uint genomeScanFastaFiles (Parameters *P, char* G, bool flagRun) {//scans fasta files. flagRun=false: check and find full size, flaRun=true: collect all the data
uint N=0;//total number of bases in the genome, including chr "spacers"
if (!flagRun && P->chrLength.size()>0)
{//previous chr records exist
P->chrStart.pop_back();//remove last record, it will be recorded again
N = P->chrStart.back()+P->chrLength.back();
P->chrLength.pop_back();//remove last record, it will be recorded again
};
ifstream fileIn;
for (uint ii=0;ii<P->genomeFastaFiles.size();ii++) {//all the input files
fileIn.open(P->genomeFastaFiles.at(ii).c_str());
if ( !fileIn.good() )
{//
ostringstream errOut;
errOut << "EXITING because of INPUT ERROR: could not open genomeFastaFile: " <<P->genomeFastaFiles.at(ii) <<"\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
};
char cc=fileIn.peek();
if ( !fileIn.good() )
{//
ostringstream errOut;
errOut << "EXITING because of INPUT ERROR: could not read from genomeFastaFile: " <<P->genomeFastaFiles.at(ii) <<"\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
};
if (cc!='>')
{
ostringstream errOut;
errOut << "EXITING because of INPUT ERROR: the file format of the genomeFastaFile: " <<P->genomeFastaFiles.at(ii) << "is not fasta:";
errOut << " the first character is " <<cc<<" , not > .\n";
errOut << " Solution: check formatting of the fasta file. Make sure the file is uncompressed (unzipped).\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
}; while(!fileIn.eof()) {//read each file until eof
string lineIn (4096,'.');
getline(fileIn,lineIn);
if (lineIn[0]=='>') {//new chromosome
if (!flagRun) {
istringstream lineInStream (lineIn);
lineInStream.ignore(1,' ');
string chrName1;
lineInStream >> chrName1;
P->chrName.push_back(chrName1);
};
if (!flagRun && P->chrStart.size()>0) P->chrLength.push_back(N-P->chrStart.at(P->chrStart.size()-1)); //true length of the chr
if (N>0) {//pad the chromosomes to bins boudnaries
N = ( (N+1)/P->genomeChrBinNbases+1 )*P->genomeChrBinNbases;
};
if (!flagRun) {
P->chrStart.push_back(N);
P->inOut->logMain << P->genomeFastaFiles.at(ii)<<" : chr # " << P->chrStart.size()-1 << " \""<<P->chrName.at(P->chrStart.size()-1)<<"\" chrStart: "<<N<<"\n"<<flush;
};
} else {//char lines
if (flagRun) lineIn.copy(G+N,lineIn.size(),0);
N += lineIn.size();
};
};
void cVehicle::readFrom(std::istream& file) {
while(file) {
std::string cur;
getline(file, cur);
std::istringstream lineIn(cur);
lineIn >> cur;
if(cur == "LOCATION")
lineIn >> loc.x >> loc.y;
else if(cur == "SECTOR")
lineIn >> sector.x >> sector.y;
else if(cur == "IN")
uint genomeScanFastaFiles (Parameters *P, char* G, bool flagRun) {//scans fasta files. flagRun=false: check and find full size, flaRun=true: collect all the data
uint N=0; //total number of bases in the genome, including chr "spacers"
ifstream fileIn;
for (uint ii=0;ii<P->genomeFastaFiles.size();ii++) {//all the input files
fileIn.open(P->genomeFastaFiles.at(ii).c_str());
if ( !fileIn.good() ) {//
ostringstream errOut;
errOut << "EXITING because of INPUT ERROR: could not open genomeFastaFile: " <<P->genomeFastaFiles.at(ii) <<"\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
};
while(!fileIn.eof()) {//read each file until eof
string lineIn (4096,'.');
getline(fileIn,lineIn);
if (lineIn[0]=='>') {//new chromosome
if (!flagRun) {
istringstream lineInStream (lineIn);
lineInStream.ignore(1,' ');
string chrName1;
lineInStream >> chrName1;
P->chrName.push_back(chrName1);
};
if (!flagRun && P->chrStart.size()>0) P->chrLength.push_back(N-P->chrStart.at(P->chrStart.size()-1)); //true length of the chr
if (N>0) {//pad the chromosomes to bins boudnaries
N = ( (N+1)/P->genomeChrBinNbases+1 )*P->genomeChrBinNbases;
};
if (!flagRun) {
P->chrStart.push_back(N);
P->inOut->logMain << P->genomeFastaFiles.at(ii)<<" : chr # " << P->chrStart.size()-1 << " \""<<P->chrName.at(P->chrStart.size()-1)<<"\" chrStart: "<<N<<"\n"<<flush;
};
} else {//char lines
if (flagRun) lineIn.copy(G+N,lineIn.size(),0);
N += lineIn.size();
};
};