Linker-Parser 0.9:

[dmitriic]

• adds EnumTokens and NumberTokens : allows using Enums as simple single-token grammar junctions, simplifying and optimizing handling of sutuations where similar non-composite tokens can occur in the same position; allows to define rule fields using types that extend java.lang.Number and have a constructor that accepts a String. Vastly simplifies implementations of java-compatible number type system. Example:

  public enum BinaryOperator implements Rule {
    @CapturePattern("-") MINUS,
    @CapturePattern("+") PLUS,
    @CapturePattern("/") SLASH,
    @CapturePattern("*") STAR;
  
    public Number apply(Number leftOperand, Number rightOperand) {
      ...
    }
  }
  public class ExampleBinaryStatement implements Rule {
    protected Number leftOperand;
    protected BinaryOperator operator;
    protected Number rightOperand;
  }
  

• adds IgnoreCase annotation allows to perform case-insensitive matching using both Pattern and Terminal matchers

• support for compilers: now parser will read rule definition from the first class in inheritance chain that implements Rule interface, ignoring all its children. This allows to reuse grammar rules while implementing different behavior for the resulting AST: one set of grammar extending tokens can be used for evaluation, another - for linting or compiling the application into a different language.

  Example (evaluating token):

  public class ExampleBinaryStatementEvaluator extends ExampleBinaryStatement implements Evaluator<Number> {
    @Override
    public Number evaluate() {
      return operator.apply(leftOperand, rightOperand)
    }
  }
  

  Example (compiling token):

  public class ExampleBinaryStatementCompiler extends ExampleBinaryStatement implements Compiler<JavaScript> {
  
    @Override
    public JavaScript compile() {
      return new JavaScript()
        .append(leftOperand)
        .append(' ')
        .append(operator)
        .append(' ')
        .append(rightOperand);
    }
  
  }