Linq xml recursive select

Question

<Nodes>
  <Node>
    <ID>1</ID>
    <TIDS>
      <TID>2</TID>
      <TID>3</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>2</ID>
    <TIDS>
      <TID>4</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>3</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>4</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>5</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>6</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>7</ID>
  </Node>
</Nodes>

i want to write query which will fselect TID and again query xml to select it's TID suppose my where condition is ID equal 1 then i want out put 2,3,4,7 in my where condition if i put ID equal 5 then put put 7 how to write recursive linq

Why should (based on what logic) it return 2,3,4 and 7 for ID=1? — Guanxi
– Guanxi, Commented Jul 25, 2013 at 17:31
@Guanxi I believe the behavior OP is looking for is that an ID of 1 returns 2 and 3, and then the lookup continues with those two IDs. So then looking up 2 will yield 4, and looking up 3 will yield 7 (continuing with 4 will also yield 7 and looking up 7 will reveal nothing). In total, having started at 1 the unique IDs the function has 'discovered' will be 2,3,4, and 7. — diceguyd30
– diceguyd30, Commented Jul 25, 2013 at 18:10

Hogan · Accepted Answer · 2013-07-25 20:23:02Z

1

 var result = xml.Elements()
   // Find element with <ID>1</ID>
   .Where(x => x.Elements().Any(d => d.Name == "ID" && d.Value == "1"))
   // Find element <TIDS>
   .Elements().Where(x => x.Name == "TIDS")
   // Find elements <TID>
   .Elements().Where(x => x.Name == "TID")
   // Select values
   .Select(x => x.Value);

note, I made the xml variable like this:

   XElement xml = XElement.Parse (@"<Nodes>
  <Node>
    <ID>1</ID>
    <TIDS>
      <TID>2</TID>
      <TID>3</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>2</ID>
    <TIDS>
      <TID>4</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>3</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>4</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>5</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>6</ID>
    <TIDS>
      <TID>7</TID>
    </TIDS>
  </Node>
  <Node>
    <ID>7</ID>
  </Node>
</Nodes>");

answered Jul 25, 2013 at 20:23

Hogan

70.7k10 gold badges83 silver badges119 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Hogan Over a year ago

@James - Thanks, I'm not sure I got it right - but I think if you put my code in a function and then called it recursively it might be what he wants.

mnsr Over a year ago

This only returns the TID values for the initial ID. It doesn't query again for each TID.

Steve Ruble · Accepted Answer · 2013-07-26 01:04:46Z

Recursive LINQ is more of a party trick than anything you'd actually want to use in production code, but here's one way to do it:

XElement nodes = /* load your xml */;

Func<XElement, int, IEnumerable<int>> query = null;

/* BAD CODE, DO NOT USE! */
query = (x,id) => 
   x.Elements("Node")
     .Where (node => (int)node.Element("ID") == id) // find our node
     .Descendants("TID")      // this will be empty in the base case
     .Select (tid => (int)tid)
     .SelectMany(tid => 
       query(x,tid)           // recurse
       .Concat(new[]{tid})    // keep the current TID if its node had no TIDs
     )
     .Distinct();

var resultOf7423 = query(nodes, 1);
var resultOf7 = query(nodes, 5);
/* END BAD CODE (I HOPE) */

This is baffling and fragile code, and you almost certainly shouldn't use it. Instead, you could create an extension method off of XElement:

public static IEnumerable<int> SelectRelatedIds(this XElement element, int id)
{
    if(element == null) throw new ArgumentNullException("element");
    return SelectRelatedIdsIterator(element, id)
                .Where(i => i != id)
                .Distinct();
}

private static IEnumerable<int> SelectRelatedIdsIterator(XElement element, int id)
{
    yield return id;

    var tids = element.Elements("Node")
                .Where (node => (int)node.Element("ID") == id)
                .Descendants("TID");

    foreach (int tid in tids)
        foreach(var i in  SelectRelatedIdsIterator(element, tid))
            yield return i;
}

This keeps the recursion contained and pretty easy to understand (at least once you get your head around iterator blocks). It's also LINQ-friendly and potentially composable in a way that the first way is not.

mnsr · Accepted Answer · 2013-07-26 02:24:40Z

0

The function GetTids returns a list of TIDs for the ID you pass in. After the initial ID is passed, it returns a list (tidIndex). Then a loop is run, and each TID value is used to query the XML and is added to the tidList.

List<string> TidList()
{
    var xml = XDocument.Load(@"C:\PathToXml\File.xml");

    var tidIndex = GetTids(xml, "1").ToList();
    var tidList = new List<string>(tidIndex);

    foreach (var tid in tidIndex)
        tidList.AddRange(GetTids(xml, tid));

    return tidList;
}

static IEnumerable<string> GetTids (XDocument xml, string id)
{
    return xml.Descendants("Node")
        .Where(x => x.Element("ID").Value == id)
        .Descendants("TID")
        .Select (s => s.Value);
}

This returns:

enter image description here

edited Jul 26, 2013 at 2:24

answered Jul 26, 2013 at 2:06

mnsr

12.4k4 gold badges55 silver badges80 bronze badges

Collectives™ on Stack Overflow

Linq xml recursive select

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related