Document* PFMatrixFormat::loadDocument( IOAdapter* io, const U2DbiRef& dbiRef, const QVariantMap& fs, U2OpStatus& os ){
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, NULL);
Q_UNUSED(opBlock);
QList<GObject*> objs;
IOAdapterFactory* iof = AppContext::getIOAdapterRegistry()->getIOAdapterFactoryById(io->getAdapterId());
TaskStateInfo siPFM;
PFMatrix m = WeightMatrixIO::readPFMatrix(iof, io->getURL().getURLString(), siPFM);
if (siPFM.hasError()) {
os.setError(tr("The file format is not PFM"));
} else {
if (m.getLength() == 0) {
os.setError(tr("Zero length or corrupted model\nMaybe model data are not enough for selected algorithm"));
}
}
CHECK_OP(os, NULL);
PFMatrixObject *mObj = PFMatrixObject::createInstance(m, QFileInfo(io->getURL().getURLString()).baseName(), dbiRef, os, fs);
CHECK_OP(os, NULL);
objs.append(mObj);
return new Document(this, io->getFactory(), io->getURL(), dbiRef, objs, fs);
}
Document* PlainTextFormat::loadDocument(IOAdapter* io, const U2DbiRef& dbiRef, const QVariantMap& fs, U2OpStatus& os){
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, NULL);
Q_UNUSED(opBlock);
QString text;
int size = io->left();
if (size > 0) {
text.reserve(size);
}
QByteArray block(BUFF_SIZE, '\0');
int blockLen = 0;
while ((blockLen = io->readBlock(block.data(), BUFF_SIZE)) > 0) {
int sizeBefore = text.length();
QString line = QString::fromLocal8Bit(block.data(), blockLen);
text.append(line);
if (text.length() != sizeBefore + blockLen) {
os.setError(L10N::errorReadingFile(io->getURL()));
break;
}
os.setProgress(io->getProgress());
}
CHECK_OP(os, NULL);
//todo: check file-readonly status?
QVariantMap hints;
hints.insert(DBI_FOLDER_HINT, fs.value(DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER));
TextObject* to = TextObject::createInstance(text, io->getURL().baseFileName(), dbiRef, os, hints);
CHECK_OP(os, NULL);
QList<GObject*> objects;
objects.append(to);
Document* d = new Document(this, io->getFactory(), io->getURL(), dbiRef, objects, fs);
return d;
}
/**
* FASTQ format specification: http://maq.sourceforge.net/fastq.shtml
*/
static void load(IOAdapter* io, const U2DbiRef& dbiRef, const QVariantMap& hints, const GUrl& docUrl, QList<GObject*>& objects, U2OpStatus& os,
int gapSize, int predictedSize, QString& writeLockReason) {
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, );
Q_UNUSED(opBlock);
writeLockReason.clear();
bool merge = gapSize!=-1;
QByteArray sequence;
QByteArray qualityScores;
QStringList headers;
QSet<QString> uniqueNames;
QVector<U2Region> mergedMapping;
QByteArray gapSequence((merge ? gapSize : 0), 0);
sequence.reserve(predictedSize);
qualityScores.reserve(predictedSize);
// for lower case annotations
GObjectReference sequenceRef;
qint64 sequenceStart = 0;
U2SequenceImporter seqImporter(hints, true);
const QString folder = hints.value(DocumentFormat::DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER).toString();
int seqNumber = 0;
int progressUpNum = 0;
const int objectsCountLimit = hints.contains(DocumentReadingMode_MaxObjectsInDoc) ? hints[DocumentReadingMode_MaxObjectsInDoc].toInt() : -1;
const bool settingsMakeUniqueName = !hints.value(DocumentReadingMode_DontMakeUniqueNames, false).toBool();
while (!os.isCoR()) {
//read header
QString sequenceName = readSequenceName(os, io, '@');
// check for eof while trying to read another FASTQ block
if (io->isEof()) {
break;
}
CHECK_OP_BREAK(os);
if(sequenceName.isEmpty()){
sequenceName = "Sequence";
}
if ((merge == false) || (seqNumber == 0)) {
QString objName = sequenceName;
if (settingsMakeUniqueName) {
objName = (merge) ? "Sequence" : TextUtils::variate(sequenceName, "_", uniqueNames);
objName.squeeze();
uniqueNames.insert(objName);
}
seqImporter.startSequence(dbiRef, folder, objName, false, os);
CHECK_OP_BREAK(os);
}
//read sequence
if (merge && sequence.length() > 0) {
seqImporter.addDefaultSymbolsBlock(gapSize,os);
sequenceStart += sequence.length();
sequenceStart+=gapSize;
CHECK_OP_BREAK(os);
}
sequence.clear();
readSequence(os, io, sequence);
MemoryLocker lSequence(os, qCeil(sequence.size()/(1000*1000)));
CHECK_OP_BREAK(os);
Q_UNUSED(lSequence);
seqImporter.addBlock(sequence.data(),sequence.length(),os);
CHECK_OP_BREAK(os);
QString qualSequenceName = readSequenceName(os, io, '+');
if (!qualSequenceName.isEmpty()) {
static const QString err = U2::FastqFormat::tr("Not a valid FASTQ file: %1, sequence name differs from quality scores name: %2 and %3");
CHECK_EXT_BREAK(sequenceName == qualSequenceName,
os.setError(err.arg(docUrl.getURLString()).arg(sequenceName).arg(qualSequenceName)));
}
// read qualities
qualityScores.clear();
readQuality(os, io, qualityScores, sequence.size());
CHECK_OP_BREAK(os);
static const QString err = U2::FastqFormat::tr("Not a valid FASTQ file: %1. Bad quality scores: inconsistent size.").arg(docUrl.getURLString());
CHECK_EXT_BREAK(sequence.length() == qualityScores.length(), os.setError(err));
seqNumber++;
progressUpNum++;
if (merge) {
headers.append(sequenceName);
mergedMapping.append(U2Region(sequenceStart, sequence.length() ));
}
else {
if (objectsCountLimit > 0 && objects.size() >= objectsCountLimit) {
os.setError(FastqFormat::tr("File \"%1\" contains too many sequences to be displayed. "
"However, you can process these data using instruments from the menu <i>Tools -> NGS data analysis</i> "
"or pipelines built with Workflow Designer.")
.arg(io->getURL().getURLString()));
break;
}
U2Sequence u2seq = seqImporter.finalizeSequenceAndValidate(os);
CHECK_OP_BREAK(os);
sequenceRef = GObjectReference(io->getURL().getURLString(), u2seq.visualName, GObjectTypes::SEQUENCE, U2EntityRef(dbiRef, u2seq.id));
U2SequenceObject* seqObj = new U2SequenceObject(u2seq.visualName, U2EntityRef(dbiRef, u2seq.id));
CHECK_EXT_BREAK(seqObj != NULL, os.setError("U2SequenceObject is NULL"));
seqObj->setQuality(DNAQuality(qualityScores));
objects << seqObj;
U1AnnotationUtils::addAnnotations(objects, seqImporter.getCaseAnnotations(), sequenceRef, NULL, hints);
}
if (PROGRESS_UPDATE_STEP == progressUpNum) {
progressUpNum = 0;
os.setProgress(io->getProgress());
}
}
CHECK_OP_EXT(os, qDeleteAll(objects); objects.clear(), );
bool emptyObjects = objects.isEmpty();
CHECK_EXT(!emptyObjects || merge, os.setError(Document::tr("Document is empty.")), );
SAFE_POINT(headers.size() == mergedMapping.size(), "headers <-> regions mapping failed!", );
if (!merge) {
return;
}
U2Sequence u2seq = seqImporter.finalizeSequenceAndValidate(os);
CHECK_OP(os,);
sequenceRef = GObjectReference(io->getURL().getURLString(), u2seq.visualName, GObjectTypes::SEQUENCE, U2EntityRef(dbiRef, u2seq.id));
U1AnnotationUtils::addAnnotations(objects, seqImporter.getCaseAnnotations(), sequenceRef, NULL, hints);
objects << new U2SequenceObject(u2seq.visualName, U2EntityRef(dbiRef, u2seq.id));
objects << DocumentFormatUtils::addAnnotationsForMergedU2Sequence(sequenceRef, dbiRef, headers, mergedMapping, hints);
if (headers.size() > 1) {
writeLockReason = DocumentFormat::MERGED_SEQ_LOCK;
}
}
void EMBLGenbankAbstractDocument::load(const U2DbiRef& dbiRef, IOAdapter* io, QList<GObject*>& objects, QVariantMap& fs, U2OpStatus& os, QString& writeLockReason) {
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, );
Q_UNUSED(opBlock);
writeLockReason.clear();
//get settings
int gapSize = qBound(-1, DocumentFormatUtils::getMergeGap(fs), 1000*1000);
bool merge = gapSize!=-1;
QScopedPointer<AnnotationTableObject> mergedAnnotations(NULL);
QStringList contigs;
QVector<U2Region> mergedMapping;
// Sequence loading is 'lazy', so, if there is no sequence, it won't be created and there is no need to remove it.
U2SequenceImporter seqImporter(fs, true);
const QString folder = fs.value(DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER).toString();
QSet<QString> usedNames;
GObjectReference sequenceRef(GObjectReference(io->getURL().getURLString(), "", GObjectTypes::SEQUENCE));
QByteArray readBuffer(ParserState::LOCAL_READ_BUFFER_SIZE, '\0');
ParserState st(isNcbiLikeFormat() ? 12 : 5, io, NULL, os);
st.buff = readBuffer.data();
TmpDbiObjects dbiObjects(dbiRef, os);
int num_sequence = 0;
qint64 sequenceStart = 0;
int sequenceSize = 0;
int fullSequenceSize = 0;
const int objectsCountLimit = fs.contains(DocumentReadingMode_MaxObjectsInDoc) ? fs[DocumentReadingMode_MaxObjectsInDoc].toInt() : -1;
for (int i=0; !os.isCoR(); i++, ++num_sequence) {
if (objectsCountLimit > 0 && objects.size() >= objectsCountLimit) {
os.setError(EMBLGenbankAbstractDocument::tr("File \"%1\" contains too many sequences to be displayed. "
"However, you can process these data using instruments from the menu <i>Tools -> NGS data analysis</i> "
"or pipelines built with Workflow Designer.")
.arg(io->getURL().getURLString()));
break;
}
//TODO: reference to a local variable??? Such a pointer will become invalid
EMBLGenbankDataEntry data;
st.entry = &data;
if (num_sequence == 0 || merge == false){
seqImporter.startSequence(dbiRef, folder, "default sequence name", false, os); //change name and circularity after finalize method
CHECK_OP(os, );
}
sequenceSize = 0;
os.setDescription(tr("Reading entry header"));
int offset = 0;
if (merge && num_sequence > 0) {
offset = gapSize;
}
if (!readEntry(&st,seqImporter,sequenceSize,fullSequenceSize,merge,offset, os)) {
break;
}
if (merge && sequenceSize > 0 && num_sequence > 0) {
sequenceStart = fullSequenceSize - sequenceSize;
sequenceStart += gapSize;
fullSequenceSize += gapSize;
}
// tolerate blank lines between records
char ch;
bool b;
while ((b = st.io->getChar(&ch)) && (ch == '\n' || ch == '\r')){}
if (b) {
st.io->skip(-1);
}
AnnotationTableObject *annotationsObject = NULL;
if (data.hasAnnotationObjectFlag) {
QString annotationName = genObjectName(usedNames, data.name, data.tags, i+1, GObjectTypes::ANNOTATION_TABLE);
QVariantMap hints;
hints.insert(DBI_FOLDER_HINT, fs.value(DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER));
if (Q_UNLIKELY(merge && NULL == mergedAnnotations)) {
mergedAnnotations.reset(new AnnotationTableObject(annotationName, dbiRef, hints));
}
annotationsObject = merge ? mergedAnnotations.data() : new AnnotationTableObject(annotationName, dbiRef, hints);
QStringList groupNames;
QMap<QString, QList<SharedAnnotationData> > groupName2Annotations;
for (int i = 0, n = data.features.size(); i < n; ++i) {
SharedAnnotationData &d = data.features[i];
if (!d->location->regions.isEmpty()) {
for (int i = 0, n = d->location->regions.size(); i < n; ++i) {
// for some reason larger numbers cannot be stored within rtree SQLite tables
if (d->location->regions[i].endPos() > 9223371036854775807LL) {
d->location->regions[i].length = 9223371036854775807LL - d->location->regions[i].startPos;
}
}
}
groupNames.clear();
d->removeAllQualifiers(GBFeatureUtils::QUALIFIER_GROUP, groupNames);
if (groupNames.isEmpty()) {
groupName2Annotations[""].append(d);
} else {
foreach(const QString &gName, groupNames) {
groupName2Annotations[gName].append(d);
}
}
CHECK_OP(os, );
}
static void load(IOAdapter* io, const U2DbiRef& dbiRef, QList<GObject*>& objects, const QVariantMap& fs, U2OpStatus& os) {
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, );
Q_UNUSED(opBlock);
U2SequenceImporter seqImporter(fs, true);
const QString folder = fs.value(DocumentFormat::DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER).toString();
QByteArray readBuffer(DocumentFormat::READ_BUFF_SIZE, '\0');
char* buff = readBuffer.data();
QBitArray ALPHAS = TextUtils::ALPHA_NUMS;
ALPHAS['-'] = true;
QByteArray seq;
QString seqName(io->getURL().baseFileName());
//reading sequence
QBuffer writer(&seq);
writer.open(QIODevice::WriteOnly);
TmpDbiObjects dbiObjects(dbiRef, os);
bool ok = true;
int len = 0;
bool isStarted = false;
int sequenceCounter = 0;
bool terminatorFound = false;
bool isSplit = fs.value((DocumentReadingMode_SequenceAsSeparateHint), false).toBool();
while (ok && !io->isEof()) {
len = io->readLine(buff, DocumentFormat::READ_BUFF_SIZE, &terminatorFound);
if (len <= 0){
continue;
}
seq.clear();
bool isSeek = writer.seek(0);
assert(isSeek); Q_UNUSED(isSeek);
if (os.isCoR()) {
break;
}
for (int i=0; i<len && ok; i++) {
char c = buff[i];
if (ALPHAS[(uchar)c]) {
ok = writer.putChar(c);
}
}
if(seq.size()>0 && isStarted == false ){
QString name = sequenceCounter == 0 ? seqName : seqName + QString("_%1").arg(sequenceCounter);
isStarted = true;
seqImporter.startSequence(dbiRef, folder, name, false, os);
}
if(isStarted){
seqImporter.addBlock(seq.data(),seq.size(),os);
}
if (seq.size()>0 && isStarted && terminatorFound && isSplit){
finishSequence(objects, io, os, dbiRef, fs, dbiObjects, seqImporter);
sequenceCounter++;
isStarted = false;
}
if (os.isCoR()) {
break;
}
os.setProgress(io->getProgress());
}
writer.close();
CHECK_OP(os, );
if (sequenceCounter == 0){
CHECK_EXT(isStarted == true, os.setError(RawDNASequenceFormat::tr("Sequence is empty")), );
}
static void load(IOAdapter* io, const U2DbiRef& dbiRef, const QVariantMap& fs, QList<GObject*>& objects,
int gapSize, QString& writeLockReason, U2OpStatus& os)
{
DbiOperationsBlock opBlock(dbiRef, os);
CHECK_OP(os, );
Q_UNUSED(opBlock);
static char fastaCommentStartChar = FastaFormat::FASTA_COMMENT_START_SYMBOL;
MemoryLocker memoryLocker(os, 1);
CHECK_OP(os, );
writeLockReason.clear();
QByteArray readBuff(DocumentFormat::READ_BUFF_SIZE + 1, 0);
char* buff = readBuff.data();
qint64 len = 0;
bool merge = gapSize != -1;
QStringList headers;
QSet<QString> uniqueNames;
QVector<U2Region> mergedMapping;
// for lower case annotations
GObjectReference sequenceRef;
//skip leading whites if present
bool lineOk = true;
static QBitArray nonWhites = ~TextUtils::WHITES;
io->readUntil(buff, DocumentFormat::READ_BUFF_SIZE, nonWhites, IOAdapter::Term_Exclude, &lineOk);
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
U2SequenceImporter seqImporter(fs, true);
const QString folder = fs.value(DocumentFormat::DBI_FOLDER_HINT, U2ObjectDbi::ROOT_FOLDER).toString();
qint64 sequenceStart = 0;
int sequenceNumber = 0;
DbiConnection con(dbiRef, os);
bool headerReaded = false;
QStringList emptySeqNames;
const int objectsCountLimit = fs.contains(DocumentReadingMode_MaxObjectsInDoc) ? fs[DocumentReadingMode_MaxObjectsInDoc].toInt() : -1;
const bool settingsMakeUniqueName = !fs.value(DocumentReadingMode_DontMakeUniqueNames, false).toBool();
while (!os.isCoR()) {
//skip start comments and read header
if(!headerReaded){
do{
len = io->readLine(buff, DocumentFormat::READ_BUFF_SIZE);
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
}while(buff[0] == fastaCommentStartChar && len > 0);
}
if (len == 0 && io->isEof()) { //end if stream
break;
}
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
CHECK_EXT_BREAK(lineOk, os.setError(FastaFormat::tr("Line is too long")));
QString headerLine = QString(QByteArray(buff+1, len-1)).trimmed();
CHECK_EXT_BREAK(buff[0] == FastaFormat::FASTA_HEADER_START_SYMBOL, os.setError(FastaFormat::tr("First line is not a FASTA header")));
//read sequence
if (sequenceNumber == 0 || !merge) {
QString objName = headerLine;
if(objName.isEmpty()){
objName = "Sequence";
}
if (settingsMakeUniqueName) {
objName = (merge) ? "Sequence" : TextUtils::variate(objName, "_", uniqueNames);
objName.squeeze();
memoryLocker.tryAcquire(2*objName.size());
CHECK_OP_BREAK(os);
uniqueNames.insert(objName);
}
seqImporter.startSequence(os, dbiRef, folder, objName, false);
CHECK_OP_BREAK(os);
sequenceRef = GObjectReference(io->getURL().getURLString(), objName, GObjectTypes::SEQUENCE);
}
if (sequenceNumber >= 1 && merge) {
seqImporter.addDefaultSymbolsBlock(gapSize, os);
sequenceStart += gapSize;
CHECK_OP_BREAK(os);
}
int sequenceLen = 0;
while (!os.isCoR()) {
do {
len = io->readLine(buff, DocumentFormat::READ_BUFF_SIZE);
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
} while (len <= 0 && !io->isEof());
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
if (len <= 0 && io->isEof()) {
break;
}
CHECK_EXT(!io->hasError(), os.setError(io->errorString()), );
buff[len] = 0;
if(buff[0] != fastaCommentStartChar && buff[0] != FastaFormat::FASTA_HEADER_START_SYMBOL){
len = TextUtils::remove(buff, len, TextUtils::WHITES);
if(len > 0){
seqImporter.addBlock(buff, len, os);
sequenceLen += len;
}
}else if( buff[0] == FastaFormat::FASTA_HEADER_START_SYMBOL){
headerReaded = true;
break;
}
CHECK_OP_BREAK(os);
os.setProgress(io->getProgress());
}
if (merge) {
memoryLocker.tryAcquire(headerLine.size());
CHECK_OP_BREAK(os);
headers.append(headerLine);
mergedMapping.append(U2Region(sequenceStart, sequenceLen));
} else {
if (objectsCountLimit > 0 && objects.size() >= objectsCountLimit) {
os.setError(FastaFormat::tr("File \"%1\" contains too many sequences to be displayed. "
"However, you can process these data using instruments from the menu <i>Tools -> NGS data analysis</i> "
"or pipelines built with Workflow Designer.")
.arg(io->getURL().getURLString()));
break;
}
memoryLocker.tryAcquire(800);
CHECK_OP_BREAK(os);
U2Sequence seq = seqImporter.finalizeSequenceAndValidate(os);
if (os.hasError() && os.getError() == U2SequenceImporter::EMPTY_SEQUENCE_ERROR) {
os.setError("");
emptySeqNames << headerLine;
continue;
}
sequenceRef.entityRef = U2EntityRef(dbiRef, seq.id);
//TODO parse header
U2StringAttribute attr(seq.id, DNAInfo::FASTA_HDR, headerLine);
con.dbi->getAttributeDbi()->createStringAttribute(attr, os);
CHECK_OP_BREAK(os);
objects << new U2SequenceObject(seq.visualName, U2EntityRef(dbiRef, seq.id));
CHECK_OP_BREAK(os);
U1AnnotationUtils::addAnnotations(objects, seqImporter.getCaseAnnotations(), sequenceRef, NULL, fs);
}
sequenceStart += sequenceLen;
sequenceNumber++;
ioLog.trace(QString("Sequence #%1 is processed").arg(sequenceNumber));
}
CHECK_OP_EXT(os, qDeleteAll(objects); objects.clear(), );
CHECK_EXT(!objects.isEmpty() || merge, os.setError(Document::tr("Document is empty.")), );
SAFE_POINT(headers.size() == mergedMapping.size(), "headers <-> regions mapping failed!", );
ioLog.trace("All sequences are processed");
if (!emptySeqNames.isEmpty()) {
QString warningMessage;
warningMessage.append(FastaFormat::tr("Loaded sequences: %1.\n").arg(sequenceNumber));
warningMessage.append(FastaFormat::tr("Skipped sequences: %1.\n").arg(emptySeqNames.size()));
warningMessage.append(FastaFormat::tr("The following sequences are empty:\n%1").arg(emptySeqNames.join(",\n")));
os.addWarning(warningMessage);
}
if (!merge) {
return;
}
U2Sequence seq = seqImporter.finalizeSequenceAndValidate(os);
CHECK_OP(os, );
sequenceRef.entityRef = U2EntityRef(dbiRef, seq.id);
U1AnnotationUtils::addAnnotations(objects, seqImporter.getCaseAnnotations(), sequenceRef, NULL, fs);
objects << new U2SequenceObject(seq.visualName, U2EntityRef(dbiRef, seq.id));
objects << DocumentFormatUtils::addAnnotationsForMergedU2Sequence( sequenceRef, dbiRef, headers, mergedMapping, fs );
if (headers.size() > 1) {
writeLockReason = QObject::tr("Document sequences were merged");
}
}
