I would like a more efficient way to remove characters from nodes in a string of XML

[thetimmer]

Can anyone offer a more ellegant/efficient solution?

I've writen a function that cleans out an XML string. The string of XML eventually gets read into a dataset like so

favoriteDataSet.ReadXml(New System.IO.StringReader(sCleanedXML))

so this function is setup to clean the offending characters without removing them if they are actually valid xml characters.

So if I sent in

<Some&Text/ToClean> blah blah blah </Some&Text/ToClean>
I would get back

<Some_Text_ToClean> blah blah blah </Some_Text_ToClean>
It works but I'd really like to learn how to implement it more efficiently.

It accepts...

sOriginal which is a long string of xml with tags like blah blah blah

asBetween which is a string of characters to look between to find the offending characters. The string of characters is pipe delimited. (ie. asBetween = "<|>")

sRemovalTokens which is a string of pipe delimited values to clean out. (ie. sRemovalTokens = "&|/|#" )

Here is the code...

 Private Function CleanXMLTags(sOriginal As String, asBetween As String, sRemovalTokens As String) As String
    Try
        Dim aBookEnds() As String = asBetween.Split("|")
        Dim aRemovalTokens() As String = sRemovalTokens.Split("|")
        Dim iStartIndex As Integer = 0
        Dim iEndIndex As Integer = 0
        Dim sCurToken As String = String.Empty
        Dim sReplaceToken As String = String.Empty
        For i As Integer = 0 To aRemovalTokens.Length - 1
            If aBookEnds.Length = 2 AndAlso Not String.Equals(aRemovalTokens(i).Trim, "", StringComparison.CurrentCultureIgnoreCase) Then
                iStartIndex = 0
                iEndIndex = 0
                While iStartIndex > -1
                    iStartIndex = sOriginal.IndexOf(aBookEnds(0), iEndIndex)
                    iEndIndex = sOriginal.IndexOf(aBookEnds(1), iEndIndex + 1)
                    If iStartIndex < 0 OrElse iEndIndex < 0 Then
                        Exit While
                    End If
                    sCurToken = sOriginal.ToString.Substring(iStartIndex, iEndIndex - iStartIndex + 1)
                    If sCurToken.ToString.StartsWith("</") Then
                        sCurToken = sCurToken.Substring(2)
                    End If
                    If sCurToken.ToString.EndsWith("/>") Then
                        sCurToken = sCurToken.Substring(0, sCurToken.Length - 2)
                    End If
                    If sCurToken.Contains(aRemovalTokens(i)) Then
                        sReplaceToken = sCurToken.Replace(aRemovalTokens(i), "_")
                        sOriginal = sOriginal.Replace(sCurToken, sReplaceToken)
                    End If
                End While
            End If
        Next
        Return sOriginal 'Cleaned
    Catch ex As Exception
        Return sOriginal
    End Try
End Function

[KGComputers]

Here's one of my reference in dealing with XML illegal characters.
Strip illegal xml characters.

Another option would be to include Regex for cleaning xml nodes.

[dbasnett]

Are you creating this, <Some&Text/ToClean> blah blah blah </Some&Text/ToClean>, or is it coming from somewhere else? What is it supposed to be when it is in this format?