[Java]: simplify next token advancement while OTF decoding composites.

Author: tmontgomery

sbe-tool/src/main/java/uk/co/real_logic/sbe/otf/OtfMessageDecoder.java

@@ -194,7 +194,7 @@ private static void decodeComposite(
  {
  listener.onBeginComposite(fieldToken, tokens, tokenIdx, toIndex);
 
- for (int i = tokenIdx + 1; i < toIndex; i++)
+ for (int i = tokenIdx + 1; i < toIndex;)
  {
  final Token typeToken = tokens.get(i);
  final int nextFieldIdx = i + typeToken.componentTokenCount();
@@ -206,25 +206,24 @@ private static void decodeComposite(
  case BEGIN_COMPOSITE:
  decodeComposite(
  fieldToken, buffer, bufferIdx + offset, tokens, i, nextFieldIdx - 1, actingVersion, listener);
- i = nextFieldIdx - 1;
  break;
 
  case BEGIN_ENUM:
  listener.onEnum(
  fieldToken, buffer, bufferIdx + offset, tokens, i, nextFieldIdx - 1, actingVersion);
- i = nextFieldIdx - 1;
  break;
 
  case BEGIN_SET:
  listener.onBitSet(
  fieldToken, buffer, bufferIdx + offset, tokens, i, nextFieldIdx - 1, actingVersion);
- i = nextFieldIdx - 1;
  break;
 
  case ENCODING:
  listener.onEncoding(typeToken, buffer, bufferIdx + offset, typeToken, actingVersion);
  break;
  }
+
+ i += typeToken.componentTokenCount();
  }
 
  listener.onEndComposite(fieldToken, tokens, tokenIdx, toIndex);