The VariantCodec library provides a generic data structure; the variant, which can be a map, collection, string or primitive (i.e. integer, double), and provides several options for serialization; binary, a C-style syntax and XML.



Why is this useful?

• It separates the data structure from the representation so you are not locked into one representation (e.g. XML).
• It is an extremely easy way to serialize data structures, from a simple property map to a complex object tree.
• It provides different serializations for different uses:
  • Binary for efficient storage,
  • Java/C-style syntax for a compact representation that is human readable,
  • XML for compatibility.
• It provides a validation/schema mechanism defined as code (ie. it separates the schema definition from the format).
• It provides an object->variant->object mapper for simple serialization.

The following snippet shows how to use the library with the C/Java style serialization format:

package org.boris.variantcodec;

import org.boris.variant.VTCollection;
import org.boris.variant.VTMap;
import org.boris.variant.codec.SourceCodec;

public class VariantCodecExample
{
    public static void main(String[] args) throws Exception {
        VTMap m = new VTMap();
        m.add("var1", "Hello World!");
        m.add("num1", 12);
        m.add("float1", 23.5f);
        m.add("binary1", new byte[] { 3, 4, 4, 5, 5 });
        VTCollection c = new VTCollection();
        c.add(1);
        c.add(2);
        c.add(3);
        m.add("coll1", c);
        System.out.println(SourceCodec.encodeFormatted(m));
    }
}

This would output the following:

{
    var1="Hello World!";
    num1=12;
    float1=23.5f;
    binary1=<0304040505>;
    coll1=[1,2,3];
}

If the above output was stored in a String, you would decode it using the following code:

VTMap m = (VTMap) SourceCodec.decode(s);
System.out.println(m.getString("var1"));

This would output the following:

Hello World!

The latest download is available from the Project Page.

The javadoc for the library can be found here.

The following shows the simple variantcodec object serializer:

import org.boris.variant.util.VariantObjectSerializer;

public class Test1
{
    public String a = "something";
    public double d = 1.8;
    public int[] c1 = { 3, 2, 1 };
    public boolean b2 = false;
    public short s1 = 234;
    public byte b3 = 12;
    public float f1 = 3.4f;

    public String toString() {
        return VariantObjectSerializer.encode(this).toString();
    }

    public static void main(String[] args) {
        System.out.println(new Test1());
    }
}

This would output the following:

{
    a="something";
    d=1.8;
    c1=[3,2,1];
    b2=false;
    s1=234s;
    b3=12b;
    f1=3.4f;
}

V0.0.3

• Added charset support to binary codec
• More consist use of reader/writer on source codec
• Built as an OSGI bundle
• Removed unused utilities

V0.0.2

• Bug fixes with source codec.
• Additional utilities.

V0.0.1

• Initial version.