Translations between textual transition systems and Petri nets

Translations between models expressed in textual transition systems and those expressed in structured Petri net notation are presented, in both directions. The translations are structure-preserving, meaning that the hierarchical structure of the systems is preserved. Furthermore, assuming non-finite data has been abstracted out of the textual transition system, then translating one model to another and then back results in a model which is identical to the original one, up to renaming and the form of Boolean expressions. Due to inherent differences between the two notations, however, some additional information is required in order to obtain this identity. The information is collected during the translation in one direction and is used in the translation back. Our translation is also semantics-preserving. That is, the original model and the translated model are bisimulation equivalent, assuming nonfinite data is abstracted. Thus, the translation preserves all temporal properties expressible in the logic CTL*. The translations are both more generally applicable and more detailed than previously considered. They are shown both for individual modules, with a collection of transitions, and for a structured system, where modules are combined in different ways.