How to parse an arbitrary XML string into a structured object

Question

I have a string of XML data whose schema is unknown. I would like to parse it into a tree structure my code can easily peruse. For example if the string is:

<foo><bar>baz</bar></foo>

I want to be able to access it with code like:

elem["foo"]["bar"]

and get baz.

EDIT: The supposed "duplicate" assumes you do know the structure / schema of the XML. As I originally stated, I do not

Possible duplicate of [How to Deserialize XML document](http://stackoverflow.com/questions/364253/how-to-deserialize-xml-document) — Donal, Jan 06 '16 at 17:10
Not a duplicate - this is about accessing a parsed XML object, not to deserialize into a (strongly typed) class. — Leonardo Herrera, Jan 06 '16 at 17:25
@Donal you can't really deserialize XML to objects if the schema is unknown. — Charles Mager, Jan 06 '16 at 17:25
Of course, there is a duplicate: http://stackoverflow.com/questions/55828/how-does-one-parse-xml-files — Leonardo Herrera, Jan 06 '16 at 17:26
@CharlesMager, he is probably assuming you are the one who added the (incorrect) link to a "possible duplicate" — JoelFan, Jan 06 '16 at 17:30
See following posting : http://stackoverflow.com/questions/28976601/recursion-parsing-xml-file-with-attributes-into-treeview-c-sharp — jdweng, Jan 06 '16 at 17:46

score 3 · Answer 1 · answered Jan 06 '16 at 17:21

It sounds pretty much like you want what LINQ to XML offers. Parse XML like so:

var doc = XDocument.Parse(@"<foo><bar>baz</bar></foo>");

Then you could query it in a similar way to your suggested syntax:

var barValue = (string)doc.Elements("foo").Elements("bar").Single()

or:

var barValue = (string)doc.Descendants("bar").Single()

See the docs for more info.

TVOHM · Answer 2 · 2016-01-06T21:24:07.997

Personally, I agree with the other answers that a LINQ to XML based solution is best. Something like:

string xml = "<root><foo><bar>baz</bar></foo></root>";
string s = XElement.Parse(xml).Element("foo").Element("bar").Value;

But if you really wanted behaviour like your example, you could write a small wrapper class such as:

EDIT: Example updated to be indexable using a multidimensional indexer^{^note}.

class MyXmlWrapper
{
    XElement _xml;

    public MyXmlWrapper(XElement xml)
    {
        _xml = xml;
    }

    public MyXmlWrapper this[string name, int index = 0]
    {
        get
        {
            return new MyXmlWrapper(_xml.Elements(name).ElementAt(index));
        }
    }

    public static implicit operator string(MyXmlWrapper xml)
    {
        return xml._xml.Value;
    }
}

And use that exactly like you wanted:

string xml = "<root><foo><bar>baz</bar></foo></root>";
MyXmlWrapper wrapper = new MyXmlWrapper(XElement.Parse(xml));
string s = wrapper["foo"]["bar"];

Edited example for returning an element from a collection:

string xml = "<root><foo><bar><baz>1</baz><baz>2</baz></bar></foo></root>";
MyXmlWrapper wrapper = new MyXmlWrapper(XElement.Parse(xml));
string baz1 = wrapper["foo"]["bar"]["baz", 0];
string baz2 = wrapper["foo"]["bar"]["baz", 1];

The OP has clarified that he wants to parse arbitrary XML files. — Leonardo Herrera, Jan 06 '16 at 17:29
Not really - we don't know what kind of XML we'll get. For example, you can't get the second `baz` element in `12`. — Leonardo Herrera, Jan 06 '16 at 19:59

score 1 · Answer 3 · answered Jan 06 '16 at 17:21

1

XDocument.Parse(string) is your friend:

https://msdn.microsoft.com/en-us/library/bb345532(v=vs.110).aspx

answered Jan 06 '16 at 17:21

Jon G

4,083
22
27

How to parse an arbitrary XML string into a structured object

3 Answers3