1: \begin{abstract}

2:   Grammar convergence is a method that helps discovering relationships

3:   between different grammars of the same language or different language

4:   versions. The key element of the method is the operational,

5:   transformation-based representation of those relationships. Given

6:   input grammars for convergence, they are transformed until they are

7:   structurally equal. The transformations are composed from

8:   primitive operators; properties of these operators and the composed

9:   chains provide quantitative and qualitative insight into the

10:   relationships between the grammars at hand.

11: 

12:   We describe a refined method for grammar convergence, and we use it

13:   in a major study, where we recover the relationships between all the

14:   grammars that occur in the different versions of the Java Language

15:   Specification (JLS). The relationships are represented as grammar

16:   transformation chains that capture all accidental or intended

17:   differences between the JLS grammars. This method is mechanized and

18:   driven by nominal and structural differences between pairs of

19:   grammars that are subject to asymmetric, binary convergence

20:   steps. 

21: 

22:   We present the underlying operator suite for grammar transformation

23:   in detail, and we illustrate the suite with many examples of

24:   transformations on the JLS grammars. We also describe the extraction

25:   effort, which was needed to make the JLS grammars amenable to

26:   automated processing. We include substantial metadata about the

27:   convergence process for the JLS so that the effort becomes

28:   reproducible and transparent.

29: %

30: \keywords{grammar convergence \and grammar transformation \and grammar recovery \and grammar extraction \and language documentation}

31: % \PACS{PACS code1 \and PACS code2 \and more}

32: % \subclass{MSC code1 \and MSC code2 \and more}

33: \end{abstract}

34: