Analysis of XML structured data

My work uses software to fill out records that are expressed as XML documents. I have the task of trawling through these XML files to pull statistics out of them. The files themselves adhere to no schema and if a form field doesn't get filled out then the XML corresponding to that field is not creat开发者_开发百科ed.

What's my best approach?

Example XML:

<Form>
    <Field>
        <id>Field ID</id>
        <value>Entered Value</value>
    </Field>
</Form>

I have been attempting to write software that I can use to query the files but have not been able to come up with anything even remotely useful.

Any thoughts appreciated.

EDIT: In terms of C#, what I would like (Though I'm sure it isn't possible) is a Dictionary that has a string as the key and the corresponding value could EITHER be a string or another Dictionary.

---

Is like this ↓ ?

XML:

<?xml version="1.0" encoding="utf-8" ?>
<Form>
 <Field>
  <id>People1</id>
  <value>C Sharp</value>
 </Field>
 <Field>
  <id>People2</id>
  <value>C Sharp</value>
 </Field>
 <Field>
   <id>People3</id>
   <value>C</value>
 </Field>

Source:

static void Main(string[] args)
    {
        var doc = XDocument.Load("test.xml");
        var result = from p in doc.Descendants("Form").Descendants("Field")
                     select new { ID = p.Element("id").Value, VALUE = p.Element("value").Value };

        foreach (var x in result)
            Console.WriteLine(x);

        var gr = from p in result
                 group p by p.VALUE into g
                 select new { Language=g.Key , Count=g.Count() };


        foreach (var x in gr)
            Console.WriteLine(string.Format("Language:{0} Count:{1}" , x.Language , x.Count));

        Console.Read();
    }

---

If the file is not too big, I would suggest perl and the XML::Simple module. This will map the XML to a perl array of hashes, and then you can simply loop through it like normal. Something like:

my $xml = XML::Simple::XmlIn( 'file.xml', force_array => [ 'Form', 'Field' ] );
my %fld_counts;
foreach my $form ( @{$xml->{Form}} )
{
    # Any start record processing...
    foreach my $fld ( @{$form->{Field}} )
    {
        my $id = $fld->{id}
        my $val = $fld->{value}
        # Do something with id/value... like...
        $fld_counts{$id}++;
    }
}

So just adjust that structure based on the stats you want to gather

---

For parsing XML I prefer using plain XmlReader. Granted, it's more verbose, but it's super efficient and transparent, at least for me. For example:

using(var xr = XmlReader.Create('your stream here'))
    while(xr.Read())
        if(xr.NodeType == XmlNodeType.Element)
            switch(xr.Name) {
                case "brands":
                    // do something here with this element,
                    // like possibly reading the whole subtree...
                    using(var xrr = xr.ReadSubtree())
                        while(xrr.Read()) {
                            // working here...
                        }
                    break;
                case "products":
                    // that is another element
                    break;
                case "some-other-element":
                    // and so on
                    break;
            } // switch