*/
#include "screenseqscommand.h"
-#include "sequence.hpp"
+#include "counttable.h"
//**********************************************************************************************************************
vector<string> ScreenSeqsCommand::setParameters(){
try {
- CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta);
- CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname);
- CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup);
- CommandParameter pqfile("qfile", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pqfile);
- CommandParameter palignreport("alignreport", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(palignreport);
- CommandParameter ptax("taxonomy", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(ptax);
- CommandParameter pstart("start", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pstart);
- CommandParameter pend("end", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pend);
- CommandParameter pmaxambig("maxambig", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pmaxambig);
- CommandParameter pmaxhomop("maxhomop", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pmaxhomop);
- CommandParameter pminlength("minlength", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pminlength);
- CommandParameter pmaxlength("maxlength", "Number", "", "-1", "", "", "",false,false); parameters.push_back(pmaxlength);
- CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
- CommandParameter pcriteria("criteria", "Number", "", "90", "", "", "",false,false); parameters.push_back(pcriteria);
- CommandParameter poptimize("optimize", "Multiple", "none-start-end-maxambig-maxhomop-minlength-maxlength", "none", "", "", "",true,false); parameters.push_back(poptimize);
- CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
- CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
+ CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none","fasta",false,true,true); parameters.push_back(pfasta);
+ CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none","name",false,false,true); parameters.push_back(pname);
+ CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none","count",false,false,true); parameters.push_back(pcount);
+ CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none","group",false,false,true); parameters.push_back(pgroup);
+ CommandParameter pqfile("qfile", "InputTypes", "", "", "none", "none", "none","qfile",false,false); parameters.push_back(pqfile);
+ CommandParameter palignreport("alignreport", "InputTypes", "", "", "none", "none", "none","alignreport",false,false); parameters.push_back(palignreport);
+ CommandParameter ptax("taxonomy", "InputTypes", "", "", "none", "none", "none","taxonomy",false,false); parameters.push_back(ptax);
+ CommandParameter pstart("start", "Number", "", "-1", "", "", "","",false,false,true); parameters.push_back(pstart);
+ CommandParameter pend("end", "Number", "", "-1", "", "", "","",false,false,true); parameters.push_back(pend);
+ CommandParameter pmaxambig("maxambig", "Number", "", "-1", "", "", "","",false,false); parameters.push_back(pmaxambig);
+ CommandParameter pmaxhomop("maxhomop", "Number", "", "-1", "", "", "","",false,false); parameters.push_back(pmaxhomop);
+ CommandParameter pminlength("minlength", "Number", "", "-1", "", "", "","",false,false); parameters.push_back(pminlength);
+ CommandParameter pmaxlength("maxlength", "Number", "", "-1", "", "", "","",false,false); parameters.push_back(pmaxlength);
+ CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors);
+ CommandParameter pcriteria("criteria", "Number", "", "90", "", "", "","",false,false); parameters.push_back(pcriteria);
+ CommandParameter poptimize("optimize", "Multiple", "none-start-end-maxambig-maxhomop-minlength-maxlength", "none", "", "", "","",true,false); parameters.push_back(poptimize);
+ CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
+ CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir);
vector<string> myArray;
for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
string ScreenSeqsCommand::getHelpString(){
try {
string helpString = "";
- helpString += "The screen.seqs command reads a fastafile and creates .....\n";
- helpString += "The screen.seqs command parameters are fasta, start, end, maxambig, maxhomop, minlength, maxlength, name, group, qfile, alignreport, taxonomy, optimize, criteria and processors.\n";
+ helpString += "The screen.seqs command reads a fastafile and screens sequences.\n";
+ helpString += "The screen.seqs command parameters are fasta, start, end, maxambig, maxhomop, minlength, maxlength, name, group, count, qfile, alignreport, taxonomy, optimize, criteria and processors.\n";
helpString += "The fasta parameter is required.\n";
helpString += "The alignreport and taxonomy parameters allow you to remove bad seqs from taxonomy and alignreport files.\n";
- helpString += "The start parameter .... The default is -1.\n";
- helpString += "The end parameter .... The default is -1.\n";
+ helpString += "The start parameter is used to set a position the \"good\" sequences must start by. The default is -1.\n";
+ helpString += "The end parameter is used to set a position the \"good\" sequences must end after. The default is -1.\n";
helpString += "The maxambig parameter allows you to set the maximum number of ambigious bases allowed. The default is -1.\n";
helpString += "The maxhomop parameter allows you to set a maximum homopolymer length. \n";
helpString += "The minlength parameter allows you to set and minimum sequence length. \n";
}
}
//**********************************************************************************************************************
+string ScreenSeqsCommand::getOutputPattern(string type) {
+ try {
+ string pattern = "";
+
+ if (type == "fasta") { pattern = "[filename],good,[extension]"; }
+ else if (type == "taxonomy") { pattern = "[filename],good,[extension]"; }
+ else if (type == "name") { pattern = "[filename],good,[extension]"; }
+ else if (type == "group") { pattern = "[filename],good,[extension]"; }
+ else if (type == "count") { pattern = "[filename],good,[extension]"; }
+ else if (type == "accnos") { pattern = "[filename],bad.accnos"; }
+ else if (type == "qfile") { pattern = "[filename],good,[extension]"; }
+ else if (type == "alignreport") { pattern = "[filename],good.align.report"; }
+ else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
+
+ return pattern;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ScreenSeqsCommand", "getOutputPattern");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
ScreenSeqsCommand::ScreenSeqsCommand(){
try {
abort = true; calledHelp = true;
outputTypes["accnos"] = tempOutNames;
outputTypes["qfile"] = tempOutNames;
outputTypes["taxonomy"] = tempOutNames;
+ outputTypes["count"] = tempOutNames;
}
catch(exception& e) {
m->errorOut(e, "ScreenSeqsCommand", "ScreenSeqsCommand");
outputTypes["accnos"] = tempOutNames;
outputTypes["qfile"] = tempOutNames;
outputTypes["taxonomy"] = tempOutNames;
+ outputTypes["count"] = 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 the user has not given a path then, add inputdir. else leave path alone.
if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
}
+
+ it = parameters.find("count");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = m->hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["count"] = inputDir + it->second; }
+ }
}
//check for required parameters
else if (namefile == "not found") { namefile = ""; }
else { m->setNameFile(namefile); }
+ countfile = validParameter.validFile(parameters, "count", true);
+ if (countfile == "not open") { countfile = ""; abort = true; }
+ else if (countfile == "not found") { countfile = ""; }
+ else { m->setCountTableFile(countfile); }
+
+ if ((namefile != "") && (countfile != "")) {
+ m->mothurOut("[ERROR]: you may only use one of the following: name or count."); m->mothurOutEndLine(); abort = true;
+ }
+
+ if ((groupfile != "") && (countfile != "")) {
+ m->mothurOut("[ERROR]: you may only use one of the following: group or count."); m->mothurOutEndLine(); abort=true;
+ }
+
alignreport = validParameter.validFile(parameters, "alignreport", true);
if (alignreport == "not open") { abort = true; }
else if (alignreport == "not found") { alignreport = ""; }
// ...at some point should added some additional type checking...
string temp;
temp = validParameter.validFile(parameters, "start", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, startPos);
+ m->mothurConvert(temp, startPos);
temp = validParameter.validFile(parameters, "end", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, endPos);
+ m->mothurConvert(temp, endPos);
temp = validParameter.validFile(parameters, "maxambig", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, maxAmbig);
+ m->mothurConvert(temp, maxAmbig);
temp = validParameter.validFile(parameters, "maxhomop", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, maxHomoP);
+ m->mothurConvert(temp, maxHomoP);
temp = validParameter.validFile(parameters, "minlength", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, minLength);
+ m->mothurConvert(temp, minLength);
temp = validParameter.validFile(parameters, "maxlength", false); if (temp == "not found") { temp = "-1"; }
- convert(temp, maxLength);
+ m->mothurConvert(temp, maxLength);
temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
m->setProcessors(temp);
- convert(temp, processors);
+ m->mothurConvert(temp, processors);
temp = validParameter.validFile(parameters, "optimize", false); //optimizing trumps the optimized values original value
if (temp == "not found"){ temp = "none"; }
if (optimize.size() == 1) { if (optimize[0] == "none") { optimize.clear(); } }
temp = validParameter.validFile(parameters, "criteria", false); if (temp == "not found"){ temp = "90"; }
- convert(temp, criteria);
+ m->mothurConvert(temp, criteria);
+
+ if (countfile == "") {
+ if (namefile == "") {
+ vector<string> files; files.push_back(fastafile);
+ parser.getNameFile(files);
+ }
+ }
}
}
if (optimize.size() != 0) { //get summary is paralellized so we need to divideFile, no need to do this step twice so I moved it here
//use the namefile to optimize correctly
if (namefile != "") { nameMap = m->readNames(namefile); }
+ else if (countfile != "") {
+ CountTable ct;
+ ct.readTable(countfile);
+ nameMap = ct.getNameMap();
+ }
getSummary(positions);
}
else {
- positions = m->divideFile(fastafile, processors);
- for (int i = 0; i < (positions.size()-1); i++) {
- lines.push_back(new linePair(positions[i], positions[(i+1)]));
- }
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
+ positions = m->divideFile(fastafile, processors);
+ for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(linePair(positions[i], positions[(i+1)])); }
+ #else
+ if(processors == 1){ lines.push_back(linePair(0, 1000)); }
+ else {
+ int numFastaSeqs = 0;
+ positions = m->setFilePosFasta(fastafile, numFastaSeqs);
+ if (positions.size() < processors) { processors = positions.size(); }
+
+ //figure out how many sequences you have to process
+ int numSeqsPerProcessor = numFastaSeqs / processors;
+ for (int i = 0; i < processors; i++) {
+ int startIndex = i * numSeqsPerProcessor;
+ if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
+ lines.push_back(linePair(positions[startIndex], numSeqsPerProcessor));
+ }
+ }
+ #endif
}
-
- string goodSeqFile = outputDir + m->getRootName(m->getSimpleName(fastafile)) + "good" + m->getExtension(fastafile);
- string badAccnosFile = outputDir + m->getRootName(m->getSimpleName(fastafile)) + "bad.accnos";
+
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastafile));
+ string badAccnosFile = getOutputFileName("accnos",variables);
+ variables["[extension]"] = m->getExtension(fastafile);
+ string goodSeqFile = getOutputFileName("fasta", variables);
+
int numFastaSeqs = 0;
set<string> badSeqNames;
numSeqsPerProcessor = numFastaSeqs / processors;
int startIndex = pid * numSeqsPerProcessor;
if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
- // cout << pid << '\t' << numSeqsPerProcessor << '\t' << startIndex << endl;
+
//align your part
driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBadAccnos, MPIPos, badSeqNames);
- //cout << pid << " done" << endl;
+
if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); return 0; }
for (int i = 1; i < processors; i++) {
-
//get bad lists
int badSize;
MPI_Recv(&badSize, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
- /*for (int j = 0; j < badSize; j++) {
- int length;
- MPI_Recv(&length, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status); //recv the length of the name
- char* buf2 = new char[length]; //make space to recieve it
- MPI_Recv(buf2, length, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status); //get name
-
- string tempBuf = buf2;
- if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
- delete buf2;
-
- badSeqNames.insert(tempBuf);
- }*/
}
}else{ //you are a child process
MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
numSeqsPerProcessor = numFastaSeqs / processors;
int startIndex = pid * numSeqsPerProcessor;
if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
- //cout << pid << '\t' << numSeqsPerProcessor << '\t' << startIndex << endl;
+
//align your part
driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBadAccnos, MPIPos, badSeqNames);
-//cout << pid << " done" << endl;
+
if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); return 0; }
//send bad list
int badSize = badSeqNames.size();
MPI_Send(&badSize, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
-
- /*
- set<string>::iterator it;
- for (it = badSeqNames.begin(); it != badSeqNames.end(); it++) {
- string name = *it;
- int length = name.length();
- char* buf2 = new char[length];
- memcpy(buf2, name.c_str(), length);
-
- MPI_Send(&length, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
- MPI_Send(buf2, length, MPI_CHAR, 0, tag, MPI_COMM_WORLD);
- }*/
}
//close files
MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
#else
-
- #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- if(processors == 1){
- numFastaSeqs = driver(lines[0], goodSeqFile, badAccnosFile, fastafile, badSeqNames);
-
- if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
-
- }else{
- processIDS.resize(0);
-
- numFastaSeqs = createProcesses(goodSeqFile, badAccnosFile, fastafile, badSeqNames);
-
- rename((goodSeqFile + toString(processIDS[0]) + ".temp").c_str(), goodSeqFile.c_str());
- rename((badAccnosFile + toString(processIDS[0]) + ".temp").c_str(), badAccnosFile.c_str());
-
- //append alignment and report files
- for(int i=1;i<processors;i++){
- m->appendFiles((goodSeqFile + toString(processIDS[i]) + ".temp"), goodSeqFile);
- m->mothurRemove((goodSeqFile + toString(processIDS[i]) + ".temp"));
-
- m->appendFiles((badAccnosFile + toString(processIDS[i]) + ".temp"), badAccnosFile);
- m->mothurRemove((badAccnosFile + toString(processIDS[i]) + ".temp"));
- }
-
- if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
-
- //read badSeqs in because root process doesnt know what other "bad" seqs the children found
- ifstream inBad;
- int ableToOpen = m->openInputFile(badAccnosFile, inBad, "no error");
-
- if (ableToOpen == 0) {
- badSeqNames.clear();
- string tempName;
- while (!inBad.eof()) {
- inBad >> tempName; m->gobble(inBad);
- badSeqNames.insert(tempName);
- }
- inBad.close();
- }
- }
- #else
- numFastaSeqs = driver(lines[0], goodSeqFile, badAccnosFile, fastafile, badSeqNames);
-
- if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
-
- #endif
-
+ if(processors == 1){ numFastaSeqs = driver(lines[0], goodSeqFile, badAccnosFile, fastafile, badSeqNames); }
+ else{ numFastaSeqs = createProcesses(goodSeqFile, badAccnosFile, fastafile, badSeqNames); }
+
+ if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
#endif
#ifdef USE_MPI
screenNameGroupFile(badSeqNames);
if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
}else if(groupfile != "") { screenGroupFile(badSeqNames); } // this screens just the group
-
+ else if (countfile != "") { screenCountFile(badSeqNames); }
+
+
if (m->control_pressed) { m->mothurRemove(goodSeqFile); return 0; }
if(alignreport != "") { screenAlignReport(badSeqNames); }
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
}
+
+ itTypes = outputTypes.find("count");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setCountTableFile(current); }
+ }
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to screen " + toString(numFastaSeqs) + " sequences.");
m->mothurOutEndLine();
set<string> badSeqGroups;
string seqName, seqList, group;
set<string>::iterator it;
-
- string goodNameFile = outputDir + m->getRootName(m->getSimpleName(namefile)) + "good" + m->getExtension(namefile);
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(namefile));
+ variables["[extension]"] = m->getExtension(namefile);
+ string goodNameFile = getOutputFileName("name", variables);
outputNames.push_back(goodNameFile); outputTypes["name"].push_back(goodNameFile);
ofstream goodNameOut; m->openOutputFile(goodNameFile, goodNameOut);
ifstream inputGroups;
m->openInputFile(groupfile, inputGroups);
-
- string goodGroupFile = outputDir + m->getRootName(m->getSimpleName(groupfile)) + "good" + m->getExtension(groupfile);
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(groupfile));
+ variables["[extension]"] = m->getExtension(groupfile);
+ string goodGroupFile = getOutputFileName("group", variables);
+
outputNames.push_back(goodGroupFile); outputTypes["group"].push_back(goodGroupFile);
ofstream goodGroupOut; m->openOutputFile(goodGroupFile, goodGroupOut);
vector<int> ambigBases;
vector<int> longHomoPolymer;
- vector<unsigned long long> positions = m->divideFile(fastafile, processors);
-
- for (int i = 0; i < (positions.size()-1); i++) {
- lines.push_back(new linePair(positions[i], positions[(i+1)]));
- }
-
+ vector<unsigned long long> positions;
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
+ positions = m->divideFile(fastafile, processors);
+ for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(linePair(positions[i], positions[(i+1)])); }
+#else
+ if(processors == 1){ lines.push_back(linePair(0, 1000)); }
+ else {
+ int numFastaSeqs = 0;
+ positions = m->setFilePosFasta(fastafile, numFastaSeqs);
+ if (positions.size() < processors) { processors = positions.size(); }
+
+ //figure out how many sequences you have to process
+ int numSeqsPerProcessor = numFastaSeqs / processors;
+ for (int i = 0; i < processors; i++) {
+ int startIndex = i * numSeqsPerProcessor;
+ if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
+ lines.push_back(linePair(positions[startIndex], numSeqsPerProcessor));
+ }
+ }
+#endif
#ifdef USE_MPI
int pid;
driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]);
#else
int numSeqs = 0;
- #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ //#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
if(processors == 1){
numSeqs = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]);
}else{
}
if (m->control_pressed) { return 0; }
- #else
- numSeqs = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]);
- if (m->control_pressed) { return 0; }
- #endif
+ //#else
+ // numSeqs = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]);
+ // if (m->control_pressed) { return 0; }
+ //#endif
#endif
sort(startPosition.begin(), startPosition.end());
sort(endPosition.begin(), endPosition.end());
}
}
/**************************************************************************************/
-int ScreenSeqsCommand::driverCreateSummary(vector<int>& startPosition, vector<int>& endPosition, vector<int>& seqLength, vector<int>& ambigBases, vector<int>& longHomoPolymer, string filename, linePair* filePos) {
+int ScreenSeqsCommand::driverCreateSummary(vector<int>& startPosition, vector<int>& endPosition, vector<int>& seqLength, vector<int>& ambigBases, vector<int>& longHomoPolymer, string filename, linePair filePos) {
try {
ifstream in;
m->openInputFile(filename, in);
- in.seekg(filePos->start);
+ in.seekg(filePos.start);
bool done = false;
int count = 0;
count++;
}
-
- #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ //if((count) % 100 == 0){ m->mothurOut("Optimizing sequence: " + toString(count)); m->mothurOutEndLine(); }
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
unsigned long long pos = in.tellg();
- if ((pos == -1) || (pos >= filePos->end)) { break; }
+ if ((pos == -1) || (pos >= filePos.end)) { break; }
#else
if (in.eof()) { break; }
#endif
/**************************************************************************************************/
int ScreenSeqsCommand::createProcessesCreateSummary(vector<int>& startPosition, vector<int>& endPosition, vector<int>& seqLength, vector<int>& ambigBases, vector<int>& longHomoPolymer, string filename) {
try {
-#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- int process = 1;
+
+ int process = 1;
int num = 0;
- processIDS.clear();
-
+ vector<int> processIDS;
+
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
+
//loop through and create all the processes you want
while (process != processors) {
int pid = fork();
m->mothurRemove(tempFilename);
}
- return num;
+
+#else
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+ //Windows version shared memory, so be careful when passing variables through the seqSumData struct.
+ //Above fork() will clone, so memory is separate, but that's not the case with windows,
+ //Taking advantage of shared memory to allow both threads to add info to vectors.
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ vector<sumData*> pDataArray;
+ DWORD dwThreadIdArray[processors-1];
+ HANDLE hThreadArray[processors-1];
+
+ //Create processor worker threads.
+ for( int i=0; i<processors-1; i++ ){
+
+ // Allocate memory for thread data.
+ sumData* tempSum = new sumData(filename, m, lines[i].start, lines[i].end, namefile, nameMap);
+ pDataArray.push_back(tempSum);
+
+ //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
+ //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
+ hThreadArray[i] = CreateThread(NULL, 0, MySumThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
+ }
+
+ //do your part
+ num = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[processors-1]);
+
+ //Wait until all threads have terminated.
+ WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
+
+ //Close all thread handles and free memory allocations.
+ for(int i=0; i < pDataArray.size(); i++){
+ num += pDataArray[i]->count;
+ for (int k = 0; k < pDataArray[i]->startPosition.size(); k++) { startPosition.push_back(pDataArray[i]->startPosition[k]); }
+ for (int k = 0; k < pDataArray[i]->endPosition.size(); k++) { endPosition.push_back(pDataArray[i]->endPosition[k]); }
+ for (int k = 0; k < pDataArray[i]->seqLength.size(); k++) { seqLength.push_back(pDataArray[i]->seqLength[k]); }
+ for (int k = 0; k < pDataArray[i]->ambigBases.size(); k++) { ambigBases.push_back(pDataArray[i]->ambigBases[k]); }
+ for (int k = 0; k < pDataArray[i]->longHomoPolymer.size(); k++) { longHomoPolymer.push_back(pDataArray[i]->longHomoPolymer[k]); }
+ CloseHandle(hThreadArray[i]);
+ delete pDataArray[i];
+ }
+
#endif
+ return num;
}
catch(exception& e) {
m->errorOut(e, "ScreenSeqsCommand", "createProcessesCreateSummary");
m->openInputFile(groupfile, inputGroups);
string seqName, group;
set<string>::iterator it;
-
- string goodGroupFile = outputDir + m->getRootName(m->getSimpleName(groupfile)) + "good" + m->getExtension(groupfile);
- outputNames.push_back(goodGroupFile); outputTypes["group"].push_back(goodGroupFile);
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(groupfile));
+ variables["[extension]"] = m->getExtension(groupfile);
+ string goodGroupFile = getOutputFileName("group", variables);
+ outputNames.push_back(goodGroupFile); outputTypes["group"].push_back(goodGroupFile);
ofstream goodGroupOut; m->openOutputFile(goodGroupFile, goodGroupOut);
while(!inputGroups.eof()){
exit(1);
}
}
+//***************************************************************************************************************
+int ScreenSeqsCommand::screenCountFile(set<string> badSeqNames){
+ try {
+ ifstream in;
+ m->openInputFile(countfile, in);
+ set<string>::iterator it;
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(countfile));
+ variables["[extension]"] = m->getExtension(countfile);
+ string goodCountFile = getOutputFileName("count", variables);
+
+ outputNames.push_back(goodCountFile); outputTypes["count"].push_back(goodCountFile);
+ ofstream goodCountOut; m->openOutputFile(goodCountFile, goodCountOut);
+
+ string headers = m->getline(in); m->gobble(in);
+ goodCountOut << headers << endl;
+
+ string name, rest; int thisTotal;
+ while (!in.eof()) {
+ if (m->control_pressed) { goodCountOut.close(); in.close(); m->mothurRemove(goodCountFile); return 0; }
+
+ in >> name; m->gobble(in);
+ in >> thisTotal; m->gobble(in);
+ rest = m->getline(in); m->gobble(in);
+
+ it = badSeqNames.find(name);
+
+ if(it != badSeqNames.end()){
+ badSeqNames.erase(it);
+ }
+ else{
+ goodCountOut << name << '\t' << thisTotal << '\t' << rest << endl;
+ }
+ }
+
+ if (m->control_pressed) { goodCountOut.close(); in.close(); m->mothurRemove(goodCountFile); return 0; }
+
+ //we were unable to remove some of the bad sequences
+ if (badSeqNames.size() != 0) {
+ for (it = badSeqNames.begin(); it != badSeqNames.end(); it++) {
+ m->mothurOut("Your count file does not include the sequence " + *it + " please correct.");
+ m->mothurOutEndLine();
+ }
+ }
+
+ in.close();
+ goodCountOut.close();
+
+ //check for groups that have been eliminated
+ CountTable ct;
+ if (ct.testGroups(goodCountFile)) {
+ ct.readTable(goodCountFile);
+ ct.printTable(goodCountFile);
+ }
+
+ if (m->control_pressed) { m->mothurRemove(goodCountFile); }
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ScreenSeqsCommand", "screenCountFile");
+ exit(1);
+ }
+}
//***************************************************************************************************************
int ScreenSeqsCommand::screenAlignReport(set<string> badSeqNames){
string seqName, group;
set<string>::iterator it;
- string goodAlignReportFile = outputDir + m->getRootName(m->getSimpleName(alignreport)) + "good" + m->getExtension(alignreport);
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(alignreport));
+ string goodAlignReportFile = getOutputFileName("alignreport", variables);
+
outputNames.push_back(goodAlignReportFile); outputTypes["alignreport"].push_back(goodAlignReportFile);
ofstream goodAlignReportOut; m->openOutputFile(goodAlignReportFile, goodAlignReportOut);
m->openInputFile(taxonomy, input);
string seqName, tax;
set<string>::iterator it;
-
- string goodTaxFile = outputDir + m->getRootName(m->getSimpleName(taxonomy)) + "good" + m->getExtension(taxonomy);
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(taxonomy));
+ variables["[extension]"] = m->getExtension(taxonomy);
+ string goodTaxFile = getOutputFileName("taxonomy", variables);
+
outputNames.push_back(goodTaxFile); outputTypes["taxonomy"].push_back(goodTaxFile);
ofstream goodTaxOut; m->openOutputFile(goodTaxFile, goodTaxOut);
ifstream in;
m->openInputFile(qualfile, in);
set<string>::iterator it;
+ map<string, string> variables;
+ variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(qualfile));
+ variables["[extension]"] = m->getExtension(qualfile);
+ string goodQualFile = getOutputFileName("qfile", variables);
- string goodQualFile = outputDir + m->getRootName(m->getSimpleName(qualfile)) + "good" + m->getExtension(qualfile);
outputNames.push_back(goodQualFile); outputTypes["qfile"].push_back(goodQualFile);
ofstream goodQual; m->openOutputFile(goodQualFile, goodQual);
}
//**********************************************************************************************************************
-int ScreenSeqsCommand::driver(linePair* filePos, string goodFName, string badAccnosFName, string filename, set<string>& badSeqNames){
+int ScreenSeqsCommand::driver(linePair filePos, string goodFName, string badAccnosFName, string filename, set<string>& badSeqNames){
try {
ofstream goodFile;
m->openOutputFile(goodFName, goodFile);
ifstream inFASTA;
m->openInputFile(filename, inFASTA);
- inFASTA.seekg(filePos->start);
+ inFASTA.seekg(filePos.start);
bool done = false;
int count = 0;
count++;
}
- #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
unsigned long long pos = inFASTA.tellg();
- if ((pos == -1) || (pos >= filePos->end)) { break; }
+ if ((pos == -1) || (pos >= filePos.end)) { break; }
#else
if (inFASTA.eof()) { break; }
#endif
int length = MPIPos[start+i+1] - MPIPos[start+i];
char* buf4 = new char[length];
- memcpy(buf4, outputString.c_str(), length);
MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
int ScreenSeqsCommand::createProcesses(string goodFileName, string badAccnos, string filename, set<string>& badSeqNames) {
try {
-#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- int process = 0;
+
+ vector<int> processIDS;
+ int process = 1;
int num = 0;
-
+
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
+
//loop through and create all the processes you want
while (process != processors) {
int pid = fork();
}
}
+ num = driver(lines[0], goodFileName, badAccnos, filename, badSeqNames);
+
//force parent to wait until all the processes are done
- for (int i=0;i<processors;i++) {
+ for (int i=0;i<processIDS.size();i++) {
int temp = processIDS[i];
wait(&temp);
}
m->openInputFile(tempFile, in);
if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
in.close(); m->mothurRemove(tempFile);
+
+ m->appendFiles((goodFileName + toString(processIDS[i]) + ".temp"), goodFileName);
+ m->mothurRemove((goodFileName + toString(processIDS[i]) + ".temp"));
+
+ m->appendFiles((badAccnos + toString(processIDS[i]) + ".temp"), badAccnos);
+ m->mothurRemove((badAccnos + toString(processIDS[i]) + ".temp"));
}
- return num;
-#endif
+ //read badSeqs in because root process doesnt know what other "bad" seqs the children found
+ ifstream inBad;
+ int ableToOpen = m->openInputFile(badAccnos, inBad, "no error");
+
+ if (ableToOpen == 0) {
+ badSeqNames.clear();
+ string tempName;
+ while (!inBad.eof()) {
+ inBad >> tempName; m->gobble(inBad);
+ badSeqNames.insert(tempName);
+ }
+ inBad.close();
+ }
+#else
+
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+ //Windows version shared memory, so be careful when passing variables through the sumScreenData struct.
+ //Above fork() will clone, so memory is separate, but that's not the case with windows,
+ //Taking advantage of shared memory to allow both threads to add info to badSeqNames.
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ vector<sumScreenData*> pDataArray;
+ DWORD dwThreadIdArray[processors-1];
+ HANDLE hThreadArray[processors-1];
+
+ //Create processor worker threads.
+ for( int i=0; i<processors-1; i++ ){
+
+ string extension = "";
+ if (i!=0) {extension += toString(i) + ".temp"; processIDS.push_back(i); }
+
+ // Allocate memory for thread data.
+ sumScreenData* tempSum = new sumScreenData(startPos, endPos, maxAmbig, maxHomoP, minLength, maxLength, filename, m, lines[i].start, lines[i].end,goodFileName+extension, badAccnos+extension);
+ pDataArray.push_back(tempSum);
+
+ //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
+ hThreadArray[i] = CreateThread(NULL, 0, MySumScreenThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
+ }
+
+ //do your part
+ num = driver(lines[processors-1], (goodFileName+toString(processors-1)+".temp"), (badAccnos+toString(processors-1)+".temp"), filename, badSeqNames);
+ processIDS.push_back(processors-1);
+
+ //Wait until all threads have terminated.
+ WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
+
+ //Close all thread handles and free memory allocations.
+ for(int i=0; i < pDataArray.size(); i++){
+ num += pDataArray[i]->count;
+ for (set<string>::iterator it = pDataArray[i]->badSeqNames.begin(); it != pDataArray[i]->badSeqNames.end(); it++) { badSeqNames.insert(*it); }
+ CloseHandle(hThreadArray[i]);
+ delete pDataArray[i];
+ }
+
+ for (int i = 0; i < processIDS.size(); i++) {
+ m->appendFiles((goodFileName + toString(processIDS[i]) + ".temp"), goodFileName);
+ m->mothurRemove((goodFileName + toString(processIDS[i]) + ".temp"));
+
+ m->appendFiles((badAccnos + toString(processIDS[i]) + ".temp"), badAccnos);
+ m->mothurRemove((badAccnos + toString(processIDS[i]) + ".temp"));
+ }
+
+#endif
+
+ return num;
+
}
catch(exception& e) {
m->errorOut(e, "ScreenSeqsCommand", "createProcesses");
projects / mothur.git / blobdiff