Splitting a path string

I have the following C# method to split a path string. The only thing I know is a fixed string in the path and I have to split the full path into two parts, first part should be one level below the fixed string and the rest should be the second part.

For example, if I have the following path:

string mainText = @"C:\Abc\Fixed\MyTemp\Japan\Tokyo";

then my firstPart would be "C:\Abc\Fixed\MyTemp" and second part "Japan\Tokyo"

I want to improve this method in terms of memory and speed.

private static void SplitPath(string mainText, out string firstPart, out string secondPart)
        {
            firstPart = string.Empty;
            secondPart = string.Empty;
            if (!string.IsNullOrEmpty(mainText))
            {
                string strConstatnt = "Fixed";
                List splitted = mainText.Split(new char[] { '\\' }).ToList();
                int indexToFixed = splitted.IndexOf(strConstatnt);
                StringBuilder sbFirst = new StringBuilder();
                StringBuilder sbSecond = new StringBuilder();
                if (indexToFixed >= 0)
                {
                    for (int i = 0; i  0)
                {
                    firstPart = sbFirst.Remove(sbFirst.Length - 1, 1).ToString();
                }
                if (sbSecond.Length > 0)
                {
                    secondPart = sbSecond.Remove(sbSecond.Length - 1, 1).ToString();
                }
            }
        }

I took the liberty to make the exercise as well to see how our approaches differ. I don't think there are many remarks in terms of performance although StringBuilder.Remove() is a red flag in that area.

All together I think my code provides a more easily readable solution which should help in terms of human-reading efficiency. In the end I pretty much abstracted the things you do under library features.

public void SplitPath(string mainText, out string firstPart, out string secondPart)
{
  const string pivot = "Fixed";
  const char delimiter = '\\';

  var entries = mainText.Split(delimiter);
  int pivotIndex = entries.Length;
  for (var i = 0; i < entries.Length; i++)
  {
      if (entries[i] == pivot)
      {
          // Take index + 2 because linq.Take/Skip is 1-based
          // Optionally take index + 1 to avoid out-of-bounds when the pivot is the last element
          pivotIndex = (i == entries.Length) ? i + 1 : i + 2;
          break;
      }
  }

  firstPart = string.Join(delimiter.ToString(), entries.Take(pivotIndex));
  secondPart = string.Join(delimiter.ToString(), entries.Skip(pivotIndex));
}

I have tested this code with inputs C:\Abc\Fixed\MyTemp\Japan\Tokyo, C:\Abc\Fixed and C:\Abc.

If you'd want to take your approach after all, a few remarks:

• Try to early return by using a negated if-statement condition.

• Put your delimiter in a variable

• Avoid magic values

• Avoid removing data from the StringBuilder

StackExchange Code Review Q#61608, answer score: 5