// Read a record from the currently opened file.
bool SamFile::ReadRecord(SamFileHeader& header,
SamRecord& record)
{
myStatus = SamStatus::SUCCESS;
if(myIsOpenForRead == false)
{
// File is not open for read
myStatus.setStatus(SamStatus::FAIL_ORDER,
"Cannot read record since the file is not open for reading");
throw(std::runtime_error("SOFTWARE BUG: trying to read a SAM/BAM record prior to opening the file."));
return(false);
}
if(myHasHeader == false)
{
// The header has not yet been read.
// TODO - maybe just read the header.
myStatus.setStatus(SamStatus::FAIL_ORDER,
"Cannot read record since the header has not been read.");
throw(std::runtime_error("SOFTWARE BUG: trying to read a SAM/BAM record prior to reading the header."));
return(false);
}
// Check to see if a new region has been set. If so, determine the
// chunks for that region.
if(myNewSection)
{
if(!processNewSection(header))
{
// Failed processing a new section. Could be an
// order issue like the file not being open or the
// indexed file not having been read.
// processNewSection sets myStatus with the failure reason.
return(false);
}
}
// Read until a record is not successfully read or there are no more
// requested records.
while(myStatus == SamStatus::SUCCESS)
{
record.setReference(myRefPtr);
record.setSequenceTranslation(myReadTranslation);
// If reading by index, this method will setup to ensure it is in
// the correct position for the next record (if not already there).
// Sets myStatus if it could not move to a good section.
// Just returns true if it is not setup to read by index.
if(!ensureIndexedReadPosition())
{
// Either there are no more records in the section
// or it failed to move to the right section, so there
// is nothing more to read, stop looping.
break;
}
// File is open for reading and the header has been read, so read the
// next record.
myInterfacePtr->readRecord(myFilePtr, header, record, myStatus);
if(myStatus != SamStatus::SUCCESS)
{
// Failed to read the record, so break out of the loop.
break;
}
// Successfully read a record, so check if we should filter it.
// First check if it is out of the section. Returns true
// if not reading by sections, returns false if the record
// is outside of the section. Sets status to NO_MORE_RECS if
// there is nothing left ot read in the section.
if(!checkRecordInSection(record))
{
// The record is not in the section.
// The while loop will detect if NO_MORE_RECS was set.
continue;
}
// Check the flag for required/excluded flags.
uint16_t flag = record.getFlag();
if((flag & myRequiredFlags) != myRequiredFlags)
{
// The record does not conatain all required flags, so
// continue looking.
continue;
}
if((flag & myExcludedFlags) != 0)
{
// The record contains an excluded flag, so continue looking.
continue;
}
//increment the record count.
myRecordCount++;
if(myStatistics != NULL)
{
// Statistics should be updated.
myStatistics->updateStatistics(record);
}
// Successfully read the record, so check the sort order.
if(!validateSortOrder(record, header))
{
// ValidateSortOrder sets the status on a failure.
return(false);
}
return(true);
} // End while loop that checks if a desired record is found or failure.
// Return true if a record was found.
return(myStatus == SamStatus::SUCCESS);
}
bool VcfFileReader::readRecord(VcfRecord& record, VcfSubsetSamples* subset)
{
myStatus = StatGenStatus::SUCCESS;
// Subset the read if there are subsets specified.
VcfSubsetSamples* subsetPtr = subset;
if((subsetPtr == NULL) && myUseSubset)
{
subsetPtr = &mySampleSubset;
}
// Check to see if a new region has been set. If so, setup for that region.
bool searchChrom = false;
if(myNewSection)
{
if(myVcfIndex != NULL)
{
// Have an index file so use
if(!processNewSection())
{
// processNewSection sets the status appropriately on failure.
return(false);
}
}
else if(myTotalRead == 0)
{
// ReadSection without an index only works if no records
// have been read.
searchChrom = true;
myNewSection = false;
}
else
{
myNewSection = false;
myStatus.setStatus(StatGenStatus::FAIL_ORDER,
"Cannot set read section with no index after reading records");
return(false);
}
}
// Keep looping until a desired record is found.
bool recordFound = false;
while(!recordFound)
{
if(!record.read(myFilePtr, mySiteOnly, myRecordDiscardRules, subsetPtr))
{
myStatus = record.getStatus();
myTotalRead += myRecordDiscardRules.getNumDiscarded();
myNumRecords += myRecordDiscardRules.getNumDiscarded();
myRecordDiscardRules.clearNumDiscarded();
return(false);
}
++myTotalRead;
myTotalRead += myRecordDiscardRules.getNumDiscarded();
// Check to see if the record is in the section.
// First check the chromosome.
if(!mySectionChrom.empty() && (mySectionChrom != record.getChromStr()))
{
if(searchChrom)
{
// Still searching for the chromosome, so continue
// to the next record.
continue;
}
// Record is not within the correct chromosome, so return failure.
myStatus = StatGenStatus::NO_MORE_RECS;
return(false);
}
searchChrom = false;
// Check if the record is after the section end if applicable.
if((mySection1BasedEndPos != -1) &&
(record.get1BasedPosition() >= mySection1BasedEndPos))
{
myStatus = StatGenStatus::NO_MORE_RECS;
return(false);
}
// Check if the record is prior to the section start if applicable.
// Determinine the VCF record end position.
// If we are not requiring overlap, then we only need to check
// the start position, but if overlap is required, then it needs
// to incrment the start by the length-1.
int numIncBases = 0;
if(mySectionOverlap)
{
// The VCF record end position is the start position + length of the
// reference string - 1.
numIncBases = record.getNumRefBases() - 1;
}
if((mySection1BasedStartPos != -1) &&
((record.get1BasedPosition() + numIncBases)
< mySection1BasedStartPos))
{
// This record is prior to the section, so keep reading.
continue;
}
++myNumRecords;
myNumRecords += myRecordDiscardRules.getNumDiscarded();
myRecordDiscardRules.clearNumDiscarded();
// Record successfully read, so check to see if it is discarded.
if((myDiscardRules & DISCARD_NON_PHASED) && !record.allPhased())
{
// Not all samples are phased, so discard this record.
continue;
}
if((myDiscardRules & DISCARD_MISSING_GT) &&
!record.hasAllGenotypeAlleles())
{
// discard missing GTs and this record had missing alleles,
// so keep reading.
continue;
}
if((myDiscardRules & DISCARD_FILTERED) &&
!(record.getFilter().passedAllFilters()))
{
// Record was filtered, so discard it.
continue;
}
if((myDiscardRules & DISCARD_MULTIPLE_ALTS) &&
(record.getNumAlts() > 1))
{
// Record had multiple alternates, so discard.
continue;
}
// Check allele counts for discarding.
if(myMinAltAlleleCount != UNSET_MIN_ALT_ALLELE_COUNT)
{
// Count the number of alternates.
int32_t altCount = 0;
for(int sampleNum = 0; sampleNum < record.getNumSamples();
sampleNum++)
{
if((myAltAlleleCountSubset != NULL) &&
!(myAltAlleleCountSubset->keep(sampleNum)))
{
// Skip this sample.
continue;
}
for(int gtNum = 0; gtNum < record.getNumGTs(sampleNum); gtNum++)
{
if(record.getGT(sampleNum, gtNum) > 0)
{
// Alternate, so increment the count.
++altCount;
}
}
}
if(altCount < myMinAltAlleleCount)
{
// Not enough alternates so continue to the next sample.
continue;
}
}
// Check to see if the minimum alternate allele count is met.
if(myMinMinorAlleleCount != UNSET_MIN_MINOR_ALLELE_COUNT)
{
// Get the number of possible alternates.
unsigned int numAlts = record.getNumAlts();
// Verify that each allele has the min count.
bool failMinorAlleleCount = false;
for(unsigned int i = 0; i <= numAlts; i++)
{
if(record.getAlleleCount(i, myMinorAlleleCountSubset)
< myMinMinorAlleleCount)
{
// Not enough of one gt, so not ok.
failMinorAlleleCount = true;
break;
}
}
if(failMinorAlleleCount)
{
// not enough alleles, so continue to the next record.
continue;
}
}
// Record was not discarded.
recordFound = true;
}
// Increment the number of kept records.
++myNumKeptRecords;
return(true);
}
