Make learned literal computation more robust (#8308)
authorAndrew Reynolds <andrew.j.reynolds@gmail.com>
Tue, 15 Mar 2022 23:32:45 +0000 (18:32 -0500)
committerGitHub <noreply@github.com>
Tue, 15 Mar 2022 23:32:45 +0000 (23:32 +0000)
commit4128af1ac9aba2512aa339b2276c5e8fc719af45
tree1b91dc75582bbd28ca15cd589a3c9bfe4bbd3cde
parente8ef445fe7b026e8bee8d76a7ad40cfcf1072517
Make learned literal computation more robust (#8308)

Simplifies our computation so that preprocessed learned literals are those exactly appearing as top-level assertions only.
src/prop/prop_engine.cpp
src/prop/prop_engine.h
src/prop/theory_proxy.cpp
src/prop/theory_proxy.h
src/prop/zero_level_learner.cpp
src/prop/zero_level_learner.h
src/smt/smt_solver.cpp
test/regress/regress0/printer/learned-lit-output.smt2