5 * Created by Pat Schloss on 12/22/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "trimflowscommand.h"
11 #include "needlemanoverlap.hpp"
14 //**********************************************************************************************************************
15 vector<string> TrimFlowsCommand::setParameters(){
17 CommandParameter pflow("flow", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pflow);
18 CommandParameter poligos("oligos", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(poligos);
19 CommandParameter pmaxhomop("maxhomop", "Number", "", "9", "", "", "",false,false); parameters.push_back(pmaxhomop);
20 CommandParameter pmaxflows("maxflows", "Number", "", "450", "", "", "",false,false); parameters.push_back(pmaxflows);
21 CommandParameter pminflows("minflows", "Number", "", "450", "", "", "",false,false); parameters.push_back(pminflows);
22 CommandParameter ppdiffs("pdiffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(ppdiffs);
23 CommandParameter pbdiffs("bdiffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(pbdiffs);
24 CommandParameter pldiffs("ldiffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(pldiffs);
25 CommandParameter psdiffs("sdiffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(psdiffs);
26 CommandParameter ptdiffs("tdiffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(ptdiffs);
27 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
28 CommandParameter psignal("signal", "Number", "", "0.50", "", "", "",false,false); parameters.push_back(psignal);
29 CommandParameter pnoise("noise", "Number", "", "0.70", "", "", "",false,false); parameters.push_back(pnoise);
30 CommandParameter pallfiles("allfiles", "Boolean", "", "t", "", "", "",false,false); parameters.push_back(pallfiles);
31 CommandParameter porder("order", "String", "", "", "", "", "",false,false); parameters.push_back(porder);
32 CommandParameter pfasta("fasta", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(pfasta);
33 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
34 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
36 vector<string> myArray;
37 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
41 m->errorOut(e, "TrimFlowsCommand", "setParameters");
45 //**********************************************************************************************************************
46 string TrimFlowsCommand::getHelpString(){
48 string helpString = "";
49 helpString += "The trim.flows command reads a flowgram file and creates .....\n";
50 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFasta).\n";
51 helpString += "For more details please check out the wiki http://www.mothur.org/wiki/Trim.flows.\n";
55 m->errorOut(e, "TrimFlowsCommand", "getHelpString");
60 //**********************************************************************************************************************
62 TrimFlowsCommand::TrimFlowsCommand(){
64 abort = true; calledHelp = true;
66 vector<string> tempOutNames;
67 outputTypes["flow"] = tempOutNames;
68 outputTypes["fasta"] = tempOutNames;
71 m->errorOut(e, "TrimFlowsCommand", "TrimFlowsCommand");
75 //**********************************************************************************************************************
77 TrimFlowsCommand::TrimFlowsCommand(string option) {
80 abort = false; calledHelp = false;
83 //allow user to run help
84 if(option == "help") { help(); abort = true; calledHelp = true; }
85 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
89 vector<string> myArray = setParameters();
91 OptionParser parser(option);
92 map<string,string> parameters = parser.getParameters();
94 ValidParameters validParameter;
95 map<string,string>::iterator it;
97 //check to make sure all parameters are valid for command
98 for (it = parameters.begin(); it != parameters.end(); it++) {
99 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
102 //initialize outputTypes
103 vector<string> tempOutNames;
104 outputTypes["flow"] = tempOutNames;
105 outputTypes["fasta"] = tempOutNames;
107 //if the user changes the input directory command factory will send this info to us in the output parameter
108 string inputDir = validParameter.validFile(parameters, "inputdir", false);
109 if (inputDir == "not found"){ inputDir = ""; }
112 it = parameters.find("flow");
113 //user has given a template file
114 if(it != parameters.end()){
115 path = m->hasPath(it->second);
116 //if the user has not given a path then, add inputdir. else leave path alone.
117 if (path == "") { parameters["flow"] = inputDir + it->second; }
120 it = parameters.find("oligos");
121 //user has given a template file
122 if(it != parameters.end()){
123 path = m->hasPath(it->second);
124 //if the user has not given a path then, add inputdir. else leave path alone.
125 if (path == "") { parameters["oligos"] = inputDir + it->second; }
131 //check for required parameters
132 flowFileName = validParameter.validFile(parameters, "flow", true);
133 if (flowFileName == "not found") {
134 flowFileName = m->getFlowFile();
135 if (flowFileName != "") { m->mothurOut("Using " + flowFileName + " as input file for the flow parameter."); m->mothurOutEndLine(); }
137 m->mothurOut("No valid current flow file. You must provide a flow file."); m->mothurOutEndLine();
140 }else if (flowFileName == "not open") { flowFileName = ""; abort = true; }
142 //if the user changes the output directory command factory will send this info to us in the output parameter
143 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
145 outputDir += m->hasPath(flowFileName); //if user entered a file with a path then preserve it
149 //check for optional parameter and set defaults
150 // ...at some point should added some additional type checking...
153 temp = validParameter.validFile(parameters, "minflows", false); if (temp == "not found") { temp = "450"; }
154 m->mothurConvert(temp, minFlows);
156 temp = validParameter.validFile(parameters, "maxflows", false); if (temp == "not found") { temp = "450"; }
157 m->mothurConvert(temp, maxFlows);
160 temp = validParameter.validFile(parameters, "oligos", true);
161 if (temp == "not found") { oligoFileName = ""; }
162 else if(temp == "not open") { abort = true; }
163 else { oligoFileName = temp; m->setOligosFile(oligoFileName); }
165 temp = validParameter.validFile(parameters, "fasta", false); if (temp == "not found"){ fasta = 0; }
166 else if(m->isTrue(temp)) { fasta = 1; }
168 temp = validParameter.validFile(parameters, "maxhomop", false); if (temp == "not found"){ temp = "9"; }
169 m->mothurConvert(temp, maxHomoP);
171 temp = validParameter.validFile(parameters, "signal", false); if (temp == "not found"){ temp = "0.50"; }
172 m->mothurConvert(temp, signal);
174 temp = validParameter.validFile(parameters, "noise", false); if (temp == "not found"){ temp = "0.70"; }
175 m->mothurConvert(temp, noise);
177 temp = validParameter.validFile(parameters, "bdiffs", false); if (temp == "not found"){ temp = "0"; }
178 m->mothurConvert(temp, bdiffs);
180 temp = validParameter.validFile(parameters, "pdiffs", false); if (temp == "not found"){ temp = "0"; }
181 m->mothurConvert(temp, pdiffs);
183 temp = validParameter.validFile(parameters, "ldiffs", false); if (temp == "not found") { temp = "0"; }
184 m->mothurConvert(temp, ldiffs);
186 temp = validParameter.validFile(parameters, "sdiffs", false); if (temp == "not found") { temp = "0"; }
187 m->mothurConvert(temp, sdiffs);
189 temp = validParameter.validFile(parameters, "tdiffs", false); if (temp == "not found") { int tempTotal = pdiffs + bdiffs + ldiffs + sdiffs; temp = toString(tempTotal); }
190 m->mothurConvert(temp, tdiffs);
192 if(tdiffs == 0){ tdiffs = bdiffs + pdiffs + ldiffs + sdiffs; }
195 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
196 m->setProcessors(temp);
197 m->mothurConvert(temp, processors);
199 flowOrder = validParameter.validFile(parameters, "order", false);
200 if (flowOrder == "not found"){ flowOrder = "TACG"; }
201 else if(flowOrder.length() != 4){
202 m->mothurOut("The value of the order option must be four bases long\n");
205 if(oligoFileName == "") { allFiles = 0; }
206 else { allFiles = 1; }
215 catch(exception& e) {
216 m->errorOut(e, "TrimFlowsCommand", "TrimFlowsCommand");
221 //***************************************************************************************************************
223 int TrimFlowsCommand::execute(){
226 if (abort == true) { if (calledHelp) { return 0; } return 2; }
228 string trimFlowFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + "trim.flow";
229 outputNames.push_back(trimFlowFileName); outputTypes["flow"].push_back(trimFlowFileName);
231 string scrapFlowFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + "scrap.flow";
232 outputNames.push_back(scrapFlowFileName); outputTypes["flow"].push_back(scrapFlowFileName);
234 string fastaFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + "flow.fasta";
236 outputNames.push_back(fastaFileName); outputTypes["fasta"].push_back(fastaFileName);
239 vector<unsigned long long> flowFilePos;
240 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
241 flowFilePos = getFlowFileBreaks();
242 for (int i = 0; i < (flowFilePos.size()-1); i++) {
243 lines.push_back(new linePair(flowFilePos[i], flowFilePos[(i+1)]));
246 ifstream in; m->openInputFile(flowFileName, in); in >> numFlows; in.close();
247 ///////////////////////////////////////// until I fix multiple processors for windows //////////////////
249 ///////////////////////////////////////// until I fix multiple processors for windows //////////////////
250 if (processors == 1) {
251 lines.push_back(new linePair(0, 1000));
254 flowFilePos = m->setFilePosEachLine(flowFileName, numFlowLines);
255 flowFilePos.erase(flowFilePos.begin() + 1); numFlowLines--;
257 //figure out how many sequences you have to process
258 int numSeqsPerProcessor = numFlowLines / processors;
259 cout << numSeqsPerProcessor << '\t' << numFlowLines << endl;
260 for (int i = 0; i < processors; i++) {
261 int startIndex = i * numSeqsPerProcessor;
262 if(i == (processors - 1)){ numSeqsPerProcessor = numFlowLines - i * numSeqsPerProcessor; }
263 lines.push_back(new linePair(flowFilePos[startIndex], numSeqsPerProcessor));
264 cout << flowFilePos[startIndex] << '\t' << numSeqsPerProcessor << endl;
269 vector<vector<string> > barcodePrimerComboFileNames;
270 if(oligoFileName != ""){
271 getOligos(barcodePrimerComboFileNames);
275 driverCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames, lines[0]);
277 createProcessesCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames);
280 if (m->control_pressed) { return 0; }
282 string flowFilesFileName;
286 set<string> namesAlreadyProcessed;
287 flowFilesFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + "flow.files";
288 m->openOutputFile(flowFilesFileName, output);
290 for(int i=0;i<barcodePrimerComboFileNames.size();i++){
291 for(int j=0;j<barcodePrimerComboFileNames[0].size();j++){
292 if (namesAlreadyProcessed.count(barcodePrimerComboFileNames[i][j]) == 0) {
294 unsigned long long size;
296 //get num bytes in file
297 pFile = fopen (barcodePrimerComboFileNames[i][j].c_str(),"rb");
298 if (pFile==NULL) perror ("Error opening file");
300 fseek (pFile, 0, SEEK_END);
306 m->mothurRemove(barcodePrimerComboFileNames[i][j]);
309 output << m->getFullPathName(barcodePrimerComboFileNames[i][j]) << endl;
310 outputNames.push_back(barcodePrimerComboFileNames[i][j]);
311 outputTypes["flow"].push_back(barcodePrimerComboFileNames[i][j]);
313 namesAlreadyProcessed.insert(barcodePrimerComboFileNames[i][j]);
320 flowFilesFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + "flow.files";
321 m->openOutputFile(flowFilesFileName, output);
323 output << m->getFullPathName(trimFlowFileName) << endl;
327 outputTypes["flow.files"].push_back(flowFilesFileName);
328 outputNames.push_back(flowFilesFileName);
330 // set fasta file as new current fastafile
331 // string current = "";
332 // itTypes = outputTypes.find("fasta");
333 // if (itTypes != outputTypes.end()) {
334 // if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); }
337 m->mothurOutEndLine();
338 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
339 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
340 m->mothurOutEndLine();
344 catch(exception& e) {
345 m->errorOut(e, "TrimSeqsCommand", "execute");
350 //***************************************************************************************************************
352 int TrimFlowsCommand::driverCreateTrim(string flowFileName, string trimFlowFileName, string scrapFlowFileName, string fastaFileName, vector<vector<string> > thisBarcodePrimerComboFileNames, linePair* line){
355 ofstream trimFlowFile;
356 m->openOutputFile(trimFlowFileName, trimFlowFile);
357 trimFlowFile.setf(ios::fixed, ios::floatfield); trimFlowFile.setf(ios::showpoint);
359 ofstream scrapFlowFile;
360 m->openOutputFile(scrapFlowFileName, scrapFlowFile);
361 scrapFlowFile.setf(ios::fixed, ios::floatfield); scrapFlowFile.setf(ios::showpoint);
364 if(fasta){ m->openOutputFile(fastaFileName, fastaFile); }
367 m->openInputFile(flowFileName, flowFile);
369 flowFile.seekg(line->start);
371 if(line->start == 0){
372 flowFile >> numFlows; m->gobble(flowFile);
373 scrapFlowFile << maxFlows << endl;
374 trimFlowFile << maxFlows << endl;
376 for(int i=0;i<thisBarcodePrimerComboFileNames.size();i++){
377 for(int j=0;j<thisBarcodePrimerComboFileNames[0].size();j++){
379 m->openOutputFile(thisBarcodePrimerComboFileNames[i][j], temp);
380 temp << maxFlows << endl;
387 FlowData flowData(numFlows, signal, noise, maxHomoP, flowOrder);
388 //cout << " driver flowdata address " << &flowData << &flowFile << endl;
392 TrimOligos trimOligos(pdiffs, bdiffs, ldiffs, sdiffs, primers, barcodes, revPrimer, linker, spacer);
396 if (m->control_pressed) { break; }
399 int currentSeqDiffs = 0;
400 string trashCode = "";
402 flowData.getNext(flowFile);
403 //cout << "driver good bit " << flowFile.good() << endl;
404 flowData.capFlows(maxFlows);
406 Sequence currSeq = flowData.getSequence();
408 if(!flowData.hasMinFlows(minFlows)){ //screen to see if sequence is of a minimum number of flows
414 int barcodeIndex = 0;
417 success = trimOligos.stripLinker(currSeq);
418 if(success > ldiffs) { trashCode += 'k'; }
419 else{ currentSeqDiffs += success; }
423 if(barcodes.size() != 0){
424 success = trimOligos.stripBarcode(currSeq, barcodeIndex);
425 if(success > bdiffs) { trashCode += 'b'; }
426 else{ currentSeqDiffs += success; }
430 success = trimOligos.stripSpacer(currSeq);
431 if(success > sdiffs) { trashCode += 's'; }
432 else{ currentSeqDiffs += success; }
436 if(numFPrimers != 0){
437 success = trimOligos.stripForward(currSeq, primerIndex);
438 if(success > pdiffs) { trashCode += 'f'; }
439 else{ currentSeqDiffs += success; }
442 if (currentSeqDiffs > tdiffs) { trashCode += 't'; }
444 if(numRPrimers != 0){
445 success = trimOligos.stripReverse(currSeq);
446 if(!success) { trashCode += 'r'; }
449 if(trashCode.length() == 0){
451 flowData.printFlows(trimFlowFile);
453 if(fasta) { currSeq.printSequence(fastaFile); }
457 m->openOutputFileAppend(thisBarcodePrimerComboFileNames[barcodeIndex][primerIndex], output);
458 output.setf(ios::fixed, ios::floatfield); trimFlowFile.setf(ios::showpoint);
460 flowData.printFlows(output);
465 flowData.printFlows(scrapFlowFile, trashCode);
469 //cout << "driver" << '\t' << currSeq.getName() << endl;
471 if((count) % 10000 == 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
473 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
474 unsigned long long pos = flowFile.tellg();
476 if ((pos == -1) || (pos >= line->end)) { break; }
478 if (flowFile.eof()) { break; }
483 if((count) % 10000 != 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
485 trimFlowFile.close();
486 scrapFlowFile.close();
488 if(fasta){ fastaFile.close(); }
492 catch(exception& e) {
493 m->errorOut(e, "TrimSeqsCommand", "driverCreateTrim");
498 //***************************************************************************************************************
500 void TrimFlowsCommand::getOligos(vector<vector<string> >& outFlowFileNames){
503 m->openInputFile(oligoFileName, oligosFile);
505 string type, oligo, group;
508 int indexBarcode = 0;
510 while(!oligosFile.eof()){
512 oligosFile >> type; m->gobble(oligosFile); //get the first column value of the row - is it a comment or a feature we are interested in?
514 if(type[0] == '#'){ //igore the line because there's a comment
515 while (!oligosFile.eof()) { char c = oligosFile.get(); if (c == 10 || c == 13){ break; } } // get rest of line if there's any crap there
517 else{ //there's a feature we're interested in
519 for(int i=0;i<type.length();i++){ type[i] = toupper(type[i]); } //make type case insensitive
521 oligosFile >> oligo; //get the DNA sequence for the feature
523 for(int i=0;i<oligo.length();i++){ //make type case insensitive and change any U's to T's
524 oligo[i] = toupper(oligo[i]);
525 if(oligo[i] == 'U') { oligo[i] = 'T'; }
528 if(type == "FORWARD"){ //if the feature is a forward primer...
531 while (!oligosFile.eof()) { // get rest of line in case there is a primer name = will have the name of the primer
532 char c = oligosFile.get();
533 if (c == 10 || c == 13){ break; }
534 else if (c == 32 || c == 9){;} //space or tab
538 //have we seen this primer already?
539 map<string, int>::iterator itPrimer = primers.find(oligo);
540 if (itPrimer != primers.end()) { m->mothurOut("primer " + oligo + " is in your oligos file already."); m->mothurOutEndLine(); }
542 primers[oligo]=indexPrimer; indexPrimer++;
543 primerNameVector.push_back(group);
546 else if(type == "REVERSE"){
547 string oligoRC = reverseOligo(oligo);
548 revPrimer.push_back(oligoRC);
550 else if(type == "BARCODE"){
553 //check for repeat barcodes
554 map<string, int>::iterator itBar = barcodes.find(oligo);
555 if (itBar != barcodes.end()) { m->mothurOut("barcode " + oligo + " is in your oligos file already."); m->mothurOutEndLine(); }
557 barcodes[oligo]=indexBarcode; indexBarcode++;
558 barcodeNameVector.push_back(group);
559 }else if(type == "LINKER"){
560 linker.push_back(oligo);
561 }else if(type == "SPACER"){
562 spacer.push_back(oligo);
565 m->mothurOut(type + " is not recognized as a valid type. Choices are forward, reverse, and barcode. Ignoring " + oligo + "."); m->mothurOutEndLine();
569 m->gobble(oligosFile);
573 if(barcodeNameVector.size() == 0 && primerNameVector[0] == ""){ allFiles = 0; }
575 //add in potential combos
576 if(barcodeNameVector.size() == 0){
578 barcodeNameVector.push_back("");
581 if(primerNameVector.size() == 0){
583 primerNameVector.push_back("");
587 outFlowFileNames.resize(barcodeNameVector.size());
588 for(int i=0;i<outFlowFileNames.size();i++){
589 outFlowFileNames[i].assign(primerNameVector.size(), "");
594 for(map<string, int>::iterator itBar = barcodes.begin();itBar != barcodes.end();itBar++){
595 for(map<string, int>::iterator itPrimer = primers.begin();itPrimer != primers.end(); itPrimer++){
597 string primerName = primerNameVector[itPrimer->second];
598 string barcodeName = barcodeNameVector[itBar->second];
600 string comboGroupName = "";
601 string fileName = "";
603 if(primerName == ""){
604 comboGroupName = barcodeNameVector[itBar->second];
605 fileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + comboGroupName + ".flow";
608 if(barcodeName == ""){
609 comboGroupName = primerNameVector[itPrimer->second];
612 comboGroupName = barcodeNameVector[itBar->second] + "." + primerNameVector[itPrimer->second];
614 fileName = outputDir + m->getRootName(m->getSimpleName(flowFileName)) + comboGroupName + ".flow";
617 outFlowFileNames[itBar->second][itPrimer->second] = fileName;
620 m->openOutputFile(fileName, temp);
626 numFPrimers = primers.size();
627 numRPrimers = revPrimer.size();
628 numLinkers = linker.size();
629 numSpacers = spacer.size();
632 catch(exception& e) {
633 m->errorOut(e, "TrimSeqsCommand", "getOligos");
637 //********************************************************************/
638 string TrimFlowsCommand::reverseOligo(string oligo){
642 for(int i=oligo.length()-1;i>=0;i--){
644 if(oligo[i] == 'A') { reverse += 'T'; }
645 else if(oligo[i] == 'T'){ reverse += 'A'; }
646 else if(oligo[i] == 'U'){ reverse += 'A'; }
648 else if(oligo[i] == 'G'){ reverse += 'C'; }
649 else if(oligo[i] == 'C'){ reverse += 'G'; }
651 else if(oligo[i] == 'R'){ reverse += 'Y'; }
652 else if(oligo[i] == 'Y'){ reverse += 'R'; }
654 else if(oligo[i] == 'M'){ reverse += 'K'; }
655 else if(oligo[i] == 'K'){ reverse += 'M'; }
657 else if(oligo[i] == 'W'){ reverse += 'W'; }
658 else if(oligo[i] == 'S'){ reverse += 'S'; }
660 else if(oligo[i] == 'B'){ reverse += 'V'; }
661 else if(oligo[i] == 'V'){ reverse += 'B'; }
663 else if(oligo[i] == 'D'){ reverse += 'H'; }
664 else if(oligo[i] == 'H'){ reverse += 'D'; }
666 else { reverse += 'N'; }
672 catch(exception& e) {
673 m->errorOut(e, "TrimFlowsCommand", "reverseOligo");
678 /**************************************************************************************************/
679 vector<unsigned long long> TrimFlowsCommand::getFlowFileBreaks() {
683 vector<unsigned long long> filePos;
684 filePos.push_back(0);
687 unsigned long long size;
689 //get num bytes in file
690 pFile = fopen (flowFileName.c_str(),"rb");
691 if (pFile==NULL) perror ("Error opening file");
693 fseek (pFile, 0, SEEK_END);
698 //estimate file breaks
699 unsigned long long chunkSize = 0;
700 chunkSize = size / processors;
702 //file too small to divide by processors
703 if (chunkSize == 0) { processors = 1; filePos.push_back(size); return filePos; }
705 //for each process seekg to closest file break and search for next '>' char. make that the filebreak
706 for (int i = 0; i < processors; i++) {
707 unsigned long long spot = (i+1) * chunkSize;
710 m->openInputFile(flowFileName, in);
713 string dummy = m->getline(in);
715 //there was not another sequence before the end of the file
716 unsigned long long sanityPos = in.tellg();
718 // if (sanityPos == -1) { break; }
719 // else { filePos.push_back(newSpot); }
720 if (sanityPos == -1) { break; }
721 else { filePos.push_back(sanityPos); }
727 filePos.push_back(size);
729 //sanity check filePos
730 for (int i = 0; i < (filePos.size()-1); i++) {
731 if (filePos[(i+1)] <= filePos[i]) { filePos.erase(filePos.begin()+(i+1)); i--; }
735 m->openInputFile(flowFileName, in);
738 //unsigned long long spot = in.tellg();
742 processors = (filePos.size() - 1);
746 catch(exception& e) {
747 m->errorOut(e, "TrimSeqsCommand", "getFlowFileBreaks");
752 /**************************************************************************************************/
754 int TrimFlowsCommand::createProcessesCreateTrim(string flowFileName, string trimFlowFileName, string scrapFlowFileName, string fastaFileName, vector<vector<string> > barcodePrimerComboFileNames){
760 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
763 //loop through and create all the processes you want
764 while (process != processors) {
768 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
772 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
774 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
775 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
776 tempBarcodePrimerComboFileNames[i][j] += toString(getpid()) + ".temp";
778 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
784 driverCreateTrim(flowFileName,
785 (trimFlowFileName + toString(getpid()) + ".temp"),
786 (scrapFlowFileName + toString(getpid()) + ".temp"),
787 (fastaFileName + toString(getpid()) + ".temp"),
788 tempBarcodePrimerComboFileNames, lines[process]);
792 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
793 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
800 m->openOutputFile(trimFlowFileName, temp);
803 m->openOutputFile(scrapFlowFileName, temp);
807 m->openOutputFile(fastaFileName, temp);
811 driverCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames, lines[0]);
813 //force parent to wait until all the processes are done
814 for (int i=0;i<processIDS.size();i++) {
815 int temp = processIDS[i];
819 //////////////////////////////////////////////////////////////////////////////////////////////////////
820 //Windows version shared memory, so be careful when passing variables through the trimFlowData struct.
821 //Above fork() will clone, so memory is separate, but that's not the case with windows,
822 //////////////////////////////////////////////////////////////////////////////////////////////////////
824 vector<trimFlowData*> pDataArray;
825 DWORD dwThreadIdArray[processors-1];
826 HANDLE hThreadArray[processors-1];
828 //Create processor worker threads.
829 for( int i=0; i<processors-1; i++ ){
830 // Allocate memory for thread data.
831 string extension = "";
832 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
834 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
836 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
837 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
838 tempBarcodePrimerComboFileNames[i][j] += extension;
840 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
847 trimFlowData* tempflow = new trimFlowData(flowFileName, (trimFlowFileName + extension), (scrapFlowFileName + extension), fastaFileName, flowOrder, tempBarcodePrimerComboFileNames, barcodes, primers, revPrimer, fasta, allFiles, lines[i]->start, lines[i]->end, m, signal, noise, numFlows, maxFlows, minFlows, maxHomoP, tdiffs, bdiffs, pdiffs, i);
848 pDataArray.push_back(tempflow);
850 //MyTrimFlowThreadFunction is in header. It must be global or static to work with the threads.
851 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
852 hThreadArray[i] = CreateThread(NULL, 0, MyTrimFlowThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
855 //using the main process as a worker saves time and memory
857 m->openOutputFile(trimFlowFileName, temp);
860 m->openOutputFile(scrapFlowFileName, temp);
864 m->openOutputFile(fastaFileName, temp);
868 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
870 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
871 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
872 tempBarcodePrimerComboFileNames[i][j] += toString(processors-1) + ".temp";
874 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
881 //do my part - do last piece because windows is looking for eof
882 int num = driverCreateTrim(flowFileName, (trimFlowFileName + toString(processors-1) + ".temp"), (scrapFlowFileName + toString(processors-1) + ".temp"), (fastaFileName + toString(processors-1) + ".temp"), tempBarcodePrimerComboFileNames, lines[processors-1]);
883 processIDS.push_back((processors-1));
885 //Wait until all threads have terminated.
886 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
888 //Close all thread handles and free memory allocations.
889 for(int i=0; i < pDataArray.size(); i++){
890 num += pDataArray[i]->count;
891 CloseHandle(hThreadArray[i]);
892 delete pDataArray[i];
898 m->mothurOutEndLine();
899 for(int i=0;i<processIDS.size();i++){
901 m->mothurOut("Appending files from process " + toString(processIDS[i])); m->mothurOutEndLine();
903 m->appendFiles((trimFlowFileName + toString(processIDS[i]) + ".temp"), trimFlowFileName);
904 m->mothurRemove((trimFlowFileName + toString(processIDS[i]) + ".temp"));
905 // m->mothurOut("\tDone with trim.flow file"); m->mothurOutEndLine();
907 m->appendFiles((scrapFlowFileName + toString(processIDS[i]) + ".temp"), scrapFlowFileName);
908 m->mothurRemove((scrapFlowFileName + toString(processIDS[i]) + ".temp"));
909 // m->mothurOut("\tDone with scrap.flow file"); m->mothurOutEndLine();
912 m->appendFiles((fastaFileName + toString(processIDS[i]) + ".temp"), fastaFileName);
913 m->mothurRemove((fastaFileName + toString(processIDS[i]) + ".temp"));
914 // m->mothurOut("\tDone with flow.fasta file"); m->mothurOutEndLine();
917 for (int j = 0; j < barcodePrimerComboFileNames.size(); j++) {
918 for (int k = 0; k < barcodePrimerComboFileNames[0].size(); k++) {
919 m->appendFiles((barcodePrimerComboFileNames[j][k] + toString(processIDS[i]) + ".temp"), barcodePrimerComboFileNames[j][k]);
920 m->mothurRemove((barcodePrimerComboFileNames[j][k] + toString(processIDS[i]) + ".temp"));
929 catch(exception& e) {
930 m->errorOut(e, "TrimFlowsCommand", "createProcessesCreateTrim");
935 //***************************************************************************************************************