]> matita.cs.unibo.it Git - helm.git/commit
unification+pullback fix. It used to saturate a coercion (generating new metas)
authorEnrico Tassi <enrico.tassi@inria.fr>
Fri, 5 Sep 2008 08:44:21 +0000 (08:44 +0000)
committerEnrico Tassi <enrico.tassi@inria.fr>
Fri, 5 Sep 2008 08:44:21 +0000 (08:44 +0000)
commit0ad07ee93a134773c16986552dc88e4e50a437ce
tree6019550f6f4ac4975493a9007e560734aaa8ff37
parent4dc282b8b71479d45704b414d1a10a27e71752f1
unification+pullback fix. It used to saturate a coercion (generating new metas)
but not do any unification steps on them nor using them to build a new term,
thus they were left in the metasenv with no chance to be closed by subsequent
calls to unification. The meets function has been specialized takin in input a
boolean for every direction (left/right) to state if the graph can grow in this
direction and it returns saturated coercions and augmented metasenv only for
the requested directions. Still unclear to me what does it mean to search a non
triangular pullback with a boolean set to false....
helm/software/components/cic_unification/cicUnification.ml
helm/software/components/cic_unification/coercGraph.ml
helm/software/components/cic_unification/coercGraph.mli
helm/software/matita/library/formal_topology/concrete_spaces.ma