Given an instance l of Languages, l.components only contains the "top-level" constructs of the language l.
In my use case, this causes trouble when checking expressivity using the DLExpressivityChecker. Consider the language EL++ whose components are [CINT, E, +, R, O, (D)]. The construct for complex roles (R) also contains reflexive chains (Rr), transitive roles (+) and role hierarchy (H). Now, having an instance checker of DLExpressivityChecker with an input ontology that only makes use of constructs for full existential restrictions (E) and role hierarchy (H), checker.isWithin(Languages.ELPLUSPLUS) returns false even though the constructs E and H are part of R and thus contained in EL++.
As a possible solution, I think that components of languages could include the sub-constructs (the set of components is currently not minimal anyway). This would also prevent potential problems with isSubLanguageOf(...) and hasAllConstructs(...).
Given an instance
lofLanguages,l.componentsonly contains the "top-level" constructs of the languagel.In my use case, this causes trouble when checking expressivity using the
DLExpressivityChecker. Consider the language EL++ whose components are[CINT, E, +, R, O, (D)]. The construct for complex roles (R) also contains reflexive chains (Rr), transitive roles (+) and role hierarchy (H). Now, having an instancecheckerofDLExpressivityCheckerwith an input ontology that only makes use of constructs for full existential restrictions (E) and role hierarchy (H),checker.isWithin(Languages.ELPLUSPLUS)returnsfalseeven though the constructs E and H are part of R and thus contained in EL++.As a possible solution, I think that components of languages could include the sub-constructs (the set of components is currently not minimal anyway). This would also prevent potential problems with
isSubLanguageOf(...)andhasAllConstructs(...).