataContext)
{
ICollection<IDeclaredElement> declaredElements = dataContext.GetData(DataConstants.DECLARED_ELEMENTS);
if (declaredElements != null || declaredElements.Any())
{
IDeclaredElement declaredElement = declaredElements.First();
if (IsCommand(declaredElement))
{
var solution = dataContext.GetData(JetBrains.ProjectModel.DataContext.DataConstants.SOLUTION);
yield return new ContextNavigation("This Command's &handler", null, NavigationActionGroup.Other, () => { GotToInheritor(solution,declaredElement); });
}
}
}
private void GotToInheritor(ISolution solution, IDeclaredElement declaredElement)
{
var inheritorsConsumer = new InheritorsConsumer();
SearchDomainFactory searchDomainFactory = solution.GetComponent<SearchDomainFactory>();
//How can I create the ITypeElement MyNameSpace.ICommandHandler<(ITypeElement)declaredElement> here?
solution.GetPsiServices().Finder.FindInheritors((ITypeElement)declaredElement, searchDomainFactory.CreateSearchDomain(solution, true), inheritorsConsumer, NullProgressIndicator.Instance);
}
private bool IsCommand(IDeclaredElement declaredElement)
{
//How can I check if my declaredElement is an implementation of ICommand here?
string className = declaredElement.ShortName;
return className.Contains("Command")
&& !className.Contains("CommandHandler");
}
}
Ok managed to work this out with a fair bit of pushing in the right direction from #CitizenMatt.
basically my solution looks like this (still needs some tidying up)
private static readonly List<HandlerMapping> HandlerMappings = new List<HandlerMapping>
{
new HandlerMapping("HandlerNavigationTest.ICommand", "HandlerNavigationTest.ICommandHandler`1", "HandlerNavigationTest"),
new HandlerMapping("HandlerNavTest2.IEvent", "HandlerNavTest2.IEventHandler`1", "HandlerNavTest2")
};
public IEnumerable<ContextNavigation> CreateWorkflow(IDataContext dataContext)
{
ICollection<IDeclaredElement> declaredElements = dataContext.GetData(DataConstants.DECLARED_ELEMENTS);
if (declaredElements != null && declaredElements.Any())
{
IDeclaredElement declaredElement = declaredElements.First();
ISolution solution = dataContext.GetData(JetBrains.ProjectModel.DataContext.DataConstants.SOLUTION);
ITypeElement handlerType = GetHandlerType(declaredElement);
if (handlerType != null)
{
yield return new ContextNavigation("&Handler", null, NavigationActionGroup.Other, () => GoToInheritor(solution, declaredElement as IClass, dataContext, handlerType));
}
}
}
private static ITypeElement GetHandlerType(IDeclaredElement declaredElement)
{
var theClass = declaredElement as IClass;
if (theClass != null)
{
foreach (IPsiModule psiModule in declaredElement.GetPsiServices().Modules.GetModules())
{
foreach (var handlerMapping in HandlerMappings)
{
IDeclaredType commandInterfaceType = TypeFactory.CreateTypeByCLRName(handlerMapping.HandledType, psiModule, theClass.ResolveContext);
ITypeElement typeElement = commandInterfaceType.GetTypeElement();
if (typeElement != null)
{
if (theClass.IsDescendantOf(typeElement))
{
IDeclaredType genericType = TypeFactory.CreateTypeByCLRName(handlerMapping.HandlerType, psiModule, theClass.ResolveContext);
ITypeElement genericTypeElement = genericType.GetTypeElement();
return genericTypeElement;
}
}
}
}
}
return null;
}
private static void GoToInheritor(ISolution solution, IClass theClass, IDataContext dataContext, ITypeElement genericHandlerType)
{
var inheritorsConsumer = new InheritorsConsumer();
var searchDomainFactory = solution.GetComponent<SearchDomainFactory>();
IDeclaredType theType = TypeFactory.CreateType(theClass);
IDeclaredType commandHandlerType = TypeFactory.CreateType(genericHandlerType, theType);
ITypeElement handlerTypeelement = commandHandlerType.GetTypeElement();
solution.GetPsiServices().Finder.FindInheritors(handlerTypeelement, searchDomainFactory.CreateSearchDomain(solution, true),
inheritorsConsumer, NullProgressIndicator.Instance);
var potentialNavigationPoints = new List<INavigationPoint>();
foreach (ITypeElement inheritedInstance in inheritorsConsumer.FoundElements)
{
IDeclaredType[] baseClasses = inheritedInstance.GetAllSuperTypes();
foreach (IDeclaredType declaredType in baseClasses)
{
if (declaredType.IsInterfaceType())
{
if (declaredType.Equals(commandHandlerType))
{
var navigationPoint = new DeclaredElementNavigationPoint(inheritedInstance);
potentialNavigationPoints.Add(navigationPoint);
}
}
}
}
if (potentialNavigationPoints.Any())
{
NavigationOptions options = NavigationOptions.FromDataContext(dataContext, "Which handler do you want to navigate to?");
NavigationManager.GetInstance(solution).Navigate(potentialNavigationPoints, options);
}
}
public class InheritorsConsumer : IFindResultConsumer<ITypeElement>
{
private const int MaxInheritors = 50;
private readonly HashSet<ITypeElement> elements = new HashSet<ITypeElement>();
public IEnumerable<ITypeElement> FoundElements
{
get { return elements; }
}
public ITypeElement Build(FindResult result)
{
var inheritedElement = result as FindResultInheritedElement;
if (inheritedElement != null)
return (ITypeElement) inheritedElement.DeclaredElement;
return null;
}
public FindExecution Merge(ITypeElement data)
{
elements.Add(data);
return elements.Count < MaxInheritors ? FindExecution.Continue : FindExecution.Stop;
}
}
And this allows me no navigate to multiple handlers if they exist. This currently relies on the interfaces for the handled type and the handler type being in the same assembly. But this seems reasonable enough for me at the moment.
I have a Window with a TabControl in it. The TabControl contains 5 different TabItems. Each TabItem has its own ViewModel associated as its DataContext, while the Window has a DataContext that has all 5 TabItem's view models as properties. The problem I am having is setup. There is a noticeable lag when I launch the Window (from my MainWindow) and I have spent a good chunk of time refactoring my code and making it faster by running things in parallel, making fewer calls to the database and running Tasks on semi-expensive operations. Everything works great, except for one TabItem and its view model. For some reason, the view does not refresh itself properly.
For instance, I have a view model called DiaryDescriptionViewModel that takes in List<SectionViewModel> and does stuff with it, with the view bound to a result collection. It works fine. My troublesome view model is called DiaryPayItemEditorViewModel and it too takes in a List<SectionViewModel> and does stuff with it, with the view bound to a result collection. Neither view models perform the work on List<SectionViewModel> on worker threads or anything. However both view models are instanced and set up in parallel, which I wouldn't think be the root of the problem.
In my DiaryPayItemEditorViewModel, I have a ObservableCollection<DiaryPayItemDetailViewModel> that a ListView is data bound to. The ListView never displays the data, even though it exists. If I take all of my view model initialization code out of a Parallel.Invoke call, then it binds and displays the data.
My assumption here is that the view is initialized (this.InitializeComponents) before the DiaryPayItemEditorViewModel is fully set up, which should be fine. Since my view models all implement INotifyPropertyChanged, the view should be notified that changes have taken place. For the life of me, I can't figure this out.
The following is the applicable source for the view Window view model (DiaryEditorViewModel), the view model who uses the same collection and works with binding (DiaryDescriptionViewModel and its child DiaryDescriptionDetailsViewModel) and then my troublesome view model (DiaryPayItemEditorViewModel and its child DiaryPayItemDetailViewModel).
DiaryEditorViewModel.cs
public class DiaryEditorViewModel : BaseChangeNotify
{
private DiaryViewModel diary;
private Project project;
private DiaryDetailsViewModel diaryDetailsViewModel;
private DiaryDescriptionViewModel diaryDescriptionViewModel;
private DiaryPayItemEditorViewModel diaryPayItemsViewModel;
private DiaryEquipmentEditorViewModel diaryEquipmentEditorViewModel;
private DiaryLaborViewModel diaryLaborViewModel;
// This is the designated constructor used by the app.
public DiaryEditorViewModel(Project project, Diary diary, UserViewModel user)
: base(user)
{
// Instance a new diary view model using the provided diary.
this.diary = new DiaryViewModel(diary, user);
this.project = project;
// Setup the repositories we will use.
var repository = new ProjectRepository();
var contractorRepository = new ContractorRepository();
// Setup the temporary collections used by the repositories.
var contractors = new List<Contractor>();
var contractorViewModels = new List<ContractorViewModel>();
var projectSections = new List<Section>();
var bidItemCollection = new List<BidItem>();
var subItemCollection = new List<SubItem>();
var sectionViewModels = new List<SectionViewModel>();
var equipmentCategories = new List<EquipmentCategory>();
var equipmentFuelTypes = new List<EquipmentFuelType>();
var equipmentList = new List<Equipment>();
var equipmentViewModels = new List<EquipmentViewModel>();
Task.Run(() =>
{
Parallel.Invoke(
// Fetch contractors for selected project.
() =>
{
contractors.AddRange(contractorRepository.GetContractorsByProjectId(diary.ProjectId));
equipmentCategories.AddRange(contractorRepository.GetEquipmentCategories());
equipmentFuelTypes.AddRange(contractorRepository.GetEquipmentFuelTypes());
equipmentList.AddRange(contractorRepository.GetEquipmentByProjectId(this.Project.ProjectId));
// Reconstruct the contractor->Equipment->FuelType & Category relationship.
contractorViewModels.AddRange(
contractors.Select(contractor =>
new ContractorViewModel(
contractor,
equipmentList.Where(equipment =>
equipment.ContractorId == contractor.ContractorId).Select(e =>
new EquipmentViewModel(
e,
contractor,
equipmentCategories.FirstOrDefault(cat =>
cat.EquipmentCategoryId == e.EquipmentCategoryId),
equipmentFuelTypes.FirstOrDefault(f =>
f.EquipmentFuelTypeId == e.EquipmentFuelTypeId))))));
},
() =>
{
// Fetch all of the Sections, Bid-Items and Sub-items for the project
projectSections.AddRange(repository.GetSectionsByProjectId(project.ProjectId));
bidItemCollection.AddRange(repository.GetBidItemsByProjectId(project.ProjectId));
subItemCollection.AddRange(repository.GetSubItemsByProjectId(project.ProjectId));
// Reconstruct the Section->BidItem->SubItem hierarchy.
sectionViewModels.AddRange(
projectSections.Select(s =>
new SectionViewModel(project, s,
bidItemCollection.Where(b => b.SectionId == s.SectionId).Select(b =>
new BidItemViewModel(project, b,
subItemCollection.Where(si => si.BidItemId == b.BidItemId))))));
}
);
// Once the parallel invocations are completed, instance all of the children view models
// using the view model collections we just set up.
Parallel.Invoke(
// Fetch contractors for selected project.
() =>
this.DiaryDetailsViewModel = new DiaryDetailsViewModel(
project,
diary,
user),
() => // This view model works just fine, with same constructor signature.
this.DiaryDescriptionViewModel = new DiaryDescriptionViewModel(
project,
diary,
user,
sectionViewModels),
() =>
this.DiaryPayItemEditorViewModel = new DiaryPayItemEditorViewModel(
project,
diary,
user,
sectionViewModels),
() => // This view model does not notify the UI of changes to its collection.
this.DiaryEquipmentEditorViewModel = new DiaryEquipmentEditorViewModel(
project,
diary,
user,
contractorViewModels),
() =>
// For the Labor view, we just pass the Contractor model collection rather than the view model collection
// since the Labor view does not need any of the additional equipment information.
this.DiaryLaborViewModel = new DiaryLaborViewModel(
project,
diary,
user,
contractors));
});
}
public Project Project
{
get
{
return this.project;
}
set
{
this.project = value;
this.OnPropertyChanged();
}
}
public DiaryViewModel Diary
{
get
{
return this.diary;
}
set
{
this.diary = value;
this.OnPropertyChanged();
}
}
public DiaryDetailsViewModel DiaryDetailsViewModel
{
get
{
return this.diaryDetailsViewModel;
}
set
{
this.diaryDetailsViewModel = value;
this.OnPropertyChanged();
}
}
public DiaryDescriptionViewModel DiaryDescriptionViewModel
{
get
{
return this.diaryDescriptionViewModel;
}
set
{
this.diaryDescriptionViewModel = value;
this.OnPropertyChanged();
}
}
public DiaryPayItemEditorViewModel DiaryPayItemEditorViewModel
{
get
{
return this.diaryPayItemsViewModel;
}
set
{
this.diaryPayItemsViewModel = value;
this.OnPropertyChanged();
}
}
public DiaryLaborViewModel DiaryLaborViewModel
{
get
{
return this.diaryLaborViewModel;
}
set
{
this.diaryLaborViewModel = value;
this.OnPropertyChanged();
}
}
public DiaryEquipmentEditorViewModel DiaryEquipmentEditorViewModel
{
get
{
return this.diaryEquipmentEditorViewModel;
}
set
{
this.diaryEquipmentEditorViewModel = value;
this.OnPropertyChanged();
}
}
}
DiaryDescriptionViewModel
This view model works just fine, with its this.DiaryDescriptions collection being bound to properly and displayed in the ListView
public class DiaryDescriptionViewModel : BaseDiaryViewModel, IDataErrorInfo
{
private ObservableCollection<DiaryDescriptionDetailsViewModel> diaryDescriptions;
private DiaryDescriptionDetailsViewModel selectedDiaryDescription;
public DiaryDescriptionViewModel()
{
}
public DiaryDescriptionViewModel(Project project, Diary diary, UserViewModel user, List<SectionViewModel> sections)
: base(project, diary, user)
{
// Restore any previously saved descriptions.
var diaryRepository = new DiaryRepository();
List<DiaryDescription> descriptions = diaryRepository.GetDiaryDescriptionsByDiaryId(diary.DiaryId);
this.ProjectSections = sections;
// Reconstruct our descriptions
this.diaryDescriptions = new ObservableCollection<DiaryDescriptionDetailsViewModel>();
foreach (DiaryDescription description in descriptions)
{
SectionViewModel section = this.GetSectionContainingBidItemId(description.BidItemId);
BidItemViewModel bidItem = section.GetBidItem(description.BidItemId);
var details = new DiaryDescriptionDetailsViewModel(description, section, bidItem);
details.PropertyChanged += ChildViewModelPropertyChanged;
this.diaryDescriptions.Add(details);
}
this.diaryDescriptions.CollectionChanged += this.DiaryDescriptionsOnCollectionChanged;
this.IsDirty = false;
}
public ObservableCollection<DiaryDescriptionDetailsViewModel> DiaryDescriptions
{
get
{
return this.diaryDescriptions;
}
set
{
if (value != null)
{
this.diaryDescriptions.CollectionChanged -= this.DiaryDescriptionsOnCollectionChanged;
this.diaryDescriptions =
new ObservableCollection<DiaryDescriptionDetailsViewModel>(
value
.OrderBy(s => s.Section.Section)
.ThenBy(i => i.BidItem.BidItem.Number));
this.diaryDescriptions.CollectionChanged += this.DiaryDescriptionsOnCollectionChanged;
}
else
{
this.diaryDescriptions = new ObservableCollection<DiaryDescriptionDetailsViewModel>();
}
this.OnPropertyChanged();
}
}
public DiaryDescriptionDetailsViewModel SelectedDiaryDescription
{
get
{
return this.selectedDiaryDescription;
}
set
{
// Always unsubscribe from events before replacing the object. Otherwise we end up with a memory leak.
if (this.selectedDiaryDescription != null)
{
this.selectedDiaryDescription.PropertyChanged -= this.ChildViewModelPropertyChanged;
}
this.selectedDiaryDescription = value;
if (value != null)
{
// If the description contains a biditem DiaryId, then we go fetch the section and biditem
// associated with the diary description.
if (value.BidItemId > 0)
{
this.selectedDiaryDescription.Section = this.GetSectionContainingBidItemId(value.BidItemId);
this.selectedDiaryDescription.BidItem = this.selectedDiaryDescription.Section.GetBidItem(value.BidItemId);
}
// Subscribe to property changed events so we can set ourself to dirty.
this.selectedDiaryDescription.PropertyChanged += this.ChildViewModelPropertyChanged;
this.selectedDiaryDescription.IsDirty = false;
}
this.OnPropertyChanged();
this.IsDirty = false;
}
}
DiaryDescriptionDetailViewModel
Working child view model.
public class DiaryDescriptionDetailsViewModel : BaseChangeNotify
{
private readonly DiaryDescription diaryDescription;
private SectionViewModel section;
private BidItemViewModel bidItem;
public DiaryDescriptionDetailsViewModel(DiaryDescription description, SectionViewModel section = null, BidItemViewModel bidItem = null)
{
this.diaryDescription = description;
if (description.BidItemId > 0)
{
this.section = section;
this.bidItem = bidItem;
}
this.IsDirty = false;
}
public DiaryDescription Description
{
get
{
return this.diaryDescription;
}
}
public int BidItemId
{
get
{
return this.diaryDescription.BidItemId;
}
}
public BidItemViewModel BidItem
{
get
{
return this.bidItem;
}
set
{
this.bidItem = value;
this.diaryDescription.BidItemId = value.BidItem.BidItemId;
this.OnPropertyChanged();
}
}
public SectionViewModel Section
{
get
{
return this.section;
}
set
{
this.section = value;
this.OnPropertyChanged();
}
}
}
DiaryPayItemEditorViewModel
And finally, the view model who is not having its collection rendered to the view.
public class DiaryPayItemEditorViewModel : BaseDiaryViewModel, IDataErrorInfo
{
private ObservableCollection<DiaryPayItemDetailViewModel> diaryPayItemDetails;
private DiaryPayItemDetailViewModel selectedDiaryPayItemDetail;
private List<DiaryPayItem> allPayItemsForSelectedBidItem;
private decimal sumOfAllPayItemsForBidItem;
public DiaryPayItemEditorViewModel()
{
}
public DiaryPayItemEditorViewModel(Project project, Diary diary, UserViewModel user, List<SectionViewModel> sections)
: base(project, diary, user)
{
this.Initialize(project, sections);
this.IsDirty = false;
}
public ObservableCollection<DiaryPayItemDetailViewModel> DiaryPayItemDetails
{
get
{
return this.diaryPayItemDetails;
}
set
{
this.diaryPayItemDetails = value;
this.OnPropertyChanged();
}
}
public DiaryPayItemDetailViewModel SelectedDiaryPayItemDetail
{
get
{
return this.selectedDiaryPayItemDetail;
}
set
{
if (this.selectedDiaryPayItemDetail != null)
{
this.selectedDiaryPayItemDetail.PropertyChanged -= this.ChildViewModelPropertyChanged;
}
if (value != null)
{
value.PropertyChanged += this.ChildViewModelPropertyChanged;
}
this.selectedDiaryPayItemDetail = value;
this.OnPropertyChanged();
}
}
private void Initialize(Project project, List<SectionViewModel> sections)
{
var repository = new DiaryRepository();
var projectRepository = new ProjectRepository();
this.DiaryPayItemDetails = new ObservableCollection<DiaryPayItemDetailViewModel>();
this.ProjectSections = sections;
// Repository calls to the database.
List<DiaryPayItem> payItems = repository.GetDiaryPayItemsByDiaryId(this.Diary.DiaryId);
var sectionItems = projectRepository.GetSectionHierarchy(project.ProjectId);
// Temporary, needs to be refined.
foreach (var diaryPayItem in payItems)
{
var subItem = sectionItems.SubItems.FirstOrDefault(sub => sub.SubItemId == diaryPayItem.SubItemId);
var bidItems =
sectionItems.BidItems.Where(bid => bid.BidItemId == subItem.BidItemId)
.Select(
bid =>
new BidItemViewModel(project, bid,
sectionItems.SubItems.Where(sub => sub.BidItemId == bid.BidItemId)));
var section = new SectionViewModel(
project,
sectionItems.Sections.FirstOrDefault(s => bidItems.Any(bid => bid.BidItem.SectionId == s.SectionId)),
bidItems);
this.DiaryPayItemDetails.Add(
new DiaryPayItemDetailViewModel(
diaryPayItem,
section,
bidItems.FirstOrDefault(bid => bid.BidItem.BidItemId == subItem.BidItemId),
subItem));
}
}
DiaryPayItemDetailViewModel - Child view model to the troublesome view model
public class DiaryPayItemDetailViewModel : BaseChangeNotify
{
private DiaryPayItem diaryPayItem;
private SectionViewModel selectedSection;
private BidItemViewModel selectedBidItem;
private SubItem selectedSubItem;
public DiaryPayItemDetailViewModel(
DiaryPayItem diaryPayItem,
SectionViewModel section,
BidItemViewModel bidItem,
SubItem subItem)
{
this.DiaryPayItem = diaryPayItem;
this.SelectedSection = section;
this.SelectedBidItem = bidItem;
this.SelectedSubItem = subItem;
}
public DiaryPayItem DiaryPayItem
{
get
{
return this.diaryPayItem;
}
set
{
this.diaryPayItem = value;
this.OnPropertyChanged();
}
}
public SectionViewModel SelectedSection
{
get
{
return this.selectedSection;
}
set
{
this.selectedSection = value;
this.OnPropertyChanged();
}
}
public BidItemViewModel SelectedBidItem
{
get
{
return this.selectedBidItem;
}
set
{
this.selectedBidItem = value;
this.OnPropertyChanged();
}
}
public SubItem SelectedSubItem
{
get
{
return this.selectedSubItem;
}
set
{
this.selectedSubItem = value;
this.DiaryPayItem.SubItemId = value.SubItemId;
this.OnPropertyChanged();
}
}
XAML for the DiaryDescription Tab Item.
<ListView ItemsSource="{Binding Path=DiaryDescriptions}"
SelectedItem="{Binding Path=SelectedDiaryDescription}">
<ListView.ItemTemplate>
<DataTemplate>
<TextBlock Text="{Binding Path=Section.SectionName}" />
</DataTemplate>
</ListView.ItemTemplate>
</ListView>
XAML for Diary Pay Items tab item.
<ListView Name="PayItemListView"
ItemsSource="{Binding Path=DiaryPayItemDetails}">
<ListView.ItemTemplate>
<DataTemplate>
<TextBlock Text="{Binding Path=SelectedBidItem.BidItem.Description}" />
</DataTemplate>
</ListView.ItemTemplate>
</ListView>
BaseChangeNotify
Lastly, to show my INotifyPropertyChanged implementation, I present my base class. It wraps all calls to the event handlers in an Application.Current.Dispatcher.Invoke() action. This forces all event handler calls to be ran on the main thread so I don't have to worry about cross-thread issues in my inherited objects.
public class BaseChangeNotify : INotifyPropertyChanged
{
private bool isDirty;
private UserViewModel user;
public BaseChangeNotify()
{
}
public BaseChangeNotify(UserViewModel user)
{
this.user = user;
}
public event PropertyChangedEventHandler PropertyChanged;
public bool IsDirty
{
get
{
return this.isDirty;
}
set
{
this.isDirty = value;
this.OnPropertyChanged();
}
}
public UserViewModel User
{
get
{
return this.user;
}
}
public virtual void OnPropertyChanged([CallerMemberName] string propertyName = "")
{
// Perform the IsDirty check so we don't get stuck in a infinite loop.
if (propertyName != "IsDirty")
{
this.IsDirty = true; // Each time a property value is changed, we set the dirty bool.
}
if (this.PropertyChanged != null)
{
// Invoke the event handlers attached by other objects.
try
{
// When unit testing, this will always be null.
if (Application.Current != null)
{
Application.Current.Dispatcher.Invoke(() =>
this.PropertyChanged(this, new PropertyChangedEventArgs(propertyName)));
}
else
{
this.PropertyChanged(this, new PropertyChangedEventArgs(propertyName));
}
}
catch (Exception)
{
throw;
}
}
}
If anyone could help me figure this out I would greatly appreciate it. I have been trying various things for the last two days and can't get it figured out. It's weird how one view model works fine, essentially performing the same kind of operation, and the other does not.
Thanks in advance.
The DiaryEditorViewModel is the view model to a DiaryEditorWindow. The DiaryPayItemEditorViewModel belongs to a user control that resides within the Window. Setting the data context in the XAML at the Window level, for the TabItem, resolved this issue. Setting the DataContext at the UserControl level caused view model to not bind properly.
I also tried setting the datacontext in the constructor, but that had the same issue. It would never bind. By setting the datacontext in the XAML of the TabItem associated with the troublesome view model, the problem was resolved. I don't understand why this is an issue. Since the view model fully implements the property changed event, I should be able to set the data context at any point, and adjust the values without a problem.
Eitherway, I have been able to solve this issue.
This is a refactoring question mainly.
I am creating some methods to go back/forward through an actions history depending on its Id/PreviousId relationship (see basic class below):
public class Action
{
public int Id { get; set; }
public int PreviousId { get; set; }
public string Title { get; set; }
}
Background Info:
I start off by getting a single action from the database. If the user selects 'GoBack', I need to get the previous action from the database and store it in a LinkedList. This means users can potentially revisit that same action (i.e. by going back then forward again) but by calling it from the LinkedList version rather than getting it from the database again. I don't want to initially retrieve all actions first either from the database. I have this functionality working but my GoBack() and GoForward() methods are pretty much identical.
I was hoping to see if there is a good way of refactoring this into a more generic method set rather than duplicating code? (Note - my code doesn't include the database calls to reduce reading so instead I've put dummy data into a List to act as my database).
Class level variables I'm referencing in the methods:
//The list I'm using to pretend to be my database containing actions
private List<Action> _actions { get; set; }
private Action _currentAction { get; set; }
private LinkedList<Action> _actionLinks { get; set; }
Here is my GoBack() method:
private void GoBack()
{
var current = _actionLinks.Find(_currentAction);
if (current == null)
return;
//If we've already stored the previous action. Just point to it
if (current.Previous != null)
{
_currentAction = current.Previous.Value;
return;
}
//We don't have this action stored so go get it from the database and cache it in the list
var previousAction = _actions.FirstOrDefault(i => i.Id == _currentAction.PreviousId);
//There are no previous actions
if(previousAction == null)
return;
_actionLinks.AddBefore(current, previousAction);
//Now reset the current action
_currentAction = previousAction;
}
Here is my GoForward() method:
private void GoForward()
{
var current = _actionLinks.Find(_currentAction);
if (current == null)
return;
//If we've already stored the next action. Just point to it
if (current.Next != null)
{
_currentAction = current.Next.Value;
return;
}
//We don't have this action stored so go get it from the database and cache it in the list
var nextAction = _actions.FirstOrDefault(i => i.PreviousId == _currentAction.Id);
//There are no further actions
if (nextAction == null)
return;
_actionLinks.AddAfter(current, nextAction);
//Now reset the current action
_currentAction = nextAction;
}
If you want to compile the code. I've added in my Constructor and BuildData method I'm using to test this:
Constructor:
public LinkListTest()
{
_actionLinks = new LinkedList<Action>();
_actions = new List<Action>();
BuildData();
//Just set current to the latest action id
_currentAction = _actions.First(i => i.Id == 6);
//Add it to the linkedlist
_actionLinks.AddFirst(_currentAction);
//Start navigating as a user would
GoBack();
GoBack();
GoForward();
GoBack();
GoForward();
GoBack();
GoBack();
}
BuildData method:
private void BuildData()
{
for (int i = 6; i >= 0; i--)
{
var action = new Action();
action.Id = i;
if (i != 0)
action.PreviousId = i - 1;
else
action.PreviousId = -1;
action.Title = string.Format("Action {0}", i);
_actions.Add(action);
}
}
Thanks in advance!
One way to de-duplicate some of the logic here is to use the visitor pattern.
using ActionListAction = System.Action<System.Collections.Generic.LinkedList<Package.Action>, System.Collections.Generic.LinkedListNode<Package.Action> ,Package.Action>;
...
private void GoBack()
{
Move(new BackwordVisitor());
}
private void GoForward()
{
Move(new ForwardVisitor());
}
private void Move(DirectionVisitor direction)
{
var current = _actionLinks.Find(_currentAction);
if (current == null)
return;
var node = direction.Pointer(current);
if (node != null)
{
_currentAction = node.Value;
return;
}
var action = _actions.FirstOrDefault(i => direction.NextSelector(i, _currentAction));
//There are no further actions
if (action == null)
return;
direction.Add(_actionLinks, current, action);
_currentAction = action;
}
private abstract class DirectionVisitor
{
public Func<LinkedListNode<Action>, LinkedListNode<Action>> Pointer { protected set; get; }
public Func<Action, Action, bool> NextSelector { protected set; get; }
public ActionListAction Add { protected set; get; }
}
private class ForwardVisitor : DirectionVisitor
{
public Forward()
{
Pointer = n => n.Next;
NextSelector = (action, current) => action.PreviousId == current.Id;
Add = (list, current, node) => list.AddAfter(current, node);
}
}
private class BackwordVisitor : DirectionVisitor
{
public Backword()
{
Pointer = n => n.Previous;
NextSelector = (action, current) => action.Id == current.PreviousId;
Add = (list, current, node) => list.AddBefore(current, node);
}
}
Since there are only two options for moving through the list, this may be overkill for this particular scenario. Passing an enum into the Move method with the direction and using conditionals may read better.
I have two property in my class: MyCountry & MyCity. I set this class to sourceobject of a property grid. I want load cities i combo when select a country. for example I have 2 Country data:
Country1
Country2
And For Country1, I have (city data)
City11
City12
City13
And For Country2, I have (city data)
city21
City22
City23
When I change select country item in propertygrid, I want load cities of it in city item. this mean, when select Country1, display City11,City12,City13 in City item and when select Country2 Display City21,Cityy22,City23 in City Item.
How can I It ?
my class is :
public class KeywordProperties
{
[TypeConverter(typeof(CountryLocationConvertor))]
public string MyCountry { get; set; }
[TypeConverter(typeof(CityLocationConvertor))]
public string MyCity { get; set; }
}
and I use below class for load countries data for display in combo :
public override StandardValuesCollection GetStandardValues(ITypeDescriptorContext context)
{
HumanRoles Db = new HumanRoles();
List<LocationsFieldSet> Items = new List<LocationsFieldSet>();
Items = Db.LoadLocations(0);
string[] LocationItems = new string[Items.Count];
int count = 0;
foreach (LocationsFieldSet Item in Items)
{
LocationItems[count] = Item.Title;
count++;
}
return new StandardValuesCollection(LocationItems);
}
public override bool GetStandardValuesExclusive(ITypeDescriptorContext context)
{
return true;
}
}
The ITypeDescriptorContext interface provides a property called Instance
which lets you access the object to which the type descriptor request is connected.
You can use this property to determine the current value of the MyCountry property
the user selected. Depending on the value you can load the cities for this country.
Furthermore, in the setter of the MyCountry property I check whether or not the
new value is different from the old one and if this is the case I reset the MyCity property
(to not get an invalid combination of country and city).
Here is a small code sample. For the sake of simplicity I only use one type converter
for both properties.
public class KeywordProperties
{
public KeywordProperties()
{
MyCountry = "Country1";
}
private string myCountry;
[TypeConverter(typeof(ObjectNameConverter))]
public string MyCountry
{
get { return myCountry; }
set
{
if (value != myCountry)
MyCity = "";
myCountry = value;
}
}
private string myCity;
[TypeConverter(typeof(ObjectNameConverter))]
public string MyCity
{
get { return myCity; }
set { myCity = value; }
}
}
public class ObjectNameConverter : StringConverter
{
public override bool GetStandardValuesSupported(ITypeDescriptorContext context)
{
return true;
}
public override TypeConverter.StandardValuesCollection GetStandardValues(ITypeDescriptorContext context)
{
KeywordProperties myKeywordProps = context.Instance as KeywordProperties;
if (context.PropertyDescriptor.Name == "MyCountry")
{
List<string> listOfCountries = new List<string>();
listOfCountries.Add("Country1");
listOfCountries.Add("Country2");
return new StandardValuesCollection(listOfCountries);
}
List<string> listOfCities = new List<string>();
if (myKeywordProps.MyCountry == "Country1")
{
listOfCities.Add("City11");
listOfCities.Add("City12");
listOfCities.Add("City13");
}
else
{
listOfCities.Add("City21");
listOfCities.Add("City22");
listOfCities.Add("City23");
}
return new StandardValuesCollection(listOfCities);
}
}
In the example above there is one side effect I do not like.
Setting the MyCountry property leads to settting also the MyCity property.
To workaround this side effect you could also use the PropertyValueChanged event
of the PropertyGrid to handle invalid country/city selections.
private void propertyGrid1_PropertyValueChanged(object s, PropertyValueChangedEventArgs e)
{
if (e.ChangedItem.Label == "MyCountry")
{
if(e.ChangedItem.Value != e.OldValue)
m.MyCity = "";
}
}
If you use this event, just repalce the code in the setter of the MyCountry property with:
myCountry = value;