Uri uri = new Uri("ms-appx:///assets/data.json");

var storageFile = await Windows.Storage.StorageFile.GetFileFromApplicationUriAsync(uri);

using (var storageStream = await storageFile.OpenReadAsync())
{
    using (Stream stream = storageStream.AsStreamForRead())
    {
        using (StreamReader reader = new StreamReader(stream))
        {
            var jsonText = reader.ReadToEnd();
            var results = JsonConvert.DeserializeObject<DemoData>(jsonText);
            this.DataContext = results;
        }
    }
}

I’ve got a file called data.json stored in an Assets folder:

And the file’s Build Action is set to Content:

Change the value passed to the Uri constructor and … the file will be loaded successfully. Once loaded, I used the strongly typed generic DeserializeObject method to convert the Json data to a class called DemoData.

public static class DbExtensions
{
    /// <summary>
    /// Determines whether the table exists
    /// </summary>
    /// <param name="connection">Existing, opened, database connection</param>
    /// <param name="tableName">The name of the table to test for.</param>
    /// <returns>True if table exists.</returns>
    public static bool TableExists(this IDbConnection connection, string tableName)
    {
        Debug.Assert(connection != null);
        Debug.Assert(!string.IsNullOrWhiteSpace(tableName));

        var cmd = connection.CreateCommand();
        cmd.CommandText = @"SELECT COUNT(*) FROM sqlite_master WHERE name=@TableName";
        var p1 = cmd.CreateParameter();
        p1.DbType = DbType.String;
        p1.ParameterName = "TableName";
        p1.Value = tableName;
        cmd.Parameters.Add(p1);

        var result = cmd.ExecuteScalar();
        return ((long)result) == 1;
    }
}

Nothing too complex … only determining that the list of tables is stored in a system table named sqlite_master. If you want to find all indexes that have dependencies on a specific table, you can instead use the column tbl_name.

If you’re using Dapper and the DapperExtensions, you might find this useful:

_connection = new System.Data.SQLite.SQLiteConnection();
DapperExtensions.DapperExtensions.SqlDialect = new DapperExtensions.Sql.SqliteDialect();
_connection.ConnectionString = new DbConnectionStringBuilder()
{
    {"Data Source", "thumbnailer.db"},
    {"Version", "3"},
    {"FailIfMissing", "False"},
}.ConnectionString;

First, it creates the SQLite connection (nuget) and then configures the DapperExtensions to use the SqliteDialect. What that means is that the extensions will generate SQL that actually works! :) Finally, in my case, I wanted the DB to be created if it didn’t exist, so I set FailIfMissing to False.

While the overall syntax of C++ lacks the refined and designed elegance of a C# application, it has come a very long way!

I’ve thrown together a simple sample. The code creates a list of strings, pushes (not add, arrgh!) strings onto the list, and then iterates through the list in three similar, but syntactically different methods below:

OK, there’s a few awesome things here that make C++ much more approachable these days:

1. a decent string class! No more hunting for a library or header file that has most of the functionality you need. (there’s a wstring as well for wchar_t needs)
2. All the common Data structure types like list, map, vector, etc… that you might need without writing them your self.
3. With using, the code doesn’t need to refer to types with namespaces (much like C#). So, instead of std::list<std:string>, it’s just list<string>, which I consider far more readable at a glance.
4. No weird casting or worrying about strings when values are passed to the push_back function. Again, it just works.
5. auto – the var of C# is called auto in C++. If I hadn’t used auto in the first iterator, the code would have declared the type as list<string>::iterator. While not terrible, auto reads well, as it’s likely that I’ll not need the detailed declaration to understand what’s being iterated upon.
6. Second option is just syntactical sugar as it internally uses a standard for loop like in the first example internally. But, you’ll note there are anonymous functions/lambdas! The square [] brackets is the capture clause of the lambda expression. Unlike C#, where the compiler automatically determines what outer scoped variables will be used by the inner lambda function, in C++, it’s necessary to explicitly declare what is required. While this is a bit annoying, there are times where this extra declaration might cause programmers to think twice about all of the variables that are required in the lambda expression. In this instance, there aren’t any variables needed, so it’s empty.
7. The last example is the most concise, and maybe a little less friendly at first and that’s mostly due to the heritage of C++ and how to make code most efficient. It’s called a range-based for statement. First, the code is using auto again and the type is a string as it’s declaring the type used within the list. The & symbol remember is a reference and the const is indicating that the value will not be changed by the inner block. These two together ideally make it so the value is observed in-place, without any need for a copy.

While the range-based for statement isn’t quite as readable as C#:

I think you might agree that the C++ version is more than passable, and nearly friendly!

(Aside, there is a shortcut initializer list syntax that can be used with non-aggregate types in C++, so an int for example that would have gotten rid of the calls to push_back).

So, I’m going to show you the pattern I applied as a replacement for the LINQ queries when I need to use a projection.

Given the following simple LINQ statement:

var query = 
    (from r in DataLayer.Database
         .GetCollection<Research>()
         .AsQueryable<Research>()
        where !r.Deleted
        select new
        {
            Id = r.Id,
            Title = r.Title,
            Created = r.Created
        }).Skip(PageSize * page).Take(PageSize);

it can be converted to a MongoCursor style search like this:

var cursor = DataLayer.Database.GetCollection<Research>()
    .FindAs<Research>(Query<Research>.NE(r => r.Deleted, true))
        .SetFields(
            Fields<Research>.Include(
                r => r.Id, 
                r => r.Title, 
                r => r.Created))
        .SetLimit(PageSize)
        .SetSkip(PageSize * page);           

I’ve attempted to format it in a similar way mostly for my own sanity. As you see, both queries first get access to the database collection. But, instead of using AsQueryable<T>, you’ll use the FindAs<T> method. The query is more verbose in the second example, although not overly so. I chose to keep it strongly typed by using the generic version Query<Research>. By doing so, it meant that I could use an Expression to set the field/property that was being queried, rather than rely on a string (I could have just passed “Deleted” as a string in the code).

By strongly typing the parameters in this way, it meant that the compile process can catch things like type mismatches (verifying that the value being compared is a Boolean for example as is the Deleted property).

Secondly, and this addresses the requirement of a result projection, I’ve included just those fields that are required by the UI rather than all of the fields in the document. One of the fields is potentially quite long (up to 1MB of text), and in this case, unnecessary in a summary list display in my web application. Here, I used the SetFields method of the MongoCursor.

The C# driver includes a static class called Fields (in a generic and non-generic form) which can be used to express the set of fields to be included/excluded. I’ll point out that there is an option to just pass in a list of strings to SetFields, but it’s not strongly typed. So again, no compile-time checking that I’ve got the property names correct. I’m going for safety here, so I chose the strongly-typed generic implementation of Fields<Research>. Then, using the Expression syntax again, I’ve selected the fields I needed as a parameter list.

Finally, I added some code to limit the result size and set the skip equivalent to the original LINQ query.

There are a number of other Query methods that you can use to build more complex operations.

For example:

var q = Query.And(
    Query<Research>.Exists(r => r.Title), 
    Query<Research>.Matches(
        r => r.Title, BsonRegularExpression.Create(new Regex("^R"))));

The above maps to the following MongoDB query:

{ "$and" : [{ "Title" : { "$exists" : true } }, { "Title" : /^R/ }] }

Title field exists and the Title field starts with an uppercase “R”.

While the Query style syntax is more verbose than the equivalent LINQ statement, the result still is expressive and very readable and maintainable.

FYI: If there’s an index on Title, then the /^R/ syntax returns the results the most efficiently in MongoDB (as it stops searching after the first character).

Take for example this query:

var query = 
    (from r in DataLayer.Database.GetCollection<Research>().AsQueryable<Research>()
        where !r.Deleted
        select new
        {
            Id = r.Id,
            Title = r.Title,
            Created = r.Created
        }).Skip(PageSize * page).Take(PageSize);

query.DebugWriteMongoQueryText();

I wanted to verify that the query was checking the Deleted property, and see if the projection was considered as part of the query, so a few lines of code later …

public static string ToMongoQueryText<TQueryable>(this IQueryable<TQueryable> query)
{
    return (query as MongoQueryable<TQueryable>).GetMongoQuery().ToString();            
}

[Conditional("DEBUG")]
public static void DebugWriteMongoQueryText<TQuerable>(this IQueryable<TQuerable> query)
{
    Debug.WriteLine("QUERY: " + query.ToMongoQueryText());
}

I added two extension methods to IQueryable. As the return type for the query in my sample is actually an anonymous type, it wasn’t something I could easily type in the Immediate Window in Visual Studio for example.

No thanks  <>f__AnonymousType2<string,string,System.DateTime>.

I added the Conditional Attribute so that the code would only execute in a Debug build of my application.

query.DebugWriteMongoQueryText();

The output of the query above was:

QUERY: { "Deleted" : { "$ne" : true } }

As you’ll see, the anonymous projection wasn’t considered. Unfortunately, that’s a known issue with the current driver. This mattered in my above example, as some of the fields could contain a large amount of data that wasn’t needed by the current display.

You might try FindName to discover it doesn’t work. That’s because it’s not recursive.

So, I created a simple extension method to do the same thing:

public static class FrameworkElementExtensions
{
    public static FrameworkElement FindDescendantByName(this FrameworkElement element, string name)
    {
        if (element == null || string.IsNullOrWhiteSpace(name)) { return null; }

        if (name.Equals(element.Name, StringComparison.OrdinalIgnoreCase))
        {
            return element;
        }
        var childCount = VisualTreeHelper.GetChildrenCount(element);
        for (int i = 0; i < childCount; i++)
        {
            var result = (VisualTreeHelper.GetChild(element, i) as FrameworkElement).FindDescendantByName(name);
            if (result != null) { return result; }
        }
        return null;
    }
}

The code above loops through the children and attempts to match by name.

A simple usage (admittedly, this is stupid and clunky as lists are often virtualized, but …)

private void Page_Loaded_1(object sender, Windows.UI.Xaml.RoutedEventArgs e)
{
    for (int i = 0; i < myPeeps.Items.Count; i++)
    {
        var element = myPeeps.ItemContainerGenerator.ContainerFromIndex(i) as FrameworkElement;
        if (element != null)
        {
            var tb = element.FindDescendantByName("tbValue") as TextBlock;
            if (tb != null)
            {
                tb.Text = i.ToString();
            }
        }
    }
}

Given this DataTemplate and Page:

<Page.Resources>
    <local:SampleDataSource x:Key="sampleData" />
    <DataTemplate x:Key="SampleDataTemplate">
        <Grid >
            <Grid.RowDefinitions>
                <RowDefinition Height="Auto"/>
                <RowDefinition Height="Auto"/>
            </Grid.RowDefinitions>
            <TextBlock HorizontalAlignment="Left" Text="{Binding Name}" />
            <TextBlock Grid.Row="1" Name="tbValue" />
        </Grid>
    </DataTemplate>
</Page.Resources>

<Grid Background="{StaticResource ApplicationPageBackgroundThemeBrush}" 
      DataContext="{StaticResource sampleData}">
    <ItemsControl x:Name="myPeeps"
        ItemsSource = "{Binding Persons}"
        ItemTemplate="{StaticResource SampleDataTemplate}" FontSize="22"/>
</Grid>

And this sample data:

public class Person {
    public string Name { get; set; }        
}    

public class SampleDataSource {
    public IEnumerable<Person> Persons { get; private set; }
    
    
    public SampleDataSource () {
        var list = new List<Person>();
        
        list.Add(new Person { Name = "Alex" });
        list.Add(new Person { Name = "Betsy" });
        list.Add(new Person { Name = "Carla" });
        list.Add(new Person { Name = "Donald" });
        list.Add(new Person { Name = "Erica" });
        list.Add(new Person { Name = "Francis" });            
        
        this.Persons = list;
    }            
}

You’ll get results like this:

static void Main(string[] args)
{            
    var allProcesses = Process.GetProcesses().Where(p =>
    {
        try
        {
            var pid = p.Id;
            var modules = p.Modules.Cast<ProcessModule>()
                .Where(m => m.FileName.Contains("System.Runtime.WindowsRuntime"));
            return modules.Count() > 0;
        }
        catch { return false; }
    }).OrderBy(p => p.MainModule.FileName).ToList();
    
    for (int i = 0; i < allProcesses.Count(); i++)
    {
        var p = allProcesses[i];
        Console.WriteLine(string.Format("{0}. {1}", i, p.MainModule.FileName));
    }
    Console.ReadKey();
}

Modern UI / Metro applications are protected and cannot be easily interrogated by a non-administrative process. While you can get some basics about all processes, a standard user process isn’t allowed to look at the loaded modules for example.

In the code above, all processes are scanned for a particular DLL. In this case, System.Runtime.WindowsRuntime. I’m not 100% confident this is the best choice … there may be a few better options (or multiple that are required). (If you know of them, please leave a comment!). It did find the Modern UI / Metro applications running in my Windows 8 VM.

Once gathered, the code just outputs the basics to the console. (The name of the host, which is WWAHost.exe apparently some times).

Next step is to learn something useful via the process object.

I’ve added code like below to solve the problem:

string appName = "";
try
{
    appName = Path.GetFileName(Assembly.GetEntryAssembly().Location);
    const string IE_EMULATION = @"Software\Microsoft\Internet Explorer\Main\FeatureControl\FEATURE_BROWSER_EMULATION";

    using (var fbeKey = Registry.CurrentUser.OpenSubKey(IE_EMULATION, true))

    { 

        fbeKey.SetValue(appName, 9000, RegistryValueKind.DWord);
 
    }
}
catch(Exception ex)
{
    MessageBox.Show(appName + "\n" + ex.ToString(), "Unexpected error setting browser mode!");
}

There’s a special registry key that must be set before IE9 loads and is navigated … it must be set to the name of your application. Oddly, not the full path to your application, just the name of your executable.

You may want to delete this registry setting when your application exits.

WPF for example, could do something like this:

<DataTemplate DataType="{x:Type local:Task}">
  <StackPanel>
    <TextBlock Text="{Binding Path=TaskName}" />
    <TextBlock Text="{Binding Path=Description}"/>
    <TextBlock Text="{Binding Path=Priority}"/>
  </StackPanel>
</DataTemplate>

When the Task type as shown above was found in a list, it would have been rendered as a StackPanel with three TextBlocks automatically. That rocked.

WinRT (Metro/Xaml, Windows 8 applications) are missing the DataType property of DataTemplates. Yes, some of you might say it’s not missing as it’s V1, but given the heritage of Windows 8 Xaml applications, I consider it missing.

While an exact duplicate of the functionality isn’t possible, it’s relatively simple to get close.

<GridView
    x:Name="itemGridView"
    ItemsSource="{Binding Source={StaticResource groupedItemsViewSource}}"
    ItemTemplateSelector="{StaticResource typedTemplateSelector}"
    SelectionMode="None"
    IsItemClickEnabled="True">

Take the GridView above for example (using the template project in Visual Studio 2012).

I’ve assigned the ItemTemplateSelector to an instance of the TypedTemplateSelector class I’ve created.

In the Resources for the Page, I added a custom DataTemplate:

I’ve added a DataTemplate with a Key called Type : SampleDataItem (without the spaces though – they’re auto converted by my WordPress theme to a squiggle face: ).

There’s nothing special about the Template, just the name. It must start with Type:.

Here the custom template selector is being constructed in the Resources:

    <local:TypedTemplateSelector x:Key="typedTemplateSelector"
                                 DefaultTemplateKey="Standard250x250ItemTemplate" />
</Page.Resources>

Here’s the class that you’d need below. It has a caching feature (IsCacheEnabled) to prevent the search from occurring more than once for a key. It’s set to True by default. When used, the object searches from the current Item through all Parents trying to match the type of the object (via the VisualTreeHelper). Only the ClassName is used as programmed below (and not the full namespace). You could easily change this behavior by removing the Split and Last calls.

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using Windows.UI.Xaml;
using Windows.UI.Xaml.Controls;
using Windows.UI.Xaml.Media;

namespace WiredPrairie.TemplateSelector
{
    public class TypedTemplateSelector : DataTemplateSelector
    {
        private Dictionary<string, DataTemplate> _cachedDataTemplates;

        /// <summary>
        /// Fallback value for DataTemplate
        /// </summary>
        public string DefaultTemplateKey { get; set; }

        /// <summary>
        /// Cache search results for a type (defaults to Enabled)
        /// </summary>
        public bool IsCacheEnabled { get; set; }

        public TypedTemplateSelector()
        {
            IsCacheEnabled = true;
        }

        protected override Windows.UI.Xaml.DataTemplate SelectTemplateCore(object item, Windows.UI.Xaml.DependencyObject container)
        {
            // grab the Type name. Type will be searched as Type:NAME as shown below
            /*
                <DataTemplate x:Key="Type:SampleDataItem">
                    <Grid HorizontalAlignment="Left" Width="250" Height="250">
                        <TextBlock Text="{Binding Title}" />
                    </Grid>
                </DataTemplate>
             */
            string key = item != null ? string.Format("Type:{0}", item.GetType().Name.Split('.').Last()) : DefaultTemplateKey;
            DataTemplate dt = GetCachedDataTemplate(key);
            try
            {
                if (dt != null) { return dt; }

                // look at all parents (visual parents)
                FrameworkElement fe = container as FrameworkElement;
                while (fe != null)
                {
                    dt = FindTemplate(fe, key);
                    if (dt != null) { return dt; }
                    // if you were to just look at logical parents,
                    // you'd find that there isn't a Parent for Items set
                    fe = VisualTreeHelper.GetParent(fe) as FrameworkElement;
                }

                dt = FindTemplate(null, key);
                return dt;
            }
            finally
            {
                if (dt != null)
                {
                    AddCachedDataTemplate(key, dt);
                }
            }
        }

        private DataTemplate GetCachedDataTemplate(string key)
        {
            if (!IsCacheEnabled) { return null; }
            VerifyCachedDataTemplateStorage();
            if (_cachedDataTemplates.ContainsKey(key))
            {
                return _cachedDataTemplates[key];
            }

            return null;
        }

        private void AddCachedDataTemplate(string key, DataTemplate dt)
        {
            if (!IsCacheEnabled) { return; }
            VerifyCachedDataTemplateStorage();
            _cachedDataTemplates[key] = dt;
        }

        /// <summary>
        /// Delay creates storage
        /// </summary>
        private void VerifyCachedDataTemplateStorage()
        {
            if (_cachedDataTemplates == null)
            {
                _cachedDataTemplates = new Dictionary<string, DataTemplate>();
            }

        }

        /// <summary>
        /// Returns a template
        /// </summary>
        /// <param name="source">Pass null to search entire app</param>
        /// <param name="key"></param>
        /// <returns></returns>
        private static DataTemplate FindTemplate(object source, string key)
        {
            var fe = source as FrameworkElement;
            object obj;
            ResourceDictionary rd = fe != null ? fe.Resources : App.Current.Resources;
            if (rd.TryGetValue(key, out obj))
            {
                DataTemplate dt = obj as DataTemplate;
                if (dt != null)
                {
                    return dt;
                }
            }
            return null;

        }
    }
}

It may have looked something like this:

<TextBox Text="{Binding Path=Value, UpdateSourceTrigger=PropertyChanged}" />

The key feature was the live updating of the property through the binding. For example, as the user typed in the TextBox above, the property Value in the bound object would have been updated as the user typed. The default behavior in Silverlight 5 and WPF is to use LostFocus: the Value is only updated when the TextBox loses focus. Sometimes, that’s OK.

In WinRT/Metro however, this option was completely removed, and no decent replacement was provided unfortunately. In searching for a reasonable work-around, I discovered a few other related missing pieces from WinRT/XAML:

• BindingOptions.GetBinding: This allowed the WPF programmer a method for retrieving a Binding object. I was interested in this option as I wanted to clone the original Binding, make it two-way, and attach to the TextChanged event so that a new binding could be updated when the text changed, with the new binding pointing to the same property as the Text property. (My solution uses the basic idea as this).
• FrameworkElement.GetBindingExpression: Again, gone. Same basic logic as above. (Probably uses same underlying support that just isn’t there).
• XamlWriter.Save: XamlWriter is completely gone. I thought maybe I’d save an instance of the TextBox to Xaml, and create a new one with adjusted bindings from the Xaml. XamlReader does exist.

You can start to feel how Xaml really is just a thin wrapper on the Windows 8 run time as you start to look around to see what remains from WPF. Even traditional Silverlight features are gone unfortunately. Maybe Windows 9?

So, I took a round-about approach to solving this problem.

 public class UpdateSourceHelper : FrameworkElement
 {
     public static string GetUpdateSourceText(DependencyObject obj)
     {
         return (string)obj.GetValue(UpdateSourceTextProperty);
     }

     public static void SetUpdateSourceText(DependencyObject obj, string value)
     {
         obj.SetValue(UpdateSourceTextProperty, value);
     }

     // Using a DependencyProperty as the backing store for UpdateSourceText.  This enables animation, styling, binding, etc...
     public static readonly DependencyProperty UpdateSourceTextProperty =
         DependencyProperty.RegisterAttached("UpdateSourceText", typeof(string), typeof(UpdateSourceHelper), new PropertyMetadata(""));
     
     public static bool GetIsEnabled(DependencyObject obj)
     {
         return (bool)obj.GetValue(IsEnabledProperty);
     }

     public static void SetIsEnabled(DependencyObject obj, bool value)
     {
         obj.SetValue(IsEnabledProperty, value);
     }

     // Using a DependencyProperty as the backing store for IsEnabled.  This enables animation, styling, binding, etc...
     public static readonly DependencyProperty IsEnabledProperty =
         DependencyProperty.RegisterAttached("IsEnabled", typeof(bool), typeof(UpdateSourceHelper), 
             new PropertyMetadata(false, 
             // property changed
             (obj, args)=> 
             {                    
                 if (obj is TextBox)
                 {
                     TextBox tb = (TextBox)obj;
                     if ((bool)args.NewValue)
                     {                                                                                      
                         tb.TextChanged += AttachedTextBoxTextChanged;
                     }
                     else
                     {
                         tb.TextChanged -= AttachedTextBoxTextChanged;
                     }
                 }
             }
         ));

     static void AttachedTextBoxTextChanged(object sender, TextChangedEventArgs e)
     {
         if (sender is TextBox)
         {
             var tb = (TextBox)sender;
             tb.SetValue(UpdateSourceHelper.UpdateSourceTextProperty, tb.Text);                
         }
     }        
 }

And here it is in use:

xmlns:local="using:WiredPrairie.Converter"

(The above namespace is needed to provide context to the Xaml parser for the attached properties. Change it to whatever you need).

<TextBox Height="Auto" Margin="0,6" Grid.Row="1" TextWrapping="Wrap" TabIndex="0" 
         Text="{Binding Value}" 
         local:UpdateSourceHelper.IsEnabled="True" 
         local:UpdateSourceHelper.UpdateSourceText="{Binding Value, Mode=TwoWay}"/>

A classic pattern in WPF for extending built in controls without subclassing was to create an attached behavior. Thankfully, this pattern still works.

The core concept is to attach to the original instance of the TextBox by using an enabling property. In this case, an attached property  called IsEnabled is added to the TextBox.

When the value of the IsEnabled property changes from the default of false, the PropertyChanged callback is called. It’s here that the simple magic happens. As long as the attached property is attached to an instance of TextBox, it wires up to the TextChanged event.

In the TextChanged event handler, the value of the attached TextBox (Text property) is set into the second binding’s value. By doing this, it then sets the datasource’s object’s bound property to the new value immediately.

The annoying part about this solution really is the extra binding that is required. As there isn’t a documented/known API for reading an existing binding, this technique requires that the developer specify the binding twice.

local:UpdateSourceHelper.UpdateSourceText="{Binding Value, Mode=TwoWay}"

In WPF, it would have been possible to declare the UpdateSourceText as defaulting to TwoWay binding. But, again, that feature has been removed. Without the TwoWay binding mode, the SetValue call in the TextChanged event handler will not update the object’s bound property. It’s one way.

I looked low and high for a more elegant solution, but couldn’t find one. If you find something better, please leave a comment!