Epsilon Flock

Epsilon Flock is a model migration language built atop EOL, for updating models in response to metamodel changes. Flock provides a rule-based transformation language for specifying model migration strategies. A conservative copying algorithm automatically migrates model values and elements that are not affected by the metamodel changes.

• Migrate models to re-establish consistency with an evolved metamodel
• Automatic copying of unaffected data
• Simple distribution of migration strategies using Eclipse extension point
• Declarative rules with imperative bodies
• Guarded rules

• Support for simultaneously accessing/modifying many models of (potentially) different metamodels
• All the usual programming constructs (while and for loops, statement sequencing, variables etc.)
• Support for those convenient first-order logic OCL operations (select, reject, collect etc.)
• Ability to create and call methods of Java objects
• Support for dynamically attaching operations to existing meta-classes and types at runtime
• Support for cached operations
• Support for extended properties
• Support for user interaction
• Ability to create reusable libraries of operations and import them from different Epsilon (not only EOL) modules

• Example: Migrate Petri net models with Epsilon Flock
• Screencast: Migrate Petri net models with Epsilon Flock
• Technical Report: Motivation and aims of Flock, along with an example of using Flock to migrate a UML class diagram.