OptionParser parser(option);
map<string,string> parameters = parser.getParameters();
- ValidParameters validParameter;
+ ValidParameters validParameter("chimera.pintail");
map<string,string>::iterator it;
//check to make sure all parameters are valid for command
int ableToOpen;
ifstream in;
- #ifdef USE_MPI
- int pid;
- MPI_Comm_size(MPI_COMM_WORLD, &processors); //set processors to the number of mpi processes running
- MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
+ ableToOpen = openInputFile(fastaFileNames[i], in, "noerror");
- if (pid == 0) {
- #endif
-
- ableToOpen = openInputFile(fastaFileNames[i], in);
+ //if you can't open it, try default location
+ if (ableToOpen == 1) {
+ if (m->getDefaultPath() != "") { //default path is set
+ string tryPath = m->getDefaultPath() + getSimpleName(fastaFileNames[i]);
+ m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
+ ableToOpen = openInputFile(tryPath, in, "noerror");
+ fastaFileNames[i] = tryPath;
+ }
+ }
in.close();
- #ifdef USE_MPI
- for (int j = 1; j < processors; j++) {
- MPI_Send(&ableToOpen, 1, MPI_INT, j, 2001, MPI_COMM_WORLD);
- }
- }else{
- MPI_Status status;
- MPI_Recv(&ableToOpen, 1, MPI_INT, 0, 2001, MPI_COMM_WORLD, &status);
- }
-
- #endif
-
if (ableToOpen == 1) {
- m->mothurOut(fastaFileNames[i] + " will be disregarded."); m->mothurOutEndLine();
+ m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
//erase from file list
fastaFileNames.erase(fastaFileNames.begin()+i);
i--;
//check for consfile
string tempConsFile = getRootName(inputDir + getSimpleName(templatefile)) + "freq";
ifstream FileTest(tempConsFile.c_str());
- if(FileTest){ m->mothurOut("I found " + tempConsFile + " in your input file directory. I will use it to save time."); m->mothurOutEndLine(); consfile = tempConsFile; FileTest.close(); }
+ if(FileTest){
+ bool GoodFile = checkReleaseVersion(FileTest, m->getVersion());
+ if (GoodFile) {
+ m->mothurOut("I found " + tempConsFile + " in your input file directory. I will use it to save time."); m->mothurOutEndLine(); consfile = tempConsFile; FileTest.close();
+ }
+ }
}
quanfile = validParameter.validFile(parameters, "quantile", true);
if (quanfile == "not open") { abort = true; }
- else if (quanfile == "not found") { quanfile = ""; }
+ else if (quanfile == "not found") { quanfile = ""; }
}
}
catch(exception& e) {
}
ifstream FileTest(tempQuan.c_str());
- if(FileTest){ m->mothurOut("I found " + tempQuan + " in your input file directory. I will use it to save time."); m->mothurOutEndLine(); quanfile = tempQuan; FileTest.close(); }
+ if(FileTest){
+ bool GoodFile = checkReleaseVersion(FileTest, m->getVersion());
+ if (GoodFile) {
+ m->mothurOut("I found " + tempQuan + " in your input file directory. I will use it to save time."); m->mothurOutEndLine(); quanfile = tempQuan; FileTest.close();
+ }
+ }
chimera = new Pintail(fastaFileNames[s], templatefile, filter, processors, maskfile, consfile, quanfile, window, increment, outputDir);
outputFileName = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + "pintail.chimeras";
accnosFileName = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + "pintail.accnos";
}
- bool hasAccnos = true;
if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; }
#ifdef USE_MPI
int pid, end, numSeqsPerProcessor;
int tag = 2001;
- vector<long> MPIPos;
- MPIWroteAccnos = false;
+ vector<unsigned long int> MPIPos;
MPI_Status status;
MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); remove(outputFileName.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; }
- for (int i = 1; i < processors; i++) {
- bool tempResult;
- MPI_Recv(&tempResult, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
- if (tempResult != 0) { MPIWroteAccnos = true; }
- }
}else{ //you are a child process
MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
MPIPos.resize(numSeqs+1);
driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, MPIPos);
if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; }
-
- MPI_Send(&MPIWroteAccnos, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
}
//close files
MPI_File_close(&outMPI);
MPI_File_close(&outMPIAccnos);
MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
-
- //delete accnos file if blank
- if (pid == 0) {
- if (!MPIWroteAccnos) {
- hasAccnos = false;
- remove(accnosFileName.c_str());
- }
- }
-
#else
-
+ vector<unsigned long int> positions = divideFile(fastaFileNames[s], processors);
+
+ for (int i = 0; i < (positions.size()-1); i++) {
+ lines.push_back(new linePair(positions[i], positions[(i+1)]));
+ }
+
//break up file
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
if(processors == 1){
- ifstream inFASTA;
- openInputFile(fastaFileNames[s], inFASTA);
- getNumSeqs(inFASTA, numSeqs);
- inFASTA.close();
-
- lines.push_back(new linePair(0, numSeqs));
+
+ numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
- driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
+ if (m->control_pressed) { remove(outputFileName.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
- if (m->control_pressed) {
- remove(outputFileName.c_str());
- remove(accnosFileName.c_str());
- for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); }
- for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
- delete chimera;
- return 0;
- }
-
- //delete accnos file if its blank
- if (isBlank(accnosFileName)) { remove(accnosFileName.c_str()); hasAccnos = false; }
-
}else{
- vector<int> positions;
processIDS.resize(0);
- ifstream inFASTA;
- openInputFile(fastaFileNames[s], inFASTA);
-
- string input;
- while(!inFASTA.eof()){
- input = getline(inFASTA);
- if (input.length() != 0) {
- if(input[0] == '>'){ long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
- }
- }
- inFASTA.close();
-
- numSeqs = positions.size();
-
- int numSeqsPerProcessor = numSeqs / processors;
-
- for (int i = 0; i < processors; i++) {
- long int startPos = positions[ i * numSeqsPerProcessor ];
- if(i == processors - 1){
- numSeqsPerProcessor = numSeqs - i * numSeqsPerProcessor;
- }
- lines.push_back(new linePair(startPos, numSeqsPerProcessor));
- }
-
- createProcesses(outputFileName, fastaFileNames[s], accnosFileName);
+ numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName);
rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
+ rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str());
//append output files
for(int i=1;i<processors;i++){
remove((outputFileName + toString(processIDS[i]) + ".temp").c_str());
}
- vector<string> nonBlankAccnosFiles;
- //delete blank accnos files generated with multiple processes
- for(int i=0;i<processors;i++){
- if (!(isBlank(accnosFileName + toString(processIDS[i]) + ".temp"))) {
- nonBlankAccnosFiles.push_back(accnosFileName + toString(processIDS[i]) + ".temp");
- }else { remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str()); }
+ //append output files
+ for(int i=1;i<processors;i++){
+ appendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName);
+ remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str());
}
-
- //append accnos files
- if (nonBlankAccnosFiles.size() != 0) {
- rename(nonBlankAccnosFiles[0].c_str(), accnosFileName.c_str());
-
- for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
- appendFiles(nonBlankAccnosFiles[h], accnosFileName);
- remove(nonBlankAccnosFiles[h].c_str());
- }
- }else{ hasAccnos = false; }
-
+
if (m->control_pressed) {
remove(outputFileName.c_str());
remove(accnosFileName.c_str());
}
#else
- ifstream inFASTA;
- openInputFile(fastaFileNames[s], inFASTA);
- getNumSeqs(inFASTA, numSeqs);
- inFASTA.close();
- lines.push_back(new linePair(0, numSeqs));
-
- driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
+ numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
- if (m->control_pressed) {
- remove(outputFileName.c_str());
- remove(accnosFileName.c_str());
- for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); }
- for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
- delete chimera;
- return 0;
- }
-
- //delete accnos file if its blank
- if (isBlank(accnosFileName)) { remove(accnosFileName.c_str()); hasAccnos = false; }
+ if (m->control_pressed) { remove(outputFileName.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
#endif
#endif
for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
outputNames.push_back(outputFileName);
- if (hasAccnos) { outputNames.push_back(accnosFileName); }
+ outputNames.push_back(accnosFileName);
m->mothurOutEndLine();
m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
}
//**********************************************************************************************************************
-int ChimeraPintailCommand::driver(linePair* line, string outputFName, string filename, string accnos){
+int ChimeraPintailCommand::driver(linePair* filePos, string outputFName, string filename, string accnos){
try {
ofstream out;
openOutputFile(outputFName, out);
ifstream inFASTA;
openInputFile(filename, inFASTA);
- inFASTA.seekg(line->start);
-
- for(int i=0;i<line->numSeqs;i++){
-
+ inFASTA.seekg(filePos->start);
+
+ bool done = false;
+ int count = 0;
+
+ while (!done) {
+
if (m->control_pressed) { return 1; }
Sequence* candidateSeq = new Sequence(inFASTA); gobble(inFASTA);
//print results
chimera->print(out, out2);
}
+ count++;
}
delete candidateSeq;
+ unsigned long int pos = inFASTA.tellg();
+ if ((pos == -1) || (pos >= filePos->end)) { break; }
+
//report progress
- if((i+1) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(i+1)); m->mothurOutEndLine(); }
+ if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
}
//report progress
- if((line->numSeqs) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(line->numSeqs)); m->mothurOutEndLine(); }
+ if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
out.close();
out2.close();
inFASTA.close();
- return 0;
+ return count;
}
catch(exception& e) {
m->errorOut(e, "ChimeraPintailCommand", "driver");
}
//**********************************************************************************************************************
#ifdef USE_MPI
-int ChimeraPintailCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, vector<long>& MPIPos){
+int ChimeraPintailCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, vector<unsigned long int>& MPIPos){
try {
MPI_Status status;
//print results
bool isChimeric = chimera->print(outMPI, outAccMPI);
- if (isChimeric) { MPIWroteAccnos = true; }
}
}
delete candidateSeq;
try {
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
int process = 0;
- // processIDS.resize(0);
+ int num = 0;
//loop through and create all the processes you want
while (process != processors) {
processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
process++;
}else if (pid == 0){
- driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename, accnos + toString(getpid()) + ".temp");
+ num = driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename, accnos + toString(getpid()) + ".temp");
+
+ //pass numSeqs to parent
+ ofstream out;
+ string tempFile = toString(getpid()) + ".temp";
+ openOutputFile(tempFile, out);
+ out << num << endl;
+ out.close();
+
exit(0);
}else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
}
wait(&temp);
}
- return 0;
+ for (int i = 0; i < processIDS.size(); i++) {
+ ifstream in;
+ string tempFile = toString(processIDS[i]) + ".temp";
+ openInputFile(tempFile, in);
+ if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
+ in.close(); remove(tempFile.c_str());
+ }
+
+ return num;
#endif
}
catch(exception& e) {