Array slices in C#

Question

How do you do it? Given a byte array:

byte[] foo = new byte[4096];

How would I get the first x bytes of the array as a separate array? (Specifically, I need it as an IEnumerable<byte>)

This is for working with Sockets. I figure the easiest way would be array slicing, similar to Perls syntax:

@bar = @foo[0..40];

Which would return the first 41 elements into the @bar array. Is there something in C# that I'm just missing, or is there some other thing I should be doing?

LINQ is an option for me (.NET 3.5), if that helps any.

Answer 1

You could use ArraySegment<T>. It's very light-weight as it doesn't copy the array:

string[] a = { "one", "two", "three", "four", "five" };
var segment = new ArraySegment<string>( a, 1, 2 );

Answer 2

Arrays are enumerable, so your foo already is an IEnumerable<byte> itself. Simply use LINQ sequence methods like Take() to get what you want out of it (don't forget to include the Linq namespace with using System.Linq;):

byte[] foo = new byte[4096];

var bar = foo.Take(41);

If you really need an array from any IEnumerable<byte> value, you could use the ToArray() method for that. That does not seem to be the case here.

Answer 3

You could use the arrays CopyTo() method.

Or with LINQ you can use Skip() and Take()...

byte[] arr = {1, 2, 3, 4, 5, 6, 7, 8};
var subset = arr.Skip(2).Take(2);

Answer 4

Starting from C# 8.0/.Net Core 3.0

Array slicing will be supported, along with the new types Index and Range being added.

Range Struct docs
Index Struct docs

Index i1 = 3;  // number 3 from beginning
Index i2 = ^4; // number 4 from end
int[] a = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 };
Console.WriteLine($"{a[i1]}, {a[i2]}"); // "3, 6"

var slice = a[i1..i2]; // { 3, 4, 5 }

Above code sample taken from the C# 8.0 blog.

note that the ^ prefix indicates counting from the end of the array. As shown in the docs example

var words = new string[]
{
                // index from start    index from end
    "The",      // 0                   ^9
    "quick",    // 1                   ^8
    "brown",    // 2                   ^7
    "fox",      // 3                   ^6
    "jumped",   // 4                   ^5
    "over",     // 5                   ^4
    "the",      // 6                   ^3
    "lazy",     // 7                   ^2
    "dog"       // 8                   ^1
};              // 9 (or words.Length) ^0

Range and Index also work outside of slicing arrays, for example with loops

Range range = 1..4; 
foreach (var name in names[range])

Will loop through the entries 1 through 4

---

_{note that at the time of writing this answer, C# 8.0 is not yet officially released
C# 8.x and .Net Core 3.x are now available in Visual Studio 2019 and onwards}

Answer 5

static byte[] SliceMe(byte[] source, int length)
{
    byte[] destfoo = new byte[length];
    Array.Copy(source, 0, destfoo, 0, length);
    return destfoo;
}

//

var myslice = SliceMe(sourcearray,41);

Answer 6

In C# 7.2, you can use Span<T>. The benefit of the new System.Memory system is that it doesn't need copying around data.

The method you need is Slice:

Span<byte> slice = foo.Slice(0, 40);

A lot of methods now support Span and IReadOnlySpan, so it will be very straightforward to use this new type.

Note that at the time of writing the Span<T> type is not defined in the the most recent version of .NET yet (4.7.1) so to use it you need to install the System.Memory package from NuGet.

Answer 7

C# 8 now (since 2019) supports Ranges which allows you to achieve Slice much easier (similar to JS syntax):

var array = new int[] { 1, 2, 3, 4, 5 };
var slice1 = array[2..^3];    // array[new Range(2, new Index(3, fromEnd: true))]
var slice2 = array[..^3];     // array[Range.EndAt(new Index(3, fromEnd: true))]
var slice3 = array[2..];      // array[Range.StartAt(2)]
var slice4 = array[..];       // array[Range.All]

You can use ranges instead of the well known LINQ functions: Skip(), Take(), Count().

Answer 8

Another possibility I haven't seen mentioned here: Buffer.BlockCopy() is slightly faster than Array.Copy(), and it has the added benefit of being able to convert on-the-fly from an array of primitives (say, short[]) to an array of bytes, which can be handy when you've got numeric arrays that you need to transmit over Sockets.

Answer 9

If you want IEnumerable<byte>, then just

IEnumerable<byte> data = foo.Take(x);

Answer 10

Here's a simple extension method that returns a slice as a new array:

public static T[] Slice<T>(this T[] arr, uint indexFrom, uint indexTo) {
    if (indexFrom > indexTo) {
        throw new ArgumentOutOfRangeException("indexFrom is bigger than indexTo!");
    }

    uint length = indexTo - indexFrom;
    T[] result = new T[length];
    Array.Copy(arr, indexFrom, result, 0, length);

    return result;
}

Then you can use it as:

byte[] slice = foo.Slice(0, 40);

Answer 11

If you don't want to add LINQ or other extensions just do:

float[] subArray = new List<float>(myArray).GetRange(0, 8).ToArray();

Answer 12

For byte arrays System.Buffer.BlockCopy will give you the very best performance.

Answer 13

byte[] foo = new byte[4096]; 

byte[] bar = foo.Take(40).ToArray();

Answer 14

You could use a wrapper around the original array (which is IList), like in this (untested) piece of code.

public class SubList<T> : IList<T>
{
    #region Fields

    private readonly int startIndex;
    private readonly int endIndex;
    private readonly int count;
    private readonly IList<T> source;

    #endregion

    public SubList(IList<T> source, int startIndex, int count)
    {
        this.source = source;
        this.startIndex = startIndex;
        this.count = count;
        this.endIndex = this.startIndex + this.count - 1;
    }

    #region IList<T> Members

    public int IndexOf(T item)
    {
        if (item != null)
        {
            for (int i = this.startIndex; i <= this.endIndex; i++)
            {
                if (item.Equals(this.source[i]))
                    return i;
            }
        }
        else
        {
            for (int i = this.startIndex; i <= this.endIndex; i++)
            {
                if (this.source[i] == null)
                    return i;
            }
        }
        return -1;
    }

    public void Insert(int index, T item)
    {
        throw new NotSupportedException();
    }

    public void RemoveAt(int index)
    {
        throw new NotSupportedException();
    }

    public T this[int index]
    {
        get
        {
            if (index >= 0 && index < this.count)
                return this.source[index + this.startIndex];
            else
                throw new IndexOutOfRangeException("index");
        }
        set
        {
            if (index >= 0 && index < this.count)
                this.source[index + this.startIndex] = value;
            else
                throw new IndexOutOfRangeException("index");
        }
    }

    #endregion

    #region ICollection<T> Members

    public void Add(T item)
    {
        throw new NotSupportedException();
    }

    public void Clear()
    {
        throw new NotSupportedException();
    }

    public bool Contains(T item)
    {
        return this.IndexOf(item) >= 0;
    }

    public void CopyTo(T[] array, int arrayIndex)
    {
        for (int i=0; i<this.count; i++)
        {
            array[arrayIndex + i] = this.source[i + this.startIndex];
        }
    }

    public int Count
    {
        get { return this.count; }
    }

    public bool IsReadOnly
    {
        get { return true; }
    }

    public bool Remove(T item)
    {
        throw new NotSupportedException();
    }

    #endregion

    #region IEnumerable<T> Members

    public IEnumerator<T> GetEnumerator()
    {
        for (int i = this.startIndex; i < this.endIndex; i++)
        {
            yield return this.source[i];
        }
    }

    #endregion

    #region IEnumerable Members

    IEnumerator IEnumerable.GetEnumerator()
    {
        return GetEnumerator();
    }

    #endregion
}

Answer 15

You can use Take extension method

var array = new byte[] {1, 2, 3, 4};
var firstTwoItems = array.Take(2);

Answer 16

This may be a solution that:

var result = foo.Slice(40, int.MaxValue);

Then the result is an IEnumerable< IEnumerable< byte>> with a first IEnumerable< byte> contains the first 40 bytes of foo, and a second IEnumerable< byte> holds the rest.

I wrote a wrapper class, the whole iteration is lazy, hope it could help:

public static class CollectionSlicer
{
    public static IEnumerable<IEnumerable<T>> Slice<T>(this IEnumerable<T> source, params int[] steps)
    {
        if (!steps.Any(step => step != 0))
        {
            throw new InvalidOperationException("Can't slice a collection with step length 0.");
        }
        return new Slicer<T>(source.GetEnumerator(), steps).Slice();
    }
}

public sealed class Slicer<T>
{
    public Slicer(IEnumerator<T> iterator, int[] steps)
    {
        _iterator = iterator;
        _steps = steps;
        _index = 0;
        _currentStep = 0;
        _isHasNext = true;
    }

    public int Index
    {
        get { return _index; }
    }

    public IEnumerable<IEnumerable<T>> Slice()
    {
        var length = _steps.Length;
        var index = 1;
        var step = 0;

        for (var i = 0; _isHasNext; ++i)
        {
            if (i < length)
            {
                step = _steps[i];
                _currentStep = step - 1;
            }

            while (_index < index && _isHasNext)
            {
                _isHasNext = MoveNext();
            }

            if (_isHasNext)
            {
                yield return SliceInternal();
                index += step;
            }
        }
    }

    private IEnumerable<T> SliceInternal()
    {
        if (_currentStep == -1) yield break;
        yield return _iterator.Current;

        for (var count = 0; count < _currentStep && _isHasNext; ++count)
        {
            _isHasNext = MoveNext();

            if (_isHasNext)
            {
                yield return _iterator.Current;
            }
        }
    }

    private bool MoveNext()
    {
        ++_index;
        return _iterator.MoveNext();
    }

    private readonly IEnumerator<T> _iterator;
    private readonly int[] _steps;
    private volatile bool _isHasNext;
    private volatile int _currentStep;
    private volatile int _index;
}

Answer 17

I do not think C# supports the Range semantics. You could write an extension method though, like:

public static IEnumerator<Byte> Range(this byte[] array, int start, int end);

But like others have said if you do not need to set a start index then Take is all you need.

Answer 18

Here is an extension function that uses a generic and behaves like the PHP function array_slice. Negative offset and length are allowed.

public static class Extensions
{
    public static T[] Slice<T>(this T[] arr, int offset, int length)
    {
        int start, end;

        // Determine start index, handling negative offset.
        if (offset < 0)
            start = arr.Length + offset;
        else
            start = offset;

        // Clamp start index to the bounds of the input array.
        if (start < 0)
            start = 0;
        else if (start > arr.Length)
            start = arr.Length;

        // Determine end index, handling negative length.
        if (length < 0)
            end = arr.Length + length;
        else
            end = start + length;

        // Clamp end index to the bounds of the input array.
        if (end < 0)
            end = 0;
        if (end > arr.Length)
            end = arr.Length;

        // Get the array slice.
        int len = end - start;
        T[] result = new T[len];
        for (int i = 0; i < len; i++)
        {
            result[i] = arr[start + i];
        }
        return result;
    }
}

Answer 19

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

namespace data_seniens
{
    class Program
    {
        static void Main(string[] args)
        {
            //new list
            float [] x=new float[]{11.25f,18.0f,20.0f,10.75f,9.50f, 11.25f, 18.0f, 20.0f, 10.75f, 9.50f };

            //variable
            float eat_sleep_area=x[1]+x[3];
            //print
            foreach (var VARIABLE in x)
            {
                if (VARIABLE < x[7])
                {
                    Console.WriteLine(VARIABLE);
                }
            }



            //keep app run
        Console.ReadLine();
        }
    }
}

Answer 20

Often when reading from streams you want to handle reading less that requested.

Array.Resize is your friend in this situation.