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","flow",false,true,true); parameters.push_back(pflow);
18 CommandParameter poligos("oligos", "InputTypes", "", "", "none", "none", "none","",false,false,true); 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,true); parameters.push_back(ppdiffs);
23 CommandParameter pbdiffs("bdiffs", "Number", "", "0", "", "", "","",false,false,true); 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,true); 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", "", "TACG", "", "", "","",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");
59 //**********************************************************************************************************************
60 string TrimFlowsCommand::getOutputPattern(string type) {
64 if (type == "flow") { pattern = "[filename],[tag],flow"; }
65 else if (type == "fasta") { pattern = "[filename],flow.fasta"; }
66 else if (type == "file") { pattern = "[filename],flow.files"; }
67 else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
72 m->errorOut(e, "TrimFlowsCommand", "getOutputPattern");
76 //**********************************************************************************************************************
78 TrimFlowsCommand::TrimFlowsCommand(){
80 abort = true; calledHelp = true;
82 vector<string> tempOutNames;
83 outputTypes["flow"] = tempOutNames;
84 outputTypes["fasta"] = tempOutNames;
85 outputTypes["file"] = tempOutNames;
88 m->errorOut(e, "TrimFlowsCommand", "TrimFlowsCommand");
92 //**********************************************************************************************************************
94 TrimFlowsCommand::TrimFlowsCommand(string option) {
97 abort = false; calledHelp = false;
100 //allow user to run help
101 if(option == "help") { help(); abort = true; calledHelp = true; }
102 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
106 vector<string> myArray = setParameters();
108 OptionParser parser(option);
109 map<string,string> parameters = parser.getParameters();
111 ValidParameters validParameter;
112 map<string,string>::iterator it;
114 //check to make sure all parameters are valid for command
115 for (it = parameters.begin(); it != parameters.end(); it++) {
116 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
119 //initialize outputTypes
120 vector<string> tempOutNames;
121 outputTypes["flow"] = tempOutNames;
122 outputTypes["fasta"] = tempOutNames;
123 outputTypes["file"] = tempOutNames;
125 //if the user changes the input directory command factory will send this info to us in the output parameter
126 string inputDir = validParameter.validFile(parameters, "inputdir", false);
127 if (inputDir == "not found"){ inputDir = ""; }
130 it = parameters.find("flow");
131 //user has given a template file
132 if(it != parameters.end()){
133 path = m->hasPath(it->second);
134 //if the user has not given a path then, add inputdir. else leave path alone.
135 if (path == "") { parameters["flow"] = inputDir + it->second; }
138 it = parameters.find("oligos");
139 //user has given a template file
140 if(it != parameters.end()){
141 path = m->hasPath(it->second);
142 //if the user has not given a path then, add inputdir. else leave path alone.
143 if (path == "") { parameters["oligos"] = inputDir + it->second; }
149 //check for required parameters
150 flowFileName = validParameter.validFile(parameters, "flow", true);
151 if (flowFileName == "not found") {
152 flowFileName = m->getFlowFile();
153 if (flowFileName != "") { m->mothurOut("Using " + flowFileName + " as input file for the flow parameter."); m->mothurOutEndLine(); }
155 m->mothurOut("No valid current flow file. You must provide a flow file."); m->mothurOutEndLine();
158 }else if (flowFileName == "not open") { flowFileName = ""; abort = true; }
160 //if the user changes the output directory command factory will send this info to us in the output parameter
161 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
163 outputDir += m->hasPath(flowFileName); //if user entered a file with a path then preserve it
167 //check for optional parameter and set defaults
168 // ...at some point should added some additional type checking...
171 temp = validParameter.validFile(parameters, "minflows", false); if (temp == "not found") { temp = "450"; }
172 m->mothurConvert(temp, minFlows);
174 temp = validParameter.validFile(parameters, "maxflows", false); if (temp == "not found") { temp = "450"; }
175 m->mothurConvert(temp, maxFlows);
178 temp = validParameter.validFile(parameters, "oligos", true);
179 if (temp == "not found") { oligoFileName = ""; }
180 else if(temp == "not open") { abort = true; }
181 else { oligoFileName = temp; m->setOligosFile(oligoFileName); }
183 temp = validParameter.validFile(parameters, "fasta", false); if (temp == "not found"){ fasta = 0; }
184 else if(m->isTrue(temp)) { fasta = 1; }
186 temp = validParameter.validFile(parameters, "maxhomop", false); if (temp == "not found"){ temp = "9"; }
187 m->mothurConvert(temp, maxHomoP);
189 temp = validParameter.validFile(parameters, "signal", false); if (temp == "not found"){ temp = "0.50"; }
190 m->mothurConvert(temp, signal);
192 temp = validParameter.validFile(parameters, "noise", false); if (temp == "not found"){ temp = "0.70"; }
193 m->mothurConvert(temp, noise);
195 temp = validParameter.validFile(parameters, "bdiffs", false); if (temp == "not found"){ temp = "0"; }
196 m->mothurConvert(temp, bdiffs);
198 temp = validParameter.validFile(parameters, "pdiffs", false); if (temp == "not found"){ temp = "0"; }
199 m->mothurConvert(temp, pdiffs);
201 temp = validParameter.validFile(parameters, "ldiffs", false); if (temp == "not found") { temp = "0"; }
202 m->mothurConvert(temp, ldiffs);
204 temp = validParameter.validFile(parameters, "sdiffs", false); if (temp == "not found") { temp = "0"; }
205 m->mothurConvert(temp, sdiffs);
207 temp = validParameter.validFile(parameters, "tdiffs", false); if (temp == "not found") { int tempTotal = pdiffs + bdiffs + ldiffs + sdiffs; temp = toString(tempTotal); }
208 m->mothurConvert(temp, tdiffs);
210 if(tdiffs == 0){ tdiffs = bdiffs + pdiffs + ldiffs + sdiffs; }
213 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
214 m->setProcessors(temp);
215 m->mothurConvert(temp, processors);
217 flowOrder = validParameter.validFile(parameters, "order", false);
218 if (flowOrder == "not found"){ flowOrder = "TACG"; }
219 else if(flowOrder.length() != 4){
220 m->mothurOut("The value of the order option must be four bases long\n");
223 if(oligoFileName == "") { allFiles = 0; }
224 else { allFiles = 1; }
233 catch(exception& e) {
234 m->errorOut(e, "TrimFlowsCommand", "TrimFlowsCommand");
239 //***************************************************************************************************************
241 int TrimFlowsCommand::execute(){
244 if (abort == true) { if (calledHelp) { return 0; } return 2; }
246 map<string, string> variables;
247 variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(flowFileName));
248 string fastaFileName = getOutputFileName("fasta",variables);
249 if(fasta){ outputNames.push_back(fastaFileName); outputTypes["fasta"].push_back(fastaFileName); }
251 variables["[tag]"] = "trim";
252 string trimFlowFileName = getOutputFileName("flow",variables);
253 outputNames.push_back(trimFlowFileName); outputTypes["flow"].push_back(trimFlowFileName);
255 variables["[tag]"] = "scrap";
256 string scrapFlowFileName = getOutputFileName("flow",variables);
257 outputNames.push_back(scrapFlowFileName); outputTypes["flow"].push_back(scrapFlowFileName);
261 vector<unsigned long long> flowFilePos;
262 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
263 flowFilePos = getFlowFileBreaks();
264 for (int i = 0; i < (flowFilePos.size()-1); i++) {
265 lines.push_back(new linePair(flowFilePos[i], flowFilePos[(i+1)]));
268 ifstream in; m->openInputFile(flowFileName, in); in >> numFlows; in.close();
269 ///////////////////////////////////////// until I fix multiple processors for windows //////////////////
271 ///////////////////////////////////////// until I fix multiple processors for windows //////////////////
272 if (processors == 1) {
273 lines.push_back(new linePair(0, 1000));
276 flowFilePos = m->setFilePosEachLine(flowFileName, numFlowLines);
277 flowFilePos.erase(flowFilePos.begin() + 1); numFlowLines--;
279 //figure out how many sequences you have to process
280 int numSeqsPerProcessor = numFlowLines / processors;
281 cout << numSeqsPerProcessor << '\t' << numFlowLines << endl;
282 for (int i = 0; i < processors; i++) {
283 int startIndex = i * numSeqsPerProcessor;
284 if(i == (processors - 1)){ numSeqsPerProcessor = numFlowLines - i * numSeqsPerProcessor; }
285 lines.push_back(new linePair(flowFilePos[startIndex], numSeqsPerProcessor));
286 cout << flowFilePos[startIndex] << '\t' << numSeqsPerProcessor << endl;
291 vector<vector<string> > barcodePrimerComboFileNames;
292 if(oligoFileName != ""){
293 getOligos(barcodePrimerComboFileNames);
297 driverCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames, lines[0]);
299 createProcessesCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames);
302 if (m->control_pressed) { return 0; }
304 string flowFilesFileName;
308 set<string> namesAlreadyProcessed;
309 variables["[tag]"] = "";
310 flowFilesFileName = getOutputFileName("file",variables);
311 m->openOutputFile(flowFilesFileName, output);
313 for(int i=0;i<barcodePrimerComboFileNames.size();i++){
314 for(int j=0;j<barcodePrimerComboFileNames[0].size();j++){
315 if (namesAlreadyProcessed.count(barcodePrimerComboFileNames[i][j]) == 0) {
317 unsigned long long size;
319 //get num bytes in file
320 pFile = fopen (barcodePrimerComboFileNames[i][j].c_str(),"rb");
321 if (pFile==NULL) perror ("Error opening file");
323 fseek (pFile, 0, SEEK_END);
329 m->mothurRemove(barcodePrimerComboFileNames[i][j]);
332 output << m->getFullPathName(barcodePrimerComboFileNames[i][j]) << endl;
333 outputNames.push_back(barcodePrimerComboFileNames[i][j]);
334 outputTypes["flow"].push_back(barcodePrimerComboFileNames[i][j]);
336 namesAlreadyProcessed.insert(barcodePrimerComboFileNames[i][j]);
343 variables["[tag]"] = "";
344 flowFilesFileName = getOutputFileName("file",variables);
345 m->openOutputFile(flowFilesFileName, output);
347 output << m->getFullPathName(trimFlowFileName) << endl;
351 outputTypes["file"].push_back(flowFilesFileName);
352 outputNames.push_back(flowFilesFileName);
354 // set fasta file as new current fastafile
355 // string current = "";
356 // itTypes = outputTypes.find("fasta");
357 // if (itTypes != outputTypes.end()) {
358 // if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); }
361 m->mothurOutEndLine();
362 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
363 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
364 m->mothurOutEndLine();
368 catch(exception& e) {
369 m->errorOut(e, "TrimSeqsCommand", "execute");
374 //***************************************************************************************************************
376 int TrimFlowsCommand::driverCreateTrim(string flowFileName, string trimFlowFileName, string scrapFlowFileName, string fastaFileName, vector<vector<string> > thisBarcodePrimerComboFileNames, linePair* line){
379 ofstream trimFlowFile;
380 m->openOutputFile(trimFlowFileName, trimFlowFile);
381 trimFlowFile.setf(ios::fixed, ios::floatfield); trimFlowFile.setf(ios::showpoint);
383 ofstream scrapFlowFile;
384 m->openOutputFile(scrapFlowFileName, scrapFlowFile);
385 scrapFlowFile.setf(ios::fixed, ios::floatfield); scrapFlowFile.setf(ios::showpoint);
388 if(fasta){ m->openOutputFile(fastaFileName, fastaFile); }
391 m->openInputFile(flowFileName, flowFile);
393 flowFile.seekg(line->start);
395 if(line->start == 0){
396 flowFile >> numFlows; m->gobble(flowFile);
397 scrapFlowFile << maxFlows << endl;
398 trimFlowFile << maxFlows << endl;
400 for(int i=0;i<thisBarcodePrimerComboFileNames.size();i++){
401 for(int j=0;j<thisBarcodePrimerComboFileNames[0].size();j++){
403 m->openOutputFile(thisBarcodePrimerComboFileNames[i][j], temp);
404 temp << maxFlows << endl;
411 FlowData flowData(numFlows, signal, noise, maxHomoP, flowOrder);
412 //cout << " driver flowdata address " << &flowData << &flowFile << endl;
416 TrimOligos trimOligos(pdiffs, bdiffs, ldiffs, sdiffs, primers, barcodes, revPrimer, linker, spacer);
420 if (m->control_pressed) { break; }
423 int currentSeqDiffs = 0;
424 string trashCode = "";
426 flowData.getNext(flowFile);
427 flowData.capFlows(maxFlows);
429 Sequence currSeq = flowData.getSequence();
430 if(!flowData.hasMinFlows(minFlows)){ //screen to see if sequence is of a minimum number of flows
436 int barcodeIndex = 0;
439 success = trimOligos.stripLinker(currSeq);
440 if(success > ldiffs) { trashCode += 'k'; }
441 else{ currentSeqDiffs += success; }
445 if (m->debug) { m->mothurOut("[DEBUG]: " + currSeq.getName() + " " + currSeq.getUnaligned() + "\n"); }
447 if(barcodes.size() != 0){
448 success = trimOligos.stripBarcode(currSeq, barcodeIndex);
449 if(success > bdiffs) { trashCode += 'b'; }
450 else{ currentSeqDiffs += success; }
454 success = trimOligos.stripSpacer(currSeq);
455 if(success > sdiffs) { trashCode += 's'; }
456 else{ currentSeqDiffs += success; }
460 if(numFPrimers != 0){
461 success = trimOligos.stripForward(currSeq, primerIndex);
462 if(success > pdiffs) { trashCode += 'f'; }
463 else{ currentSeqDiffs += success; }
466 if (currentSeqDiffs > tdiffs) { trashCode += 't'; }
468 if(numRPrimers != 0){
469 success = trimOligos.stripReverse(currSeq);
470 if(!success) { trashCode += 'r'; }
473 if(trashCode.length() == 0){
475 flowData.printFlows(trimFlowFile);
477 if(fasta) { currSeq.printSequence(fastaFile); }
481 m->openOutputFileAppend(thisBarcodePrimerComboFileNames[barcodeIndex][primerIndex], output);
482 output.setf(ios::fixed, ios::floatfield); trimFlowFile.setf(ios::showpoint);
484 flowData.printFlows(output);
489 flowData.printFlows(scrapFlowFile, trashCode);
493 //cout << "driver" << '\t' << currSeq.getName() << endl;
495 if((count) % 10000 == 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
497 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
498 unsigned long long pos = flowFile.tellg();
500 if ((pos == -1) || (pos >= line->end)) { break; }
502 if (flowFile.eof()) { break; }
507 if((count) % 10000 != 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); }
509 trimFlowFile.close();
510 scrapFlowFile.close();
512 if(fasta){ fastaFile.close(); }
516 catch(exception& e) {
517 m->errorOut(e, "TrimSeqsCommand", "driverCreateTrim");
522 //***************************************************************************************************************
524 void TrimFlowsCommand::getOligos(vector<vector<string> >& outFlowFileNames){
527 m->openInputFile(oligoFileName, oligosFile);
529 string type, oligo, group;
532 int indexBarcode = 0;
534 while(!oligosFile.eof()){
536 oligosFile >> type; m->gobble(oligosFile); //get the first column value of the row - is it a comment or a feature we are interested in?
538 if(type[0] == '#'){ //igore the line because there's a comment
539 while (!oligosFile.eof()) { char c = oligosFile.get(); if (c == 10 || c == 13){ break; } } // get rest of line if there's any crap there
541 else{ //there's a feature we're interested in
543 for(int i=0;i<type.length();i++){ type[i] = toupper(type[i]); } //make type case insensitive
545 oligosFile >> oligo; //get the DNA sequence for the feature
547 for(int i=0;i<oligo.length();i++){ //make type case insensitive and change any U's to T's
548 oligo[i] = toupper(oligo[i]);
549 if(oligo[i] == 'U') { oligo[i] = 'T'; }
552 if(type == "FORWARD"){ //if the feature is a forward primer...
555 while (!oligosFile.eof()) { // get rest of line in case there is a primer name = will have the name of the primer
556 char c = oligosFile.get();
557 if (c == 10 || c == 13){ break; }
558 else if (c == 32 || c == 9){;} //space or tab
562 //have we seen this primer already?
563 map<string, int>::iterator itPrimer = primers.find(oligo);
564 if (itPrimer != primers.end()) { m->mothurOut("primer " + oligo + " is in your oligos file already."); m->mothurOutEndLine(); }
566 primers[oligo]=indexPrimer; indexPrimer++;
567 primerNameVector.push_back(group);
570 else if(type == "REVERSE"){
571 string oligoRC = reverseOligo(oligo);
572 revPrimer.push_back(oligoRC);
574 else if(type == "BARCODE"){
577 //check for repeat barcodes
578 map<string, int>::iterator itBar = barcodes.find(oligo);
579 if (itBar != barcodes.end()) { m->mothurOut("barcode " + oligo + " is in your oligos file already."); m->mothurOutEndLine(); }
581 barcodes[oligo]=indexBarcode; indexBarcode++;
582 barcodeNameVector.push_back(group);
583 }else if(type == "LINKER"){
584 linker.push_back(oligo);
585 }else if(type == "SPACER"){
586 spacer.push_back(oligo);
589 m->mothurOut(type + " is not recognized as a valid type. Choices are forward, reverse, and barcode. Ignoring " + oligo + "."); m->mothurOutEndLine();
593 m->gobble(oligosFile);
597 if(barcodeNameVector.size() == 0 && primerNameVector[0] == ""){ allFiles = 0; }
599 //add in potential combos
600 if(barcodeNameVector.size() == 0){
602 barcodeNameVector.push_back("");
605 if(primerNameVector.size() == 0){
607 primerNameVector.push_back("");
611 outFlowFileNames.resize(barcodeNameVector.size());
612 for(int i=0;i<outFlowFileNames.size();i++){
613 outFlowFileNames[i].assign(primerNameVector.size(), "");
618 for(map<string, int>::iterator itBar = barcodes.begin();itBar != barcodes.end();itBar++){
619 for(map<string, int>::iterator itPrimer = primers.begin();itPrimer != primers.end(); itPrimer++){
621 string primerName = primerNameVector[itPrimer->second];
622 string barcodeName = barcodeNameVector[itBar->second];
624 string comboGroupName = "";
625 string fileName = "";
627 map<string, string> variables;
628 variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(flowFileName));
630 if(primerName == ""){
631 comboGroupName = barcodeNameVector[itBar->second];
632 variables["[tag]"] = comboGroupName;
633 fileName = getOutputFileName("flow", variables);
636 if(barcodeName == ""){
637 comboGroupName = primerNameVector[itPrimer->second];
640 comboGroupName = barcodeNameVector[itBar->second] + "." + primerNameVector[itPrimer->second];
642 variables["[tag]"] = comboGroupName;
643 fileName = getOutputFileName("flow", variables);
646 outFlowFileNames[itBar->second][itPrimer->second] = fileName;
649 m->openOutputFile(fileName, temp);
655 numFPrimers = primers.size();
656 numRPrimers = revPrimer.size();
657 numLinkers = linker.size();
658 numSpacers = spacer.size();
661 catch(exception& e) {
662 m->errorOut(e, "TrimSeqsCommand", "getOligos");
666 //********************************************************************/
667 string TrimFlowsCommand::reverseOligo(string oligo){
671 for(int i=oligo.length()-1;i>=0;i--){
673 if(oligo[i] == 'A') { reverse += 'T'; }
674 else if(oligo[i] == 'T'){ reverse += 'A'; }
675 else if(oligo[i] == 'U'){ reverse += 'A'; }
677 else if(oligo[i] == 'G'){ reverse += 'C'; }
678 else if(oligo[i] == 'C'){ reverse += 'G'; }
680 else if(oligo[i] == 'R'){ reverse += 'Y'; }
681 else if(oligo[i] == 'Y'){ reverse += 'R'; }
683 else if(oligo[i] == 'M'){ reverse += 'K'; }
684 else if(oligo[i] == 'K'){ reverse += 'M'; }
686 else if(oligo[i] == 'W'){ reverse += 'W'; }
687 else if(oligo[i] == 'S'){ reverse += 'S'; }
689 else if(oligo[i] == 'B'){ reverse += 'V'; }
690 else if(oligo[i] == 'V'){ reverse += 'B'; }
692 else if(oligo[i] == 'D'){ reverse += 'H'; }
693 else if(oligo[i] == 'H'){ reverse += 'D'; }
695 else { reverse += 'N'; }
701 catch(exception& e) {
702 m->errorOut(e, "TrimFlowsCommand", "reverseOligo");
707 /**************************************************************************************************/
708 vector<unsigned long long> TrimFlowsCommand::getFlowFileBreaks() {
712 vector<unsigned long long> filePos;
713 filePos.push_back(0);
716 unsigned long long size;
718 //get num bytes in file
719 pFile = fopen (flowFileName.c_str(),"rb");
720 if (pFile==NULL) perror ("Error opening file");
722 fseek (pFile, 0, SEEK_END);
727 //estimate file breaks
728 unsigned long long chunkSize = 0;
729 chunkSize = size / processors;
731 //file too small to divide by processors
732 if (chunkSize == 0) { processors = 1; filePos.push_back(size); return filePos; }
734 //for each process seekg to closest file break and search for next '>' char. make that the filebreak
735 for (int i = 0; i < processors; i++) {
736 unsigned long long spot = (i+1) * chunkSize;
739 m->openInputFile(flowFileName, in);
742 string dummy = m->getline(in);
744 //there was not another sequence before the end of the file
745 unsigned long long sanityPos = in.tellg();
747 // if (sanityPos == -1) { break; }
748 // else { filePos.push_back(newSpot); }
749 if (sanityPos == -1) { break; }
750 else { filePos.push_back(sanityPos); }
756 filePos.push_back(size);
758 //sanity check filePos
759 for (int i = 0; i < (filePos.size()-1); i++) {
760 if (filePos[(i+1)] <= filePos[i]) { filePos.erase(filePos.begin()+(i+1)); i--; }
764 m->openInputFile(flowFileName, in);
767 //unsigned long long spot = in.tellg();
771 processors = (filePos.size() - 1);
775 catch(exception& e) {
776 m->errorOut(e, "TrimSeqsCommand", "getFlowFileBreaks");
781 /**************************************************************************************************/
783 int TrimFlowsCommand::createProcessesCreateTrim(string flowFileName, string trimFlowFileName, string scrapFlowFileName, string fastaFileName, vector<vector<string> > barcodePrimerComboFileNames){
789 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
792 //loop through and create all the processes you want
793 while (process != processors) {
797 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
801 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
803 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
804 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
805 tempBarcodePrimerComboFileNames[i][j] += toString(getpid()) + ".temp";
807 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
813 driverCreateTrim(flowFileName,
814 (trimFlowFileName + toString(getpid()) + ".temp"),
815 (scrapFlowFileName + toString(getpid()) + ".temp"),
816 (fastaFileName + toString(getpid()) + ".temp"),
817 tempBarcodePrimerComboFileNames, lines[process]);
821 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
822 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
829 m->openOutputFile(trimFlowFileName, temp);
832 m->openOutputFile(scrapFlowFileName, temp);
836 m->openOutputFile(fastaFileName, temp);
840 driverCreateTrim(flowFileName, trimFlowFileName, scrapFlowFileName, fastaFileName, barcodePrimerComboFileNames, lines[0]);
842 //force parent to wait until all the processes are done
843 for (int i=0;i<processIDS.size();i++) {
844 int temp = processIDS[i];
848 //////////////////////////////////////////////////////////////////////////////////////////////////////
849 //Windows version shared memory, so be careful when passing variables through the trimFlowData struct.
850 //Above fork() will clone, so memory is separate, but that's not the case with windows,
851 //////////////////////////////////////////////////////////////////////////////////////////////////////
853 vector<trimFlowData*> pDataArray;
854 DWORD dwThreadIdArray[processors-1];
855 HANDLE hThreadArray[processors-1];
857 //Create processor worker threads.
858 for( int i=0; i<processors-1; i++ ){
859 // Allocate memory for thread data.
860 string extension = "";
861 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
863 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
865 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
866 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
867 tempBarcodePrimerComboFileNames[i][j] += extension;
869 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
876 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);
877 pDataArray.push_back(tempflow);
879 //MyTrimFlowThreadFunction is in header. It must be global or static to work with the threads.
880 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
881 hThreadArray[i] = CreateThread(NULL, 0, MyTrimFlowThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
884 //using the main process as a worker saves time and memory
886 m->openOutputFile(trimFlowFileName, temp);
889 m->openOutputFile(scrapFlowFileName, temp);
893 m->openOutputFile(fastaFileName, temp);
897 vector<vector<string> > tempBarcodePrimerComboFileNames = barcodePrimerComboFileNames;
899 for(int i=0;i<tempBarcodePrimerComboFileNames.size();i++){
900 for(int j=0;j<tempBarcodePrimerComboFileNames[0].size();j++){
901 tempBarcodePrimerComboFileNames[i][j] += toString(processors-1) + ".temp";
903 m->openOutputFile(tempBarcodePrimerComboFileNames[i][j], temp);
910 //do my part - do last piece because windows is looking for eof
911 int num = driverCreateTrim(flowFileName, (trimFlowFileName + toString(processors-1) + ".temp"), (scrapFlowFileName + toString(processors-1) + ".temp"), (fastaFileName + toString(processors-1) + ".temp"), tempBarcodePrimerComboFileNames, lines[processors-1]);
912 processIDS.push_back((processors-1));
914 //Wait until all threads have terminated.
915 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
917 //Close all thread handles and free memory allocations.
918 for(int i=0; i < pDataArray.size(); i++){
919 num += pDataArray[i]->count;
920 CloseHandle(hThreadArray[i]);
921 delete pDataArray[i];
927 m->mothurOutEndLine();
928 for(int i=0;i<processIDS.size();i++){
930 m->mothurOut("Appending files from process " + toString(processIDS[i])); m->mothurOutEndLine();
932 m->appendFiles((trimFlowFileName + toString(processIDS[i]) + ".temp"), trimFlowFileName);
933 m->mothurRemove((trimFlowFileName + toString(processIDS[i]) + ".temp"));
934 // m->mothurOut("\tDone with trim.flow file"); m->mothurOutEndLine();
936 m->appendFiles((scrapFlowFileName + toString(processIDS[i]) + ".temp"), scrapFlowFileName);
937 m->mothurRemove((scrapFlowFileName + toString(processIDS[i]) + ".temp"));
938 // m->mothurOut("\tDone with scrap.flow file"); m->mothurOutEndLine();
941 m->appendFiles((fastaFileName + toString(processIDS[i]) + ".temp"), fastaFileName);
942 m->mothurRemove((fastaFileName + toString(processIDS[i]) + ".temp"));
943 // m->mothurOut("\tDone with flow.fasta file"); m->mothurOutEndLine();
946 for (int j = 0; j < barcodePrimerComboFileNames.size(); j++) {
947 for (int k = 0; k < barcodePrimerComboFileNames[0].size(); k++) {
948 m->appendFiles((barcodePrimerComboFileNames[j][k] + toString(processIDS[i]) + ".temp"), barcodePrimerComboFileNames[j][k]);
949 m->mothurRemove((barcodePrimerComboFileNames[j][k] + toString(processIDS[i]) + ".temp"));
958 catch(exception& e) {
959 m->errorOut(e, "TrimFlowsCommand", "createProcessesCreateTrim");
964 //***************************************************************************************************************