The Feature Flags Pattern

I was listening to Episode 1101 of the podcast Dot Net Rocks. Jez Humble was talking about the concept of Feature Toggles or Feature Flags. Feature Switches? While the term has some opinionated definitions, the concept that I found most interesting was the idea of deploying software with the new features initially disabled, or switched off by some mechanism. After you think the feature is ready for production, switch on the feature. If there is an issue, you don't have to roll back the version or deploy another release, just switch the feature back off again and replace its .dll. Didn't get enough debug information? Switch the feature back on and let a few crash reports trickle in and then switch it back off. If you feature is particularly processor or network intensive, you can perform load testing by slowly releasing the features to select clients or only part of the population/userbase.

While I personally choose an SQL table for my approach to storing the toggle switches (internal business app), one could use the application's .config file. An application could pull the settings from a .config file on a networked drive as a way of controlling multiple application instances by making one changed to a centralized location. Below I show a minimalist implementation by creating a Dictionary from the <appSettings> in a App.config file.

Behold:

public static Dictionary<string, bool> GetFeatureFlags()
{
  return ConfigurationManager.AppSettings.AllKeys.ToDictionary(s => s, IsFlagSet);
}

public static bool IsFlagSet(string settingName)
{
  bool result = false;
  bool.TryParse(ConfigurationManager.AppSettings[settingName], out result);
  return result;
}

Of course with a dictionary you have to be careful that you don't try an access a column that does not exist with the indexer, so you might be better off using IsFlagSet(string), which will never throw. Although this is of limited use (AppSettings is already a NameValueCollecion), perhaps you can make use of this generic function I wrote that uses generics to convert AppSettings into a dictionary with the type of the value being the generic:

public static Dictionary<string, T> GetDictionary<T>()
{
 return ConfigurationManager.AppSettings.AllKeys.ToDictionary<string, string, T>(s => s, GetSetting<T>);
}

public static T GetSetting<T>(string settingName)
{
  try
  {
    T result = (T) Convert.ChangeType(
      ConfigurationManager.AppSettings[settingName],
      typeof (T)
    );
    if (result != null)
    {
      return result;
    }
  }
  catch { }

  return default(T);
}

Please note that swallowing an exception ("catch { }") is typically considered poor form. In this particular scenario, I am aware of the possible exceptions that can be thrown by this code and I want the code to return the default(T) in this scenario and never throw.

EditLabel UserControl

EditLabel is editable Label control, who's text can be modified at run-time by double-clicking on it. Under the hood, it contains a Label control and TextBox control. It works like a Label control until you double-click it with the mouse. When double-clicked it enters 'edit-mode' where a TextBox overlays the Label. The control will AutoSize to the characters you type in the TextBox, updating the Label's size at the same time. You can leave the control's edit mode by hitting either enter, the escape key, or letting the control lose focus. After editing, the control acts like a label again, with the updated text. You can check out the code for the EditLabel project here on my GitHub.

The EditLabel isn't too interesting by itself, but I wanted to blog about it for all the interesting attributes that I never knew about until I started getting into designing controls, and controlling how controls are displayed in the designer window in visual studio.

The first thing my EditLabel control needed was and event that the parent form could subscribe to know when the text changed. Creating the event simply isn't enough. If you want the event to show up in the Visual Studio's Form Designer when you click on events, you need to add the Browsable and EditorBrowsable attribute. This will make is show up in the designer. You can control the description that appears as well as what category it shows up under when viewing events/properties in the Categorize view as opposed to Alphabetical view by using the Description and Category attributes respectively:

[Category("Property Changed")]
[Description("Event raised when the value of the Text property is changed on Control.")]
[Browsable(true), EditorBrowsable(EditorBrowsableState.Always)]
public new event EventHandler TextChanged;

There is also the notion of default events. That is, the event for which a handler is added when you double click the control in the designer. For example, when you double click a TextBox, it automatically adds the TextChanged event, instead of, say, the KeyDown event. To make an event the default event, add the DefaultEvent attribute at the control's class-level declaration:

[DefaultEvent("TextChanged")]
public partial class EditLabel : UserControl
{
   ...

Another thing I wanted was the Text property. Like a Label or a TextBox control, I wanted a public Text property that could be edited at design time. However, every time I re-built the project, the text would get reset to the default. As it turns out, if you want modifications to properties to be persisted, it must be serialized to the MainForm.Designer.cs, or whatever the auto-generated Designer.cs file is. In order to tell the designer to serialize changes made at design time to Designer.cs file, you have to add the DesignerSerializationVisibility attribute above the property. Here is the attributes and declaration for the Text property:

[Category("Appearance")]
[Description("The text associated with the control.")]
[Browsable(true), EditorBrowsable(EditorBrowsableState.Always)]
[DesignerSerializationVisibility(DesignerSerializationVisibility.Visible)]
public override string Text
{
   get { return ctrlLabel.Text; }
   set { ChangeText(value); }
}

You can view the full EditLabel.cs here, or view the entire EditLabel project here.

Graceful Console Exit

When writing console apps sometimes you want to delay the closing of the console window to give the user time to read the output. Often, if something went wrong, I would simply report the error to the standard output and immediately exit the application. However that did not work too well for my users; the exit message was not displayed long enough for them to read so they knew how to correct the issue. What I needed was something that give a user enough time to read the output, but not prevent the process from exiting if it was being ran as a scheduled/automated task. The solution I came up with was a countdown timer that would block execution from proceeding until it counted back from some number to zero.

Lets view the code, shall we:

public static void CountdownTimer(int Seconds = 5)
{
    Console.ResetColor();
    Console.WriteLine(Environment.NewLine);
    Console.Write("The application will now terminate...   ");

    if (Seconds < 1) return;
    if (Seconds > 9) Seconds = 9;

    for (int counter = Seconds * 10; counter > 0; counter--)
    {
        if (Console.KeyAvailable)
            return;

        if (counter % 10 == 0)
            Console.Write("\b{0}", (counter / 10));

        System.Threading.Thread.Sleep(100);
    }
}

I print each number as the timer counts down. Each time I print a new number, I place a backspace "\b" before it to erase the previous number. To avoid having to keep track how many character a digit prints to use the same number of backspaces, I just kept the function simple by limiting the number of seconds to a max of 9.
I check for Console.KeyAvailable to detect key strokes to skip the rest of the countdown and exit immediately. The reason I did 10 Sleeps of 100ms per second was to make the exit upon key press more responsive.

Validate all input parameters in one line/Check several objects for empty or null in a single method call.

Often my methods start with several guarding clauses. That is, conditional if statements that check for null or empty parameters and immediately return if they are. This is also known as defensive programming. Usually I am not concerned with this code; indeed I identify these blocks as validation code and overlook this code entirely. It was not until recently that I noticed this as an area that I was repeating myself and could be put in a reusable function.

Lets look at the code:

/// <summary>
/// Checks the parameters for empty, nulls, or invalid states.
/// </summary>
/// <returns>True if the params are null, empty, contains an array or object that is null or empty, contains a blank, whitespace, null or empty string, or contains DataTable that does not pass a call to IsValidDatatable().</returns>
public static bool ContainsNullOrEmpty(params object[] Items)
{
    if (Items == null || Items.Length < 1)
        return true;
    
    foreach (object item in Items)
    {
        if (item == null)
            return true;
        
        if (item is string)
        {
            if (string.IsNullOrWhiteSpace(item as String))
                return true;
        }
        else if (item is DataTable)
        {
            if (!IsValidDatatable(item as DataTable))
                return true;
        }
        
        if (item.GetType().IsArray)
        {
            bool isEmpty = true;
            foreach (object itm in (Array)item)
            {
                if (ContainsEmptyOrNulls(itm))
                    return true;
                
                isEmpty = false;
            }
            if (isEmpty)
                return true;
        }
    }

    return false;
}

My approach above uses the params keyword. By using params, I can pass in any number of parameters (including zero, although that doesn't help us in this context). By using an object instead of a generic type I can pass multiple different types in one method call. If I used generics, I would have to have one method call for each type that I wanted to validate.

The idea is to cram all of your common validation logic into this method, and call it everywhere to increase the readability of your business logic by not cluttering it up with validation logic. Notice how this function also checks for empty or white-space strings, as well as calling a custom IsValidDatatable(DataTable) function. If you have several functions that return an int of -1 or 0 upon failure, you might want to add another conditional to check if (item is int) and then if the value of the integer reflects an erroneous state.

Another cool feature it that if the item is an an array, or even an array of nested arrays, it will still check every item in those arrays. Notice how I get the item type, then check the Type.IsArray property boolean. If it is true, I cast the object as an System.Array, then call ContainsEmptyOrNulls() recursively in a foreach loop. We can return true right away on the first null condition met, but we must be careful not to return on a false condition and to instead let the false conditions fall through and continue on, in the case that there is a null later or in another array.

Enjoy!

EntityJustWorks - C# class/object/entity to SQL database/script/command mapper.

Note by author:

   Since writing this, I have expanded on this idea quite a bit. I have written a lightweight ORM class library that I call EntityJustWorks.
   The full project can be found on GitHub or CodePlex.


   EntityJustWorks not only goes from a class to DataTable (below), but also provides:

• Populate the public properties of a class from a DataTable

• DataTable to class as a C# code file (using CodeDOM)

• DataTable to class Assembly (using Reflection.Emit)

• C# class to a SQL 'CREATE TABLE' or 'INSERT INTO' script

• SQL 'SELECT' statement to a List<T> of populated classes, each one resembling a row

Security Warning:
This library generates dynamic SQL, and has functions that generate SQL and then immediately executes it. While it its true that all strings funnel through the function Helper.EscapeSingleQuotes, this can be defeated in various ways and only parameterized SQL should be considered SAFE. If you have no need for them, I recommend stripping semicolons ; and dashes --. Also there are some Unicode characters that can be interpreted as a single quote or may be converted to one when changing encodings. Additionally, there are Unicode characters that can crash .NET code, but mainly controls (think TextBox). You almost certainly should impose a white list:
string clean = new string(dirty.Where(c => "abcdefghijklmnopqrstuvwxyz0123456789.,\"_ !@".Contains(c)).ToArray());

PLEASE USE the SQLScript.StoredProcedure and DatabaseQuery.StoredProcedure classes to generate SQL for you, as the scripts it produces is parameterized. All of the functions can be altered to generate parameterized instead of sanitized scripts. Ever since people have started using this, I have been maintaining backwards compatibility. However, I may break this in the future, as I do not wish to teach one who is learning dangerous/bad habits. This project is a few years old, and its already showing its age. What is probably needed here is a total re-write, deprecating this version while keep it available for legacy users after slapping big warnings all over the place. This project was designed to generate the SQL scripts for standing up a database for a project, using only MY input as data. This project was never designed to process a USER'S input.! Even if the data isn't coming from an adversary, client/user/manually entered data is notoriously inconsistent. Please do not use this code on any input that did not come from you, without first implementing parameterization. Again, please see the SQLScript.StoredProcedure class for inspiration on how to do that.

    In this post, I bring the wagons 'round full circle and use what I have shown you in older posts to make a class that can create (or insert into) a table in an SQL database from a C# class object at run-time. The only prior knowledge needed would be the connection string. The name of the the table will be the name of the class Type, and the tables's columns are created from the class's public properties, with their names being the property names.

    If you want to take a data-first approach, you can generate the code to C# classes that mirrors a table in a relational database given a connection string and the table name, or use the new Emit class and generate a dynamic assembly at run-time that contains a class with public auto-properties matching the DataColumns that you can use to store DataRows in. Similarly, the name of class name will be the same name as the table, with each of the table's columns being represented by a public auto property on the class of the same name and type. An improvement here would be to create the class inside a dynamic assembly using System.Reflection.Emit. Anywho, I have been holding on to this code for too long I just want to get it out of the door.

    For those of you who just want the code:
      GitHub  (Download zip)
      BitBucket (Download)
      code.msdn.microsoft.com

    Going from a query to a DataTable is as simple as a call to SQLAdapter.Fill(). For populating the public properties of a class from a DataTable, see this post I made. To build a DataTable who's columns match the public properties of a class, check out this post here. Going from a DataTable to a class as a C# code file is covered in this previous post.

    Going from a DataTable to SQL CREATE TABLE or INSERT INTO scripts that can be passed to ExecuteNonQuery() to execute the command, see below.

    Here is the INSERT INTO code. This code takes advantage of some functions in the Helper class that converts a DataTable's ColumnNames and Row's Values as a List of String, which is then passed into a function that takes a list of string and formats it with a delimiter string and prefix and postfix strings so I can easily format the list to look like the COLUMNS and VALUES part of an SQL INSERT INTO statement. For example: INSERT INTO ([FullName], [Gender], [BirthDate]) VALUES ('John Doe', 'M', '10/3/1981''). If you want to view how the functions RowToColumnString(), RowToValueString(), and ListToDelimitedString() work, click here to view Helper.cs.

public static string InsertInto<T>(params T[] ClassObjects) where T : class
{
   DataTable table = Map.ClassToDatatable<T>(ClassObjects);
   return InsertInto(table);   // We don't need to check IsValidDatatable() because InsertInto does
}

public static string InsertInto(DataTable Table)
{
   if (!Helper.IsValidDatatable(Table))
      return string.Empty;

   StringBuilder result = new StringBuilder();
   foreach (DataRow row in Table.Rows)
   {
      if (row == null || row.Table.Columns.Count < 1 || row.ItemArray.Length < 1)
         return string.Empty;

      string columns = Helper.RowToColumnString(row);
      string values = Helper.RowToValueString(row);

      if (string.IsNullOrWhiteSpace(columns) || string.IsNullOrWhiteSpace(values))
         return string.Empty;

      result.AppendFormat("INSERT INTO [{0}] {1} VALUES {2}", Table.TableName, columns, values);
   }

   return result.ToString();
}

    Here is the CREATE TABLE CODE. A very important function here is the GetDataTypeString() member, which converts the Column.DataType to the SQL equivalent as a string, which is used to craft a CREATE TABLE script. I use static strings formatted to work with String.Format to replace the crucial bits in the CREATE TABLE command:

public static string CreateTable<T>(params T[] ClassObjects) where T : class
{
 DataTable table = Map.ClassToDatatable<T>(ClassObjects);
 return Script.CreateTable(table);
}

public static string CreateTable(DataTable Table)
{
   if (Helper.IsValidDatatable(Table, IgnoreRows: true))
      return string.Empty;

   StringBuilder result = new StringBuilder();
   result.AppendFormat("CREATE TABLE [{0}] ({1}", Table.TableName, Environment.NewLine);

   bool FirstTime = true;
   foreach (DataColumn column in Table.Columns.OfType<DataColumn>())
   {
      if (FirstTime) FirstTime = false;
      else
         result.Append(",");

      result.AppendFormat("[{0}] {1} {2}NULL{3}",
         column.ColumnName,
         GetDataTypeString(column.DataType),
         column.AllowDBNull ? "" : "NOT ",
         Environment.NewLine
      );
   }
   result.AppendFormat(") ON [PRIMARY]{0}GO", Environment.NewLine);

   return result.ToString();
}

private static string GetDataTypeString(Type DataType)
{
   switch (DataType.Name)
   {
      case "Boolean": return "[bit]";
      case "Char": return "[char]";
      case "SByte": return "[tinyint]";
      case "Int16": return "[smallint]";
      case "Int32": return "[int]";
      case "Int64": return "[bigint]";
      case "Byte": return "[tinyint] UNSIGNED";
      case "UInt16": return "[smallint] UNSIGNED";
      case "UInt32": return "[int] UNSIGNED";
      case "UInt64": return "[bigint] UNSIGNED";
      case "Single": return "[float]";
      case "Double": return "[double]";
      case "Decimal": return "[decimal]";
      case "DateTime": return "[datetime]";
      case "Guid": return "[uniqueidentifier]";
      case "Object": return "[variant]";
      case "String": return "[nvarchar](250)";
      default: return "[nvarchar](MAX)";
   }
}

...


    Here is definition of all the public members:

namespace EntityJustWorks.SQL
{
   public static class Convert
   {
      public static string SQLTableToCSharp(string ConnectionString, string TableName);
      public static bool ClassToSQLTable(string ConnectionString, params T[] ClassCollection);
   }
   
   public static class Code
   {
      public static string SQLTableToCSharp(string ConnectionString, string TableName);
      public static string DatatableToCSharp(DataTable Table);
   }

   public static class Script
   {
      public static string InsertInto<T>(params T[] ClassObjects);
      public static string InsertInto(DataTable Table);
      public static string CreateTable<T>(params T[] ClassObjects);
      public static string CreateTable(DataTable Table);
   }
   
   public static class Map
   {
      public static IList<T> DatatableToClass<T>(DataTable Table);
      public static DataTable ClassToDatatable<T>(params T[] ClassCollection);
      public static DataTable ClassToDatatable<T>();
   }

   public static class Query
   {
      public static IList<T> QueryToClass<T>(string ConnectionString, string FormatString_Query,
                                                            params object[] FormatString_Parameters);
      public static DataTable QueryToDataTable(string ConnectionString, string FormatString_Query,
                                                            params object[] FormatString_Parameters);
      public static T QueryToScalarType<T>(string ConnectionString, string FormatString_Query,
                                                            params object[] FormatString_Parameters);
      public static int ExecuteNonQuery(string ConnectionString, string FormatString_Command,
                                                            params object[] FormatString_Parameters);
   }
   
   public static class Helper
   {
      public static bool IsValidDatatable(DataTable Table, bool IgnoreRows = false);
      public static bool IsCollectionEmpty<T>(IEnumerable<T> Input);
      public static bool IsNullableType(Type Input);
   }
}

    Please feel free to comment, either on my GitHub or my blog, with criticisms or suggestions.