C# - Raise the bar with functional & immutable constructs (Dutch)
Download as PPTX, PDF2 likes1,239 views
The document discusses the advantages of using functional and immutable programming constructs in C#, highlighting concepts like tuples, functions, and partial application. It emphasizes the importance of immutability for easier reasoning and performance benefits in concurrent programming. Numerous examples and GitHub links are provided to illustrate these concepts and implementations.
C# - Raise the bar with functional & immutable constructs (Dutch)
- 2. C# Raise the bar with Functional & Immutable constructs
- 3. O, Kan dat ook?
- 4. Today • Func is cool • Tuple is your friend • Immutable is a life saver • Practice makes perfect • Action!
- 5. Action public delegate void Action<in T1, in T2>(T1 arg1, T2 arg2);
- 6. Func public delegate TResult Func<in T1, in T2, out TResult>(T1 arg1, T2 arg2);
- 7. Tuple public class Tuple<T1, T2> : IStructuralEquatable, IStructuralComparable, IComparable, Ituple
- 9. Functions • Data in – Data out • Func in – Data out • Data in – Func out • Func in – Func out
- 10. Functions: Data in – Data out x => 2 * x
- 11. Functions: Func in – Data out • LINQ: Where + more • Example: Object initialize • GitHub: https://github.com/NForza/Functional/tree/ma ster/Samples/ObjectInitialization • Example: Aggregate • GitHub: https://github.com/NForza/Functional/tree/ma ster/Samples/Aggregate
- 12. Functions: Data in – Func out x => y => x + y
- 13. Functions: Func in – Func out • Example: Partial application • Example: Memoize
- 14. Partial Application var f = (x, y) => x + y; var p = x => y => f(x, y);
- 16. Memoization • Pure functions • Long running operations
- 17. Memoization Cache Now: • Unlimited size, unlimited time Ideas: • Limited size, unlimited time • Unlimited size, limited time • Don’t use the cache unless real function takes > 1 sec.
- 19. Abstractions • 1 or n • for • Is Index needed? • Is List/IList the correct abstraction (do we need to change the list?) • foreach • What & How are together • I hear functional abstraction • Return: IEnumerable, IDictionary, IGroup
- 20. Abstraction Examples • Interactive Extensions • NuGet: https://www.nuget.org/packages/Ix-Main/ • NForza Extensions • GitHub: https://github.com/NForza/Functional • NuGet: Soon! • Weighted Average • GitHub: https://github.com/NForza/Functional/blob/master /Sources/NForza.Functional/EnumerableExtensions. cs#L563-601
- 21. Miscellaneous Examples • Defer (make Lazy) • GitHub: https://github.com/NForza/Functional/tree/ma ster/Samples/Defer • Bindings • GitHub: https://github.com/rickbeerendonk/Binding • NuGet: https://www.nuget.org/packages/Beerendonk.B inding
- 22. Immutable: Why? • Easier reasoning • Equality Has Meaning (not in .NET’s immutable colleactions) • Sharing = Easy, no defensive copying required! No ToArray() / ToList() • Concurrent Programming: No locks!
- 23. Immutable? • Int • String • Tuple • IEnumerable • IReadOnlyCollection/List/Dictionary • IList
- 24. Immutable Collections • ImmutableList<T> • ImmutableDictionary<TKey, TValue> • ImmutableSortedDictionary<TKey, TValue> • ImmutableHashSet<T> • ImmutableSortedSet<T> • ImmutableStack<T> • ImmutableQueue<T>
- 25. Immutable Collections • Demo: https://github.com/rickbeerendonk/Immuta bleDemo • NuGet: https://msdn.microsoft.com/en- us/library/dn769092(v=vs.110).aspx
- 26. Immutable Coll.: Performance Mutable (amortized) Mutable (worst case) Immutable Stack.Push O(1) O(n) O(1) Queue.Enqueue e O(1) O(n) O(1) List.Add O(1) O(n) O(log n) HashSet.Add O(1) O(n) O(log n) SortedSet.Add O(log n) O(n) O(log n) Dictionary.Add O(1) O(n) O(log n) SortedDictionar y.Add O(log n) O(n log n) O(log n)
- 27. Immutable Coll.: Memory • Worse: Per Item • Better: No reserved space • Better: Immediate shrink after delete
- 28. Why? • Less code • Reusable • More fun!