Método Dottrace y Optimización con índiceOF

Question

public static string[] GetStringInBetween(string strBegin, string strEnd, string strSource, bool includeBegin, bool includeEnd)
        {
            string[] result = { "", "" };
            int iIndexOfBegin = strSource.IndexOf(strBegin);

            if (iIndexOfBegin != -1)
            {
                // include the Begin string if desired
                if (includeBegin)
                    iIndexOfBegin -= strBegin.Length;

                strSource = strSource.Substring(iIndexOfBegin + strBegin.Length);
                int iEnd = strSource.IndexOf(strEnd);

                if (iEnd != -1)
                {
                    // include the End string if desired
                    if (includeEnd)
                        iEnd += strEnd.Length;

                    result[0] = strSource.Substring(0, iEnd);

                    // advance beyond this segment
                    if (iEnd + strEnd.Length < strSource.Length)
                        result[1] = strSource.Substring(iEnd + strEnd.Length);
                }
            }

            return result;
        }

Uso:

string[] result = null;
result = HtmlHelper.GetStringInBetween(bits[0], bits[1], tagValuePair.Value, true, true);

Estoy usando Dottrace y este método usa el 33% de mi CPU.¿Cómo puedo optimizarlo?Because de mi choque de aplicación o me sale de la memoria.¿Es inteligente que este método sea estático?

Dottrace Show 30% Uso de CPU sobre esto:

System.String.IndexOf(String, Int32, Int32, StringComparison)

Editar:

    GetStringInBetween(string strBegin, string strEnd, string strSource, bool includeBegin, bool includeEnd)

strBegin = "<td class=\"m92_t_col2\">"
strEnd = "</td>"
strSource = "xxxxxxxx<td class=\"m92_t_col2\">Di. 31.01.12</td>xxxxxxxxxxxxxx
includeBegin = true
includeEnd = true

then i will get result
result[0] = "<td class=\"m92_t_col2\">Di. 31.01.12</td>"

Espero que esto ayude a lo que este método hace.Trate de encontrar cadena entre Strebegin y Strend ...

Answer 1

Copying part of the string (your first SubString call) just to keep searching in it is bad for performance. Instead, keep your original input string but use the overload on IndexOf that takes a start index and then adjust your index calculation for extracting the result accordingly.

Also, knowing that these strings are not localized, you might gain some by using an ordinal comparer in IndexOf.

Something along the lines of

public static string[] GetStringInBetween(string strBegin, string strEnd, string strSource, bool includeBegin, bool includeEnd)
{
    string[] result = { "", "" };
    int iIndexOfBegin = strSource.IndexOf(strBegin, StringComparison.Ordinal);

    if (iIndexOfBegin != -1)
    {
        int iEnd = strSource.IndexOf(strEnd, iIndexOfBegin, StringComparison.Ordinal);

        if (iEnd != -1)
        {
            result[0] = strSource.Substring(
                iIndexOfBegin + (includeBegin ? 0 : strBegin.Length), 
                iEnd + (includeEnd ? strEnd.Length : 0) - iIndexOfBegin);

            // advance beyond this segment
            if (iEnd + strEnd.Length < strSource.Length)
                result[1] = strSource.Substring(iEnd + strEnd.Length);
        }
    }

    return result;
}

Answer 2

For your sample input, it appears you are working with HTML fragments.

I suggest using the HTML Agility Pack to parse the HTML - it exposes the result in an easy to query fashion, using LINQ to XML or XPath type syntax. It is a fast and efficient library.