- move to jbuilder
- use a functorial heap (with indices embedded in lit/var)
- update Vec with optims from mc2
- change semantics of Vec.shrink
- use new Log module
When analyzing an mcst conflict clause and looking at a semantic
propagation in the trail, the last resolved clause was looked at again,
which caused an invalid history to be generated (the computation of the
backtrack clause was not affected because the second resolution of the clause
was basically a no-op thanks to the 'seen' field), thus it did not
introduce any soundness bug, just a faulty clause history which was
caught during proof expansion.
Remove the automatic clausification encoding of the coq backend, which
was slow because of the 'tauto' tactic. It was particularly a problem
for long problems, since the tauto tactic would take longer each time
because of the new hypotheses. The buren is now on the hands of the
functor argument, which hopefully should be able to do it better.
Uses a more complete tactic to go from or-separated clause to the
negation-implication encoding of clauses used by the coq backend.
Also uses a better suffix for temporary clauses than "_or".
Formerly, clauses introduced by the theory were left as is, but it
should instead be 'clausified' i.e transformed into what msat expects
(which is an encoding of clauses).
Before, atoms printed in the dot backend could either be in a html
label, or in a simple label, which caused some problems for escaping
special characters such as newlines. This commit fixes that problem by
having all labels be html labels in the dot output.
Following a bugfix for doublon elimination, doublons are now eliminated
in add_clause, and as such, in case no simplification is made, the
clause without doublons should be used (and not init). This was a typo
missed during the refactoring of doublon elimination.
