StringReader

In C# there are many ways to separate a string into lines. With StringReader, we read lines from a string individually in the order they appear.

The StringReader type enables us to access string data through a stream-oriented interface. It can replace Split() method calls, and sometimes improve performance.

Example

StringReader is a class found in the System.IO namespace. The instance constructor receives a string parameter. This is stored internally as a field in the StringReader.

using System;
using System.IO;

class Program
{
    const string _input = "Dot Net Perls\nA B\nC D";

    static void Main()
    {
        // Creates new StringReader instance from System.IO
        using (StringReader reader = new StringReader(_input))
        {
            // Loop over the lines in the string.
            int count = 0;
            string line;
            while ((line = reader.ReadLine()) != null)
            {
                count++;
                Console.WriteLine("Line {0}: {1}", count, line);
            }
        }
    }
}
Line 1: Dot Net Perls
Line 2: A B
Line 3: C D

Benchmark

How efficient is StringReader in comparison to the Split method? We ran a simple benchmark that separates into lines a string literal with several newline characters in it.

using System;
using System.Diagnostics;
using System.IO;

class Program
{
    static int Method1(string data)
    {
        int count = 0;
        using (StringReader reader = new StringReader(data))
        {
            while (true)
            {
                string line = reader.ReadLine();
                if (line == null)
                {
                    break;
                }
                count += line.Length;
            }
        }
        return count;
    }

    static int Method2(string data)
    {
        int count = 0;
        string[] lines = data.Split(new char[] { '\n' });
        foreach (string line in lines)
        {
            count += line.Length;
        }
        return count;
    }

    static void Main()
    {
        const string data = "Dot\nNet\nPerls\nA B C\nD E F";
        Console.WriteLine(Method1(data));
        Console.WriteLine(Method2(data));
        const int max = 1000000;
        // Version 1: use StringReader.
        var s1 = Stopwatch.StartNew();
        for (int i = 0; i < max; i++)
        {
            Method1(data);
        }
        s1.Stop();
        // Version 2: use Split.
        var s2 = Stopwatch.StartNew();
        for (int i = 0; i < max; i++)
        {
            Method2(data);
        }
        s2.Stop();
        Console.WriteLine(((double)(s1.Elapsed.TotalMilliseconds * 1000000) / max).ToString("0.00 ns"));
        Console.WriteLine(((double)(s2.Elapsed.TotalMilliseconds * 1000000) / max).ToString("0.00 ns"));
    }
}
21
21
161.42 ns    StringReader
67.73 ns     Split

Read blocks

Read() can be used to populate a character array (a block) based on a range of the StringReader buffer. The buffer must be allocated before you call the method.

Internals

StringReader is a class with a member field string, an integer position, and an integer length. All the logic on the StringReader is based on the values of these three fields.

ReadLine internals

The ReadLine method on StringReader simply scans for the next newline starting at the current position, and then returns a substring based on the field string.

StringReader provides methods that are stream-oriented, enabling you to use a string as a file stream. StringReader is useful for looping over the lines in a string.