Tukukkk

I wrote a small program to test the speed of two separate extension methods I created, detailed in the below screenshot:

public static class Extensions

{

    public static List<string> ParseEmails(this string emails, char charsToSplitOn)

    {

        List<string> list = new List<string>();

        foreach (string email in emails)

        {

            string splitArr = email.Replace(" ", "").Split(charsToSplitOn, StringSplitOptions.RemoveEmptyEntries);

            foreach (string item in splitArr)

            {

                list.Add(item);

            }

        }

        return list;

    }



    public static string ParseEmails2(this string emails, char charsToSplitOn)

    {

        string str = string.Empty;

        foreach (string item in emails)

        {

            str += item + ';';

        }

        return str.Replace(" ", "").Split(charsToSplitOn, StringSplitOptions.RemoveEmptyEntries);

    }

}

The Main method initializes a Stopwatch class to track the time each method takes to execute iterations amount of times.

The first method, ParseEmails has a for loop within another for loop, while the second method 'ParseEmails2' has only one for loop.

I would expect that then the second method, ParseEmails2, would be faster, as, from my understanding, this is done in O(n) time, whereas my first method, ParseEmails, is done in O(n^2) time.

If this is true, then shouldn't my results point to ParseEmails2 being the faster of the two methods?

You have to be precise about what you mean by O(n^2). What is n? If n corresponds to the number of emails (in the emails array), then the first method runs on the scale of O(n), because you iterate over all emails once.

In the second method you use string concatenation, which means that you create a new string each run of the loop resulting in a complexity of O(n^2).

Prefer using StringBuilder or Lists over concatenating strings inside a loop.

Well...you don't seem to understand the purpose of the O-notation, O is less about the performance and more about the consistancy of performance. O(n²) means a parabola and O(n) a straight line, but it doesn't tell you about the actual time. So once you get to many elements, the O(n) algorithm should be faster. A reason for that could be that you use string's += and + operators which use the copy on write-semantic of C# string.

ParseEmails2 is not actually O(n), because

str += item + ';';

contains a loop.

you should check this link. string concatenation vs builder performance
String concatenation vs String Builder. Performance

The string builder will most likely be marginally faster in this case, but really it's probably not going to be enough to fret about. The number of times you are re-creating the string (because strings are immutable and joining forces a new string to be created from the two existing ones).

Because of memory allocation on very step will cause more time.