The appearance of the list is as expected;

The only code behind required is in initialising the selected items. Note the data required for populating the CheckedList is a generic Dictionary<int, string> which is set in the DataContext. A second Dictionary<int, string> allows for getting and setting the checked items.

The XAML is as follows. The ControlTemplate is responsible for setting the background of the selected items to transparent. The ListBox’s SelectionMode is Multiple. The text displayed for each checkbox (content property) is from the bound Dictionary’s Value property. The CheckBox’s IsChecked property binds to the ListBoxItem’s IsSelected property. Note the ListBoxItem is implicitly created when the ListBox binds to its data.

1. The first (and preferred) is to use LINQ queries with the CSLA LinqBindingList to enable filtering of the underlying collections. The LinqBindingList acts as a view over a business objects collection so binding UI elements to a LinqBindingList is really the same as binding to the actual business objects collection. Though when I initialised the LinqBindingList using its constructor and set it as the datasource for XamDataGrid a databinding exception occurred (attempting to access an item outside the bounds of an array).
2. The second option was the CSLA FilteredBindingList which is described as obsolete (replaced by LinqBindingList). This also acts as a view over the underlying business objects collection with filtering based on properties/values. This worked well in the attempts I made to use it.
3. The third option was to use the XamDataGrid’s built in filtering. This is quite sophisticated but after some initial investigation I found that this functionality is not available for the Express edition.

An important part of the GoalBook application is to explore and implement new technology, so I was determined to use option 1. After some binging (and googling) I found a post on the CSLA forum suggesting to look at the CSLA unit testing solution for examples of initialising the LinqBindingList. So I downloaded and looked into this and found that the initialisation was always being done by casting the result of a LINQ query.

GoalBook currently uses FilteredBindingList as its implementation for filtering and search.

Or in C#;

This will implicitly apply a margin of zero to each paragraph when it is created.

Also, In the Xaml declaration for RichTextBox set IsDocumentEnabled = “True”.

Initially I looked at the .NET Web Browser Control with its Html editing capability (through MSHTML). Then I considered the Wpf Rich TextBox (which uses FlowDocument). Both of these options have thrown up design challenges.

The Web Browser Control provides full compatibility with the Toodledo notes format, but is an interop solution. It also requires the use of a template to format the printed output. While the Wpf Rich TextBox provides only partial compatibility (ordered and bulleted lists are different in Html) with the Toodledo notes format and requires a considerable amount of parsing and conversion of the notes text.

My preference has been to use the Wpf Rich TextBox because of its use of FlowDocument (which is compatible with the printing functionality in GoalBook) and because it is a pure Wpf solution.

The FlowDocumentPaginator creates and returns DocumentPage instances with the header and footer located at predefined offsets.

To achieve this user defined Folder ordering in the Infragistics XamDatagrid it is necessary to implement a custom SortComparer and GroupByComparer because the XamDatagrid will sort the Folders alphabetically by default.

The SortComparer and GroupByComparer plug into the XamDatagrid’s Field settings and are implemented using the standard IComparer interface. When each Folder name is presented for sorting, it’s corresponding order field is obtained and the comparison is performed using this numeric value rather than the Folder name.

The result that I can see is the GoalBook code is cleaner and more professional and that is the result I was after. StyleCop is like a mentor constantly reviewing and pointing out where improvements are required and when used in a team provides an independent agreed coding standard.