I an using DataContractSerializer
and i want to separate data of same object into multiple files.
[DataContract]
public class TestObj
{
[DataMember]
protected double field1 = 0.0;
[DataMember]
protected double field2 = 0.0;
}
Specifically I want to save field1
in one XML file and field2
in a different XML file. Is there any way to do that using data contract serialization?
This is how I am serializing currently:
DataContractSerializer serializaer = new DataContractSerializer(GetType(), null,
0x7FFFFFFF /*maxItemsInObjectGraph*/,
false /*ignoreExtensionDataObject*/,
true /*preserveObjectReferences : this is most important to me */,
null /*dataContractSurrogate*/);
var fs = File.Open(fName, FileMode.Create);
serializaer.WriteObject(fs, this);
fs.Dispose();
return true;
I can suggest using Custom Xml Writer paired with serializer.
public class CustomWriter : XmlTextWriter
{
public CustomWriter(TextWriter writer) : base(writer) { }
public CustomWriter(Stream stream, Encoding encoding) : base(stream, encoding) { }
public CustomWriter(string filename, Encoding encoding) : base(filename, encoding) { }
public List<string> FieldList { get; set; }
private string _localName;
public override void WriteStartElement(string prefix, string localName, string ns)
{
if (!FieldList.Contains(localName))
base.WriteStartElement(prefix, localName, ns);
else
_localName = localName;
}
public override void WriteString(string text)
{
if (!FieldList.Contains(_localName))
base.WriteString(text);
}
public override void WriteEndElement()
{
if (!FieldList.Contains(_localName))
base.WriteEndElement();
else
_localName = null;
}
}
Use it as follows:
var data = new TestObj();
var serializer = new DataContractSerializer(
typeof(TestObj), null, 0x7FFFFFFF, false, true, null);
using (var writer = new CustomWriter(Console.Out)) // Specify filename or stream instead
{
writer.Formatting = Formatting.Indented;
writer.FieldList = new List<string> { "field1", "field3" }; // Specify fields to ignore
serializer.WriteObject(writer, data);
}
Just specify a list of fields that should be ignored in the FieldList property.
Of course, with this way, the DataContractSerializer will intermediate create xml with all the elements contained in the class. And only after that our custom writer will filter them. But it will happen on the fly, very quickly and effectively.
Once you serialise the whole object you can then split it using Linq to XML.
Oh, I even found an example how to do it at How to Split an XML file into multiple XML Files.
User contributions licensed under CC BY-SA 3.0