X-Git-Url: https://git.donarmstrong.com/?p=mothur.git;a=blobdiff_plain;f=chimeracheckcommand.cpp;h=31a36f73c6d0339d775dd6a1a4efa8e523c0f14f;hp=625314a8f6a622618b5af87d8bc5889ee3fe8a3e;hb=c48d91112209b841444923670dca5454da0e2a4d;hpb=a61980e36d61937acccadc41be260165a1e389f4 diff --git a/chimeracheckcommand.cpp b/chimeracheckcommand.cpp index 625314a..31a36f7 100644 --- a/chimeracheckcommand.cpp +++ b/chimeracheckcommand.cpp @@ -8,26 +8,107 @@ */ #include "chimeracheckcommand.h" -#include "chimeracheckrdp.h" +#include "referencedb.h" -//*************************************************************************************************************** +//********************************************************************************************************************** +vector ChimeraCheckCommand::setParameters(){ + try { + CommandParameter ptemplate("reference", "InputTypes", "", "", "none", "none", "none","",false,true,true); parameters.push_back(ptemplate); + CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none","chimera",false,true,true); parameters.push_back(pfasta); + CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none","",false,false,true); parameters.push_back(pname); + CommandParameter psvg("svg", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(psvg); + CommandParameter pincrement("increment", "Number", "", "10", "", "", "","",false,false); parameters.push_back(pincrement); + CommandParameter pksize("ksize", "Number", "", "7", "", "", "","",false,false); parameters.push_back(pksize); + CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors); + CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir); + CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir); + CommandParameter psave("save", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(psave); + vector myArray; + for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } + return myArray; + } + catch(exception& e) { + m->errorOut(e, "ChimeraCheckCommand", "setParameters"); + exit(1); + } +} +//********************************************************************************************************************** +string ChimeraCheckCommand::getHelpString(){ + try { + string helpString = ""; + helpString += "The chimera.check command reads a fastafile and referencefile and outputs potentially chimeric sequences.\n"; + helpString += "This command was created using the algorythms described in CHIMERA_CHECK version 2.7 written by Niels Larsen. \n"; + helpString += "The chimera.check command parameters are fasta, reference, processors, ksize, increment, svg and name.\n"; + helpString += "The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required unless you have a valid current fasta file. \n"; + helpString += "You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n"; + helpString += "The reference parameter allows you to enter a reference file containing known non-chimeric sequences, and is required. \n"; + helpString += "The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n"; +#ifdef USE_MPI + helpString += "When using MPI, the processors parameter is set to the number of MPI processes running. \n"; +#endif + helpString += "The increment parameter allows you to specify how far you move each window while finding chimeric sequences, default is 10.\n"; + helpString += "The ksize parameter allows you to input kmersize, default is 7. \n"; + helpString += "The svg parameter allows you to specify whether or not you would like a svg file outputted for each query sequence, default is False.\n"; + helpString += "The name parameter allows you to enter a file containing names of sequences you would like .svg files for.\n"; + helpString += "You may enter multiple name files by separating their names with dashes. ie. fasta=abrecovery.svg.names-amzon.svg.names \n"; + helpString += "If the save parameter is set to true the reference sequences will be saved in memory, to clear them later you can use the clear.memory command. Default=f."; + helpString += "The chimera.check command should be in the following format: \n"; + helpString += "chimera.check(fasta=yourFastaFile, reference=yourTemplateFile, processors=yourProcessors, ksize=yourKmerSize) \n"; + helpString += "Example: chimera.check(fasta=AD.fasta, reference=core_set_aligned,imputed.fasta, processors=4, ksize=8) \n"; + helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n"; + return helpString; + } + catch(exception& e) { + m->errorOut(e, "ChimeraCheckCommand", "getHelpString"); + exit(1); + } +} +//********************************************************************************************************************** +string ChimeraCheckCommand::getOutputPattern(string type) { + try { + string pattern = ""; + + if (type == "chimera") { pattern = "[filename],chimeracheck.chimeras"; } + else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; } + + return pattern; + } + catch(exception& e) { + m->errorOut(e, "ChimeraCheckCommand", "getOutputPattern"); + exit(1); + } +} +//********************************************************************************************************************** +ChimeraCheckCommand::ChimeraCheckCommand(){ + try { + abort = true; calledHelp = true; + setParameters(); + vector tempOutNames; + outputTypes["chimera"] = tempOutNames; + } + catch(exception& e) { + m->errorOut(e, "ChimeraCheckCommand", "ChimeraCheckCommand"); + exit(1); + } +} +//*************************************************************************************************************** ChimeraCheckCommand::ChimeraCheckCommand(string option) { try { - abort = false; + abort = false; calledHelp = false; + ReferenceDB* rdb = ReferenceDB::getInstance(); //allow user to run help - if(option == "help") { help(); abort = true; } + if(option == "help") { help(); abort = true; calledHelp = true; } + else if(option == "citation") { citation(); abort = true; calledHelp = true;} else { - //valid paramters for this command - string Array[] = {"fasta","processors","increment","template","ksize","svg", "name","outputdir","inputdir" }; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + vector myArray = setParameters(); OptionParser parser(option); map parameters = parser.getParameters(); - ValidParameters validParameter; + ValidParameters validParameter("chimera.check"); map::iterator it; //check to make sure all parameters are valid for command @@ -35,67 +116,219 @@ ChimeraCheckCommand::ChimeraCheckCommand(string option) { if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } } + vector tempOutNames; + outputTypes["chimera"] = tempOutNames; + //if the user changes the input directory command factory will send this info to us in the output parameter string inputDir = validParameter.validFile(parameters, "inputdir", false); if (inputDir == "not found"){ inputDir = ""; } else { - string path; - it = parameters.find("fasta"); + it = parameters.find("reference"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + string 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; } + if (path == "") { parameters["reference"] = inputDir + it->second; } } + } + + //check for required parameters + fastafile = validParameter.validFile(parameters, "fasta", false); + if (fastafile == "not found") { + //if there is a current fasta file, use it + string filename = m->getFastaFile(); + if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); } + else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; } + }else { + m->splitAtDash(fastafile, fastaFileNames); - it = parameters.find("template"); - //user has given a template file - if(it != parameters.end()){ - path = hasPath(it->second); - //if the user has not given a path then, add inputdir. else leave path alone. - if (path == "") { parameters["template"] = inputDir + it->second; } + //go through files and make sure they are good, if not, then disregard them + for (int i = 0; i < fastaFileNames.size(); i++) { + + bool ignore = false; + if (fastaFileNames[i] == "current") { + fastaFileNames[i] = m->getFastaFile(); + if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); } + else { + m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true; + //erase from file list + fastaFileNames.erase(fastaFileNames.begin()+i); + i--; + } + } + + if (!ignore) { + + + if (inputDir != "") { + 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]; } + } + + int ableToOpen; + ifstream in; + + 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() + m->getSimpleName(fastaFileNames[i]); + m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + fastaFileNames[i] = tryPath; + } + } + + //if you can't open it, try default location + if (ableToOpen == 1) { + if (m->getOutputDir() != "") { //default path is set + string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]); + m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + fastaFileNames[i] = tryPath; + } + } + + in.close(); + + if (ableToOpen == 1) { + m->mothurOut("Unable to open " + fastaFileNames[i] +". It will be disregarded."); m->mothurOutEndLine(); + //erase from file list + fastaFileNames.erase(fastaFileNames.begin()+i); + i--; + }else { + m->setFastaFile(fastaFileNames[i]); + } + } } - it = parameters.find("name"); - //user has given a template file - if(it != parameters.end()){ - path = hasPath(it->second); - //if the user has not given a path then, add inputdir. else leave path alone. - if (path == "") { parameters["name"] = inputDir + it->second; } - } + //make sure there is at least one valid file left + if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; } } - - - //check for required parameters - fastafile = validParameter.validFile(parameters, "fasta", true); - if (fastafile == "not open") { abort = true; } - else if (fastafile == "not found") { fastafile = ""; m->mothurOut("fasta is a required parameter for the chimera.check command."); m->mothurOutEndLine(); abort = true; } //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(fastafile); //if user entered a file with a path then preserve it + outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; } + + namefile = validParameter.validFile(parameters, "name", false); + if (namefile == "not found") { namefile = ""; } + else { + m->splitAtDash(namefile, nameFileNames); + + //go through files and make sure they are good, if not, then disregard them + for (int i = 0; i < nameFileNames.size(); i++) { + + bool ignore = false; + if (nameFileNames[i] == "current") { + nameFileNames[i] = m->getNameFile(); + if (nameFileNames[i] != "") { m->mothurOut("Using " + nameFileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); } + else { + m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true; + //erase from file list + nameFileNames.erase(nameFileNames.begin()+i); + i--; + } + } + + if (!ignore) { + + if (inputDir != "") { + string path = m->hasPath(nameFileNames[i]); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { nameFileNames[i] = inputDir + nameFileNames[i]; } + } + + int ableToOpen; + ifstream in; + + ableToOpen = m->openInputFile(nameFileNames[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() + m->getSimpleName(nameFileNames[i]); + m->mothurOut("Unable to open " + nameFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + nameFileNames[i] = tryPath; + } + } + + //if you can't open it, try default location + if (ableToOpen == 1) { + if (m->getOutputDir() != "") { //default path is set + string tryPath = m->getOutputDir() + m->getSimpleName(nameFileNames[i]); + m->mothurOut("Unable to open " + nameFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + nameFileNames[i] = tryPath; + } + } + + in.close(); + + if (ableToOpen == 1) { + m->mothurOut("Unable to open " + nameFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); + //erase from file list + nameFileNames.erase(nameFileNames.begin()+i); + i--; + }else { + m->setNameFile(nameFileNames[i]); + } + } + } + + //make sure there is at least one valid file left + if (nameFileNames.size() != 0) { + if (nameFileNames.size() != fastaFileNames.size()) { + m->mothurOut("Different number of valid name files and fasta files, aborting command."); m->mothurOutEndLine(); + abort = true; + } + } } - templatefile = validParameter.validFile(parameters, "template", true); - if (templatefile == "not open") { abort = true; } - else if (templatefile == "not found") { templatefile = ""; m->mothurOut("template is a required parameter for the chimera.check command."); m->mothurOutEndLine(); abort = true; } + string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); } + m->setProcessors(temp); + m->mothurConvert(temp, processors); + + temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; } + save = m->isTrue(temp); + rdb->save = save; + if (save) { //clear out old references + rdb->clearMemory(); + } + + //this has to go after save so that if the user sets save=t and provides no reference we abort + templatefile = validParameter.validFile(parameters, "reference", true); + if (templatefile == "not found") { + //check for saved reference sequences + if (rdb->referenceSeqs.size() != 0) { + templatefile = "saved"; + }else { + m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required."); + m->mothurOutEndLine(); + abort = true; + } + }else if (templatefile == "not open") { abort = true; } + else { if (save) { rdb->setSavedReference(templatefile); } } - namefile = validParameter.validFile(parameters, "name", true); - if (namefile == "not open") { abort = true; } - else if (namefile == "not found") { namefile = ""; } - - string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found") { temp = "1"; } - convert(temp, processors); temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found") { temp = "7"; } - convert(temp, ksize); + m->mothurConvert(temp, ksize); temp = validParameter.validFile(parameters, "svg", false); if (temp == "not found") { temp = "F"; } - svg = isTrue(temp); + svg = m->isTrue(temp); + if (nameFileNames.size() != 0) { svg = true; } temp = validParameter.validFile(parameters, "increment", false); if (temp == "not found") { temp = "10"; } - convert(temp, increment); + m->mothurConvert(temp, increment); } } catch(exception& e) { @@ -103,232 +336,167 @@ ChimeraCheckCommand::ChimeraCheckCommand(string option) { exit(1); } } -//********************************************************************************************************************** - -void ChimeraCheckCommand::help(){ - try { - - m->mothurOut("The chimera.check command reads a fastafile and templatefile and outputs potentially chimeric sequences.\n"); - m->mothurOut("This command was created using the algorythms described in CHIMERA_CHECK version 2.7 written by Niels Larsen. \n"); - m->mothurOut("The chimera.check command parameters are fasta, template, processors, ksize, increment, svg and name.\n"); - m->mothurOut("The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required. \n"); - m->mothurOut("The template parameter allows you to enter a template file containing known non-chimeric sequences, and is required. \n"); - m->mothurOut("The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n"); - #ifdef USE_MPI - m->mothurOut("When using MPI, the processors parameter is set to the number of MPI processes running. \n"); - #endif - m->mothurOut("The increment parameter allows you to specify how far you move each window while finding chimeric sequences, default is 10.\n"); - m->mothurOut("The ksize parameter allows you to input kmersize, default is 7. \n"); - m->mothurOut("The svg parameter allows you to specify whether or not you would like a svg file outputted for each query sequence, default is False.\n"); - m->mothurOut("The name parameter allows you to enter a file containing names of sequences you would like .svg files for.\n"); - m->mothurOut("The chimera.check command should be in the following format: \n"); - m->mothurOut("chimera.check(fasta=yourFastaFile, template=yourTemplateFile, processors=yourProcessors, ksize=yourKmerSize) \n"); - m->mothurOut("Example: chimera.check(fasta=AD.fasta, template=core_set_aligned,imputed.fasta, processors=4, ksize=8) \n"); - m->mothurOut("Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n\n"); - } - catch(exception& e) { - m->errorOut(e, "ChimeraCheckCommand", "help"); - exit(1); - } -} - -//*************************************************************************************************************** - -ChimeraCheckCommand::~ChimeraCheckCommand(){ /* do nothing */ } - //*************************************************************************************************************** int ChimeraCheckCommand::execute(){ try{ - if (abort == true) { return 0; } - - int start = time(NULL); - - chimera = new ChimeraCheckRDP(fastafile, templatefile, namefile, svg, increment, ksize, outputDir); - - if (m->control_pressed) { delete chimera; return 0; } - - string outputFileName = outputDir + getRootName(getSimpleName(fastafile)) + "chimeracheck.chimeras"; + if (abort == true) { if (calledHelp) { return 0; } return 2; } - #ifdef USE_MPI - - int pid, end, numSeqsPerProcessor; - int tag = 2001; - vector MPIPos; + for (int i = 0; i < fastaFileNames.size(); i++) { + + m->mothurOut("Checking sequences from " + fastaFileNames[i] + " ..." ); m->mothurOutEndLine(); - MPI_Status status; - MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are - MPI_Comm_size(MPI_COMM_WORLD, &processors); - - MPI_File inMPI; - MPI_File outMPI; - - int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY; - int inMode=MPI_MODE_RDONLY; + int start = time(NULL); - //char* outFilename = new char[outputFileName.length()]; - //memcpy(outFilename, outputFileName.c_str(), outputFileName.length()); + string thisNameFile = ""; + if (nameFileNames.size() != 0) { thisNameFile = nameFileNames[i]; } - char outFilename[1024]; - strcpy(outFilename, outputFileName.c_str()); + chimera = new ChimeraCheckRDP(fastaFileNames[i], templatefile, thisNameFile, svg, increment, ksize, outputDir); - //char* inFileName = new char[fastafile.length()]; - //memcpy(inFileName, fastafile.c_str(), fastafile.length()); + if (m->control_pressed) { delete chimera; return 0; } - char inFileName[1024]; - strcpy(inFileName, fastafile.c_str()); - - MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer - MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI); + if (outputDir == "") { outputDir = m->hasPath(fastaFileNames[i]); }//if user entered a file with a path then preserve it + map variables; + variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[i])); + string outputFileName = getOutputFileName("chimera", variables); + outputNames.push_back(outputFileName); outputTypes["chimera"].push_back(outputFileName); - //delete outFilename; - //delete inFileName; + #ifdef USE_MPI + + int pid, numSeqsPerProcessor; + int tag = 2001; + vector MPIPos; + + MPI_Status status; + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + MPI_Comm_size(MPI_COMM_WORLD, &processors); - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); delete chimera; return 0; } + MPI_File inMPI; + MPI_File outMPI; + + int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY; + int inMode=MPI_MODE_RDONLY; + + char outFilename[1024]; + strcpy(outFilename, outputFileName.c_str()); - if (pid == 0) { //you are the root process - MPIPos = setFilePosFasta(fastafile, numSeqs); //fills MPIPos, returns numSeqs + char inFileName[1024]; + strcpy(inFileName, fastaFileNames[i].c_str()); + + MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer + MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI); - //send file positions to all processes - MPI_Bcast(&numSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //send numSeqs - MPI_Bcast(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //send file pos + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } outputTypes.clear(); delete chimera; return 0; } - //figure out how many sequences you have to align - numSeqsPerProcessor = numSeqs / processors; - if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; } - int startIndex = pid * numSeqsPerProcessor; - - //align your part - driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, MPIPos); - - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); remove(outputFileName.c_str()); delete chimera; return 0; } + if (pid == 0) { //you are the root process + MPIPos = m->setFilePosFasta(fastaFileNames[i], numSeqs); //fills MPIPos, returns numSeqs + + //send file positions to all processes + for(int j = 1; j < processors; j++) { + MPI_Send(&numSeqs, 1, MPI_INT, j, tag, MPI_COMM_WORLD); + MPI_Send(&MPIPos[0], (numSeqs+1), MPI_LONG, j, tag, MPI_COMM_WORLD); + } + + //figure out how many sequences you have to align + numSeqsPerProcessor = numSeqs / processors; + int startIndex = pid * numSeqsPerProcessor; + if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; } + - //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); + //align your part + driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, MPIPos); + + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } outputTypes.clear(); delete chimera; return 0; } + + //wait on chidren + for(int j = 1; j < processors; j++) { + char buf[5]; + MPI_Recv(buf, 5, MPI_CHAR, j, tag, MPI_COMM_WORLD, &status); + } + }else{ //you are a child process + MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status); + MPIPos.resize(numSeqs+1); + MPI_Recv(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status); + + //figure out how many sequences you have to align + numSeqsPerProcessor = numSeqs / processors; + int startIndex = pid * numSeqsPerProcessor; + if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; } + + //align your part + driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, MPIPos); + + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } outputTypes.clear(); delete chimera; return 0; } + + //tell parent you are done. + char buf[5]; + strcpy(buf, "done"); + MPI_Send(buf, 5, MPI_CHAR, 0, tag, MPI_COMM_WORLD); } - }else{ //you are a child process - MPI_Bcast(&numSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //get numSeqs - MPIPos.resize(numSeqs+1); - MPI_Bcast(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //get file positions - - //figure out how many sequences you have to align - numSeqsPerProcessor = numSeqs / processors; - if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; } - int startIndex = pid * numSeqsPerProcessor; - - //align your part - driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, MPIPos); - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); delete chimera; return 0; } - - //tell parent you are done. - char buf[4]; - strcpy(buf, "done"); - MPI_Send(buf, 4, MPI_CHAR, 0, tag, MPI_COMM_WORLD); - } + //close files + MPI_File_close(&inMPI); + MPI_File_close(&outMPI); + MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case + #else - //close files - MPI_File_close(&inMPI); - MPI_File_close(&outMPI); - #else - - //break up file - #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) - if(processors == 1){ - ifstream inFASTA; - openInputFile(fastafile, inFASTA); - numSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); - inFASTA.close(); - - lines.push_back(new linePair(0, numSeqs)); - - driver(lines[0], outputFileName, fastafile); - - if (m->control_pressed) { - remove(outputFileName.c_str()); - for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); - delete chimera; - return 0; - } - - }else{ - vector positions; - processIDS.resize(0); - - ifstream inFASTA; - openInputFile(fastafile, inFASTA); + + + //break up file + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) + vector positions = m->divideFile(fastaFileNames[i], processors); + + for (int s = 0; s < (positions.size()-1); s++) { + lines.push_back(new linePair(positions[s], positions[(s+1)])); + } + + if(processors == 1){ + numSeqs = driver(lines[0], outputFileName, fastaFileNames[i]); + + if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } for (int j = 0; j < lines.size(); j++) { delete lines[j]; } outputTypes.clear(); lines.clear(); delete chimera; return 0; } + + }else{ + processIDS.resize(0); + + numSeqs = createProcesses(outputFileName, fastaFileNames[i]); - 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); } + rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str()); + + //append output files + for(int j=1;jappendFiles((outputFileName + toString(processIDS[j]) + ".temp"), outputFileName); + m->mothurRemove((outputFileName + toString(processIDS[j]) + ".temp")); } - } - 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; + + if (m->control_pressed) { + for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } outputTypes.clear(); + for (int j = 0; j < lines.size(); j++) { delete lines[j]; } lines.clear(); + delete chimera; + return 0; } - lines.push_back(new linePair(startPos, numSeqsPerProcessor)); } + + #else + lines.push_back(new linePair(0, 1000)); + numSeqs = driver(lines[0], outputFileName, fastaFileNames[i]); - - createProcesses(outputFileName, fastafile); + if (m->control_pressed) { for (int j = 0; j < lines.size(); j++) { delete lines[j]; } lines.clear(); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } outputTypes.clear(); delete chimera; return 0; } + #endif + #endif + delete chimera; + for (int j = 0; j < lines.size(); j++) { delete lines[j]; } lines.clear(); - rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str()); - - //append output files - for(int i=1;icontrol_pressed) { - remove(outputFileName.c_str()); - for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); - delete chimera; - return 0; - } - } + m->mothurOutEndLine(); m->mothurOut("This method does not determine if a sequence is chimeric, but allows you to make that determination based on the IS values."); m->mothurOutEndLine(); + m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine(); - #else - ifstream inFASTA; - openInputFile(fastafile, inFASTA); - numSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); - inFASTA.close(); - lines.push_back(new linePair(0, numSeqs)); - - driver(lines[0], outputFileName, fastafile); - - if (m->control_pressed) { - remove(outputFileName.c_str()); - for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); - delete chimera; - return 0; - } - #endif - #endif - delete chimera; - for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); - - m->mothurOutEndLine(); m->mothurOut("This method does not determine if a sequence is chimeric, but allows you to make that determination based on the IS values."); m->mothurOutEndLine(); + } m->mothurOutEndLine(); m->mothurOut("Output File Names: "); m->mothurOutEndLine(); - m->mothurOut(outputFileName); m->mothurOutEndLine(); + for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } m->mothurOutEndLine(); - m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine(); - + return 0; } @@ -339,23 +507,26 @@ int ChimeraCheckCommand::execute(){ } //********************************************************************************************************************** -int ChimeraCheckCommand::driver(linePair* line, string outputFName, string filename){ +int ChimeraCheckCommand::driver(linePair* filePos, string outputFName, string filename){ try { ofstream out; - openOutputFile(outputFName, out); + m->openOutputFile(outputFName, out); ofstream out2; ifstream inFASTA; - openInputFile(filename, inFASTA); + m->openInputFile(filename, inFASTA); + + inFASTA.seekg(filePos->start); + + bool done = false; + int count = 0; + + while (!done) { - inFASTA.seekg(line->start); - - for(int i=0;inumSeqs;i++){ - if (m->control_pressed) { return 1; } - Sequence* candidateSeq = new Sequence(inFASTA); gobble(inFASTA); + Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA); if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file //find chimeras @@ -368,16 +539,23 @@ int ChimeraCheckCommand::driver(linePair* line, string outputFName, string filen } delete candidateSeq; + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) + unsigned long long pos = inFASTA.tellg(); + if ((pos == -1) || (pos >= filePos->end)) { break; } + #else + if (inFASTA.eof()) { break; } + #endif + //report progress - if((i+1) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(i+1)); m->mothurOutEndLine(); } + if((count) % 100 == 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count) + "\n"); } } //report progress - if((line->numSeqs) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(line->numSeqs)); m->mothurOutEndLine(); } + if((count) % 100 != 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count) + "\n"); } out.close(); inFASTA.close(); - return 0; + return count; } catch(exception& e) { m->errorOut(e, "ChimeraCheckCommand", "driver"); @@ -386,7 +564,7 @@ int ChimeraCheckCommand::driver(linePair* line, string outputFName, string filen } //********************************************************************************************************************** #ifdef USE_MPI -int ChimeraCheckCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, vector& MPIPos){ +int ChimeraCheckCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, vector& MPIPos){ try { MPI_File outAccMPI; MPI_Status status; @@ -408,7 +586,7 @@ int ChimeraCheckCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File istringstream iss (tempBuf,istringstream::in); delete buf4; - Sequence* candidateSeq = new Sequence(iss); gobble(iss); + Sequence* candidateSeq = new Sequence(iss); m->gobble(iss); if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file //find chimeras @@ -420,10 +598,10 @@ int ChimeraCheckCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File delete candidateSeq; //report progress - if((i+1) % 100 == 0){ cout << "Processing sequence: " << (i+1) << endl; m->mothurOutJustToLog("Processing sequence: " + toString(i+1) + "\n"); } + if((i+1) % 100 == 0){ cout << "Processing sequence: " << (i+1) << endl; } } //report progress - if(num % 100 != 0){ cout << "Processing sequence: " << num << endl; m->mothurOutJustToLog("Processing sequence: " + toString(num) + "\n"); } + if(num % 100 != 0){ cout << "Processing sequence: " << num << endl; } return 0; } @@ -438,9 +616,9 @@ int ChimeraCheckCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File int ChimeraCheckCommand::createProcesses(string outputFileName, string filename) { try { -#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) int process = 0; - // processIDS.resize(0); + int num = 0; //loop through and create all the processes you want while (process != processors) { @@ -450,9 +628,21 @@ int ChimeraCheckCommand::createProcesses(string outputFileName, 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){ - driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename); + num = driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename); + + //pass numSeqs to parent + ofstream out; + string tempFile = outputFileName + toString(getpid()) + ".num.temp"; + m->openOutputFile(tempFile, out); + out << num << endl; + out.close(); + exit(0); - }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); } + }else { + m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); + for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); } + exit(0); + } } //force parent to wait until all the processes are done @@ -461,7 +651,15 @@ int ChimeraCheckCommand::createProcesses(string outputFileName, string filename) wait(&temp); } - return 0; + for (int i = 0; i < processIDS.size(); i++) { + ifstream in; + string tempFile = outputFileName + toString(processIDS[i]) + ".num.temp"; + m->openInputFile(tempFile, in); + if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; } + in.close(); m->mothurRemove(tempFile); + } + + return num; #endif } catch(exception& e) {