void Car::addNewCost(QDate date, unsigned int distance, QString description, double price)
{
Cost *cost = new Cost(date,distance,description,price,CREATE_NEW_EVENT,this);
_costlist.append(cost);
qSort(_costlist.begin(), _costlist.end(), sortCostByDistance);
cost->save();
emit costsChanged();
}
void PathLengthDirectInfSampler::sampleUniform(State* statePtr, const Cost& maxCost)
{
//Check if a solution path has been found
if (std::isfinite(maxCost.value()) == false)
{
//We don't have a solution yet, we sample from our basic sampler instead...
baseSampler_->sampleUniform(statePtr);
}
else //We have a solution
{
//Set the new transverse diameter
phsPtr_->setTransverseDiameter(maxCost.value());
//Check whether the problem domain (i.e., StateSpace) or PHS has the smaller measure. Sample the smaller directly and reject from the larger.
if (informedSubSpace_->getMeasure() <= phsPtr_->getPhsMeasure())
{
//The PHS is larger than the subspace, just sample from the subspace directly.
//Variables
//The informed subset of the sample as a vector
std::vector<double> informedVector(informedSubSpace_->getDimension());
//Sample from the state space until the sample is in the PHS
do
{
//Generate a random sample
baseSampler_->sampleUniform(statePtr);
//Is there an extra "uninformed" subspace to trim off before comparing to the PHS?
if ( StateSampler::space_->isCompound() == false )
{
//No, space_ == informedSubSpace_
informedSubSpace_->copyToReals(informedVector, statePtr);
}
else
{
//Yes, we need to do some work to extract the subspace
informedSubSpace_->copyToReals(informedVector, statePtr->as<CompoundState>()->components[informedIdx_]);
}
}
//Check if the informed state is in the PHS
while ( phsPtr_->isInPhs(informedSubSpace_->getDimension(), &informedVector[0]) == false );
}
else
{
//The PHS has a smaller volume than the subspace.
//Sample from within the PHS until the sample is in the state space
do
{
this->sampleUniformIgnoreBounds(statePtr, maxCost);
}
while ( StateSampler::space_->satisfiesBounds(statePtr) == false );
}
}
}
Cost IntersectQP::cost(OperationContext &context, QueryExecutionContext &qec) const
{
Cost result;
Vector::const_iterator it = args_.begin();
if(it != args_.end()) {
result = (*it)->cost(context, qec);
for(++it; it != args_.end(); ++it) {
result.intersectOp((*it)->cost(context, qec));
}
}
return result;
}
ompl::base::Cost ompl::base::MultiOptimizationObjective::stateCost(const State *s) const
{
Cost c = identityCost();
for (std::vector<Component>::const_iterator comp = components_.begin();
comp != components_.end();
++comp)
{
c = Cost(c.value() + comp->weight * (comp->objective->stateCost(s).value()));
}
return c;
}
static void* Cost_optimizeA(void* object){
pthread_setname_np(pthread_self(), "Athread");
Cost* cost = (Cost*)object;
set_affinity(3);
cost->running_a = true;
while(cost->running_a) {
cost->optimizeA();
if(allDie) {
return 0;
}
}
}
static void* Cost_optimizeQD(void* object){
pthread_setname_np(pthread_self(), "QDthread");
Cost* cost = (Cost*)object;
set_affinity(2);
cost->running_qd = true;
while(cost->running_a) { // A is not done
cost->optimizeQD();
if(allDie) {
return 0;
}
}
cost->running_qd = false;
}
inline double ticksToAvgPrice(Lots lots, Cost cost, const Contr& contr) noexcept
{
double ticks = 0;
if (lots != 0_lts) {
ticks = fractToReal(cost.count(), lots.count());
}
return ticks * contr.priceInc();
}
Cost NegativeNodePredicateFilterQP::cost(OperationContext &context, QueryExecutionContext &qec) const
{
Cost cost = arg_->cost(context, qec);
Cost predCost = pred_->cost(context, qec);
// The predicate expression is run once for every argument key
// cost.pages += ceil(cost.keys * predCost.pages);
cost.pagesForKeys += cost.keys * predCost.totalPages();
// Take a token key away from the result, because it is likely to
// return less nodes than the argument - we just don't know how many less.
if(cost.keys > 1) cost.keys -= 1;
// Add a token page to the result, because having a predicate takes more
// time than not
cost.pagesOverhead += 1;
return cost;
}
Track::Track(Cost cost){
rows=cost.rows;
cols=cost.cols;
baseImage=lastFrame=thisFrame=cost.baseImage;
cameraMatrix=Mat(cost.cameraMatrix);
depth=cost.depthMap();
PToLie(Mat(cost.basePose),basePose);
pose=basePose.clone();
}
void solve_nonlinear () {
for (size_t i = 0; i < signals.rows(); ++i) {
const Math::Vector<cost_value_type> signal (signals.row(i));
Math::Vector<cost_value_type> values (tensors.row(i));
cost.set_voxel (&signal, &values);
Math::Vector<cost_value_type> x (cost.size());
cost.init (x);
//Math::check_function_gradient (cost, x, 1e-10, true);
Math::GradientDescent<Cost> optim (cost);
try { optim.run (10000, 1e-8); }
catch (Exception& E) {
E.display();
}
//x = optim.state();
//Math::check_function_gradient (cost, x, 1e-10, true);
cost.get_values (values, optim.state());
}
}
ompl::base::Cost ompl::base::StateCostIntegralObjective::motionCost(const State *s1, const State *s2) const
{
if (interpolateMotionCost_)
{
Cost totalCost = this->identityCost();
int nd = si_->getStateSpace()->validSegmentCount(s1, s2);
State *test1 = si_->cloneState(s1);
Cost prevStateCost = this->stateCost(test1);
if (nd > 1)
{
State *test2 = si_->allocState();
for (int j = 1; j < nd; ++j)
{
si_->getStateSpace()->interpolate(s1, s2, (double)j / (double)nd, test2);
Cost nextStateCost = this->stateCost(test2);
totalCost = Cost(totalCost.value() +
this->trapezoid(prevStateCost, nextStateCost, si_->distance(test1, test2)).value());
std::swap(test1, test2);
prevStateCost = nextStateCost;
}
si_->freeState(test2);
}
// Lastly, add s2
totalCost = Cost(totalCost.value() +
this->trapezoid(prevStateCost, this->stateCost(s2), si_->distance(test1, s2)).value());
si_->freeState(test1);
return totalCost;
}
return this->trapezoid(this->stateCost(s1), this->stateCost(s2), si_->distance(s1, s2));
}
double PathLengthDirectInfSampler::getInformedMeasure(const Cost& currentCost) const
{
//Variable
//The measure of the informed set
double informedMeasure;
//The informed measure is then the measure of the PHS for the given cost:
informedMeasure = phsPtr_->getPhsMeasure(currentCost.value());
//And if the space is compound, further multiplied by the measure of the uniformed subspace
if ( StateSampler::space_->isCompound() == true )
{
informedMeasure = informedMeasure*uninformedSubSpace_->getMeasure();
}
//Return the smaller of the two measures
return std::min(StateSampler::space_->getMeasure(), informedMeasure);
}
void PathLengthDirectInfSampler::sampleUniformIgnoreBounds(State* statePtr, const Cost& maxCost)
{
//Variable
//The informed subset of the sample as a vector
std::vector<double> informedVector(informedSubSpace_->getDimension());
//Set the new transverse diameter
phsPtr_->setTransverseDiameter(maxCost.value());
//Sample the ellipse
rng_.uniformProlateHyperspheroid(phsPtr_, informedSubSpace_->getDimension(), &informedVector[0]);
//If there is an extra "uninformed" subspace, we need to add that to the state before converting the raw vector representation into a state....
if ( StateSampler::space_->isCompound() == false )
{
//No, space_ == informedSubSpace_
//Copy into the state pointer
informedSubSpace_->copyFromReals(statePtr, informedVector);
}
else
{
//Yes, we need to also sample the uninformed subspace
//Variables
//A state for the uninformed subspace
State* uninformedState = uninformedSubSpace_->allocState();
//Copy the informed subspace into the state pointer
informedSubSpace_->copyFromReals(statePtr->as<CompoundState>()->components[informedIdx_], informedVector);
//Sample the uniformed subspace
uninformedSubSampler_->sampleUniform(uninformedState);
//Copy the informed subspace into the state pointer
uninformedSubSpace_->copyState(statePtr->as<CompoundState>()->components[uninformedIdx_], uninformedState);
//Free the state
uninformedSubSpace_->freeState(uninformedState);
}
}
ompl::base::Cost ompl::base::OptimizationObjective::combineCosts(Cost c1, Cost c2) const
{
return Cost(c1.value() + c2.value());
}
std::vector<std::pair<std::vector<uint>, std::vector<SentenceInfo> aStar::Suchalgorithmus(char* eingabe, PTree<PTree <Cost> >* blacktree, Lexicon* eLex, Lexicon* fLex){
igzstream in(eingabe);
aStar::flex=fLex;
elex=eLex;
schwarz=blacktree;
string token,line;
unsigned int lineNumber = 0;
while(getline(in,line)){
istringstream ist(line); //Einlesen des Satzes
vector<unsigned int> sentence_id;
vector<HypothesisNode> Knoten;
Knoten.push_back(HypothesisNode());//initialisiert den ersten Knoten
Cost startkosten(0);
Knoten[0].setBestcost(startkosten);
//cout << "start kosten " << Knoten[0].getBestcost().cost() << endl;
int aktPos=0; //merkt sich, wieviele Wörter schon eingelesen wurden
while ( ist >> token){
Word word_id_french=flex->getWord_or_add(token); // das word zum Wort (mit 2 Bits Sprache)
unsigned int id_french= word_id_french.wordId(); //die id ohne sprachbits
sentence_id.push_back(id_french);
aktPos++;
HypothesisNode knoten_next= HypothesisNode();//initialisiert den nächsten Knoten mit den bisherigen Kosten
for (int laengePhrase=1; laengePhrase<5; laengePhrase++){
int posPhraseStart=aktPos-laengePhrase; //gibt die Pos. für den Knoten, auf dem die Phrase beginnt
if (posPhraseStart < 0) break; //wir befinden uns am Satzanfang und es gibt keine Phrasen
vector<unsigned int> fphrase;
for (int i=posPhraseStart; i<aktPos; i++){
fphrase.push_back(sentence_id[i]);
}
PTree<PTree <Cost> >* schwarzRest=schwarz->traverse(fphrase);
if (!schwarzRest) continue; //wenn es die französische Phrase nicht gibt, nächste überprüfen
PTree <Cost>* blauBaum=&schwarzRest->c;
if (blauBaum){
int counter=0; //nur fürs Programmieren, damit alle Fehler ausgemerzt werden
for (PTree<Cost>::iterator it=blauBaum->begin(); it!=blauBaum->end(); it++){
//if (counter++==10) continue;
vector<unsigned int> ephrase=it->phrase();
Cost relfreq = it->c;
if (relfreq.cost() == 1./0. ) continue;
double cost_till_aktPos=Knoten[posPhraseStart].getBestcost().cost();
if (cost_till_aktPos+prune > knoten_next.getBestcost().cost()) continue; //pruning ergibt, das ist eine schlecht Übersetzung
if(cost_till_aktPos+relfreq.cost() < knoten_next.getBestcost().cost()) knoten_next.setBestcost(Knoten[posPhraseStart].getBestcost()+relfreq.cost());
PartialTranslation* Kante= new PartialTranslation(relfreq,ephrase,&knoten_next,posPhraseStart);
Knoten[posPhraseStart].add_PartialTranslation_to_Inbound(Kante);
knoten_next.add_PartialTranslation_to_Outbound(Kante);
}
}
}
if (knoten_next.getOutbound().size() == 0){
//zuerst in das englische Lexicon einfügen
string word_string = flex->getString(sentence_id[aktPos-1]);
unsigned int id_english=elex->getWord_or_add(word_string);
//dann die Kante anlegen, dabei sollen die Kosten niedrig sein, da sie sowieso genutzt werden muss, kann sie auch direkt exploriert werden
PartialTranslation* Kante= new PartialTranslation(Cost(0),vector<unsigned int>{id_english},&Knoten[aktPos],aktPos-1);
knoten_next.setBestcost(Knoten.back().getBestcost());
knoten_next.add_PartialTranslation_to_Outbound(Kante);
Knoten.back().add_PartialTranslation_to_Inbound(Kante);
}
Knoten.push_back(knoten_next); //letzter Knoten (node_next) hat keine eingehenden Kanten
}
//dotGraph(Knoten, elex);
aStar astar(Knoten);
astar.lineNumber = lineNumber;
std::vector<SentenceInfo> sentenceInfos = astar.print(astar.search(), sentence_id, schwarz);
for(unsigned int i = 0; i < Knoten.size(); i++){
HypothesisNode& hnode = Knoten[i];
for(unsigned int j = 0; j < hnode.getOutbound().size(); j++)
delete hnode.getOutbound()[j];
}
std::pair<std::vector<uint>, std::vector<SentenceInfo> > pair_tmp;
pair_tmp.first = sentence_id;
pair_tmp.second = sentenceInfos;
nBestLists.push_back(pair_tmp);
lineNumber ++;
}
return nBestLists;
}
bool ompl::base::OptimizationObjective::isCostBetterThan(Cost c1, Cost c2) const
{
return c1.value() < c2.value();
}
void addCost (const Cost& c)
{
mCosts [static_cast <int> (c.getLoadType ())] = c;
}
int main(int argc, char* argv[]) {
vector<Core> core;
core.push_back(Core(0, 1));
core.push_back(Core(1, 4));
core.push_back(Core(4, 6));
core.push_back(Core(3, 0));
core.push_back(Core(4, 3));
double** bandwidth;
double** latency;
bandwidth = new double*[5];
latency = new double*[5];
for (int i = 0; i < 5; i++) {
bandwidth[i] = new double[5];
memset(bandwidth[i], 0, sizeof(double) * 5);
latency[i] = new double[5];
memset(latency[i], 0, sizeof(double) * 5);
}
bandwidth[0][3] = 10;
bandwidth[0][4] = 20;
bandwidth[1][0] = 30;
bandwidth[1][4] = 40;
bandwidth[2][3] = 25;
bandwidth[3][1] = 40;
bandwidth[3][2] = 50;
bandwidth[4][0] = 10;
bandwidth[4][1] = 30;
latency[0][3] = 10;
latency[0][4] = 10;
latency[1][0] = 10;
latency[1][4] = 10;
latency[2][3] = 10;
latency[3][1] = 10;
latency[3][2] = 10;
latency[4][0] = 10;
latency[4][1] = 10;
Network network;
network.init(7, 5);
Coordinate from = { 0, 1 };
Coordinate to = { 3, 0 };
network.addCore(from, 0); //core1
network.addCore(to, 3); //core4
network.changeConnection(from, to, ADD); //1-4
to.x = 4;
to.y = 3;
network.addCore(to, 4); //core5
network.changeConnection(from, to, ADD); //1-5
from.x = 1;
from.y = 4;
network.addCore(from, 1); //core2
to.x = 0;
to.y = 1;
network.changeConnection(from, to, ADD); //2-1
to.x = 4;
to.y = 3;
network.changeConnection(from, to, ADD); //2-5
from.x = 4;
from.y = 6;
network.addCore(from, 2); //core3
to.x = 3;
to.y = 0;
network.changeConnection(from, to, ADD); //3-4
from.x = 3;
from.y = 0;
to.x = 1;
to.y = 4;
network.changeConnection(from, to, ADD); //4-2
to.x = 4;
to.y = 6;
network.changeConnection(from, to, ADD); //4-3
from.x = 4;
from.y = 3;
to.x = 0;
to.y = 1;
network.changeConnection(from, to, ADD); //5-1
to.x = 1;
to.y = 4;
network.changeConnection(from, to, ADD); //5-2
Cost cost;
cost.init(1, 1, 0.2, 0.04);
cost.initCost(bandwidth, latency, core, 1, network);
network.updateUtilization(bandwidth, core);
network.showDiagram();
cost.printCost();
cout << endl;
int changedCore = 3;
Coordinate newPos = { 4, 6 };
int swapCore = network.getCoreIndex(newPos);
assert(swapCore == 2);
Coordinate oldPos = core[changedCore].getPosition();
//remove old cost (compaction, slack, proximity)
cost.updateCost(bandwidth, latency, 1, core, REMOVE, changedCore,
swapCore);
//remove all connections from the changed core
network.changeAllConnections(bandwidth, core, changedCore, REMOVE);
//remove all connections from the old position of the swap core
network.changeAllConnections(bandwidth, core, swapCore, REMOVE);
//add the overlap
if (bandwidth[changedCore][swapCore] != 0) {
network.changeConnection(core[changedCore].getPosition(),
core[swapCore].getPosition(), ADD);
}
if (bandwidth[swapCore][changedCore] != 0) {
network.changeConnection(core[swapCore].getPosition(),
core[changedCore].getPosition(), ADD);
}
//place core on new pos
core[swapCore].setPosition(oldPos);
network.addCore(oldPos, swapCore);
core[changedCore].setPosition(newPos);
network.addCore(newPos, changedCore);
//add all connections of changedCore
network.changeAllConnections(bandwidth, core, changedCore, ADD);
//add all connections of swap core
network.changeAllConnections(bandwidth, core, swapCore, ADD);
//remove overlap
if (bandwidth[changedCore][swapCore] != 0) {
network.changeConnection(core[changedCore].getPosition(),
core[swapCore].getPosition(), REMOVE);
}
if (bandwidth[swapCore][changedCore] != 0) {
network.changeConnection(core[swapCore].getPosition(),
core[changedCore].getPosition(), REMOVE);
}
//calculate new cost (compaction, slack, proximity)
cost.updateCost(bandwidth, latency, 1, core, ADD, changedCore,
swapCore);
//calculate new cost
//cost.initCost(bandwidth, latency, core, LINK_LATENCY, network);
cost.calculateCost(bandwidth, core, network);
network.updateUtilization(bandwidth, core);
//network.printNetwork();
network.showDiagram();
cost.printCost();
cout << endl;
return 0;
}
QueryPlan *IntersectQP::optimize(OptimizationContext &opt)
{
// Optimize the arguments
vector<QueryPlan*> newArgs;
Vector::iterator it;
for(it = args_.begin(); it != args_.end(); ++it) {
QueryPlan *arg = (*it)->optimize(opt);
if(arg->getType() == type_) {
// Merge IntersectQP arguments into this object
const Vector &args = ((OperationQP*)arg)->getArgs();
Vector::const_iterator aend = args.end();
for(Vector::const_iterator ai = args.begin();
ai != aend; ++ai) newArgs.push_back(*ai);
} else {
newArgs.push_back(arg);
}
}
args_.clear();
std::copy(newArgs.begin(), newArgs.end(), back_inserter(args_));
removeSupersets(opt);
// Return if we only have one argument
if(args_.size() == 1) return args_[0];
// Pull forward filters
for(it = args_.begin(); it != args_.end(); ++it) {
switch((*it)->getType()) {
case VALUE_FILTER:
case PREDICATE_FILTER:
case NODE_PREDICATE_FILTER:
case NEGATIVE_NODE_PREDICATE_FILTER:
// We can't skip NUMERIC_PREDICATE_FILTER, because changing it's input will change
// the cardinality of it's input, and cause it to return the wrong results.
// case NUMERIC_PREDICATE_FILTER:
// case DOC_EXISTS:
case LEVEL_FILTER: {
string before = logBefore(this);
FilterQP *filter = (FilterQP*)*it;
*it = filter->getArg();
filter->setArg(this);
logTransformation(opt.getLog(), "Filter pulled forward", before, filter);
return filter->optimize(opt);
}
default: break;
}
}
// Identify potential pairs of ValueQP for a RangeQP
newArgs.clear();
for(it = args_.begin(); it != args_.end(); ++it) {
if(isLessThanOrGreaterThan(*it)) {
for(Vector::iterator it2 = it + 1;
it2 != args_.end(); ++it2) {
if(isLessThanOrGreaterThan(*it2)) {
QueryPlan *range = createRange((ValueQP*)*it, (ValueQP*)*it2);
if(range != 0) {
logTransformation(opt.getLog(), "Merged into range",
logIntersectBefore(*it, *it2), range);
newArgs.push_back(range);
}
}
}
}
newArgs.push_back((*it)->optimize(opt));
}
args_.clear();
std::copy(newArgs.begin(), newArgs.end(), back_inserter(args_));
// Return if we only have one argument
if(args_.size() == 1) return args_[0];
// Try to pull forward document indexes, so they will combine
if(opt.getPhase() < OptimizationContext::ALTERNATIVES) {
string before = logBefore(this);
QueryPlan *result = PullForwardDocumentJoin().run(this);
if(result != 0) {
logTransformation(opt.getLog(), "Pull forward document join", before, result);
return result->optimize(opt);
}
}
if(opt.getPhase() < OptimizationContext::REMOVE_REDUNDENTS)
return this;
// TBD remove the need for QueryExecutionContext here - jpcs
QueryExecutionContext qec(GET_CONFIGURATION(opt.getContext())->getQueryContext(),
/*debugging*/false);
qec.setContainerBase(opt.getContainerBase());
qec.setDynamicContext(opt.getContext());
// Sort the arguments based on how many keys we think they'll return
sort(args_.begin(), args_.end(), keys_compare_less(opt.getOperationContext(), qec));
// Remove document index joins if they are unlikely to be useful.
it = args_.begin();
QueryPlan *leftMost = *it;
Cost lCost = leftMost->cost(opt.getOperationContext(), qec);
newArgs.clear();
newArgs.push_back(*it);
for(++it; it != args_.end(); ++it) {
if(StructuralJoinQP::isDocumentIndex(*it, /*toBeRemoved*/true) &&
StructuralJoinQP::isSuitableForDocumentIndexComparison(leftMost)) {
Cost itCost = (*it)->cost(opt.getOperationContext(), qec);
// TBD Calculate these constants? - jpcs
static const double KEY_RATIO_THRESHOLD = 2.0;
static const double PAGES_PER_KEY_FACTOR = 2.0;
if((itCost.keys / lCost.keys) > KEY_RATIO_THRESHOLD ||
(itCost.totalPages() / itCost.keys) > (lCost.totalPages() * PAGES_PER_KEY_FACTOR / lCost.keys)) {
string before = logIntersectBefore(leftMost, *it);
logTransformation(opt.getLog(), "Remove document join", before, leftMost);
leftMost->logCost(qec, lCost, 0);
(*it)->logCost(qec, itCost, 0);
continue;
}
}
newArgs.push_back(*it);
}
args_.clear();
std::copy(newArgs.begin(), newArgs.end(), back_inserter(args_));
// Return if we only have one argument
if(args_.size() == 1) return args_[0];
return this;
}
void AnalyzeInput::evaltrajs(vector<Trajectory*> &trajs){
//******Calculate Cost, Acceleration, and Distance traveled per time step*******************************
Cost evalcost;
this->DR = 0;
this->C = 0;
for (int i = 0; i < trajs.size(); i++){
trajs[i]->R = 0;
trajs[i]->C = 0;
double Cmin = 100;
double Cmax = 0;
double Rmin = 1E9;
double Rmax = 0;
for (int j = 0; j < trajs[i]->particles.size()-1; j++){
Particle* P = trajs[i]->particles[j];
Particle* Pn = trajs[i]->particles[j+1];
P->dr = distance(P,Pn);
if(P->dr > Rmax)
Rmax = P->dr;
if(P->dr < Rmin)
Rmin = P->dr;
trajs[i]->R += P->dr;
Pn->vel = P->dr * FRAMESPERSEC;
if (j>1){
Particle* Pm = trajs[i]->particles[j-1];
P->accel = abs(Pn->vel - Pm->vel) * FRAMESPERSEC/2.0;
}
if (j >= COST_TRAJ_SIZE - 1){
int s = (j) - (COST_TRAJ_SIZE -1);
Trajectory tr;
for (int x = s; x <= j; x++){
tr.particles.push_back(trajs[i]->particles[x]);
}
Pn->cost = evalcost.cost(&tr, Pn);
}else{
Pn->cost = 0.0;
}
if(Pn->cost > Cmax)
Cmax = Pn->cost;
if(Pn->cost < Cmin && Pn->cost != 0)
Cmin = Pn->cost;
trajs[i]->C += Pn->cost;
}
trajs[i]->Cl = Cmin;
trajs[i]->Ch = Cmax;
trajs[i]->C = trajs[i]->C / ((double) trajs[i]->particles.size());
this->C += trajs[i]->C;
trajs[i]->Rl = Rmin;
trajs[i]->Rh = Rmax;
trajs[i]->R = trajs[i]->R / ((double) trajs[i]->particles.size());
this->DR += trajs[i]->R;
}
this->C = this->C / trajs.size();
this->DR = this->DR / trajs.size();
}
QueryPlan *DescendantOrSelfJoinQP::optimize(OptimizationContext &opt)
{
XPath2MemoryManager *mm = opt.getMemoryManager();
QueryPlan *qp = StructuralJoinQP::optimize(opt);
if(qp != this) return qp;
if(findType(left_) == ImpliedSchemaNode::METADATA) {
switch(right_->getType()) {
case DESCENDANT_OR_SELF: {
DescendantOrSelfJoinQP *rsj = (DescendantOrSelfJoinQP*)right_;
if(findType(rsj->left_) == ImpliedSchemaNode::METADATA) {
string before = logBefore(this);
left_ = new (mm) IntersectQP(left_, rsj->left_, 0, mm);
left_->setLocationInfo(rsj);
right_ = rsj->right_;
flags_ |= rsj->flags_;
logTransformation(opt.getLog(), "Combine document join", before, this);
return optimize(opt);
}
break;
}
default: {
if(findType(right_) == ImpliedSchemaNode::METADATA) {
string before = logBefore(this);
QueryPlan *result = new (mm) IntersectQP(left_, right_, flags_, mm);
result->setLocationInfo(this);
logTransformation(opt.getLog(), "Combine document join", this, result);
return result->optimize(opt);
}
break;
}
}
}
if(opt.getPhase() < OptimizationContext::REMOVE_REDUNDENTS)
return this;
if(findType(left_) == ImpliedSchemaNode::METADATA) {
switch(right_->getType()) {
case STEP: {
string before = logBefore(this);
StepQP *step = (StepQP*)right_;
right_ = step->getArg();
step->setArg(this);
logTransformation(opt.getLog(), "Push back document join", before, step);
return step->optimize(opt);
}
case DESCENDANT:
case DESCENDANT_OR_SELF:
case ATTRIBUTE:
case CHILD:
case ATTRIBUTE_OR_CHILD: {
string before = logBefore(this);
StructuralJoinQP *sj = (StructuralJoinQP*)right_;
right_ = sj->getLeftArg();
sj->setLeftArg(this);
logTransformation(opt.getLog(), "Push back document join", before, sj);
return sj->optimize(opt);
}
case ANCESTOR:
case ANCESTOR_OR_SELF:
case PARENT:
case PARENT_OF_ATTRIBUTE:
case PARENT_OF_CHILD: {
string before = logBefore(this);
StructuralJoinQP *sj = (StructuralJoinQP*)right_;
right_ = sj->getRightArg();
sj->setRightArg(this);
logTransformation(opt.getLog(), "Push back document join", before, sj);
return sj->optimize(opt);
}
case EXCEPT: {
string before = logBefore(this);
// Add to both arguments of ExceptQP
ExceptQP *ex = (ExceptQP*)right_;
right_ = ex->getLeftArg();
ex->setLeftArg(this);
QueryPlan *copy = new (mm) DescendantOrSelfJoinQP(left_->copy(mm), ex->getRightArg(), flags_, mm);
copy->setLocationInfo(this);
ex->setRightArg(copy);
logTransformation(opt.getLog(), "Push back document join", before, ex);
return ex->optimize(opt);
}
default: break;
}
}
// TBD remove the need for QueryExecutionContext here - jpcs
QueryExecutionContext qec(GET_CONFIGURATION(opt.getContext())->getQueryContext(),
/*debugging*/false);
qec.setContainerBase(opt.getContainerBase());
qec.setDynamicContext(opt.getContext());
// Remove this document index join if it's unlikely to be useful
if(isDocumentIndex(left_, /*toBeRemoved*/true) && isSuitableForDocumentIndexComparison(right_)) {
Cost rCost = right_->cost(opt.getOperationContext(), qec);
Cost lCost = left_->cost(opt.getOperationContext(), qec);
// TBD Calculate these constants? - jpcs
static const double KEY_RATIO_THRESHOLD = 2.0;
static const double KEYS_PER_PAGE_THRESHOLD = 2.0;
if((lCost.keys / rCost.keys) > KEY_RATIO_THRESHOLD ||
(lCost.keys / lCost.totalPages()) > KEYS_PER_PAGE_THRESHOLD) {
logTransformation(opt.getLog(), "Remove document join", this, right_);
right_->logCost(qec, rCost, 0);
left_->logCost(qec, lCost, 0);
return right_;
}
}
return this;
}