bool
Genome::getSizeFromFile(const char *fileName, unsigned *nBases, unsigned *nContigs)
{
GenericFile *file;
unsigned localNBases, localnContigs;
if (!openFileAndGetSizes(fileName,&file,nBases ? nBases : &localNBases, nContigs ? nContigs : &localnContigs)) {
return false;
}
file->close();
delete file;
return true;
}
const Genome *
Genome::loadFromFile(const char *fileName, unsigned chromosomePadding, unsigned i_minOffset, unsigned length)
{
GenericFile *loadFile;
unsigned nBases,nContigs;
if (!openFileAndGetSizes(fileName,&loadFile,&nBases,&nContigs)) {
//
// It already printed an error. Just fail.
//
return NULL;
}
if (0 == length) {
length = nBases - i_minOffset;
} else {
//
// Don't let length go beyond nBases.
//
length = __min(length,nBases - i_minOffset);
}
Genome *genome = new Genome(nBases,length, chromosomePadding);
genome->nBases = nBases;
genome->nContigs = genome->maxContigs = nContigs;
genome->contigs = new Contig[nContigs];
genome->minOffset = i_minOffset;
if (i_minOffset >= nBases) {
WriteErrorMessage("Genome::loadFromFile: specified minOffset %u >= nBases %u\n",i_minOffset,nBases);
}
genome->maxOffset = i_minOffset + length;
static const unsigned contigNameBufferSize = 512;
char contigNameBuffer[contigNameBufferSize];
unsigned n;
size_t contigSize;
char *curName;
for (unsigned i = 0; i < nContigs; i++) {
if (NULL == loadFile->gets(contigNameBuffer, contigNameBufferSize)){
WriteErrorMessage("Unable to read contig description\n");
delete genome;
return NULL;
}
for (n = 0; n < contigNameBufferSize; n++){
if (contigNameBuffer[n] == ' ') {
contigNameBuffer[n] = '\0';
break;
}
}
genome->contigs[i].beginningOffset = atoi(contigNameBuffer);
contigNameBuffer[n] = ' ';
n++; // increment n so we start copying at the position after the space
contigSize = strlen(contigNameBuffer + n) - 1; //don't include the final \n
genome->contigs[i].name = new char[contigSize + 1];
genome->contigs[i].nameLength = (unsigned)contigSize;
curName = genome->contigs[i].name;
for (unsigned pos = 0; pos < contigSize; pos++) {
curName[pos] = contigNameBuffer[pos + n];
}
curName[contigSize] = '\0';
}
//
// Skip over the miserable \n that gets left in the file.
//
/* char newline;
if (1 != fread(&newline,1,1,loadFile)) {
WriteErrorMessage("Genome::loadFromFile: Unable to read expected newline\n");
delete genome;
return NULL;
}
if (newline != 10) {
WriteErrorMessage("Genome::loadFromFile: Expected newline to be 0x0a, got 0x%02x\n",newline);
delete genome;
return NULL;
}
*/
if (0 != loadFile->advance(i_minOffset)) {
WriteErrorMessage("Genome::loadFromFile: _fseek64bit failed\n");
soft_exit(1);
}
size_t retval;
if (length != (retval = loadFile->read(genome->bases,length))) {
WriteErrorMessage("Genome::loadFromFile: fread of bases failed; wanted %u, got %d\n", length, retval);
loadFile->close();
delete loadFile;
delete genome;
return NULL;
}
loadFile->close();
delete loadFile;
genome->fillInContigLengths();
genome->sortContigsByName();
return genome;
}
const Genome *
Genome::loadFromFile(const char *fileName, unsigned chromosomePadding, GenomeLocation minLocation, GenomeDistance length, bool map)
{
GenericFile *loadFile;
GenomeDistance nBases;
unsigned nContigs;
if (!openFileAndGetSizes(fileName, &loadFile, &nBases, &nContigs, map)) {
//
// It already printed an error. Just fail.
//
return NULL;
}
GenomeLocation maxLocation(nBases);
if (0 == length) {
length = maxLocation - minLocation;
} else {
//
// Don't let length go beyond nBases.
//
length = __min(length, maxLocation - minLocation);
maxLocation = minLocation + length;
}
Genome *genome = new Genome(nBases, length, chromosomePadding);
genome->nBases = nBases;
genome->nContigs = genome->maxContigs = nContigs;
genome->contigs = new Contig[nContigs];
genome->minLocation = minLocation;
if (GenomeLocationAsInt64(minLocation) >= nBases) {
WriteErrorMessage("Genome::loadFromFile: specified minOffset %u >= nBases %u\n", GenomeLocationAsInt64(minLocation), nBases);
soft_exit(-1);
}
genome->maxLocation = maxLocation;
static const unsigned contigNameBufferSize = 512;
char contigNameBuffer[contigNameBufferSize];
unsigned n;
size_t contigSize;
char *curName;
for (unsigned i = 0; i < nContigs; i++) {
if (NULL == loadFile->gets(contigNameBuffer, contigNameBufferSize)){
WriteErrorMessage("Unable to read contig description\n");
delete genome;
return NULL;
}
for (n = 0; n < contigNameBufferSize; n++){
if (contigNameBuffer[n] == ' ') {
contigNameBuffer[n] = '\0';
break;
}
}
_int64 contigStart;
if (1 != sscanf(contigNameBuffer, "%lld", &contigStart)) {
WriteErrorMessage("Unable to parse contig start in genome file '%s', '%s%'\n", fileName, contigNameBuffer);
soft_exit(1);
}
genome->contigs[i].beginningLocation = GenomeLocation(contigStart);
contigNameBuffer[n] = ' ';
n++; // increment n so we start copying at the position after the space
contigSize = strlen(contigNameBuffer + n) - 1; //don't include the final \n
genome->contigs[i].name = new char[contigSize + 1];
genome->contigs[i].nameLength = (unsigned)contigSize;
curName = genome->contigs[i].name;
for (unsigned pos = 0; pos < contigSize; pos++) {
curName[pos] = contigNameBuffer[pos + n];
}
curName[contigSize] = '\0';
}
if (0 != loadFile->advance(GenomeLocationAsInt64(minLocation))) {
WriteErrorMessage("Genome::loadFromFile: _fseek64bit failed\n");
soft_exit(1);
}
size_t readSize;
if (map) {
GenericFile_map *mappedFile = (GenericFile_map *)loadFile;
genome->bases = (char *)mappedFile->mapAndAdvance(length, &readSize);
genome->mappedFile = mappedFile;
mappedFile->prefetch();
} else {
readSize = loadFile->read(genome->bases, length);
loadFile->close();
delete loadFile;
loadFile = NULL;
}
if (length != readSize) {
WriteErrorMessage("Genome::loadFromFile: fread of bases failed; wanted %u, got %d\n", length, readSize);
delete loadFile;
delete genome;
return NULL;
}
genome->fillInContigLengths();
genome->sortContigsByName();
return genome;
}