/*! \fn bool BamWriter::Open(const std::string& filename,
const SamHeader& samHeader,
const RefVector& referenceSequences)
\brief Opens a BAM file for writing.
This is an overloaded function.
Will overwrite the BAM file if it already exists.
\param[in] filename name of output BAM file
\param[in] samHeader header data, wrapped in SamHeader object
\param[in] referenceSequences list of reference entries
\return \c true if opened successfully
\sa Close(), IsOpen(), BamReader::GetHeader(), BamReader::GetReferenceData()
*/
bool BamWriter::Open(const std::string& filename,
const SamHeader& samHeader,
const RefVector& referenceSequences)
{
d->SetParallel(m_numThreads);
return d->Open(filename, samHeader.ToString(), referenceSequences);
}
// makes a virtual, unified header for all the bam files in the multireader
string BamMultiReaderPrivate::GetHeaderText(void) const {
// N.B. - right now, simply copies all header data from first BAM,
// and then appends RG's from other BAM files
// TODO: make this more intelligent wrt other header lines/fields
// if no readers open
const size_t numReaders = m_readers.size();
if ( numReaders == 0 ) return string();
// retrieve first reader's header
const MergeItem& firstItem = m_readers.front();
const BamReader* reader = firstItem.Reader;
if ( reader == 0 ) return string();
SamHeader mergedHeader = reader->GetHeader();
// Add filename to read ID name so we know
// which file it came from
for (SamReadGroupIterator it = mergedHeader.ReadGroups.Begin();
it != mergedHeader.ReadGroups.End(); it++) {
it->ID = it->ID + "-" + reader->GetFilename();
}
// iterate over any remaining readers (skipping the first)
for ( size_t i = 1; i < numReaders; ++i ) {
const MergeItem& item = m_readers.at(i);
const BamReader* reader = item.Reader;
if ( reader == 0 ) continue;
// retrieve current reader's header
SamHeader currentHeader = reader->GetHeader();
// Add filename to read ID name so we know
// which file it came from
for (SamReadGroupIterator it = currentHeader.ReadGroups.Begin();
it != currentHeader.ReadGroups.End(); it++) {
it->ID = it->ID + "-" + reader->GetFilename();
}
// append current reader's RG entries to merged header
// N.B. - SamReadGroupDictionary handles duplicate-checking
mergedHeader.ReadGroups.Add(currentHeader.ReadGroups);
// TODO: merge anything else??
}
// return stringified header
return mergedHeader.ToString();
}
//{{{bool sort_inter_chrom_bam(string in_file_name,
bool sort_inter_chrom_bam(string in_file_name,
string out_file_name)
{
// open input BAM file
BamReader reader;
if ( !reader.Open(in_file_name) ) {
cerr << "sort ERROR: could not open " <<
in_file_name << " for reading... Aborting." << endl;
return false;
}
SamHeader header = reader.GetHeader();
if ( !header.HasVersion() )
header.Version = Constants::SAM_CURRENT_VERSION;
string header_text = header.ToString();
RefVector ref = reader.GetReferenceData();
// set up alignments buffer
BamAlignment al;
vector<BamAlignment> buffer;
buffer.reserve( (size_t)(SORT_DEFAULT_MAX_BUFFER_COUNT*1.1) );
bool bufferFull = false;
int buff_count = 0;
// iterate through file
while ( reader.GetNextAlignment(al)) {
// check buffer's usage
bufferFull = ( buffer.size() >= SORT_DEFAULT_MAX_BUFFER_COUNT );
// store alignments until buffer is "full"
if ( !bufferFull )
buffer.push_back(al);
// if buffer is "full"
else {
// so create a sorted temp file with current buffer contents
// then push "al" into fresh buffer
create_sorted_temp_file(buffer,
out_file_name,
buff_count,
header_text,
ref);
++buff_count;
buffer.push_back(al);
}
}
// handle any leftover buffer contents
if ( !buffer.empty() ) {
create_sorted_temp_file(buffer,
out_file_name,
buff_count,
header_text,
ref);
++buff_count;
}
reader.Close();
return merge_sorted_files(out_file_name, buff_count, header_text, ref);
/*
for (int i = 0; i < buff_count; ++i) {
stringstream temp_name;
temp_name << out_file_name << i;
}
*/
}
// generates mutiple sorted temp BAM files from single unsorted BAM file
bool SortTool::SortToolPrivate::GenerateSortedRuns(void) {
// open input BAM file
BamReader reader;
if ( !reader.Open(m_settings->InputBamFilename) ) {
cerr << "bamtools sort ERROR: could not open " << m_settings->InputBamFilename
<< " for reading... Aborting." << endl;
return false;
}
// get basic data that will be shared by all temp/output files
SamHeader header = reader.GetHeader();
header.SortOrder = ( m_settings->IsSortingByName
? Constants::SAM_HD_SORTORDER_QUERYNAME
: Constants::SAM_HD_SORTORDER_COORDINATE );
m_headerText = header.ToString();
m_references = reader.GetReferenceData();
// set up alignments buffer
BamAlignment al;
vector<BamAlignment> buffer;
buffer.reserve( (size_t)(m_settings->MaxBufferCount*1.1) );
bool bufferFull = false;
// if sorting by name, we need to generate full char data
// so can't use GetNextAlignmentCore()
if ( m_settings->IsSortingByName ) {
// iterate through file
while ( reader.GetNextAlignment(al)) {
// check buffer's usage
bufferFull = ( buffer.size() >= m_settings->MaxBufferCount );
// store alignments until buffer is "full"
if ( !bufferFull )
buffer.push_back(al);
// if buffer is "full"
else {
// push any unmapped reads into buffer,
// don't want to split these into a separate temp file
if ( !al.IsMapped() )
buffer.push_back(al);
// "al" is mapped, so create a sorted temp file with current buffer contents
// then push "al" into fresh buffer
else {
CreateSortedTempFile(buffer);
buffer.push_back(al);
}
}
}
}
// sorting by position, can take advantage of GNACore() speedup
else {
// iterate through file
while ( reader.GetNextAlignmentCore(al) ) {
// check buffer's usage
bufferFull = ( buffer.size() >= m_settings->MaxBufferCount );
// store alignments until buffer is "full"
if ( !bufferFull )
buffer.push_back(al);
// if buffer is "full"
else {
// push any unmapped reads into buffer,
// don't want to split these into a separate temp file
if ( !al.IsMapped() )
buffer.push_back(al);
// "al" is mapped, so create a sorted temp file with current buffer contents
// then push "al" into fresh buffer
else {
CreateSortedTempFile(buffer);
buffer.push_back(al);
}
}
}
}
// handle any leftover buffer contents
if ( !buffer.empty() )
CreateSortedTempFile(buffer);
// close reader & return success
reader.Close();
return true;
}
/*! \fn bool BamWriter::Open(const std::string& filename,
const SamHeader& samHeader,
const RefVector& referenceSequences)
\brief Opens a BAM file for writing.
This is an overloaded function.
Will overwrite the BAM file if it already exists.
\param filename name of output BAM file
\param samHeader header data, wrapped in SamHeader object
\param referenceSequences list of reference entries
\return \c true if opened successfully
\sa Close(), IsOpen(), BamReader::GetHeader(), BamReader::GetReferenceData()
*/
bool BamWriter::Open(const std::string& filename,
const SamHeader& samHeader,
const RefVector& referenceSequences)
{
return d->Open(filename, samHeader.ToString(), referenceSequences);
}