int Convert::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String outFile = "";
String refFile = "";
bool lshift = false;
bool noeof = false;
bool params = false;
bool useBases = false;
bool useEquals = false;
bool useOrigSeq = false;
bool recover = false;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_STRINGPARAMETER("in", &inFile)
LONG_STRINGPARAMETER("out", &outFile)
LONG_STRINGPARAMETER("refFile", &refFile)
LONG_PARAMETER("lshift", &lshift)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("recover", &recover)
LONG_PARAMETER("params", ¶ms)
LONG_PARAMETER_GROUP("SequenceConversion")
EXCLUSIVE_PARAMETER("useBases", &useBases)
EXCLUSIVE_PARAMETER("useEquals", &useEquals)
EXCLUSIVE_PARAMETER("useOrigSeq", &useOrigSeq)
LONG_PHONEHOME(VERSION)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
// parameters start at index 2 rather than 1.
inputParameters.Read(argc, argv, 2);
// If no eof block is required for a bgzf file, set the bgzf file type to
// not look for it.
if(noeof)
{
// Set that the eof block is not required.
BgzfFileType::setRequireEofBlock(false);
}
// Check to see if the in file was specified, if not, report an error.
if(inFile == "")
{
printUsage(std::cerr);
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--in is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
if(outFile == "")
{
printUsage(std::cerr);
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--out is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
// Check to see if the ref file was specified.
// Open the reference.
GenomeSequence* refPtr = NULL;
if(refFile != "")
{
refPtr = new GenomeSequence(refFile);
}
SamRecord::SequenceTranslation translation;
if((useBases) && (refPtr != NULL))
{
translation = SamRecord::BASES;
}
else if((useEquals) && (refPtr != NULL))
{
translation = SamRecord::EQUAL;
}
else
{
useOrigSeq = true;
translation = SamRecord::NONE;
}
if(params)
{
inputParameters.Status();
}
// Open the input file for reading.
SamFile samIn;
if(recover) samIn.setAttemptRecovery(true);
samIn.OpenForRead(inFile);
// Open the output file for writing.
SamFile samOut;
samOut.OpenForWrite(outFile);
samOut.SetWriteSequenceTranslation(translation);
samOut.SetReference(refPtr);
// Read the sam header.
SamFileHeader samHeader;
samIn.ReadHeader(samHeader);
// Write the sam header.
samOut.WriteHeader(samHeader);
SamRecord samRecord;
// Set returnStatus to success. It will be changed
// to the failure reason if any of the writes fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(1) {
try {
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(samHeader, samRecord))
{
// left shift if necessary.
if(lshift)
{
samRecord.shiftIndelsLeft();
}
// Successfully read a record from the file, so write it.
if(!samOut.WriteRecord(samHeader, samRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOut.GetStatusMessage());
returnStatus = samOut.GetStatus();
}
}
break;
} catch (std::runtime_error e) {
std::cerr << "Caught runtime error: " << e.what() << "\n";
if(!recover) {
std::cerr << "Corrupted BAM file detected - consider using --recover option.\n";
break;
}
std::cerr << "Attempting to resync at next good BGZF block and BAM record.\n";
// XXX need to resync SamFile stream here
bool rc = samIn.attemptRecoverySync(checkSignature, SIGNATURE_LENGTH);
if(rc) {
std::cerr << "Successful resync - some data lost.\n";
continue; // succeeded
}
std::cerr << "Failed to re-sync on data stream.\n";
break; // failed to resync
}
}
std::cerr << std::endl << "Number of records read = " <<
samIn.GetCurrentRecordCount() << std::endl;
std::cerr << "Number of records written = " <<
samOut.GetCurrentRecordCount() << std::endl;
if(refPtr != NULL)
{
delete(refPtr);
}
// Since the reads were successful, return the status based
// on the status of the writes. If any failed, return
// their failure status.
return(returnStatus);
}
int WriteRegion::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String outFile = "";
String indexFile = "";
String readName = "";
String bed = "";
myStart = UNSPECIFIED_INT;
myEnd = UNSPECIFIED_INT;
myPrevStart = UNSPECIFIED_INT;
myPrevEnd = UNSPECIFIED_INT;
myRefID = UNSET_REF;
myRefName.Clear();
myPrevRefName.Clear();
myBedRefID = SamReferenceInfo::NO_REF_ID;
bool lshift = false;
bool noeof = false;
bool params = false;
myWithinReg = false;
myWroteReg = false;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inFile)
LONG_STRINGPARAMETER("out", &outFile)
LONG_PARAMETER_GROUP("Optional Region Parameters")
LONG_STRINGPARAMETER("bamIndex", &indexFile)
LONG_STRINGPARAMETER("refName", &myRefName)
LONG_INTPARAMETER("refID", &myRefID)
LONG_INTPARAMETER("start", &myStart)
LONG_INTPARAMETER("end", &myEnd)
LONG_STRINGPARAMETER("bed", &bed)
LONG_PARAMETER("withinReg", &myWithinReg)
LONG_STRINGPARAMETER("readName", &readName)
LONG_PARAMETER_GROUP("Optional Other Parameters")
LONG_PARAMETER("lshift", &lshift)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
inputParameters.Read(argc-1, &(argv[1]));
// If no eof block is required for a bgzf file, set the bgzf file type to
// not look for it.
if(noeof)
{
// Set that the eof block is not required.
BgzfFileType::setRequireEofBlock(false);
}
// Check to see if the in file was specified, if not, report an error.
if(inFile == "")
{
usage();
// mandatory argument was not specified.
inputParameters.Status();
std::cerr << "Missing mandatory argument: --in" << std::endl;
return(-1);
}
if(outFile == "")
{
usage();
// mandatory argument was not specified.
inputParameters.Status();
std::cerr << "Missing mandatory argument: --out" << std::endl;
return(-1);
}
if(indexFile == "")
{
// In file was not specified, so set it to the in file
// + ".bai"
indexFile = inFile + ".bai";
}
if(myRefID != UNSET_REF && myRefName.Length() != 0)
{
std::cerr << "Can't specify both refID and refName" << std::endl;
inputParameters.Status();
return(-1);
}
if(myRefID != UNSET_REF && bed.Length() != 0)
{
std::cerr << "Can't specify both refID and bed" << std::endl;
inputParameters.Status();
return(-1);
}
if(myRefName.Length() != 0 && bed.Length() != 0)
{
std::cerr << "Can't specify both refName and bed" << std::endl;
inputParameters.Status();
return(-1);
}
if(!bed.IsEmpty())
{
myBedFile = ifopen(bed, "r");
}
if(params)
{
inputParameters.Status();
}
// Open the file for reading.
mySamIn.OpenForRead(inFile);
// Open the output file for writing.
SamFile samOut;
samOut.OpenForWrite(outFile);
// Open the bam index file for reading if a region was specified.
if((myRefName.Length() != 0) || (myRefID != UNSET_REF) || (myBedFile != NULL))
{
mySamIn.ReadBamIndex(indexFile);
}
// Read & write the sam header.
mySamIn.ReadHeader(mySamHeader);
samOut.WriteHeader(mySamHeader);
// Read the sam records.
SamRecord samRecord;
// Track the status.
int numSectionRecords = 0;
// Set returnStatus to success. It will be changed
// to the failure reason if any of the writes fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(getNextSection())
{
// Keep reading records until they aren't anymore.
while(mySamIn.ReadRecord(mySamHeader, samRecord))
{
if(!readName.IsEmpty())
{
// Check for readname.
if(strcmp(samRecord.getReadName(), readName.c_str()) != 0)
{
// not a matching read name, so continue to the next record.
continue;
}
}
// Check to see if the read has already been processed.
if(myPrevEnd != UNSPECIFIED_INT)
{
// Because we already know that the bed was sorted,
// we know that the previous section started before
// this one, so if the previous end is greater than
// this record's end position we know that it
// was already written in the previous section.
// Note: can't be equal to the previous end since
// the end range was exclusive, while
// get0BasedAlignmentEnd is inclusive.
// myPrevEnd is reset by getNextSection when a new
// chromosome is hit.
if(samRecord.get0BasedAlignmentEnd() < myPrevEnd)
{
// This record was already written.
continue;
}
}
// Shift left if applicable.
if(lshift)
{
samRecord.shiftIndelsLeft();
}
// Successfully read a record from the file, so write it.
samOut.WriteRecord(mySamHeader, samRecord);
++numSectionRecords;
}
myWroteReg = true;
}
if(myBedFile != NULL)
{
ifclose(myBedFile);
}
std::cerr << "Wrote " << outFile << " with " << numSectionRecords
<< " records.\n";
return(returnStatus);
}
