Giters

aqez / LINQ-notes

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

title author date
LINQ
Tony Dwire
2021-02-03
 
                _                       _  _  _           _           _             _  _  _  _
               (_)                     (_)(_)(_)         (_) _       (_)          _(_)(_)(_)(_)_
               (_)                        (_)            (_)(_)_     (_)         (_)          (_)
               (_)                        (_)            (_)  (_)_   (_)         (_)          (_)
               (_)                        (_)            (_)    (_)_ (_)         (_)     _    (_)
               (_)                        (_)            (_)      (_)(_)         (_)    (_) _ (_)
               (_) _  _  _  _           _ (_) _          (_)         (_)         (_)_  _  _(_) _
               (_)(_)(_)(_)(_)         (_)(_)(_)         (_)         (_)           (_)(_)(_)  (_)

  • What is LINQ

  • Before and After LINQ

  • Using complex LINQ

  • Writing our own LINQ-style extension methods

What is LINQ?

Linq is a collection of interfaces and methods in the System.Linq namespace that make it easy to work with collections of data.

  • LINQ stands for Language Integrated Query
  • Many extension methods for IEnumerable<T> interface
  • Also introduces some other interfaces, such as IGrouping, IQueryable, etc.
  • All of these interfaces implement IEnumerable<T> so we will stick to looking at that for this video
  • Extremely useful for working with collections - maybe the most useful collection library in existence

There are two main ways of using the LINQ functionality. One is via the query language:

var numbers = new int[] { 1, 2, 3, 4, 5 };

var odd =
    from number in numbers
    where number % 2 == 0
    select number;

Console.WriteLine(odd); // { 1, 3, 5 }

The other is via the extension methods that the System.Linq namespace provides:

var numbers = new int[] { 1, 2, 3, 4, 5 };

var odd = numbers.Where(n => n % 2 == 0);

Console.WriteLine(odd); // { 1, 3, 5 }

IEnumerable

IEnumerable<T> is a forward iterating, read only interface over a collection of T's

Example IEnumerable

int[] numbers = new int[] { 1, 2, 3, 4, 5 };

IEnumerable<int> enumerable = numbers; // This gives us a way to look at numbers in a forward iterating, read-only way

// Alternatively, we could write it like this:
IEnumerable<int> enumerable = new int[] { 1, 2, 3, 4, 5 };

Sum

This may sound limiting, but what we can do with this concept goes very deep. For instance, we can get a sum:

IEnumerable<int> enumerable = new int[] { 1, 2, 3, 4, 5 };

int sum = 0;
var enumerator = enumerable.GetEnumerator();

while(enumerator.MoveNext())
{
    sum += enumerator.Current;
}

Assert.Equal(15, sum);

LINQ methods

The extension methods in the System.Linq namespace give us a set of easy to perform, nicely wrapped common operations that we can use on collections implementing IEnumerable<T> that we do not have to write ourselves. In this, and the next few slides we will go over the more common ones and the way they let us transform our code.

As an example of what we will be looking for, they can transform our code from something like this:

For Each

IEnumerable<Person> allPeople = GetSomePeople();
List<string> fourOrMoreLetters = new List<string>();
foreach(var person in allPeople)
{
    if(person.Name.Length > 4)
    {
        fourOrMoreLetters.Add(person.Name);
    }
}

To code that looks like this:

LINQ

var fourOrMoreLetters = GetSomePeople().Select(person => person.Name).Where(name => name.Length > 4);

Where

Where is probably the most used extension method for me personally. It is used to filter the source and produce a new IEnumerable<T> that only produces elements that pass the predicate's test.

Given the following class and array:

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

var people = new Person[]
{
    new Person { Name = "Bob", Age = 30 },
    new Person { Name = "Jack", Age = 50 },
    new Person { Name = "Beth", Age = 42 }
};

Get all of the people with a name starting with "B"

Before

var startsWithB = new List<Person>();
foreach(var person in people)
{
    if(person.Name.StartsWith("B"))
    {
        startsWithB.Add(person);
    }
}

After

var startsWithB = people.Where(person => person.Name.StartsWith("B"));

Select

Select is the second most common one for me. It is 'map' from other languages. It is used to take one sequence of elements and turn it into another sequence of elements, one for each input.

Given a collection of Person, select their names

Before

var names = new List<string>();
foreach(var person in people)
{
    names.Add(person.Name);
}

After

var names = people.Select(person => person.Name);

It is also important to note that the input and output of this method do not have to be particularly related, you could return a random number for each one, for instance.

Select double the age of each Person

var doubledAges = people.Select(person => person.Age * 2); // [ 60, 100, 84 ]

Select the third letter of each Person's name

var thirdLetters = people.Select(person => person.Name[2]); // [ 'b', 'c', 't' ]

Any, All

The Any and All extension methods determine if any or all of the elements satisfy some predicate.

Determine if all the people are under age 30

Before

bool all30OrUnder = true;
foreach(var person in people)
{
    if(person.Age > 30)
    {
        all30OrUnder = false;
        break;
    }
}

After

// Using `All`
bool all30OrUnder = people.All(person => person.Age <= 30);

// Using `Any`
bool all30OrUnder = !people.Any(person => person.Age > 30);

Min, Max, Average, Sum

These extension methods provide an easy way to get the min, max, average or sum of some predicate over a collection. They have 'default predicate' implementations for the built in numeric types, and other types can specify their own predicate.

Given a list of people, get the average age

Before

int sum = 0;
int count = 0;

foreach(var person in people)
{
    sum += person.Age;
    count += 1;
}

int average = sum / count;

After

double average = people.Average(person => person.Age);

Note, for collections of numeric builtin types, you don't have to supply a predicate.

// Don't do this:
double average = (new int[] { 1, 2, 3, 4, 5 }).Average(x => x);

// Instead do this:
double average = (new int[] { 1, 2, 3, 4, 5 }).Average();

First, FirstOrDefault, Last, LastOrDefault, Count

These give simple access to the beginning and ending elements of a collection as well as counting items in a collection. They can also be given a predicate to determine the first/last/count where some condition is true.

Get the last Person in the collection where the person's name starts with 'B'

Before

Person lastStartingWithB = null;
foreach(var person in people)
{
    if(person.Name.StartsWith("B"))
    {
        lastStartingWithB = person;
    }
}

After

var last = people.LastOrDefault(person => person.Name.StartsWith("B"));

Count the people who have more than 5 letters in their name

Before

int moreThan5Letters = 0;
foreach(var person in people)
{
    if(person.Name.Length > 5)
    {
        moreThan5Letters += 1;
    }
}

After

int moreThan5Letters = people.Count(person => person.Name.Length > 5);

Cast, OfType

Cast and OfType are similar but not exactly the same. Cast will go through each item in a collection and try to cast it to the destination type, and if it fails it will throw an exception. OfType will go through each item in a collection, determine if it can be casted, and return only those that can be casted as the destination type.

Given a list of Vehicles, turn them all into Trucks

Before

List<Truck> trucks = new List<Truck>();
foreach(Vehicle vehicle in vehicles)
{
    trucks.Add((Truck)vehicle);
}

After

List<Truck> trucks = vehicles.Cast<Truck>();

Given a list of Vehicles, get the Helicopters

Before

List<Helicopter> helicopters = new List<Helicopter>();
foreach(Vehicle vehicle in vehicles)
{
    if(vehicle is Helicopter helicopter)
    {
        helicopters.Add(helicopter);
    }
}

After

List<Helicopter> = vehicles.OfType<Helicopter>();

OrderBy, OrderByDescending

OrderBy and OrderByDescending allow for easy sorting of collections with a specified predicate that returns the thing to be sorted by.

Sort a collection of Person by age

Before

List<Person> orderedByAge = new List<Person>();

foreach (var person in people)
{
    int insertIndex = 0;
    for (int i = 0; i < orderedByAge.Count; i++)
    {
        if (person.Age <= orderedByAge[i].Age)
        {
            break;
        }
        insertIndex += 1;
    }
    orderedByAge.Insert(insertIndex, person);
}

After

List<Person> orderedByAge = people.OrderBy(person => person.Age);

GroupBy

Allows easy grouping of items based on some 'key'.

Group people by the first letter of their name

Before

Dictionary<char, List<Person>> grouped = new Dictionary<char, List<Person>>();

foreach(var person in people)
{
    if(!grouped.ContainsKey(person.Name[0]))
    {
        grouped.Add(person.Name[0], new List<Person>());
    }

    grouped[person.Name[0]].Add(person);
}

After

var grouped = people.GroupBy(person => person.Name[0]);

Combining LINQ methods to get stuff done

What if we wanted to get the letter of the alphabet with the oldest (on average) people that have that letter as the first letter of their name?

Workflow would be..

  1. Group everyone by the first letter of their name
  2. Average the ages for each group
  3. Order the groups by the average age
  4. Take the first one
var people = new Person[]
{
    new Person { Name = "Tony", Age = 32 },
    new Person { Name = "Tabatha", Age = 48 },
    new Person { Name = "Bob", Age = 38 }
};

var oldestAverageLetter = people.GroupBy(person => person.Name[0])
                                .Select(group => new { Letter = group.Key, SummedAges = group.Average(person => person.Age) })
                                .OrderByDescending(group => group.AverageAge)
                                .First();

Console.WriteLine(highestLetterAndAge); // { Letter = "T", AverageAge = 40 }

Predicates as Separate Methods

The predicates we have seen so far have all been defined as lambdas. This is not the only way to do it though. Anything that can be used as a delegate matching the required type for the LINQ method can be used.

var youngEnoughPeople = people.Where(IsYoungEnough);


bool IsYoungEnough(Person person)
{
    return person.Age < 20;
}
Func<Person, bool> oldEnough = person => person.Age > 100;

var oldEnoughPeople = people.Where(oldEnough);

About