#include "filterseqscommand.h"
#include "sequence.hpp"
+//**********************************************************************************************************************
+vector<string> FilterSeqsCommand::getValidParameters(){
+ try {
+ string Array[] = {"fasta", "trump", "soft", "hard", "vertical", "outputdir","inputdir", "processors"};
+ vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "FilterSeqsCommand", "getValidParameters");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+FilterSeqsCommand::FilterSeqsCommand(){
+ try {
+ abort = true; calledHelp = true;
+ vector<string> tempOutNames;
+ outputTypes["fasta"] = tempOutNames;
+ outputTypes["filter"] = tempOutNames;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "FilterSeqsCommand", "FilterSeqsCommand");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+vector<string> FilterSeqsCommand::getRequiredParameters(){
+ try {
+ string Array[] = {"fasta"};
+ vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "FilterSeqsCommand", "getRequiredParameters");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+vector<string> FilterSeqsCommand::getRequiredFiles(){
+ try {
+ vector<string> myArray;
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "FilterSeqsCommand", "getRequiredFiles");
+ exit(1);
+ }
+}
/**************************************************************************************/
-
FilterSeqsCommand::FilterSeqsCommand(string option) {
try {
- abort = false;
+ abort = false; calledHelp = false;
filterFileName = "";
//allow user to run help
- if(option == "help") { help(); abort = true; }
+ if(option == "help") { help(); abort = true; calledHelp = true; }
else {
//valid paramters for this command
OptionParser parser(option);
map<string,string> parameters = parser.getParameters();
- ValidParameters validParameter;
+ ValidParameters validParameter("filter.seqs");
map<string,string>::iterator it;
//check to make sure all parameters are valid for command
if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
}
+ //initialize outputTypes
+ vector<string> tempOutNames;
+ outputTypes["fasta"] = tempOutNames;
+ outputTypes["filter"] = 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 = ""; }
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; }
}
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; }
}
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]; }
}
- int ableToOpen;
ifstream in;
- ableToOpen = openInputFile(fastafileNames[i], in);
+ 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() + 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(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--;
}else{
- string simpleName = getSimpleName(fastafileNames[i]);
+ string simpleName = m->getSimpleName(fastafileNames[i]);
filterFileName += simpleName.substr(0, simpleName.find_first_of('.'));
}
in.close();
//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
// ...at some point should added some additional type checking...
string temp;
+ hard = validParameter.validFile(parameters, "hard", true); if (hard == "not found") { hard = ""; }
+ else if (hard == "not open") { hard = ""; abort = true; }
+
temp = validParameter.validFile(parameters, "trump", false); if (temp == "not found") { temp = "*"; }
trump = temp[0];
temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; }
convert(temp, processors);
- hard = validParameter.validFile(parameters, "hard", true); if (hard == "not found") { hard = ""; }
- else if (hard == "not open") { abort = true; }
-
- vertical = validParameter.validFile(parameters, "vertical", false); if (vertical == "not found") { vertical = "T"; }
+ vertical = validParameter.validFile(parameters, "vertical", false);
+ if (vertical == "not found") {
+ if ((hard == "") && (trump == '*') && (soft == 0)) { vertical = "T"; } //you have not given a hard file or set the trump char.
+ else { vertical = "F"; }
+ }
numSeqs = 0;
-
}
}
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 .... The default is ....\n");
- m->mothurOut("The vertical parameter .... The default is T.\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, soft=yourSoft, hard=yourHard, vertical=yourVertical) \n");
- m->mothurOut("Example filter.seqs(fasta=abrecovery.fasta, trump=..., soft=..., hard=..., vertical=T).\n");
+ m->mothurOut("filter.seqs(fasta=yourFastaFile, trump=yourTrump) \n");
+ m->mothurOut("Example filter.seqs(fasta=abrecovery.fasta, trump=.).\n");
m->mothurOut("Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFasta).\n\n");
}
int FilterSeqsCommand::execute() {
try {
- if (abort == true) { return 0; }
+ if (abort == true) { if (calledHelp) { return 0; } return 2; }
ifstream inFASTA;
- openInputFile(fastafileNames[0], inFASTA);
+ m->openInputFile(fastafileNames[0], inFASTA);
Sequence testSeq(inFASTA);
alignmentLength = testSeq.getAlignLength();
m->mothurOutEndLine(); m->mothurOutEndLine();
- if (m->control_pressed) { return 0; }
+ if (m->control_pressed) { outputTypes.clear(); return 0; }
#ifdef USE_MPI
int pid;
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);
+ outputNames.push_back(filterFile); outputTypes["filter"].push_back(filterFile);
#ifdef USE_MPI
}
if(filter[i] == '1'){ filteredLength++; }
}
- if (m->control_pressed) { for(int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; }
+ if (m->control_pressed) { outputTypes.clear(); for(int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; }
m->mothurOutEndLine();
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 pid, numSeqsPerProcessor, num;
int tag = 2001;
- vector<long>MPIPos;
+ vector<unsigned long int>MPIPos;
MPI_Status status;
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
MPI_File outMPI;
- MPI_File tempMPI;
MPI_File inMPI;
int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
int inMode=MPI_MODE_RDONLY;
- //char* outFilename = new char[filteredFasta.length()];
- //memcpy(outFilename, filteredFasta.c_str(), filteredFasta.length());
-
char outFilename[1024];
strcpy(outFilename, filteredFasta.c_str());
-
- //char* inFileName = new char[fastafileNames[s].length()];
- //memcpy(inFileName, fastafileNames[s].c_str(), fastafileNames[s].length());
-
+
char inFileName[1024];
strcpy(inFileName, fastafileNames[s].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);
-
- //delete inFileName;
- //delete outFilename;
if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); 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
//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
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);
MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
#else
+ vector<unsigned long int> 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)]));
+ }
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
if(processors == 1){
- ifstream inFASTA;
- int numFastaSeqs;
- openInputFile(fastafileNames[s], inFASTA);
- getNumSeqs(inFASTA, numFastaSeqs);
- inFASTA.close();
-
- lines.push_back(new linePair(0, numFastaSeqs));
-
+ int numFastaSeqs = driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]);
numSeqs += numFastaSeqs;
-
- driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]);
}else{
- setLines(fastafileNames[s]);
- createProcessesRunFilter(filter, fastafileNames[s]);
+ int numFastaSeqs = createProcessesRunFilter(filter, fastafileNames[s]);
+ numSeqs += numFastaSeqs;
rename((fastafileNames[s] + toString(processIDS[0]) + ".temp").c_str(), filteredFasta.c_str());
//append fasta files
for(int i=1;i<processors;i++){
- appendFiles((fastafileNames[s] + toString(processIDS[i]) + ".temp"), filteredFasta);
+ m->appendFiles((fastafileNames[s] + toString(processIDS[i]) + ".temp"), filteredFasta);
remove((fastafileNames[s] + toString(processIDS[i]) + ".temp").c_str());
}
}
if (m->control_pressed) { return 1; }
#else
- ifstream inFASTA;
- int numFastaSeqs;
- openInputFile(fastafileNames[s], inFASTA);
- getNumSeqs(inFASTA, numFastaSeqs);
- inFASTA.close();
-
- lines.push_back(new linePair(0, numFastaSeqs));
-
+ int numFastaSeqs = driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]);
numSeqs += numFastaSeqs;
-
- driverRunFilter(filter, filteredFasta, fastafileNames[s], lines[0]);
if (m->control_pressed) { return 1; }
#endif
#endif
- outputNames.push_back(filteredFasta);
+ outputNames.push_back(filteredFasta); outputTypes["fasta"].push_back(filteredFasta);
}
return 0;
}
#ifdef USE_MPI
/**************************************************************************************/
-int FilterSeqsCommand::driverMPIRun(int start, int num, MPI_File& inMPI, MPI_File& outMPI, vector<long>& MPIPos) {
+int FilterSeqsCommand::driverMPIRun(int start, int num, MPI_File& inMPI, MPI_File& outMPI, vector<unsigned long int>& MPIPos) {
try {
string outputString = "";
int count = 0;
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();
}
#endif
/**************************************************************************************/
-int FilterSeqsCommand::driverRunFilter(string F, string outputFilename, string inputFilename, linePair* line) {
+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(line->start);
-
- for(int i=0;i<line->num;i++){
+ in.seekg(filePos->start);
+
+ bool done = false;
+ int count = 0;
+
+ while (!done) {
if (m->control_pressed) { in.close(); out.close(); return 0; }
- Sequence seq(in);
+ Sequence seq(in); m->gobble(in);
if (seq.getName() != "") {
string align = seq.getAligned();
string filterSeq = "";
}
out << '>' << seq.getName() << endl << filterSeq << endl;
- }
- gobble(in);
+ count++;
+ }
+
+ #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(); }
}
+ //report progress
+ if((count) % 100 != 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
+
+
out.close();
in.close();
- return 0;
+ return count;
}
catch(exception& e) {
m->errorOut(e, "FilterSeqsCommand", "driverRunFilter");
try {
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
int process = 0;
- int exitCommand = 1;
+ int num = 0;
processIDS.clear();
//loop through and create all the processes you want
process++;
}else if (pid == 0){
string filteredFasta = filename + toString(getpid()) + ".temp";
- driverRunFilter(F, filteredFasta, filename, lines[process]);
+ num = driverRunFilter(F, filteredFasta, filename, lines[process]);
+
+ //pass numSeqs to parent
+ ofstream out;
+ string tempFile = filename + 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
for (int i=0;i<processors;i++) {
int temp = processIDS[i];
wait(&temp);
+ }
+
+ for (int i = 0; i < processIDS.size(); i++) {
+ ifstream 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());
}
+
- return exitCommand;
+ return num;
#endif
}
catch(exception& e) {
F.setLength(alignmentLength);
- if(trump != '*' || isTrue(vertical) || soft != 0){
+ if(trump != '*' || m->isTrue(vertical) || soft != 0){
F.initialize();
}
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();
#ifdef USE_MPI
int pid, numSeqsPerProcessor, num;
int tag = 2001;
- vector<long> MPIPos;
+ vector<unsigned long int> MPIPos;
MPI_Status status;
MPI_File inMPI;
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
MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
#else
+ vector<unsigned long int> 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)]));
+ }
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
if(processors == 1){
- ifstream inFASTA;
- int numFastaSeqs;
- openInputFile(fastafileNames[s], inFASTA);
- getNumSeqs(inFASTA, numFastaSeqs);
- inFASTA.close();
-
+ int numFastaSeqs = driverCreateFilter(F, fastafileNames[s], lines[0]);
numSeqs += numFastaSeqs;
-
- lines.push_back(new linePair(0, numFastaSeqs));
-
- driverCreateFilter(F, fastafileNames[s], lines[0]);
}else{
- setLines(fastafileNames[s]);
- createProcessesCreateFilter(F, fastafileNames[s]);
+ int numFastaSeqs = createProcessesCreateFilter(F, fastafileNames[s]);
+ numSeqs += numFastaSeqs;
}
if (m->control_pressed) { return filterString; }
#else
- ifstream inFASTA;
- int numFastaSeqs;
- openInputFile(fastafileNames[s], inFASTA);
- getNumSeqs(inFASTA, numFastaSeqs);
- inFASTA.close();
-
+ int numFastaSeqs = driverCreateFilter(F, fastafileNames[s], lines[0]);
numSeqs += numFastaSeqs;
-
- lines.push_back(new linePair(0, numFastaSeqs));
-
- driverCreateFilter(F, fastafileNames[s], lines[0]);
if (m->control_pressed) { return filterString; }
#endif
#endif
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
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
}
}
/**************************************************************************************/
-int FilterSeqsCommand::driverCreateFilter(Filters& F, string filename, linePair* line) {
+int FilterSeqsCommand::driverCreateFilter(Filters& F, string filename, linePair* filePos) {
try {
ifstream in;
- openInputFile(filename, in);
+ m->openInputFile(filename, in);
- in.seekg(line->start);
-
- for(int i=0;i<line->num;i++){
+ in.seekg(filePos->start);
+
+ bool done = false;
+ int count = 0;
+
+ while (!done) {
if (m->control_pressed) { in.close(); return 1; }
- Sequence seq(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++;
}
+ #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((i+1) % 100 == 0){ m->mothurOut(toString(i+1)); m->mothurOutEndLine(); }
+ if((count) % 100 == 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
}
-
//report progress
- if((line->num) % 100 != 0){ m->mothurOut(toString(line->num)); m->mothurOutEndLine(); }
-
+ if((count) % 100 != 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
in.close();
- return 0;
+ return count;
}
catch(exception& e) {
m->errorOut(e, "FilterSeqsCommand", "driverCreateFilter");
}
#ifdef USE_MPI
/**************************************************************************************/
-int FilterSeqsCommand::MPICreateFilter(int start, int num, Filters& F, MPI_File& inMPI, vector<long>& MPIPos) {
+int FilterSeqsCommand::MPICreateFilter(int start, int num, Filters& F, MPI_File& inMPI, vector<unsigned long int>& MPIPos) {
try {
MPI_Status status;
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
int FilterSeqsCommand::createProcessesCreateFilter(Filters& F, string filename) {
try {
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- int process = 0;
- int exitCommand = 1;
+ int process = 1;
+ int num = 0;
processIDS.clear();
//loop through and create all the processes you want
while (process != processors) {
- int pid = vfork();
+ int pid = fork();
if (pid > 0) {
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){
- driverCreateFilter(F, filename, lines[process]);
+ //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;
+ 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("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
+ for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
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;i<processors;i++) {
+ for (int i=0;i<(processors-1);i++) {
int temp = processIDS[i];
wait(&temp);
}
- return exitCommand;
-#endif
- }
- catch(exception& e) {
- m->errorOut(e, "FilterSeqsCommand", "createProcessesCreateFilter");
- exit(1);
- }
-}
-/**************************************************************************************************/
-
-int FilterSeqsCommand::setLines(string filename) {
- try {
-
- vector<long int> positions;
- bufferSizes.clear();
-
- ifstream inFASTA;
- openInputFile(filename, inFASTA);
+ //parent reads in and combines Filter info
+ for (int i = 0; i < processIDS.size(); i++) {
+ string tempFilename = filename + toString(processIDS[i]) + "filterValues.temp";
+ ifstream in;
+ m->openInputFile(tempFilename, in);
- string input;
- while(!inFASTA.eof()){
- input = getline(inFASTA);
+ int temp, tempNum;
+ string tempFilterString;
- if (input.length() != 0) {
- if(input[0] == '>'){ long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
- }
- }
- inFASTA.close();
-
- int numFastaSeqs = positions.size();
-
- FILE * pFile;
- long size;
-
- //get num bytes in file
- pFile = fopen (filename.c_str(),"rb");
- if (pFile==NULL) perror ("Error opening file");
- else{
- fseek (pFile, 0, SEEK_END);
- size=ftell (pFile);
- fclose (pFile);
- }
-
- numSeqs += numFastaSeqs;
-
- int numSeqsPerProcessor = numFastaSeqs / processors;
-
- for (int i = 0; i < processors; i++) {
+ in >> tempNum; m->gobble(in); num += tempNum;
- long int startPos = positions[ i * numSeqsPerProcessor ];
- if(i == processors - 1){
- numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor;
- bufferSizes.push_back(size - startPos);
- }else{
- long int myEnd = positions[ (i+1) * numSeqsPerProcessor ];
- bufferSizes.push_back(myEnd-startPos);
- }
- lines.push_back(new linePair(startPos, numSeqsPerProcessor));
+ 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());
}
- return 0;
+ return num;
+#endif
}
catch(exception& e) {
- m->errorOut(e, "FilterSeqsCommand", "setLines");
+ m->errorOut(e, "FilterSeqsCommand", "createProcessesCreateFilter");
exit(1);
}
}