XML Parser handling invalid XML : The Official Microsoft ASP.NET Forums

Reply

m.koopman

Participant

1372 Points

294 Posts

XML Parser handling invalid XML

Jan 04, 2012 02:20 PM|LINK

Hey,

I've got a question regarding teh XmlTextReader object. Im creating a parser object that is supposed to be able to deal with invalid XML or invalid XML Elements. So everything is working fine, EXCEPT the XmlTextReader.

P.S. I'm dealing with 3rd-party xml files so I have no grip on the what they submit. And I can't use the regular XDocument cause this object going to throw an exception when it runs into the problem.

This is the case:

I managed to get my custom parser working, even if it runs into a problem. So what i do right now is that I complete the current parsed elements and then return the so far parsed elements.

However, when the XmlTextReader goes into the error state I'm simply not able to get it out of that state no matter what. This is exactly my problem, cause I know where I should continue, but I can't continue cause the XmlTextReader stays in the error state.

I tried using the skip and readto... methods to skip this line of XML messing up my parser.

So if anyone has the solution for this problem, you would do me a great favour. If not I will have to write my own XML Text Reader.

So to clearify my problem, here is my code:

public class Parser
    {
        public delegate void OnParseError       (Exception exception, XElement element);
        public delegate void OnParseComplete    (XElement result);

        public static void Parse(string xml, OnParseError onError, OnParseComplete onComplete)
        {
            List<XElement>  elements    = new List<XElement>();
            Stack<XElement> elemstack   = new Stack<XElement>();
            XmlTextReader   reader      = new XmlTextReader(new MemoryStream(Encoding.ASCII.GetBytes(xml)));

            bool done = false;

            while (!done)
            {
                try
                {
                    while (reader.Read())
                    {
                        switch (reader.NodeType)
                        {
                            case XmlNodeType.Attribute:
                                if (elemstack.Count > 0)
                                    elemstack.Peek().SetAttributeValue(XName.Get(reader.Name, reader.NamespaceURI), reader.Value);
                                break;
                            case XmlNodeType.Element:
                                XElement startElement = new XElement(XName.Get(reader.Name, reader.NamespaceURI));
                                elemstack.Push(startElement);
                                break;
                            case XmlNodeType.EndElement:
                                if(reader.Name != elemstack.Peek().Name)
                                    reader.Skip();
                                else
                                {
                                    var endElement = elemstack.Pop();
                                    if (elemstack.Count > 0)
                                        elemstack.Peek().Add(endElement);
                                    elements.Add(endElement);
                                }
                                break;
                            case XmlNodeType.Text:
                                elemstack.Peek().Value = reader.Value;
                                break;
                            case XmlNodeType.CDATA:
                            case XmlNodeType.Comment:
                            case XmlNodeType.Document:
                            case XmlNodeType.DocumentFragment:
                            case XmlNodeType.DocumentType:
                            case XmlNodeType.EndEntity:
                            case XmlNodeType.Entity:
                            case XmlNodeType.EntityReference:
                            case XmlNodeType.None:
                            case XmlNodeType.Notation:
                            case XmlNodeType.ProcessingInstruction:
                            case XmlNodeType.SignificantWhitespace:
                            case XmlNodeType.Whitespace:
                            case XmlNodeType.XmlDeclaration:
                            default:
                                break;
                        }
                    }

                    done = true;
                    onComplete(elements.Last());
                }
                catch (XmlException ex)
                {
                    if (elemstack.Count > 0)
                    {
                        var errElem = elemstack.Pop();
                        onError(ex, errElem);
                        while (elemstack.Count > 0)
                        {
                            var endElement = elemstack.Pop();
                            if (elemstack.Count > 0) { elemstack.Peek().Add(endElement); }
                            elements.Add(endElement);
                        }
                        onComplete(elements.Last());
                    }
                    else
                        onError(ex, null);
                }
            }
        }
    }

Matthijs Koopman

Please mark my reply as answer if you found it help full

Edited by m.koopman on Jan 04, 2012 02:25 PM
Moved by bbcompent1 on Jan 04, 2012 04:54 PM

Reply

ignatandrei

All-Star

137716 Points

22159 Posts

Moderator

MVP

Re: XML Parser handling invalid XML

Jan 04, 2012 09:42 PM|LINK

m.koopman

. Im creating a parser object that is supposed to be able to deal with invalid XML or invalid XML Elements. So everything is working fine, EXCEPT the XmlTextReader.

P.S. I'm dealing with 3rd-party xml files so I have no grip on the what they submit.

If are not VALID XML, then can not be parsed.Period.

How do you parse

?

That said, the only reccomandation is make your parser to verify if it is Valid XML. If not, parse best or throw error ( maybe a human will parse faster to find error?)

If want automatically, look at http://htmlagilitypack.codeplex.com/ to see how it parses invalid( from an XML point of view) HTML and makes an XML.

My .NET blog : http://msprogrammer.serviciipeweb.ro/ |2012 Programmer tools

Reply

m.koopman

Participant

1372 Points

294 Posts

Re: XML Parser handling invalid XML

Jan 05, 2012 07:11 AM|LINK

My parser will try to parse as many as posible, e.g. whenthe parser fails in parsing an RSS feed in the third item, it should skip that item and continue with the others.

The only thing i need to know is if it is possible to resume the XMLTextReader after it went into an error state. If not I'll write my own.

Matthijs Koopman

Please mark my reply as answer if you found it help full