bool getNextAlignment(BamTools::BamAlignment &alignment, BamTools::BamReader &bamReader, const std::map<std::string, int> &groupID, std::vector< BamTools::BamAlignment > &alignmentSample, std::map<std::string, int> &wellIndex, unsigned int nSample) {
if(nSample > 0) {
// We are randomly sampling, so next read should come from the sample that was already taken from the bam file
if(alignmentSample.size() > 0) {
alignment = alignmentSample.back();
alignmentSample.pop_back();
alignment.BuildCharData();
return(true);
} else {
return(false);
}
} else {
// No random sampling, so we're either returning everything or we're looking for specific read names
bool storeRead = false;
while(bamReader.GetNextAlignment(alignment)) {
if(groupID.size() > 0) {
std::string thisReadGroupID = "";
if( !alignment.GetTag("RG", thisReadGroupID) || (groupID.find(thisReadGroupID)==groupID.end()) );
continue;
}
storeRead=true;
if(wellIndex.size() > 0) {
// We are filtering by position, so check if we should skip or keep the read
int thisCol,thisRow;
if(1 != ion_readname_to_rowcol(alignment.Name.c_str(), &thisRow, &thisCol))
std::cerr << "Error parsing read name: " << alignment.Name << "\n";
std::stringstream wellIdStream;
wellIdStream << thisCol << ":" << thisRow;
std::map<std::string, int>::iterator wellIndexIter;
wellIndexIter = wellIndex.find(wellIdStream.str());
if(wellIndexIter != wellIndex.end()) {
// If the read ID matches we should keep, unless its a duplicate
if(wellIndexIter->second >= 0) {
storeRead=true;
wellIndexIter->second=-1;
} else {
storeRead=false;
std::cerr << "WARNING: found extra instance of readID " << wellIdStream.str() << ", keeping only first\n";
}
} else {
// read ID is not one we should keep
storeRead=false;
}
}
if(storeRead)
break;
}
return(storeRead);
}
}
