CODEWITHSUNDEEP
c#.netxmlnetwork-programming
ygoe
ygoe

Reputation: 20374

Read entire elements from an XML network stream

I am writing a network server in C# .NET 4.0. There is a network TCP/IP connection over which I can receive complete XML elements. They arrive regularly and I need to process them immediately. Each XML element is a complete XML document in itself, so it has an opening element, several sub-nodes and a closing element. There is no single root element for the entire stream. So when I open the connection, what I get is like this:

<status>
    <x>123</x>
    <y>456</y>
</status>

Then some time later it continues:

<status>
    <x>234</x>
    <y>567</y>
</status>

And so on. I need a way to read the complete XML string until a status element is complete. I don't want to do that with plain text reading methods because I don't know in what formatting the data arrives. I can in no way wait until the entire stream is finished, as is often described elsewhere. I have tried using the XmlReader class but its documentation is weird, the methods don't work out, the first element is lost and after sending the second element, an XmlException occurs because there are two root elements.

Upvotes: 6

Views: 5947

Answers (5)

unbob
unbob

Reputation: 381

Not substantially different from dtb's solution, but linqier

static IEnumerable<XDocument> GetDocs(Stream xmlStream)
{
    var xmlSettings = new XmlReaderSettings() { ConformanceLevel = ConformanceLevel.Fragment };
    using (var xmlReader = XmlReader.Create(xmlStream, xmlSettings))
    {
        var xmlPathNav = new XPathDocument(xmlReader).CreateNavigator();
        foreach (var selectee in xmlPathNav.Select("/*").OfType<XPathNavigator>())
            yield return XDocument.Load(selectee.ReadSubtree());
    }
}

I ran into a similar problem in PowerShell, but the asker's question was in C#, so I've attempted to translate it (and verified that it works). Here is where I found the clue that got me over the last little bumps (". . .The way the XPathDocument does its magic is by creating a “transparent” root node, and holding the fragments from it. I say it’s transparent because your XPath queries can use the root node axis and still get properly resolved to the fragments. . .")

The fragments of XML I'm working with happen to be smallish. If you had bigger chunks, you'd probably want to look into XStreamingElement - it can add a lot of complexity but also greatly decrease memory usage when dealing with large volumes of XML.

Upvotes: 0

Anastasiosyal
Anastasiosyal

Reputation: 6626

You could use XElement.Load which is meant more for streaming of Xml Element fragments that is new in .net 3.5 and also supports reading directly from a stream.

Have a look at System.Xml.Linq

I think that you may well still have to add some control logic so as to partition the messages you are receiving, but you may as well give it a go.

Upvotes: 1

dtb
dtb

Reputation: 217293

Try this:

var settings = new XmlReaderSettings
{
    ConformanceLevel = ConformanceLevel.Fragment
};

using (var reader = XmlReader.Create(stream, settings))
{
    while (!reader.EOF)
    {
        reader.MoveToContent();

        var doc = XDocument.Load(reader.ReadSubtree());

        Console.WriteLine("X={0}, Y={1}",
            (int)doc.Root.Element("x"),
            (int)doc.Root.Element("y"));

        reader.ReadEndElement();
    }
}

Upvotes: 9

dillenmeister
dillenmeister

Reputation: 1647

If you change the "conformance level" to "fragment", it might work with the XmlReader.

This is a (slightly modified) example from MSDN:

XmlReaderSettings settings = new XmlReaderSettings();
settings.ConformanceLevel = ConformanceLevel.Fragment;
XmlReader reader = XmlReader.Create(streamOfXmlFragments, settings);

Upvotes: 1

Svarog
Svarog

Reputation: 2208

I'm not sure there's anything built-in that does that. I'd open a string builder, fill it until I see a </status> tag, and then parse it using the ordinary XmlDocument.

Upvotes: 0

Related Questions