/*_________________________________________________________________________________________________
|
| simplifyDB : [void] -> [bool]
|
| Description:
| Simplify all constraints according to the current top-level assigment (redundant constraints
| may be removed altogether).
|________________________________________________________________________________________________@*/
void Solver::simplifyDB(void)
{
if (!ok) return; // GUARD (public method)
assert(decisionLevel == 0);
Constr* confl;
if ((confl = propagate()) != NULL){
if (doDeriv) {
analyzeFinalDeriv(confl);
}
ok = false;
return; }
if (nAssigns() == last_simplify)
return;
last_simplify = nAssigns();
for (int type = 0; type < 2; type++){
vec<Constr*>& cs = type ? (vec<Constr*>&)learnts : constrs;
int j = 0;
for (int i = 0; i < cs.size(); i++){
if (cs[i]->simplify(*this))
cs[i]->remove(*this);
else
cs[j++] = cs[i];
}
cs.shrink(cs.size()-j);
}
}
/*_________________________________________________________________________________________________
|
| simplify : [void] -> [bool]
|
| Description:
| Simplify the clause database according to the current top-level assigment. Currently, the only
| thing done here is the removal of satisfied clauses, but more things can be put here.
|________________________________________________________________________________________________@*/
bool MiniSATP::simplify()
{
remove_satisfied = false;
assert(decisionLevel() == 0);
// Modified Lines
// if (!ok || propagate() != NULL)
// return ok = false;
if (!ok) return false;
Clause * confl = propagate( );
if ( confl != NULL )
{
fillExplanation( confl );
return ok = false;
}
if (nAssigns() == simpDB_assigns || (simpDB_props > 0))
return true;
// Remove satisfied clauses:
removeSatisfied(learnts);
if (remove_satisfied) // Can be turned off.
removeSatisfied(clauses);
// Remove fixed variables from the variable heap:
order_heap.filter(VarFilter(*this));
simpDB_assigns = nAssigns();
simpDB_props = clauses_literals + learnts_literals; // (shouldn't depend on stats really, but it will do for now)
return true;
}
/*_________________________________________________________________________________________________
|
| simplifyDB : [void] -> [bool]
|
| Description:
| Simplify the clause database according to the current top-level assigment. Currently, the only
| thing done here is the removal of satisfied clauses, but more things can be put here.
|________________________________________________________________________________________________@*/
void Solver::simplifyDB()
{
if (!ok) return; // GUARD (public method)
assert(decisionLevel() == 0);
if (propagate() != NULL){
ok = false;
return; }
if (nAssigns() == simpDB_assigns || simpDB_props > 0) // (nothing has changed or preformed a simplification too recently)
return;
// Clear watcher lists:
for (int i = simpDB_assigns; i < nAssigns(); i++){
Lit p = trail[i];
vec<GClause>& ws = watches[index(~p)];
for (int j = 0; j < ws.size(); j++)
if (ws[j].isLit())
if (removeWatch(watches[index(~ws[j].lit())], GClause_new(p))) // (remove binary GClause from "other" watcher list)
n_bin_clauses--;
watches[index( p)].clear(true);
watches[index(~p)].clear(true);
}
// Remove satisfied clauses:
for (int type = 0; type < 2; type++){
vec<Clause*>& cs = type ? learnts : clauses;
int j = 0;
for (int i = 0; i < cs.size(); i++){
if (!locked(cs[i]) && simplify(cs[i])) // (the test for 'locked()' is currently superfluous, but without it the reason-graph is not correctly maintained for decision level 0)
remove(cs[i]);
else
cs[j++] = cs[i];
}
cs.shrink(cs.size()-j);
}
simpDB_assigns = nAssigns();
simpDB_props = stats.clauses_literals + stats.learnts_literals; // (shouldn't depend on 'stats' really, but it will do for now)
}
/*_________________________________________________________________________________________________
|
| simplify : [void] -> [bool]
|
| Description:
| Simplify the clause database according to the current top-level assigment. Currently, the only
| thing done here is the removal of satisfied clauses, but more things can be put here.
|________________________________________________________________________________________________@*/
bool Solver::simplify()
{
assert(decisionLevel() == 0);
if (!ok || propagate() != NULL)
return ok = false;
if (nAssigns() == simpDB_assigns || (simpDB_props > 0))
return true;
// Remove satisfied clauses:
removeSatisfied(learnts);
if (remove_satisfied) // Can be turned off.
removeSatisfied(clauses);
// Remove fixed variables from the variable heap:
order_heap.filter(VarFilter(*this));
simpDB_assigns = nAssigns();
simpDB_props = clauses_literals + learnts_literals; // (shouldn't depend on stats really, but it will do for now)
return true;
}
/*_________________________________________________________________________________________________
|
| search : (nof_conflicts : int) (nof_learnts : int) (params : const SearchParams&) -> [lbool]
|
| Description:
| Search for a model the specified number of conflicts, keeping the number of learnt clauses
| below the provided limit. NOTE! Use negative value for 'nof_conflicts' or 'nof_learnts' to
| indicate infinity.
|
| Output:
| 'l_True' if a partial assigment that is consistent with respect to the clauseset is found. If
| all variables are decision variables, this means that the clause set is satisfiable. 'l_False'
| if the clause set is unsatisfiable. 'l_Undef' if the bound on number of conflicts is reached.
|________________________________________________________________________________________________@*/
lbool Solver::search(int nof_conflicts, int nof_learnts, const SearchParams& params)
{
if (!ok) return l_False; // GUARD (public method)
assert(root_level == decisionLevel);
stats.starts++;
int conflictC = 0;
var_decay = 1 / params.var_decay;
cla_decay = 1 / params.clause_decay;
model.clear();
for (;;){
Constr* confl = propagate();
if (confl != NULL){
// CONFLICT
//if (verbosity >= 2) printf(L_IND"**CONFLICT**\n", L_ind);
stats.conflicts++; conflictC++;
vec<Lit> learnt_clause;
int backtrack_level;
if (decisionLevel == root_level) {
if (doDeriv) {
analyzeFinalDeriv(confl);
}
return l_False;
}
Deriv* deriv = new Deriv();
derivs.push_back(deriv);
analyze(confl, learnt_clause, backtrack_level, deriv);
cancelUntil(max(backtrack_level, root_level));
record(learnt_clause, deriv);
varDecayActivity(); claDecayActivity();
}else{
// NO CONFLICT
if (nof_conflicts >= 0 && conflictC >= nof_conflicts){
// Reached bound on number of conflicts:
progress_estimate = progressEstimate();
propQ.clear();
cancelUntil(root_level);
return l_Undef; }
if (decisionLevel == 0)
// Simplify the set of problem clauses:
simplifyDB(), assert(ok);
if (nof_learnts >= 0 && learnts.size()-nAssigns() >= nof_learnts)
// Reduce the set of learnt clauses:
reduceDB();
// New variable decision:
stats.decisions++;
Var next = order.select(params.random_var_freq);
if (next == var_Undef){
// Model found:
model.growTo(nVars);
for (int i = 0; i < nVars; i++) model[i] = (value(i) == l_True);
cancelUntil(root_level);
return l_True;
}
// VERY important for the Y-variables to try positive polarity first
// - follows the value ordering heuristic of aiming for solutions (as opposed to fail-first var order)
check(assume(Lit(next)));
//check(assume(~Lit(next)));
}
}
}
/*_________________________________________________________________________________________________
|
| search : (nof_conflicts : int) (nof_learnts : int) (params : const SearchParams&) -> [lbool]
|
| Description:
| Search for a model the specified number of conflicts, keeping the number of learnt clauses
| below the provided limit. NOTE! Use negative value for 'nof_conflicts' or 'nof_learnts' to
| indicate infinity.
|
| Output:
| 'l_True' if a partial assigment that is consistent with respect to the clauseset is found. If
| all variables are decision variables, this means that the clause set is satisfiable. 'l_False'
| if the clause set is unsatisfiable. 'l_Undef' if the bound on number of conflicts is reached.
|________________________________________________________________________________________________@*/
lbool Solver::search(int nof_conflicts, int nof_learnts)
{
assert(ok);
int backtrack_level;
int conflictC = 0;
vec<Lit> learnt_clause;
starts++;
for (;;){
#ifdef _DEBUGSEARCH
for(int k=0; k<decisionLevel(); ++k)
std::cout << " ";
std::cout << "propagate" << std::endl;
#endif
Clause* confl = propagate();
if (confl != NULL){
// CONFLICT
conflicts++; conflictC++;
if (decisionLevel() <= init_level) {
#ifdef _DEBUGSEARCH
std::cout << "infeasible" << std::endl;
#endif
return l_False;
}
learnt_clause.clear();
analyze(confl, learnt_clause, backtrack_level);
cancelUntil(backtrack_level);
#ifdef _DEBUGSEARCH
for(int k=0; k<decisionLevel(); ++k)
std::cout << " ";
std::cout << "backtrack to " << decisionLevel() << std::endl;
#endif
assert(value(learnt_clause[0]) == l_Undef);
if (learnt_clause.size() == 1){
uncheckedEnqueue(learnt_clause[0]);
#ifdef _DEBUGSEARCH
for(int k=0; k<decisionLevel(); ++k)
std::cout << " ";
std::cout << "deduce " ;
printLit(learnt_clause[0]);
std::cout << std::endl;
#endif
}else{
Clause* c = Clause_new(learnt_clause, true);
#ifdef _DEBUGSEARCH
for(int k=0; k<decisionLevel(); ++k)
std::cout << " ";
std::cout << "deduce ";
printClause(*c);
std::cout << std::endl;
#endif
learnts.push(c);
attachClause(*c);
claBumpActivity(*c);
uncheckedEnqueue(learnt_clause[0], c);
}
varDecayActivity();
claDecayActivity();
}else{
// NO CONFLICT
if (nof_conflicts >= 0 && conflictC >= nof_conflicts){
// Reached bound on number of conflicts:
progress_estimate = progressEstimate();
cancelUntil(0);
return l_Undef; }
// Simplify the set of problem clauses:
if (decisionLevel() == 0 && !simplify())
return l_False;
if (nof_learnts >= 0 && learnts.size()-nAssigns() >= nof_learnts)
// Reduce the set of learnt clauses:
reduceDB();
Lit next = lit_Undef;
while (decisionLevel() < assumptions.size()){
// Perform user provided assumption:
Lit p = assumptions[decisionLevel()];
if (value(p) == l_True){
// Dummy decision level:
newDecisionLevel();
}else if (value(p) == l_False){
analyzeFinal(~p, conflict);
return l_False;
}else{
next = p;
break;
}
}
if (next == lit_Undef){
// New variable decision:
decisions++;
next = pickBranchLit(polarity_mode, random_var_freq);
if (next == lit_Undef)
// Model found:
return l_True;
}
#ifdef _DEBUGSEARCH
for(int k=0; k<decisionLevel(); ++k)
std::cout << " ";
std::cout << "branch on " ;
printLit(next);
std::cout << std::endl;
#endif
// Increase decision level and enqueue 'next'
assert(value(next) == l_Undef);
newDecisionLevel();
uncheckedEnqueue(next);
}
}
}
/*_________________________________________________________________________________________________
|
| search : (nof_conflicts : int) (nof_learnts : int) (params : const SearchParams&) -> [lbool]
|
| Description:
| Search for a model the specified number of conflicts, keeping the number of learnt clauses
| below the provided limit. NOTE! Use negative value for 'nof_conflicts' or 'nof_learnts' to
| indicate infinity.
|
| Output:
| 'l_True' if a partial assigment that is consistent with respect to the clauseset is found. If
| all variables are decision variables, this means that the clause set is satisfiable. 'l_False'
| if the clause set is unsatisfiable. 'l_Undef' if the bound on number of conflicts is reached.
|________________________________________________________________________________________________@*/
lbool MiniSATP::search(int nof_conflicts, int nof_learnts)
{
assert(ok);
int backtrack_level;
int conflictC = 0;
vec<Lit> learnt_clause;
starts++;
bool first = true;
for (;;){
Clause* confl = propagate();
if (confl != NULL){
// CONFLICT
conflicts++; conflictC++;
if (decisionLevel() == 0)
{
// Added Lines
fillExplanation(confl);
return l_False;
}
first = false;
learnt_clause.clear();
analyze(confl, learnt_clause, backtrack_level);
cancelUntil(backtrack_level);
assert(value(learnt_clause[0]) == l_Undef);
if (learnt_clause.size() == 1){
uncheckedEnqueue(learnt_clause[0]);
}else{
Clause* c = Clause_new(learnt_clause, true);
learnts.push(c);
attachClause(*c);
claBumpActivity(*c);
uncheckedEnqueue(learnt_clause[0], c);
}
varDecayActivity();
claDecayActivity();
}else{
// NO CONFLICT
if (nof_conflicts >= 0 && conflictC >= nof_conflicts){
// Reached bound on number of conflicts:
progress_estimate = progressEstimate();
cancelUntil(0);
return l_Undef; }
// Simplify the set of problem clauses:
if (decisionLevel() == 0 && !simplify())
{
return l_False;
}
if (nof_learnts >= 0 && learnts.size()-nAssigns() >= nof_learnts)
// Reduce the set of learnt clauses:
reduceDB();
Lit next = lit_Undef;
while (decisionLevel() < assumptions.size()){
// Perform user provided assumption:
Lit p = assumptions[decisionLevel()];
if (value(p) == l_True){
// Dummy decision level:
newDecisionLevel();
}else if (value(p) == l_False){
analyzeFinal(~p, conflict);
return l_False;
}else{
next = p;
break;
}
}
if (next == lit_Undef){
// New variable decision:
decisions++;
next = pickBranchLit(polarity_mode, random_var_freq);
if (next == lit_Undef)
{
// Added Line
// Clear explanation vector if satisfiable
explanation.clear( );
// Model found:
return l_True;
}
}
// Increase decision level and enqueue 'next'
assert(value(next) == l_Undef);
newDecisionLevel();
uncheckedEnqueue(next);
}
}
}
/*_________________________________________________________________________________________________
|
| search : (nof_conflicts : int) (nof_learnts : int) (params : const SearchParams&) -> [lbool]
|
| Description:
| Search for a model the specified number of conflicts, keeping the number of learnt clauses
| below the provided limit. NOTE! Use negative value for 'nof_conflicts' or 'nof_learnts' to
| indicate infinity.
|
| Output:
| 'l_True' if a partial assigment that is consistent with respect to the clauseset is found. If
| all variables are decision variables, this means that the clause set is satisfiable. 'l_False'
| if the clause set is unsatisfiable. 'l_Undef' if the bound on number of conflicts is reached.
|________________________________________________________________________________________________@*/
lbool Solver::search(int nof_conflicts, int nof_learnts, const SearchParams& params)
{
if (!ok) return l_False; // GUARD (public method)
assert(root_level == decisionLevel());
stats.starts++;
int conflictC = 0;
var_decay = 1 / params.var_decay;
cla_decay = 1 / params.clause_decay;
model.clear();
for (;;){
Clause* confl = propagate();
if (confl != NULL){
// CONFLICT
stats.conflicts++; conflictC++;
vec<Lit> learnt_clause;
int backtrack_level;
if (decisionLevel() == root_level){
// Contradiction found:
analyzeFinal(confl);
return l_False; }
analyze(confl, learnt_clause, backtrack_level);
cancelUntil(max(backtrack_level, root_level));
newClause(learnt_clause, true);
if (learnt_clause.size() == 1) level[var(learnt_clause[0])] = 0; // (this is ugly (but needed for 'analyzeFinal()') -- in future versions, we will backtrack past the 'root_level' and redo the assumptions)
varDecayActivity();
claDecayActivity();
}else{
// NO CONFLICT
if (nof_conflicts >= 0 && conflictC >= nof_conflicts){
// Reached bound on number of conflicts:
progress_estimate = progressEstimate();
cancelUntil(root_level);
return l_Undef; }
if (decisionLevel() == 0)
// Simplify the set of problem clauses:
simplifyDB(), assert(ok);
if (nof_learnts >= 0 && learnts.size()-nAssigns() >= nof_learnts)
// Reduce the set of learnt clauses:
reduceDB();
// New variable decision:
stats.decisions++;
Var next = order.select(params.random_var_freq);
if (next == var_Undef){
// Model found:
model.growTo(nVars());
for (int i = 0; i < nVars(); i++) model[i] = value(i);
cancelUntil(root_level);
return l_True;
}
check(assume(~Lit(next)));
}
}
}
/*_________________________________________________________________________________________________
|
| search : (nof_conflicts : int) (nof_learnts : int) (params : const SearchParams&) -> [lbool]
|
| Description:
| Search for a model the specified number of conflicts, keeping the number of learnt clauses
| below the provided limit. NOTE! Use negative value for 'nof_conflicts' or 'nof_learnts' to
| indicate infinity.
|
| Output:
| 'l_True' if a partial assigment that is consistent with respect to the clauseset is found. If
| all variables are decision variables, this means that the clause set is satisfiable. 'l_False'
| if the clause set is unsatisfiable. 'l_Undef' if the bound on number of conflicts is reached.
|________________________________________________________________________________________________@*/
lbool Solver::search(int nof_conflicts, int nof_learnts)
{
int backtrack_level;
int conflictC = 0;
vec<Lit> learnt_clause;
starts++;
// bool first = true;
for (;;){
Clause* confl = propagate();
if (confl != NULL){
// CONFLICT
conflicts++; conflictC++;
if (decisionLevel() == 0) return l_False;
// first = false;
learnt_clause.clear();
analyze(confl, learnt_clause, backtrack_level);
cancelUntil(backtrack_level);
#ifdef __PRINT
char c1 = sign(learnt_clause[0]) ? '-' : '+';
char c2 = polarity[var(learnt_clause[0])] == 0 ? '-' : '+';
if (decisionLevel() < minDecisionLevel &&
original_activity[var(learnt_clause[0])] > 0) {
printf("Conflict record: ");
printLit(learnt_clause[0]);
printf(" .%d.\t.%d. %c .%c .%g\n", decisionLevel(), trail.size(),
c1, c2, original_activity[var(learnt_clause[0])]);
if (decisionLevel() == 0) {
minDecisionLevel = (unsigned)(-1);
} else {
minDecisionLevel = decisionLevel();
}
}
#endif
if (learnt_clause.size() == 1){
uncheckedEnqueue(learnt_clause[0]);
}else{
Clause* c = Clause::Clause_new(learnt_clause, true);
learnts.push(c);
attachClause(*c);
claBumpActivity(*c);
uncheckedEnqueue(learnt_clause[0], c);
}
#ifdef _MINISAT_DEFAULT_VSS
varDecayActivity();
#endif
claDecayActivity();
}else{
// NO CONFLICT
if (nof_conflicts >= 0 && conflictC >= nof_conflicts){
// Reached bound on number of conflicts:
progress_estimate = progressEstimate();
// cancelUntil(0);
return l_Undef; }
// Simplify the set of problem clauses:
if (decisionLevel() == 0 && !simplify())
return l_False;
if (nof_learnts >= 0 && learnts.size()-nAssigns() >= nof_learnts)
// Reduce the set of learnt clauses:
reduceDB();
Lit next = lit_Undef;
while (decisionLevel() < assumptions.size()){
// Perform user provided assumption:
Lit p = assumptions[decisionLevel()];
if (value(p) == l_True){
// Dummy decision level:
newDecisionLevel();
}else if (value(p) == l_False){
analyzeFinal(~p, conflict);
return l_False;
}else{
next = p;
break;
}
}
if (next == lit_Undef){
// New variable decision:
decisions++;
next = pickBranchLit(polarity_mode, random_var_freq);
if (next == lit_Undef)
// Model found:
return l_True;
#ifdef __PRINT
printf("Decision: ");
printLit(next);
printf("\t.%f.\t.%d.\t.%d.", activity[var(next)], decisionLevel(), trail.size());
printf("\n");
#endif
}
// Increase decision level and enqueue 'next'
newDecisionLevel();
uncheckedEnqueue(next);
}
//#ifdef __PRINT
// printTrail();
//#endif
}
}