Read nodes of a xml file in C#

Your log file is not an XML document. Since an XML document must have one and only one root element, it’s a series of XML documents concatenated together. Such a series of documents can be read by XmlReader by setting XmlReaderSettings.ConformanceLevel == ConformanceLevel.Fragment. Having done so, you can read through the file and deserialize each root element individually using XmlSerializer as follows:

static List<ApplicationLogEventObject> ReadEvents(string fileName)
{
    return ReadObjects<ApplicationLogEventObject>(fileName);
}

static List<T> ReadObjects<T>(string fileName)
{
    var list = new List<T>();

    var serializer = new XmlSerializer(typeof(T));
    var settings = new XmlReaderSettings { ConformanceLevel = ConformanceLevel.Fragment };
    using (var textReader = new StreamReader(fileName))
    using (var xmlTextReader = XmlReader.Create(textReader, settings))
    {
        while (xmlTextReader.Read())
        {   // Skip whitespace
            if (xmlTextReader.NodeType == XmlNodeType.Element) 
            {
                using (var subReader = xmlTextReader.ReadSubtree())
                {
                    var logEvent = (T)serializer.Deserialize(subReader);
                    list.Add(logEvent);
                }
            }
        }
    }

    return list;            
}

Using the following version of ApplicationLogEventObject:

public class ApplicationLogEventObject
{
    public string EventType { get; set; }

    [XmlElement("DateStamp")]
    public string DateStampString { 
        get
        {
            // Replace with culturally invariant desired formatting.
            return DateStamp.ToString(CultureInfo.InvariantCulture);
        }
        set
        {
            DateStamp = Convert.ToDateTime(value, CultureInfo.InvariantCulture);
        }
    }

    [XmlIgnore]
    public DateTime DateStamp { get; set; }

    public string ShortDescription { get; set; }
    public string LongDescription { get; set; }
}

Sample .Net fiddle.

Notes:

  • The <DateStamp> element values 10/13/2016 11:15:00 AM are not in the correct format for dates and times in XML, which is ISO 8601. Thus I introduced a surrogate string DateStampString property to manually handle the conversion from and to your desired format, and then marked the original DateTime property with XmlIgnore.

  • Using ReadSubtree() prevents the possibility of reading past the end of each root element when the XML is not indented.

  • According to the documentation for XmlTextReader:

    Starting with the .NET Framework 2.0, we recommend that you use the System.Xml.XmlReader class instead.

    Thus I recommend replacing use of that type with XmlReader.

  • The child nodes of your <ApplicationLogEventObject> are elements not attributes, so XmlReader.GetAttribute() was not an appropriate method to use to read them.

  • Given that your log files are not formatting their times in ISO 8601, you should at least make sure they are formatted in a culturally invariant format so that log files can be exchanged between computers with different regional settings. Doing your conversions using CultureInfo.InvariantCulture ensures this.

Leave a Comment