How To Write Elegant Code with C# Pattern Matching
Anton Martyniuk
Posted on July 30, 2024
Introduced in C# 8.0 Pattern Matching, provides an elegant way to write more expressive and concise code.
Pattern Matching is evolving in every new C# version.
In this post, you will explore how to use Pattern Matching and how to write elegant code in C# by using Pattern Matching.
On my website: antondevtips.com I share .NET and Architecture best practices.
Subscribe to become a better developer.
What Is Pattern Matching
Pattern matching in C# is used to check a value against a pattern.
It can be used in various scenarios, such as type checking, deconstructing tuples, and working with different data structures.
Pattern matching in C# includes the following components:
- Type Patterns
- Type Patterns with Nullable Types
- Constant Patterns
- Property Patterns
- Positional Patterns
- List Patterns
Type Patterns
Type patterns are used to check if a value is of a specific type.
They can be used with the is
keyword.
You can use this pattern to assign a matched type to a variable:
public void Process(object obj)
{
if (obj is string s)
{
Console.WriteLine($"String: {s}");
}
else if (obj is int i)
{
Console.WriteLine($"Integer: {i}");
}
}
In this example, we check if obj
is a string
or an integer
and handle each case accordingly.
Type Patterns with Nullable Types
Type patterns can also be used with nullable types.
You can use the is
and is not
keywords to check if a value is null
or not:
public void CheckForNull(object? obj)
{
if (obj is null)
{
Console.WriteLine("Object is null.");
}
else if (obj is not null)
{
Console.WriteLine("Object is not null.");
}
}
You can use the is
keyword to check if a value is of a specific type, including nullable types:
public void ProcessNullable(int? number)
{
if (number is int value)
{
Console.WriteLine($"Number is {value}.");
}
else
{
Console.WriteLine("Number is null.");
}
}
Constant Patterns
Constant patterns are used to compare a value with a constant.
Together with switch expressions - pattern matching looks very elegant:
public string GetDayType(int day)
{
return day switch
{
0 => "Sunday",
1 => "Monday",
2 => "Tuesday",
3 => "Wednesday",
4 => "Thursday",
5 => "Friday",
6 => "Saturday",
_ => "Invalid day"
};
}
Here, we use a switch expression with constant patterns to return the name of the day based on the integer value.
Range Constant Patterns
You can use and
, or
pattern combinators to create range comparisons for numbers:
public string EvaluateNumber(int number)
{
return number switch
{
> 0 and <= 10 => "Number is between 1 and 10",
> 10 and <= 20 => "Number is between 11 and 20",
> 20 and <= 30 => "Number is between 21 and 30",
_ => "Number is out of range"
};
}
public string CheckSpecialNumber(int number)
{
return number switch
{
5 or 23 or 42 => "Special number",
_ => "Regular number"
};
}
You can use not
to exclude a range of numbers:
public string ExcludeNumber(int number)
{
return number switch
{
not (>= 10 and <= 20) => "Number is not between 10 and 20",
_ => "Number is between 10 and 20"
};
}
Such pattern matching is not limited to only numbers, you can also use it to check characters:
public string EvaluateLetter(char letter)
{
return letter switch
{
>= 'A' and <= 'Z' => "Uppercase letter",
>= 'a' and <= 'z' => "Lowercase letter",
_ => "Not a letter"
};
}
public string CheckVowel(char letter)
{
return letter switch
{
'A' or 'E' or 'I' or 'O' or 'U' or 'a' or 'e' or 'i' or 'o' or 'u' => "Vowel",
_ => "Consonant"
};
}
You can use not
to exclude a range of characters:
public string ExcludeSpecialCharacter(char character)
{
return character switch
{
not (>= '0' and <= '9') and not (>= 'A' and <= 'Z') and not (>= 'a' and <= 'z') => "Special character",
_ => "Alphanumeric character"
};
}
Enum Constant Patterns
You can use pattern matching on enums. Let's define a OrderStatus
enum:
public enum OrderStatus
{
Pending,
Processing,
Shipped,
Delivered,
Cancelled,
Returned
}
You can use and
, or
, not
for enums as well:
public string GetOrderStatusMessage(OrderStatus status)
{
return status switch
{
OrderStatus.Pending or OrderStatus.Processing => "Order is in progress.",
OrderStatus.Shipped or OrderStatus.Delivered => "Order is on its way or has been delivered.",
OrderStatus.Cancelled or OrderStatus.Returned => "Order has been cancelled or returned.",
_ => "Unknown order status."
};
}
You can negate enum patterns by using not
.
You can check if an order is finished in the following way:
public bool IsOrderFinished(OrderStatus status)
{
return status switch
{
not (OrderStatus.Delivered or OrderStatus.Cancelled or OrderStatus.Returned) => true,
_ => false
};
}
When using not
with or
, it's important to understand how the logic is applied.
The not
operator negates the entire expression that follows it.
If you need to check if an order is not Shipped
and not Delivered
, put parentheses for all expressions that should be negated with not
:
not (OrderStatus.Shipped or OrderStatus.Delivered)
If you miss the parentheses, the following interpretation will be incorrect:
not OrderStatus.Shipped or OrderStatus.Delivered
Because it will check if an order is not Shipped
or order is Delivered
, instead of checking "is not" for both of the options.
Property Patterns
Property patterns are used to match an object based on its properties.
public record Person(string Name, int Age);
public void DisplayPersonInfo(Person person)
{
if (person is { Name: "Anton", Age: 30 })
{
Console.WriteLine("Anton is 30 years old.");
}
}
In this example, we check if the Person
object has the name "Anton" and age 30.
Let's explore another code example that checks if a person is not null, has a dedicated Name and Age:
public void DisplayPersonInfo(Person? person)
{
if (person != null && person.Name == "Anton" && person.Age == 30)
{
Console.WriteLine("Anton is 30 years old.");
}
}
With pattern matching, you can make this code more concise and readable:
public void DisplayPersonInfo(Person? person)
{
if (person is { Name: "Anton", Age: 30 })
{
Console.WriteLine("Anton is 30 years old.");
}
}
In case you need to use the person object that matches the expression, you can assign a variable in pattern matching:
public void DisplayPersonInfo(Person? person)
{
if (person is { Name: "Anton", Age: 30 } p)
{
Console.WriteLine($"{p.Name} is {p.Age} years old.");
}
}
Here, you can use p
variable to access Name
and Age
properties of a Person.
You can also use negate expressions in pattern matching to simplify the code.
Instead of using a few logical expressions to match a person who is not "Anton" or is null:
public void DisplayPersonInfo(Person? person)
{
if (person is null || (person.Name != "Anton" && person.Age != 30))
{
Console.WriteLine("User is not found.");
}
}
You can use is not
expression that will match all persons who are not "Anton", including null persons:
public void DisplayPersonInfo(Person? person)
{
if (person is not { Name: "Anton", Age: 30 })
{
// Either user is null or not Anton
Console.WriteLine("User is not found.");
}
}
You can also check for null
and not null
using empty property pattern matching:
public void CheckForNull(object? obj)
{
if (obj is {})
{
Console.WriteLine("Object is not null.");
}
else if (obj is not {})
{
Console.WriteLine("Object is null.");
}
}
Positional Patterns
Positional patterns are used to deconstruct a value into its parts and match them.
For example, we can deconstruct a Point object into its x
and y
coordinates and handle different cases:
public readonly struct Point
{
public int X { get; }
public int Y { get; }
public Point(int x, int y) => (X, Y) = (x, y);
public void Deconstruct(out int x, out int y) => (x, y) = (X, Y);
}
public void DisplayPointInfo(Point point)
{
if (point is (0, 0))
{
Console.WriteLine("Point is at the origin.");
}
else if (point is (int x, int y))
{
Console.WriteLine($"Point is at ({x}, {y}).");
}
}
List Patterns
List patterns, introduced in C# 11, enable pattern matching on sequences.
For example, you can match a list of integers against different patterns:
public void AnalyzeList(List<int> numbers)
{
switch (numbers)
{
case []:
Console.WriteLine("The list is empty.");
break;
case [1, 2, 3]:
Console.WriteLine("The list contains 1, 2, 3.");
break;
case [var first, var second, ..]:
Console.WriteLine($"The list starts with {first} and {second}.");
break;
default:
Console.WriteLine("The list has a different pattern.");
break;
}
}
Pattern Matching On Multiple Values
You can use Pattern Matching for multiple values in switch expressions.
For example, calculating count of days for a shipment to be delivered, based on destination and weight:
public int CalculateDeliveryDays(string destination, decimal weight)
{
return (destination, weight) switch
{
("Local", <= 1.0m) => 1,
("Local", <= 5.0m) => 2,
("Local", > 5.0m) => 3,
("International", <= 1.0m) => 5,
("International", <= 5.0m) => 7,
("International", > 5.0m) => 10,
_ => 10 // Default case for unspecified scenarios
};
}
Pattern Matching with Switch Expressions and When Clauses
Pattern Matching with switch expressions become even more powerful when combined with when
clauses.
The when
keyword allows you to add additional conditions to each case, enabling more control over your logic.
Let's explore an example involving calculation of a discount based on category, price and a member vs. non-member client:
public decimal CalculateDiscount(string category, decimal price, bool isMember)
{
return (category, price) switch
{
// 5% discount on electronics under $100
("Electronics", <= 100.0m) when !isMember => price * 0.05m,
// 10% discount on electronics under $500
("Electronics", <= 500.0m) when !isMember => price * 0.10m,
// 15% discount on electronics over $500
("Electronics", > 500.0m) when !isMember => price * 0.15m,
// 7% discount for members on electronics under $100
("Electronics", > 100.0m) when isMember => price * 0.07m,
// 12% discount for members on electronics under $500
("Electronics", <= 500.0m) when isMember => price * 0.12m,
// 18% discount for members on electronics over $500
("Electronics", > 500.0m) when isMember => price * 0.18m,
_ => 0.0m // No discount for unspecified scenarios
};
}
Here a member gets an additional discount when compared to a regular customer.
With a when
clause you can even discard all the patterns and write the logical expression to make the pattern matching:
public enum MemberType
{
Bronze,
Silver,
Gold
}
public decimal CalculateDiscount(string category, decimal price, MemberType memberType)
{
return (category, price) switch
{
// 3% discount for Bronze members on electronics
_ when memberType is MemberType.Bronze && category == "Electronics" => price * 0.03m,
// 5% discount for Silver members on electronics
_ when memberType is MemberType.Silver && category == "Electronics" => price * 0.05m,
// 8% discount for Gold members on electronics
_ when memberType is MemberType.Gold && category == "Electronics" => price * 0.08m,
_ => 0.0m
};
}
One of my favourite examples of how elegant code you can write with pattern matching and switch expressions is to identify the current temperature:
public string ClassifyTemperatureImproved(int temperature)
{
return temperature switch
{
< 0 => "Freezing",
>= 0 and < 10 => "Cold",
>= 10 and < 20 => "Cool",
>= 20 and < 30 => "Warm",
>= 30 => "Hot"
};
}
Summary
Pattern Matching together with switch expressions is a powerful addition to C#, allowing developers to write cleaner, more concise code.
You can reduce the complexity of your conditional logic, making your code more readable and less error-prone.
Whether you are checking number ranges, converting enum values, calculating discounts, or handling complex logic, pattern matching helps you write more elegant code.
On my website: antondevtips.com I share .NET and Architecture best practices.
Subscribe to become a better developer.
Posted on July 30, 2024
Join Our Newsletter. No Spam, Only the good stuff.
Sign up to receive the latest update from our blog.