A Language Workbench based on GLL and Term Rewriting

I dedicate this work to my parents, for their encouragement and support at every step during my Master’s studies. One of the main difficulties in developing domain-specific languages is the use of deterministic parsing technologies in language workbenches. These deterministic parsers do not naturally support modularity, and impose re-strictions on defining the grammar of a language. For example, LL(k) parsers cannot deal with left recursion or ambiguities, which have to be removed by rewriting the grammar. These modifications usually lead to a gram-mar definition which is not readable and may cause maintenance problems. More importantly, the parse trees from a modified grammar may be signif-icantly different from the ones of the original, ambiguous grammar. This may cause problems in processing parse trees, for example, for mapping to EMF models. In this thesis, we present a new language workbench based on the GL