X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=filterseqscommand.cpp;h=a12058e4785d00e1c21b84ea453ce82c4f62797d;hb=f663afa231c9bc1b5e18e0ea3bdd2b2ee784f5b2;hp=1ebac0eba5a7e7fc54d85b24dc705c0f8a8314b5;hpb=284fd95c611ccc3b1a7875c4dacfca06d1f50ed6;p=mothur.git diff --git a/filterseqscommand.cpp b/filterseqscommand.cpp index 1ebac0e..a12058e 100644 --- a/filterseqscommand.cpp +++ b/filterseqscommand.cpp @@ -44,7 +44,7 @@ FilterSeqsCommand::FilterSeqsCommand(string option) { it = parameters.find("fasta"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["fasta"] = inputDir + it->second; } } @@ -52,7 +52,7 @@ FilterSeqsCommand::FilterSeqsCommand(string option) { it = parameters.find("hard"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["hard"] = inputDir + it->second; } } @@ -62,25 +62,25 @@ FilterSeqsCommand::FilterSeqsCommand(string option) { fasta = validParameter.validFile(parameters, "fasta", false); if (fasta == "not found") { m->mothurOut("fasta is a required parameter for the filter.seqs command."); m->mothurOutEndLine(); abort = true; } else { - splitAtDash(fasta, fastafileNames); + m->splitAtDash(fasta, fastafileNames); //go through files and make sure they are good, if not, then disregard them for (int i = 0; i < fastafileNames.size(); i++) { if (inputDir != "") { - string path = hasPath(fastafileNames[i]); + string path = m->hasPath(fastafileNames[i]); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { fastafileNames[i] = inputDir + fastafileNames[i]; } } ifstream in; - int ableToOpen = openInputFile(fastafileNames[i], in, "noerror"); + int ableToOpen = m->openInputFile(fastafileNames[i], in, "noerror"); //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]); + string tryPath = m->getDefaultPath() + m->getSimpleName(fastafileNames[i]); m->mothurOut("Unable to open " + fastafileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine(); - ableToOpen = openInputFile(tryPath, in, "noerror"); + ableToOpen = m->openInputFile(tryPath, in, "noerror"); fastafileNames[i] = tryPath; } } @@ -92,7 +92,7 @@ FilterSeqsCommand::FilterSeqsCommand(string option) { fastafileNames.erase(fastafileNames.begin()+i); i--; }else{ - string simpleName = getSimpleName(fastafileNames[i]); + string simpleName = m->getSimpleName(fastafileNames[i]); filterFileName += simpleName.substr(0, simpleName.find_first_of('.')); } in.close(); @@ -106,7 +106,7 @@ FilterSeqsCommand::FilterSeqsCommand(string option) { //if the user changes the output directory command factory will send this info to us in the output parameter outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; - outputDir += hasPath(fastafileNames[0]); //if user entered a file with a path then preserve it + outputDir += m->hasPath(fastafileNames[0]); //if user entered a file with a path then preserve it } } //check for optional parameter and set defaults @@ -147,13 +147,14 @@ void FilterSeqsCommand::help(){ try { m->mothurOut("The filter.seqs command reads a file containing sequences and creates a .filter and .filter.fasta file.\n"); - m->mothurOut("The filter.seqs command parameters are fasta, trump, soft, hard and vertical. \n"); + m->mothurOut("The filter.seqs command parameters are fasta, trump, soft, hard, processors and vertical. \n"); m->mothurOut("The fasta parameter is required. You may enter several fasta files to build the filter from and filter, by separating their names with -'s.\n"); m->mothurOut("For example: fasta=abrecovery.fasta-amazon.fasta \n"); m->mothurOut("The trump parameter .... The default is ...\n"); m->mothurOut("The soft parameter .... The default is ....\n"); m->mothurOut("The hard parameter allows you to enter a file containing the filter you want to use.\n"); m->mothurOut("The vertical parameter removes columns where all sequences contain a gap character. The default is T.\n"); + m->mothurOut("The processors parameter allows you to specify the number of processors to use. The default is 1.\n"); m->mothurOut("The filter.seqs command should be in the following format: \n"); m->mothurOut("filter.seqs(fasta=yourFastaFile, trump=yourTrump) \n"); m->mothurOut("Example filter.seqs(fasta=abrecovery.fasta, trump=.).\n"); @@ -174,7 +175,7 @@ int FilterSeqsCommand::execute() { if (abort == true) { return 0; } ifstream inFASTA; - openInputFile(fastafileNames[0], inFASTA); + m->openInputFile(fastafileNames[0], inFASTA); Sequence testSeq(inFASTA); alignmentLength = testSeq.getAlignLength(); @@ -198,8 +199,12 @@ int FilterSeqsCommand::execute() { ofstream outFilter; - string filterFile = outputDir + filterFileName + ".filter"; - openOutputFile(filterFile, outFilter); + //prevent giantic file name + string filterFile; + if (fastafileNames.size() > 3) { filterFile = outputDir + "merge.filter"; } + else { filterFile = outputDir + filterFileName + ".filter"; } + + m->openOutputFile(filterFile, outFilter); outFilter << filter << endl; outFilter.close(); outputNames.push_back(filterFile); @@ -254,7 +259,7 @@ int FilterSeqsCommand::filterSequences() { for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); - string filteredFasta = outputDir + getRootName(getSimpleName(fastafileNames[s])) + "filter.fasta"; + string filteredFasta = outputDir + m->getRootName(m->getSimpleName(fastafileNames[s])) + "filter.fasta"; #ifdef USE_MPI int pid, start, end, numSeqsPerProcessor, num; int tag = 2001; @@ -283,7 +288,7 @@ int FilterSeqsCommand::filterSequences() { if (pid == 0) { //you are the root process - MPIPos = setFilePosFasta(fastafileNames[s], num); //fills MPIPos, returns numSeqs + MPIPos = m->setFilePosFasta(fastafileNames[s], num); //fills MPIPos, returns numSeqs numSeqs += num; //send file positions to all processes @@ -305,8 +310,8 @@ int FilterSeqsCommand::filterSequences() { //wait on chidren for(int i = 1; i < processors; i++) { - char buf[4]; - MPI_Recv(buf, 4, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status); + char buf[5]; + MPI_Recv(buf, 5, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status); } }else { //you are a child process @@ -326,11 +331,11 @@ int FilterSeqsCommand::filterSequences() { if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); return 0; } - char buf[4]; + char buf[5]; strcpy(buf, "done"); //tell parent you are done. - MPI_Send(buf, 4, MPI_CHAR, 0, tag, MPI_COMM_WORLD); + MPI_Send(buf, 5, MPI_CHAR, 0, tag, MPI_COMM_WORLD); } MPI_File_close(&outMPI); @@ -338,7 +343,7 @@ int FilterSeqsCommand::filterSequences() { MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case #else - vector positions = divideFile(fastafileNames[s], processors); + vector positions = m->divideFile(fastafileNames[s], processors); for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); @@ -355,14 +360,14 @@ int FilterSeqsCommand::filterSequences() { //append fasta files for(int i=1;iappendFiles((fastafileNames[s] + toString(processIDS[i]) + ".temp"), filteredFasta); remove((fastafileNames[s] + toString(processIDS[i]) + ".temp").c_str()); } } if (m->control_pressed) { return 1; } #else - numFastaSeqs = driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]); + int numFastaSeqs = driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]); numSeqs += numFastaSeqs; if (m->control_pressed) { return 1; } @@ -400,7 +405,7 @@ int FilterSeqsCommand::driverMPIRun(int start, int num, MPI_File& inMPI, MPI_Fil istringstream iss (tempBuf,istringstream::in); delete buf4; - Sequence seq(iss); gobble(iss); + Sequence seq(iss); m->gobble(iss); if (seq.getName() != "") { string align = seq.getAligned(); @@ -456,10 +461,10 @@ int FilterSeqsCommand::driverMPIRun(int start, int num, MPI_File& inMPI, MPI_Fil int FilterSeqsCommand::driverRunFilter(string F, string outputFilename, string inputFilename, linePair* filePos) { try { ofstream out; - openOutputFile(outputFilename, out); + m->openOutputFile(outputFilename, out); ifstream in; - openInputFile(inputFilename, in); + m->openInputFile(inputFilename, in); in.seekg(filePos->start); @@ -470,7 +475,7 @@ int FilterSeqsCommand::driverRunFilter(string F, string outputFilename, string i if (m->control_pressed) { in.close(); out.close(); return 0; } - Sequence seq(in); gobble(in); + Sequence seq(in); m->gobble(in); if (seq.getName() != "") { string align = seq.getAligned(); string filterSeq = ""; @@ -485,8 +490,12 @@ int FilterSeqsCommand::driverRunFilter(string F, string outputFilename, string i count++; } - unsigned long int pos = in.tellg(); - if ((pos == -1) || (pos >= filePos->end)) { break; } + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + unsigned long int pos = in.tellg(); + if ((pos == -1) || (pos >= filePos->end)) { break; } + #else + if (in.eof()) { break; } + #endif //report progress if((count) % 100 == 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); } @@ -527,8 +536,8 @@ int FilterSeqsCommand::createProcessesRunFilter(string F, string filename) { //pass numSeqs to parent ofstream out; - string tempFile = toString(getpid()) + ".temp"; - openOutputFile(tempFile, out); + string tempFile = filename + toString(getpid()) + ".num.temp"; + m->openOutputFile(tempFile, out); out << num << endl; out.close(); @@ -544,8 +553,8 @@ int FilterSeqsCommand::createProcessesRunFilter(string F, string filename) { for (int i = 0; i < processIDS.size(); i++) { ifstream in; - string tempFile = toString(processIDS[i]) + ".temp"; - openInputFile(tempFile, in); + string tempFile = filename + toString(processIDS[i]) + ".num.temp"; + m->openInputFile(tempFile, in); if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; } in.close(); remove(tempFile.c_str()); } @@ -570,7 +579,7 @@ string FilterSeqsCommand::createFilter() { F.setLength(alignmentLength); - if(trump != '*' || isTrue(vertical) || soft != 0){ + if(trump != '*' || m->isTrue(vertical) || soft != 0){ F.initialize(); } @@ -578,7 +587,7 @@ string FilterSeqsCommand::createFilter() { else { F.setFilter(string(alignmentLength, '1')); } numSeqs = 0; - if(trump != '*' || isTrue(vertical) || soft != 0){ + if(trump != '*' || m->isTrue(vertical) || soft != 0){ for (int s = 0; s < fastafileNames.size(); s++) { for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); @@ -604,7 +613,7 @@ string FilterSeqsCommand::createFilter() { if (m->control_pressed) { MPI_File_close(&inMPI); return 0; } if (pid == 0) { //you are the root process - MPIPos = setFilePosFasta(fastafileNames[s], num); //fills MPIPos, returns numSeqs + MPIPos = m->setFilePosFasta(fastafileNames[s], num); //fills MPIPos, returns numSeqs numSeqs += num; //send file positions to all processes @@ -646,7 +655,7 @@ string FilterSeqsCommand::createFilter() { MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case #else - vector positions = divideFile(fastafileNames[s], processors); + vector positions = m->divideFile(fastafileNames[s], processors); for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); @@ -662,7 +671,7 @@ string FilterSeqsCommand::createFilter() { if (m->control_pressed) { return filterString; } #else - numFastaSeqs = driverCreateFilter(F, fastafileNames[s], lines[0]); + int numFastaSeqs = driverCreateFilter(F, fastafileNames[s], lines[0]); numSeqs += numFastaSeqs; if (m->control_pressed) { return filterString; } #endif @@ -679,7 +688,7 @@ string FilterSeqsCommand::createFilter() { MPI_Comm_rank(MPI_COMM_WORLD, &pid); - if(trump != '*' || isTrue(vertical) || soft != 0){ + if(trump != '*' || m->isTrue(vertical) || soft != 0){ if (pid == 0) { //only one process should output the filter @@ -727,11 +736,10 @@ string FilterSeqsCommand::createFilter() { if (pid == 0) { //only one process should output the filter #endif + F.setNumSeqs(numSeqs); - - if(isTrue(vertical) == 1) { F.doVertical(); } + if(m->isTrue(vertical) == 1) { F.doVertical(); } if(soft != 0) { F.doSoft(); } - filterString = F.getFilter(); #ifdef USE_MPI @@ -766,7 +774,7 @@ int FilterSeqsCommand::driverCreateFilter(Filters& F, string filename, linePair* try { ifstream in; - openInputFile(filename, in); + m->openInputFile(filename, in); in.seekg(filePos->start); @@ -777,18 +785,22 @@ int FilterSeqsCommand::driverCreateFilter(Filters& F, string filename, linePair* if (m->control_pressed) { in.close(); return 1; } - Sequence seq(in); gobble(in); + Sequence seq(in); m->gobble(in); if (seq.getName() != "") { if (seq.getAligned().length() != alignmentLength) { m->mothurOut("Sequences are not all the same length, please correct."); m->mothurOutEndLine(); m->control_pressed = true; } - if(trump != '*'){ F.doTrump(seq); } - if(isTrue(vertical) || soft != 0){ F.getFreqs(seq); } + if(trump != '*') { F.doTrump(seq); } + if(m->isTrue(vertical) || soft != 0) { F.getFreqs(seq); } cout.flush(); count++; } - unsigned long int pos = in.tellg(); - if ((pos == -1) || (pos >= filePos->end)) { break; } + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + unsigned long int pos = in.tellg(); + if ((pos == -1) || (pos >= filePos->end)) { break; } + #else + if (in.eof()) { break; } + #endif //report progress if((count) % 100 == 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); } @@ -833,7 +845,7 @@ int FilterSeqsCommand::MPICreateFilter(int start, int num, Filters& F, MPI_File& if (seq.getAligned().length() != alignmentLength) { cout << "Alignment length is " << alignmentLength << " and sequence " << seq.getName() << " has length " << seq.getAligned().length() << ", please correct." << endl; exit(1); } if(trump != '*'){ F.doTrump(seq); } - if(isTrue(vertical) || soft != 0){ F.getFreqs(seq); } + if(m->isTrue(vertical) || soft != 0){ F.getFreqs(seq); } cout.flush(); //report progress @@ -856,7 +868,7 @@ int FilterSeqsCommand::MPICreateFilter(int start, int num, Filters& F, MPI_File& int FilterSeqsCommand::createProcessesCreateFilter(Filters& F, string filename) { try { #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) - int process = 0; + int process = 1; int num = 0; processIDS.clear(); @@ -868,45 +880,63 @@ int FilterSeqsCommand::createProcessesCreateFilter(Filters& F, string filename) 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){ + //reset child's filter counts to 0; + F.a.clear(); F.a.resize(alignmentLength, 0); + F.t.clear(); F.t.resize(alignmentLength, 0); + F.g.clear(); F.g.resize(alignmentLength, 0); + F.c.clear(); F.c.resize(alignmentLength, 0); + F.gap.clear(); F.gap.resize(alignmentLength, 0); + num = driverCreateFilter(F, filename, lines[process]); //write out filter counts to file filename += toString(getpid()) + "filterValues.temp"; ofstream out; - openOutputFile(filename, out); + m->openOutputFile(filename, out); out << num << endl; + out << F.getFilter() << endl; for (int k = 0; k < alignmentLength; k++) { out << F.a[k] << '\t'; } out << endl; for (int k = 0; k < alignmentLength; k++) { out << F.t[k] << '\t'; } out << endl; for (int k = 0; k < alignmentLength; k++) { out << F.g[k] << '\t'; } out << endl; for (int k = 0; k < alignmentLength; k++) { out << F.c[k] << '\t'; } out << endl; for (int k = 0; k < alignmentLength; k++) { out << F.gap[k] << '\t'; } out << endl; - + + //cout << F.getFilter() << endl; out.close(); exit(0); }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); } } + //parent do your part + num = driverCreateFilter(F, filename, lines[0]); + //force parent to wait until all the processes are done - for (int i=0;iopenInputFile(tempFilename, in); int temp, tempNum; - in >> tempNum; gobble(in); num += tempNum; - for (int k = 0; k < alignmentLength; k++) { in >> temp; F.a[k] += temp; } gobble(in); - for (int k = 0; k < alignmentLength; k++) { in >> temp; F.t[k] += temp; } gobble(in); - for (int k = 0; k < alignmentLength; k++) { in >> temp; F.g[k] += temp; } gobble(in); - for (int k = 0; k < alignmentLength; k++) { in >> temp; F.c[k] += temp; } gobble(in); - for (int k = 0; k < alignmentLength; k++) { in >> temp; F.gap[k] += temp; } gobble(in); + string tempFilterString; + + in >> tempNum; m->gobble(in); num += tempNum; + + in >> tempFilterString; + F.mergeFilter(tempFilterString); + + for (int k = 0; k < alignmentLength; k++) { in >> temp; F.a[k] += temp; } m->gobble(in); + for (int k = 0; k < alignmentLength; k++) { in >> temp; F.t[k] += temp; } m->gobble(in); + for (int k = 0; k < alignmentLength; k++) { in >> temp; F.g[k] += temp; } m->gobble(in); + for (int k = 0; k < alignmentLength; k++) { in >> temp; F.c[k] += temp; } m->gobble(in); + for (int k = 0; k < alignmentLength; k++) { in >> temp; F.gap[k] += temp; } m->gobble(in); in.close(); remove(tempFilename.c_str());