These PR extends the relax command by adding three new options:
--exclude-named-classes: if true, A EquivalentTo B between named classes are excluded from relax
--include-subclass-of: if true, A Subclass A and R some B is relaxed the same way as A EquivalentTo A and R some B, see @cmungall request in #1183
--enforce-obo-format: If true, only simple existentials and named subsumptions are asserted during the process
I like all the functionality, but the --enforce-obo-format is really clunky. I would be happy if someone made a better suggestion; I also dont like --exclude-nested.
The only "breaking" change in this PR is that the CLI gets --exclude-named-classes by default as true. Previously, before this option existed, A=B was relaxed to A sub B and B sub A. No one I personally now has ever said that this is a good idea, so we changed the default, as discussed in #1060.
Fixes #1060, #1183
docs/have been added/updatedmvn verifysays all tests passmvn sitesays all JavaDocs correctCHANGELOG.mdhas been updatedThese PR extends the relax command by adding three new options:
--exclude-named-classes: if true, A EquivalentTo B between named classes are excluded fromrelax--include-subclass-of: if true,A Subclass A and R some Bisrelaxedthe same way asA EquivalentTo A and R some B, see @cmungall request in #1183--enforce-obo-format: If true, only simple existentials and named subsumptions are asserted during the processI like all the functionality, but the
--enforce-obo-formatis really clunky. I would be happy if someone made a better suggestion; I also dont like--exclude-nested.The only "breaking" change in this PR is that the CLI gets
--exclude-named-classesby default astrue. Previously, before this option existed, A=B was relaxed to A sub B and B sub A. No one I personally now has ever said that this is a good idea, so we changed the default, as discussed in #1060.