// ValidateReaders checks that all the readers point to BAM files representing
// alignments against the same set of reference sequences, and that the
// sequences are identically ordered. If these checks fail the operation of
// the multireader is undefined, so we force program exit.
void BamMultiReader::ValidateReaders(void) const {
int firstRefCount = readers.front().first->GetReferenceCount();
BamTools::RefVector firstRefData = readers.front().first->GetReferenceData();
for (vector<pair<BamReader*, BamAlignment*> >::const_iterator it = readers.begin(); it != readers.end(); ++it) {
BamReader* reader = it->first;
BamTools::RefVector currentRefData = reader->GetReferenceData();
BamTools::RefVector::const_iterator f = firstRefData.begin();
BamTools::RefVector::const_iterator c = currentRefData.begin();
if (reader->GetReferenceCount() != firstRefCount || firstRefData.size() != currentRefData.size()) {
cerr << "ERROR: mismatched number of references in " << reader->GetFilename()
<< " expected " << firstRefCount
<< " reference sequences but only found " << reader->GetReferenceCount() << endl;
exit(1);
}
// this will be ok; we just checked above that we have identically-sized sets of references
// here we simply check if they are all, in fact, equal in content
while (f != firstRefData.end()) {
if (f->RefName != c->RefName || f->RefLength != c->RefLength) {
cerr << "ERROR: mismatched references found in " << reader->GetFilename()
<< " expected: " << endl;
for (BamTools::RefVector::const_iterator a = firstRefData.begin(); a != firstRefData.end(); ++a)
cerr << a->RefName << " " << a->RefLength << endl;
cerr << "but found: " << endl;
for (BamTools::RefVector::const_iterator a = currentRefData.begin(); a != currentRefData.end(); ++a)
cerr << a->RefName << " " << a->RefLength << endl;
exit(1);
}
++f; ++c;
}
}
}
NewGenomeFile::NewGenomeFile(const BamTools::RefVector &refVector)
: _maxId(-1)
{
for (size_t i = 0; i < refVector.size(); ++i) {
QuickString chrom = refVector[i].RefName;
CHRPOS length = refVector[i].RefLength;
_maxId++;
_chromSizeIds[chrom] = pair<CHRPOS, CHRPOS>(length, _maxId);
}
}
NewGenomeFile::NewGenomeFile(const BamTools::RefVector &refVector)
: _maxId(-1)
{
size_t i = 0;
for (; i < refVector.size(); ++i) {
string chrom = refVector[i].RefName;
CHRPOS length = refVector[i].RefLength;
_maxId++;
_chromSizeIds[chrom] = pair<CHRPOS, CHRPOS>(length, _maxId);
_chromList.push_back(chrom);
}
// Special: BAM files can have unmapped reads, which show as no chromosome, or an empty chrom string.
// Add in an empty chrom so these don't error.
_maxId++;
_chromSizeIds[""] = pair<CHRPOS, int>(0, _maxId);
_chromList.push_back("");
}
RegionCoverage::RegionCoverage( const BamTools::RefVector& references )
: m_contigList(NULL)
, m_bcovRegion(NULL)
, m_rcovRegion(NULL)
, m_lastRegionAssigned(NULL)
, m_numAuxFields(0)
, m_ncovDepths(0)
{
m_numRefContigs = references.size();
if( m_numRefContigs ) {
m_contigList = new TargetContig*[m_numRefContigs];
for( size_t i = 0; i < m_numRefContigs; ++i ) {
m_contigList[i] = new TargetContig( references[i].RefName, references[i].RefLength );
m_contigIdx[ references[i].RefName ] = i;
}
}
// initial values force set up on first call to 'iterator'
m_bcovContigIdx = m_rcovContigIdx = m_numRefContigs;
m_bcovRegionPos = 0;
}