1 #ifndef SCREENSEQSCOMMAND_H
2 #define SCREENSEQSCOMMAND_H
8 * Created by Pat Schloss on 6/3/09.
9 * Copyright 2009 Patrick D. Schloss. All rights reserved.
13 #include "command.hpp"
14 #include "sequence.hpp"
16 class ScreenSeqsCommand : public Command {
19 ScreenSeqsCommand(string);
21 ~ScreenSeqsCommand() {}
23 vector<string> setParameters();
24 string getCommandName() { return "screen.seqs"; }
25 string getCommandCategory() { return "Sequence Processing"; }
26 string getHelpString();
27 string getCitation() { return "http://www.mothur.org/wiki/Screen.seqs"; }
28 string getDescription() { return "enables you to keep sequences that fulfill certain user defined criteria"; }
31 void help() { m->mothurOut(getHelpString()); }
37 unsigned long long start;
38 unsigned long long end;
39 linePair(unsigned long long i, unsigned long long j) : start(i), end(j) {}
42 vector<linePair> lines;
44 int screenNameGroupFile(set<string>);
45 int screenGroupFile(set<string>);
46 int screenAlignReport(set<string>);
47 int screenQual(set<string>);
48 int screenTaxonomy(set<string>);
50 int driver(linePair, string, string, string, set<string>&);
51 int createProcesses(string, string, string, set<string>&);
54 int driverMPI(int, int, MPI_File&, MPI_File&, MPI_File&, vector<unsigned long long>&, set<string>&);
58 string fastafile, namefile, groupfile, alignreport, outputDir, qualfile, taxonomy;
59 int startPos, endPos, maxAmbig, maxHomoP, minLength, maxLength, processors, criteria;
60 vector<string> outputNames;
61 vector<string> optimize;
62 map<string, int> nameMap;
65 int getSummary(vector<unsigned long long>&);
66 int createProcessesCreateSummary(vector<int>&, vector<int>&, vector<int>&, vector<int>&, vector<int>&, string);
67 int driverCreateSummary(vector<int>&, vector<int>&, vector<int>&, vector<int>&, vector<int>&, string, linePair);
70 /**************************************************************************************************/
71 //custom data structure for threads to use.
72 // This is passed by void pointer so it can be any data type
73 // that can be passed using a single void pointer (LPVOID).
75 vector<int> startPosition;
76 vector<int> endPosition;
77 vector<int> seqLength;
78 vector<int> ambigBases;
79 vector<int> longHomoPolymer;
80 string filename, namefile;
81 unsigned long long start;
82 unsigned long long end;
85 map<string, int> nameMap;
89 sumData(string f, MothurOut* mout, unsigned long long st, unsigned long long en, string nf, map<string, int> nam) {
99 /**************************************************************************************************/
100 //custom data structure for threads to use.
101 // This is passed by void pointer so it can be any data type
102 // that can be passed using a single void pointer (LPVOID).
103 struct sumScreenData {
104 int startPos, endPos, maxAmbig, maxHomoP, minLength, maxLength;
105 unsigned long long start;
106 unsigned long long end;
109 string goodFName, badAccnosFName, filename;
110 set<string>* badSeqNames;
114 sumScreenData(int s, int e, int a, int h, int minl, int maxl, string f, MothurOut* mout, unsigned long long st, unsigned long long en, string gf, string bf, set<string>* bn) {
133 /**************************************************************************************************/
134 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
136 static DWORD WINAPI MySumThreadFunction(LPVOID lpParam){
138 pDataArray = (sumData*)lpParam;
142 pDataArray->m->openInputFile(pDataArray->filename, in);
144 //print header if you are process 0
145 if ((pDataArray->start == 0) || (pDataArray->start == 1)) {
147 }else { //this accounts for the difference in line endings.
148 in.seekg(pDataArray->start-1); pDataArray->m->gobble(in);
151 pDataArray->count = pDataArray->end;
152 for(int i = 0; i < pDataArray->end; i++){ //end is the number of sequences to process
154 if (pDataArray->m->control_pressed) { in.close(); pDataArray->count = 1; return 1; }
156 Sequence current(in); pDataArray->m->gobble(in);
158 if (current.getName() != "") {
161 if (pDataArray->namefile != "") {
162 //make sure this sequence is in the namefile, else error
163 map<string, int>::iterator it = pDataArray->nameMap.find(current.getName());
165 if (it == pDataArray->nameMap.end()) { pDataArray->m->mothurOut("[ERROR]: " + current.getName() + " is not in your namefile, please correct."); pDataArray->m->mothurOutEndLine(); pDataArray->m->control_pressed = true; }
166 else { num = it->second; }
169 //for each sequence this sequence represents
170 for (int i = 0; i < num; i++) {
171 pDataArray->startPosition.push_back(current.getStartPos());
172 pDataArray->endPosition.push_back(current.getEndPos());
173 pDataArray->seqLength.push_back(current.getNumBases());
174 pDataArray->ambigBases.push_back(current.getAmbigBases());
175 pDataArray->longHomoPolymer.push_back(current.getLongHomoPolymer());
185 catch(exception& e) {
186 pDataArray->m->errorOut(e, "ScreenSeqsCommand", "MySumThreadFunction");
191 /**************************************************************************************************/
193 static DWORD WINAPI MySumScreenThreadFunction(LPVOID lpParam){
194 sumScreenData* pDataArray;
195 pDataArray = (sumScreenData*)lpParam;
200 pDataArray->m->openOutputFile(pDataArray->goodFName, goodFile);
202 ofstream badAccnosFile;
203 pDataArray->m->openOutputFile(pDataArray->badAccnosFName, badAccnosFile);
206 pDataArray->m->openInputFile(pDataArray->filename, in);
208 //print header if you are process 0
209 if ((pDataArray->start == 0) || (pDataArray->start == 1)) {
211 }else { //this accounts for the difference in line endings.
212 in.seekg(pDataArray->start-1); pDataArray->m->gobble(in);
215 pDataArray->count = pDataArray->end;
216 for(int i = 0; i < pDataArray->end; i++){ //end is the number of sequences to process
218 if (pDataArray->m->control_pressed) { in.close(); badAccnosFile.close(); goodFile.close(); pDataArray->count = 1; return 1; }
220 Sequence currSeq(in); pDataArray->m->gobble(in);
222 if (currSeq.getName() != "") {
223 bool goodSeq = 1; // innocent until proven guilty
224 if(goodSeq == 1 && pDataArray->startPos != -1 && pDataArray->startPos < currSeq.getStartPos()) { goodSeq = 0; }
225 if(goodSeq == 1 && pDataArray->endPos != -1 && pDataArray->endPos > currSeq.getEndPos()) { goodSeq = 0; }
226 if(goodSeq == 1 && pDataArray->maxAmbig != -1 && pDataArray->maxAmbig < currSeq.getAmbigBases()) { goodSeq = 0; }
227 if(goodSeq == 1 && pDataArray->maxHomoP != -1 && pDataArray->maxHomoP < currSeq.getLongHomoPolymer()) { goodSeq = 0; }
228 if(goodSeq == 1 && pDataArray->minLength != -1 && pDataArray->minLength > currSeq.getNumBases()) { goodSeq = 0; }
229 if(goodSeq == 1 && pDataArray->maxLength != -1 && pDataArray->maxLength < currSeq.getNumBases()) { goodSeq = 0; }
232 currSeq.printSequence(goodFile);
235 badAccnosFile << currSeq.getName() << endl;
236 pDataArray->badSeqNames->insert(currSeq.getName());
241 if((i+1) % 100 == 0){ pDataArray->m->mothurOut("Processing sequence: " + toString(i+1)); pDataArray->m->mothurOutEndLine(); }
244 if((pDataArray->count) % 100 != 0){ pDataArray->m->mothurOut("Processing sequence: " + toString(pDataArray->count)); pDataArray->m->mothurOutEndLine(); }
250 badAccnosFile.close();
255 catch(exception& e) {
256 pDataArray->m->errorOut(e, "ScreenSeqsCommand", "MySumScreenThreadFunction");
263 /**************************************************************************************************/