Bitcoin Classic introduced the Compact Message Format as a very simple but powerful format to encode and decode any type of messages.

The compact message format is a key/value-pair based format. Each key/value pair is called a token and a message is build up of a series of tokens.

Take this as a short example, a message with 3 tokens. Each having a name and a value.

   Name=Paris
   Population=2229621
   Area=105.6

In all object oriented languages there are very similar constructions available to create or parse the messages. Please refer to the API docs of your language bindings for details. I'll give a generic example here, to explain the concepts.

For creation of messages we use the builder pattern in the form of the MessageBuilder class.

The MessageBuilder class has a series of add() methods each of which appends a token to your message.

  byte[] bytes = new byte[100];
  MessageBuilder builder = new MessageBuilder(bytes);
  builder.add(City.Name, "Paris");
  builder.add(City.Population, 2229621);
  builder.add(City.Area, 105.6);
  builder.close();

This allows really easy to read and understand code.

The MessageParser is using more of a SOX parser approach where you call MessageParser.Next() and then you can ask the parser for the tag-is and the actual value.

  MessageParser parser = new MessageParser(inputMessage);
  while (parser.next() == MessageParser.FoundTag) {
     if (parser.tag() == City.Population) {
         int population = parser.data.toInt();
         break;
     }
  }

At this time there are implementations for;

• C++ which depend on Qt
• C++ with boost
• C# Should work with any version, the project assumes 4.5
• Java
• Python
• C