//retrieve candidate result set by the var_sig in the _query.
void VSTree::retrieve(SPARQLquery& _query)
{
Util::logging("IN retrieve");
//debug
// {
// VNode* temp_ptr = this->getLeafNodeByEntityID(473738);
// stringstream _ss;
//
// for (int i=0;i<temp_ptr->getChildNum();i++)
// if (temp_ptr->getChildEntry(i).getEntityId() == 473738)
// {
// _ss << "entity id=473738 entry sig:" << endl;
// _ss << "entity id=473738 leaf node line: " << temp_ptr->getFileLine() << endl;
// _ss << Signature::BitSet2str(temp_ptr->getChildEntry(i).getEntitySig().entityBitSet) << endl;
// break;
// }
//
// _ss << "leaf node sig:" << endl;
// _ss << Signature::BitSet2str(temp_ptr->getEntry().getEntitySig().entityBitSet) << endl;
//
// temp_ptr = temp_ptr->getFather(*(this->node_buffer));
// while (temp_ptr != NULL)
// {
// _ss << "line=" << temp_ptr->getFileLine() << endl;
// _ss << Signature::BitSet2str(temp_ptr->getEntry().getEntitySig().entityBitSet) << endl;
// temp_ptr = temp_ptr->getFather(*(this->node_buffer));
// }
// Util::logging(_ss.str());
// }
vector<BasicQuery*>& queryList = _query.getBasicQueryVec();
// enumerate each BasicQuery and retrieve their variables' mapping entity in the VSTree.
vector<BasicQuery*>::iterator iter=queryList.begin();
for(; iter != queryList.end(); iter++)
{
int varNum = (*iter)->getVarNum();
for (int i = 0; i < varNum; i++)
{
//debug
{
std::stringstream _ss;
_ss << "retrieve of var: " << i << endl;
Util::logging(_ss.str());
}
bool flag = (*iter)->isLiteralVariable(i);
const EntityBitSet& entityBitSet = (*iter)->getVarBitSet(i);
IDList* idListPtr = &( (*iter)->getCandidateList(i) );
this->retrieveEntity(entityBitSet, idListPtr);
#ifdef DEBUG_VSTREE
stringstream _ss;
_ss << "total num: " << this->entry_num << endl;
_ss << "candidate num: " << idListPtr->size() << endl;
_ss << endl;
_ss << "isExist 473738: " << (idListPtr->isExistID(473738)?"true":"false") <<endl;
_ss << "isExist 473472: " << (idListPtr->isExistID(473472)?"true":"false") <<endl;
_ss << "isExist 473473: " << (idListPtr->isExistID(473473)?"true":"false") <<endl;
Util::logging(_ss.str());
#endif
//the basic query should end if one non-literal var has no candidates
if(idListPtr->size() == 0 && !flag)
{
break;
}
}
}
Util::logging("OUT retrieve");
}
int printClusters(NodeSet roots, IDList orphanNodes,
string clusterListName, string clusterName,
vector<float> cutoffs)
{
TreeNode *tempNode = 0;
NodeSetIter setIter;
NodeList nodeList, tempList;
NodeListIter nodeIt, tempIt;
IDList OTU;
IDListIter it;
unsigned int size, numOTUs;
FILE *clusterListFile, *clusterFile;
clusterListFile = fopen(clusterListName.c_str(),"wb");
clusterFile = fopen(clusterName.c_str(),"wb");
if(clusterListFile == NULL|| clusterFile == NULL)
{
cout << "Cannot open output files. Skipped" << endl;
return 0;
}
printf("\n");
vector<float>::iterator c;
float distLevel;
for(c = cutoffs.begin(); c != cutoffs.end(); c++)
{
distLevel = *(c);
numOTUs = 0;
nodeList.clear();
// extract the valid nodes for each distance level
for(setIter=roots.begin(); setIter!=roots.end(); ++setIter)
{
tempNode=0;
if(*setIter != 0)
{
if((*setIter)->dist < distLevel || fabs((*setIter)->dist-distLevel) < EPSILON)
{
nodeList.push_front(*setIter);
continue;
}
tempList.push_front(*setIter);
while (tempList.size()!=0)
{
tempIt=tempList.begin();
tempNode=(*tempIt);
tempList.pop_front();
if (tempNode->left->dist < distLevel || fabs(tempNode->left->dist-distLevel) < EPSILON)
nodeList.push_front(tempNode->left);
else
tempList.push_front(tempNode->left);
if (tempNode->right->dist < distLevel || fabs(tempNode->right->dist-distLevel) < EPSILON)
nodeList.push_front(tempNode->right);
else
tempList.push_front(tempNode->right);
}
}
tempList.clear();
}
fprintf(clusterListFile," %.6f ", distLevel);
fprintf(clusterFile," %.6f ", distLevel);
// write the nodeList to file
tempList.clear();
for(nodeIt=nodeList.begin(); nodeIt!=nodeList.end(); ++nodeIt)
{
// clean up and initialize
fprintf(clusterFile,"|");
tempNode=0;
size=0;
OTU.clear();
tempList.push_front(*nodeIt);
while(tempList.size()!=0)
{
tempIt=tempList.begin();
tempNode=(*tempIt);
tempList.pop_front();
if(tempNode->left==0 && tempNode->right==0)
{
OTU.push_back(tempNode->ID);
size+=tempNode->numMembers;
}
if (tempNode->right!=0)
tempList.push_front(tempNode->right);
if(tempNode->left!=0 )
tempList.push_front(tempNode->left);
}
tempList.clear();
// print to clusterFile
it=OTU.begin();
fprintf(clusterFile,"%u",(*it));
++it;
for(;it!=OTU.end(); ++it)
fprintf(clusterFile," %u",(*it));
fprintf(clusterListFile, "%d ", size);
++numOTUs;
}
for (it=orphanNodes.begin(); it != orphanNodes.end(); ++it) {
fprintf(clusterFile,"|%u",(*it));
fprintf(clusterListFile, "1 ");
}
numOTUs += orphanNodes.size();
fprintf(clusterFile,"|\n");
fprintf(clusterListFile, "\n");
printf("Dist: %.6f. numOTUs: %u. numSingletons: %lu\n", distLevel, numOTUs, orphanNodes.size());
}
printf("\n");
OTU.clear();
fclose(clusterListFile);
fclose(clusterFile);
return 1;
}
