Introduction

This renewed AGG manual corresponds to the AGG version 1.2.5 and latter. The AGG version 1.2.4 is mainly covered by this manual, too. The installation instructions of the AGG tool can be found in a README file of the AGG package that you can download from URL http://tfs.cs.tu-berlin.de/agg. The AGG package also contains graph grammar examples, two of which we will use for our explanations.

The aim of this manual is to give a short but sufficient knowledge to understand how to specify a graph grammar:

1. to define and validate a set of transformation rules,
2. to perform transformations of the host graph,
3. to analyze the specified graph transformation system.

The AGG language is a rule based visual language supporting an algebraic approach to graph transformation. It aims at specifying and rapid prototyping applications with complex, graph structured data. The AGG environment is designed as a tool to edit directed, typed and attributed graphs and to define a graph grammar (i.e. a start (host) graph plus a set of transformation rules) as input for the graph transformation engine of the system. Having an AGG graph grammar at hand, it may be validated using AGG's analysis techniques, namely critical pair analysis, consistency checking and termination criteria for Layered Graph Transformation Systems (LGTS).

The main characteristics of the AGG language can be described as follow:

• Complex data structures are modeled as graphs which may be typed over a type graph.
• AGG graphs may be attributed by Java objects and types. Basic data types as well as object classes already available in Java class libraries may be used.
• Moreover, new Java classes may be included.
• The system's behavior is specified by graph rules using an if-then description style.
• Moreover, AGG features rules with Negative Application Conditions to express requirements for non-existence of substructures.
• The graph rules may be attributed by Java expressions which are evaluated during rule applications.
• Rules may have attribute conditions being boolean Java expressions.
• Application of a graph rule transforms the structure graph.
• Application of several rules sequentially shows an application scenario.
• A consistency control structure can be defined as global graph consistency conditions (constraints). They describe invariants on graphs. Additionaly, these global consistency conditions can be transformed into Post Application Conditions for an individual rule.

With AGG it is possible to define typed attributed graph transformation with node type inheritance. This means that the attributed type graph can be enriched by an inheritance relation between nodes. Each node type can have only one direct ancestor from which it inherits the attribute and edge types. Rules using this feature are equivalent to a number of concrete rules, resulting from the substitution of the ancestor nodes by the nodes in their inheritance clan. Therefore, rules become more compact and suitable for their use in combination with object-oriented modelling.

The AGG language and design concepts are described in

• ''Introduction to the Language Concepts of AGG'' by Michael Rudolf and Gabriele Taentzer (http://tfs.cs.tu-berlin.de/agg/intro.ps.zip).

• ''Concepts and Implementation of an Interpreter for Attributed Graphtransformation'' Diploma thesis of Michael Rudolf (in german) (http://tfs.cs.tu-berlin.de/agg/Diplomarbeiten/Rudolf.ps.gz).
• ''Design and Implementation of an Attribute Manager for Conditional and Distributed Graph Transformation'' Diploma thesis of Boris Melamed (http://tfs.cs.tu-berlin.de/agg/Diplomarbeiten/Melamed.ps.gz).

The AGG environment provides graphical editors for graphs, rules and graph consistency constraints and an integrated textual editor for Java expressions. Moreover, visual interpretation and step by step transformation of attributed graph grammars is supported. While step means performing direct derivations (transformation) for user-selected productions (rules) and occurrences (matches), a whole transformation sequence is executed in the interpretation mode. Furthermore, analysis tools for graph transformation systems are available.

In the following, first the editing and afterwards the interpretation and validation facilities are presented for the sample applications.