Get XML node from large XML file with XmlReader?

Question

How can I efficiently get XML node from large XML file?

Currently I have this code to load XML file and getting NodeList. This worked fine when that file was small (containing only that node). However, when I merged it with another XML file, Loading that new entire file cases my Windows Mobile device to crash (out of memory)

XmlDocument xmlDocument = new XmlDocument();
xmlDocument.Load(configFile);
XmlNodeList nodeList = xmlDocument.GetElementsByTagName("appsettings");

My XML is looks like this (I want to get appsettings which controls visibility of buttons in my app):

<site>
<settings>
<appsettings>
<add key="0" type="Home" value="True"/>
<add key="1" type="About" value="False"/>
<add key="2" type="Contact" value="True"/>
</appsettings>
...........
</settings>
</site>

How can I search for this node with XmlReader and load it into nodeList? I have tried to use

It's 13MB, it has some data for the app, unrelated to appsetting tag, but I have to have it there now. — 22332112, Oct 25 '13 at 17:18
Does anyone else think that a 12MB xml settings file is oversized for a process that has to live with less than 32MB process space? I would switch to a binary file, objects and serialze/ de-serialize. — josef, Oct 26 '13 at 03:27

score 1 · Accepted Answer · answered Oct 25 '13 at 18:01

1

You should use an XmlReader. Something like:

using (var reader = XmlReader.Create("file://path/to/file"))
{
  reader.ReadToDescendant("appsettings");
  while (reader.Read() && reader.Name == "add")
  {
    yield return new
      {
        Key = reader.GetAttribute("key"),
        Type = reader.GetAttribute("type"),
        Value = reader.GetAttribute("value")
      };
  }
}

answered Oct 25 '13 at 18:01

Reda

2,289
17
19

Thanks. I just had to set IgnoreWhitespace to true in addition. – 22332112 Oct 25 '13 at 20:43

score 0 · Answer 2 · answered Oct 25 '13 at 18:01

Have you tried LINQ to XML? (The old model is soooo 40s :P)

maybe something like this:

var xmlDoc = XDocument.Load(@"E:\Xml.xml");
var query = xmlDoc.Root.Element("settings").Descendants().First();

We know that the "appsettings" node always reside on a specific level. So we take the root and query for the settings node. Then we take the decsendants of that node and take the first of it. Which in the current case is "appsettings".

I know there is a LINQ way to make this even better, but this is what I could think of out of my head with no visual studio or LINQPad.

Hope this helps.

Get XML node from large XML file with XmlReader?

2 Answers2