Simon Cruanes
7ef673ca30
refactor mini-cc
2020-12-22 14:59:30 -05:00
Simon Cruanes
35e5e30e93
fix(smtlib): handle "xor"
2020-12-22 14:59:21 -05:00
Simon Cruanes
7cb97a89c5
make type abstract
2020-12-22 12:36:02 -05:00
Simon Cruanes
1890f7af13
try to tighten type constraints
2020-12-22 12:30:01 -05:00
Simon Cruanes
f05de5f741
fix warnings
2020-12-22 12:17:32 -05:00
Simon Cruanes
6c1b7df301
chore: try to fix build again
2020-12-22 11:53:12 -05:00
Simon Cruanes
14bb5898f0
Merge branch 'wip-fix-lra' into wip-lra-simplex-unsat-core
2020-12-22 11:46:09 -05:00
Simon Cruanes
5ca4730968
try to fix build
2020-12-22 11:06:59 -05:00
Simon Cruanes
3b3a2e1caf
use simplex conflicts in LRA
2020-12-22 10:55:55 -05:00
Guillaume Bury
4be726db43
Add unsat core explanations to the simplex
2020-12-22 14:16:23 +01:00
Simon Cruanes
9d579af235
wip: make lazyness of model explicit
2020-12-21 16:15:49 -05:00
Simon Cruanes
b3a7acf95b
feat(LRA): handle congruence closure and theory combination in LRA
...
- merges in the CC are handled by adding corresponding equalities
locally
- theory combination pushes the decision `a=b` into the SAT solver
if a,b have the same model values and are not provably equal
in the CC already.
- also, fix model construction
2020-11-17 18:24:09 -05:00
Simon Cruanes
6417bbdd80
feat(util): add Backtrack_stack.iter
2020-11-17 18:23:58 -05:00
Simon Cruanes
9839e5a36b
feat(core): add push_decision, for model-based th combination
2020-11-17 18:23:16 -05:00
Simon Cruanes
60fe3506d5
feat(core): add CC.explain_eq to explain why two nodes were merged
2020-11-17 18:22:51 -05:00
Simon Cruanes
6a0731eeb1
refactor: improve model production in FM
2020-11-17 17:03:12 -05:00
Simon Cruanes
db1c50f7ed
feat(LRA): expose model after fourier-motzkin returns "SAT"
2020-11-17 16:52:49 -05:00
Simon Cruanes
764695bb65
feat: pass a sign along with the formula in acts_add_decision_lit
2020-11-14 18:58:40 -05:00
Simon Cruanes
77b33346f5
fix(tycheck): handle proper unary minus
2020-11-14 13:17:23 -05:00
Simon Cruanes
103c320577
fix(lra): fixed case splitting on a != b
2020-11-14 00:23:45 -05:00
Simon Cruanes
37089adb17
feat: add -t option to main
2020-11-14 00:23:35 -05:00
Simon Cruanes
b2ab465cff
fix(LRA): track explanations properly when rewriting with an equality
2020-11-13 23:58:14 -05:00
Simon Cruanes
a6f6a99fb3
fix: small perf improvement
2020-11-13 23:28:39 -05:00
Simon Cruanes
39ed753b38
add tests for simplex
2020-11-13 23:28:32 -05:00
Simon Cruanes
5ff0fff85b
wip: refactor(lra): import Simplex from funarith, replace FM with it
2020-11-13 22:35:59 -05:00
Simon Cruanes
4097ed9dc2
refactor: account for @gbury's review
2020-11-13 11:57:57 -05:00
Simon Cruanes
919c1e6011
fix: only push atoms that don't have a value into next_decisions
2020-11-13 11:57:57 -05:00
Simon Cruanes
2286a72437
fix: in final-check, resume if there are new decisions to do
2020-11-13 11:57:57 -05:00
Simon Cruanes
79b1585804
feat: allow the theory to ask for some literals to be decided on
...
it's part of the actions provided to the theory.
close #19
2020-11-13 11:57:57 -05:00
Simon Cruanes
d9bf16dfde
feat: allow to set the default polarity of variables at creation time
2020-11-13 11:57:48 -05:00
Simon Cruanes
df25e84a01
fix(LRA): fix bug in FM resolution; add more comments
2020-11-12 19:23:45 -05:00
Simon Cruanes
0fc5b279d1
fix(LRA): invalid normalization
2020-11-12 18:46:43 -05:00
Simon Cruanes
3e703cf89e
refactor: better debug in LRA
2020-11-12 18:46:34 -05:00
Simon Cruanes
9a3e387405
refactor: explicit recursion in preprocessing
...
if a preprocessor fires, it's up to it to preprocess subterms. rewriting
is now from the root, not the leaves on.
Use that in LRA to rewrite under linear expressions.
2020-11-12 18:21:37 -05:00
Simon Cruanes
51be2f52e8
style
2020-10-12 02:06:16 -04:00
Simon Cruanes
534fc45783
wip: better progress bar
2020-10-11 23:30:22 -04:00
Simon Cruanes
f556fe58ab
test: clarify where tests belong
2020-10-10 18:54:57 -04:00
Simon Cruanes
3449404964
refactor: move base-term to sidekick-arith
2020-10-10 17:43:42 -04:00
Simon Cruanes
349d884664
chore: add sidekick-arith library, depends on zarith
2020-10-10 17:18:20 -04:00
Simon Cruanes
7a74ec45b7
fix
2020-10-10 16:39:36 -04:00
Simon Cruanes
e9931b377c
remove bloody warn-error
2020-10-10 16:00:21 -04:00
Simon Cruanes
dc320bf0c9
fix(tycheck): handle n-ary +/-
2020-10-10 14:34:05 -04:00
Simon Cruanes
7c3c88d6f6
feat(lra): bugfixes
2020-10-10 14:33:54 -04:00
Simon Cruanes
93b56618f1
wip: first implem of Fourier Motzkin
2020-10-10 01:22:22 -04:00
Simon Cruanes
9783c3ae1b
wip: reimplement a fourier motzkin module, from scratch
2020-10-10 00:00:20 -04:00
Simon Cruanes
c67e44e654
detail
2020-10-09 23:59:38 -04:00
Simon Cruanes
581c7eff0b
wip
2020-10-09 22:09:15 -04:00
Simon Cruanes
7e6800811f
wip: LRA: process all lits during final check
2020-10-04 22:28:09 -04:00
Simon Cruanes
fabdb27dfe
wip: feat(lra): preprocess by renaming lits/terms and storing defs
2020-10-04 21:37:03 -04:00
Simon Cruanes
ac6ca7d584
wip: properly typecheck and build LRA terms
2020-10-04 00:32:52 -04:00