Thinktecture.Runtime.Extensions.Json 8.4.0

This library provides some interfaces, classes, Roslyn Source Generators, Roslyn Analyzers and Roslyn CodeFixes for implementation of Smart Enums, Value Objects and Discriminated Unions.

• Requirements
• Migrations
• Smart Enum
• Value Objects
• Discriminated Unions
  • Ad hoc unions
  • Regular unions

Documentation

See wiki for more documentation.

Requirements

• Version 8:
  • C# 11 (or higher) for generated code
  • SDK 8.0.400 (or higher) for building projects
• Version 7:
  • C# 11 (or higher) for generated code
  • SDK 7.0.401 (or higher) for building projects

Migrations

• Migration from v7 to v8
• Migration from v6 to v7

Ideas and real-world use cases

Smart Enums:

• Shipping Method
• CSV-Importer-Type

Value objects:

• Open-ended End Date
• (Always-positive) Amount

Smart Enums

Smart Enums provide a powerful alternative to traditional C# enums, offering type-safety, extensibility, and rich behavior. Unlike regular C# enums which are limited to numeric values and lack extensibility, Smart Enums can:

• Use any type as the underlying type (e.g., strings, integers) or none at all
• Include additional fields, properties and behavior
• Use polymorphism to define custom behavior for each value
• Prevent creation of invalid values
• Integrate seamlessly with JSON serializers, MessagePack, Entity Framework Core and ASP.NET Core

Install: Install-Package Thinktecture.Runtime.Extensions

Documentation: Smart Enums

Some of the Key Features are:

• Choice between always-valid and maybe-valid Smart Enum
• Reflection-free iteration over all items
• Fast lookup/conversion from underlying type to Smart Enum and vice versa
• Allows custom properties and methods
• Exhaustive pattern matching with Switch/Map methods
• Provides appropriate constructor, based on the specified properties/fields
• Proper implementation of Equals, GetHashCode, ToString and equality operators
• Provides implementation of IComparable, IComparable<T>, IFormattable, IParsable<T> and comparison operators <, <=, >, >= (if applicable to the underlying type)
• Custom comparer and equality comparer

Roslyn Analyzers and CodeFixes help the developers to implement the Smart Enums correctly

Provides support for:

• JSON (System.Text.Json and Newtonsoft)
• Minimal Api Parameter Binding and ASP.NET Core Model Binding
• Entity Framework Core
• MessagePack

Definition of a Smart Enum with custom properties and methods.

[SmartEnum<string>]
public partial class ShippingMethod
{
   public static readonly ShippingMethod Standard = new(
      "STANDARD",
      basePrice: 5.99m,
      weightMultiplier: 0.5m,
      estimatedDays: 5,
      requiresSignature: false);

   public static readonly ShippingMethod Express = new(
      "EXPRESS",
      basePrice: 15.99m,
      weightMultiplier: 0.75m,
      estimatedDays: 2,
      requiresSignature: true);

   public static readonly ShippingMethod NextDay = new(
      "NEXT_DAY",
      basePrice: 29.99m,
      weightMultiplier: 1.0m,
      estimatedDays: 1,
      requiresSignature: true);

   private readonly decimal _basePrice;
   private readonly decimal _weightMultiplier;
   private readonly int _estimatedDays;

   public bool RequiresSignature { get; }

   public decimal CalculatePrice(decimal orderWeight)
   {
      return _basePrice + (orderWeight * _weightMultiplier);
   }

   public DateTime GetEstimatedDeliveryDate()
   {
      return DateTime.Today.AddDays(_estimatedDays);
   }
}

Behind the scenes a Roslyn Source Generator generates additional code. Some of the features that are now available are ...

Basic Operations

[SmartEnum<string>]
public partial class ProductType
{
    // The source generator creates a private constructor
    public static readonly ProductType Groceries = new("Groceries");
}

// Enumeration over all defined items
IReadOnlyList<ProductType> allTypes = ProductType.Items;

// Value retrieval
ProductType productType = ProductType.Get("Groceries");        // Get by key (throws if not found)
ProductType productType = (ProductType)"Groceries";            // Same as above but by using a cast
bool found = ProductType.TryGet("Groceries", out var productType);  // Safe retrieval (returns false if not found)

// Validation with detailed error information
ValidationError? error = ProductType.Validate("Groceries", null, out ProductType? productType);

// IParsable<T> (useful for Minimal APIs)
bool parsed = ProductType.TryParse("Groceries", null, out ProductType? parsedType);

// IFormattable (e.g. for numeric keys)
string formatted = ProductGroup.Fruits.ToString("000", CultureInfo.InvariantCulture);  // "001"

// IComparable
int comparison = ProductGroup.Fruits.CompareTo(ProductGroup.Vegetables);
bool isGreater = ProductGroup.Fruits > ProductGroup.Vegetables;  // Comparison operators

Type Conversion and Equality

// Implicit conversion to key type
string key = ProductType.Groceries;  // Returns "Groceries"

// Equality comparison
bool equal = ProductType.Groceries.Equals(ProductType.Groceries);
bool equal = ProductType.Groceries == ProductType.Groceries;  // Operator overloading
bool notEqual = ProductType.Groceries != ProductType.Housewares;

// Methods inherited from Object
int hashCode = ProductType.Groceries.GetHashCode();
string key = ProductType.Groceries.ToString();  // Returns "Groceries"

// TypeConverter
var converter = TypeDescriptor.GetConverter(typeof(ProductType));
string? keyStr = (string?)converter.ConvertTo(ProductType.Groceries, typeof(string));
ProductType? converted = (ProductType?)converter.ConvertFrom("Groceries");

Pattern Matching with Switch/Map

All Switch/Map methods are exhaustive by default ensuring all cases are handled correctly.

ProductType productType = ProductType.Groceries;

// Execute different actions based on the enum value (void return)
productType.Switch(
    groceries: () => Console.WriteLine("Processing groceries order"),
    housewares: () => Console.WriteLine("Processing housewares order")
);

// Transform enum values into different types
string department = productType.Switch(
    groceries: () => "Food and Beverages",
    housewares: () => "Home and Kitchen"
);

// Direct mapping to values - clean and concise
decimal discount = productType.Map(
    groceries: 0.05m,    // 5% off groceries
    housewares: 0.10m    // 10% off housewares
);

For optimal performance Smart Enums provide overloads that prevent closures.

ILogger logger = ...;

// Prevent closures by passing the parameter as first method argument
productType.Switch(logger,
    groceries: static l => l.LogInformation("Processing groceries order"),
    housewares: static l => l.LogInformation("Processing housewares order")
);

// Use a tuple to pass multiple values
var context = (Logger: logger, OrderId: "123");

productType.Switch(context,
    groceries: static ctx => ctx.Logger.LogInformation("Processing groceries order {OrderId}", ctx.OrderId),
    housewares: static ctx => ctx.Logger.LogInformation("Processing housewares order {OrderId}", ctx.OrderId)
);

Value Objects

Install: Install-Package Thinktecture.Runtime.Extensions

Documentation: Value Objects

Value objects help solve several common problems in software development:

1. Type Safety: Prevent mixing up different concepts that share the same primitive type

   // Problem: Easy to accidentally swap parameters
   void ProcessOrder(int customerId, int orderId) { ... }
   ProcessOrder(orderId, customerId); // Compiles but wrong!
   
   // Solution: Value objects make it type-safe
   [ValueObject<int>]
   public partial struct CustomerId { }
   
   [ValueObject<int>]
   public partial struct OrderId { }
   
   void ProcessOrder(CustomerId customerId, OrderId orderId) { ... }
   ProcessOrder(orderId, customerId); // Won't compile!
   

2. Built-in Validation: Ensure data consistency at creation time

   [ValueObject<decimal>]
   public partial struct Amount
   {
       static partial void ValidateFactoryArguments(ref ValidationError? validationError, ref decimal value)
       {
           if (value < 0)
           {
               validationError = new ValidationError("Amount cannot be negative");
               return;
           }
           // Normalize to two decimal places
           value = Math.Round(value, 2);
       }
   }
   
   var amount = Amount.Create(100.50m);     // Success: 100.50
   var invalid = Amount.Create(-50m);       // Throws ValidationException
   

3. Immutability: Prevent accidental modifications and ensure thread safety

4. Complex Value Objects: Encapsulate multiple related values with validation

   [ComplexValueObject]
   public sealed partial class DateRange
   {
       public DateOnly Start { get; }
       public DateOnly End { get; }
   
       static partial void ValidateFactoryArguments(
           ref ValidationError? validationError,
           ref DateOnly start,
           ref DateOnly end)
       {
           if (end < start)
           {
               validationError = new ValidationError(
                   $"End date '{end}' cannot be before start date '{start}'");
               return;
           }
   
           // Ensure dates are not in the past
           var today = DateOnly.FromDateTime(DateTime.Today);
           if (start < today)
           {
               validationError = new ValidationError("Start date cannot be in the past");
               return;
           }
       }
   
       public int DurationInDays => End.DayNumber - Start.DayNumber + 1;
   
       public bool Contains(DateOnly date) => date >= Start && date <= End;
   }
   
   // Usage
   var range = DateRange.Create(
       start: DateOnly.FromDateTime(DateTime.Today),
       end: DateOnly.FromDateTime(DateTime.Today.AddDays(7))
   );
   
   Console.WriteLine(range.DurationInDays);    // 8
   Console.WriteLine(range.Contains(range.Start));  // true
   

Key Features:

• Two types of value objects:
  • Simple value objects (wrapper around a single value with validation)
  • Complex value objects (multiple properties representing a single concept)
• Comprehensive validation support with descriptive error messages
• Framework integration:
  • JSON serialization (System.Text.Json and Newtonsoft.Json)
  • Entity Framework Core support
  • ASP.NET Core Model Binding
  • MessagePack serialization
• Rich feature set:
  • Type conversion and comparison operators
  • Custom equality comparison
  • Proper implementation of standard interfaces (IComparable, IFormattable, etc.)
  • Configurable null and empty string handling
• Development support:
  • Roslyn Analyzers and CodeFixes for correct implementation
  • Logging for debugging and insights

For more examples and detailed documentation, see the wiki.

Discriminated Unions

Install: Install-Package Thinktecture.Runtime.Extensions

Documentation: Discriminated Unions

Discriminated unions are a powerful feature that allows a type to hold a value that could be one of several different types. They provide type safety, exhaustive pattern matching, and elegant handling of complex domain scenarios. Key benefits include:

• Type-safe representation of values that can be one of several types
• Exhaustive pattern matching ensuring all cases are handled
• Elegant modeling of domain concepts with multiple states
• Clean handling of success/failure scenarios without exceptions

The library provides two types of unions to suit different needs:

Ad hoc unions

Perfect for simple scenarios where you need to combine a few types quickly. Features:

• Type-safe combination of up to 5 different types
• Implicit conversions and type checking
• Exhaustive pattern matching with Switch/Map methods
• Built-in equality comparison
• Support for class, struct, or ref struct implementations

// Quick combination of types
[Union<string, int>]
public partial class TextOrNumber;

// Create and use the union
TextOrNumber value = "Hello";           // Implicit conversion
TextOrNumber number = 42;               // Works with any defined type

// Type-safe access
if (value.IsString)
{
    string text = value.AsString;       // Type-safe access
    Console.WriteLine(text);
}

// Exhaustive pattern matching
var result = value.Switch(
    @string: text => $"Text: {text}",
    int32: num => $"Number: {num}"
);

// Custom property names for clarity
[Union<string, int>(T1Name = "Text", T2Name = "Number")]
public partial class BetterNamed;
// Now use .IsText, .IsNumber, .AsText, .AsNumber

Regular unions

Ideal for modeling domain concepts and complex hierarchies. Features:

• Inheritance-based approach for complex scenarios
• Support for both classes and records
• Integration with value objects
• Generic type support
• Exhaustive pattern matching

Perfect for modeling domain concepts:

// Model domain concepts clearly
[Union]
public partial record OrderStatus
{
    public record Pending : OrderStatus;
    public record Processing(DateTime StartedAt) : OrderStatus;
    public record Completed(DateTime CompletedAt, string TrackingNumber) : OrderStatus;
    public record Cancelled(string Reason) : OrderStatus;
}

// Generic result type for error handling
[Union]
public partial record Result<T>
{
    public record Success(T Value) : Result<T>;
    public record Failure(string Error) : Result<T>;

    // Convenient implicit conversions
    public static implicit operator Result<T>(T value) => new Success(value);
    public static implicit operator Result<T>(string error) => new Failure(error);
}

// Usage
Result<int> result = await GetDataAsync();
var message = result.Switch(
    success: s => $"Got value: {s.Value}",
    failure: f => $"Error: {f.Error}"
);