X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=removelineagecommand.cpp;h=5f5435f8679a9b65bb4799dae9dff30555818d3c;hb=f06fdb807822f8e06db003ed809c87250905cfc8;hp=22453a46078a7e99b33c07351722913fbb88f854;hpb=ec1b5bc7460ac8ad40f54f0729900d9ed98e7292;p=mothur.git diff --git a/removelineagecommand.cpp b/removelineagecommand.cpp index 22453a4..5f5435f 100644 --- a/removelineagecommand.cpp +++ b/removelineagecommand.cpp @@ -211,7 +211,7 @@ RemoveLineageCommand::RemoveLineageCommand(string option) { if (taxons[0] == '\'') { taxons = taxons.substr(1); } if (taxons[(taxons.length()-1)] == '\'') { taxons = taxons.substr(0, (taxons.length()-1)); } } - + m->splitAtChar(taxons, listOfTaxons, '-'); if ((fastafile == "") && (namefile == "") && (groupfile == "") && (alignfile == "") && (listfile == "") && (taxfile == "")) { m->mothurOut("You must provide one of the following: fasta, name, group, alignreport, taxonomy or listfile."); m->mothurOutEndLine(); abort = true; } @@ -541,14 +541,18 @@ int RemoveLineageCommand::readTax(){ bool wroteSomething = false; - bool taxonsHasConfidence = false; - vector< map > searchTaxons; - string noConfidenceTaxons = taxons; - int hasConPos = taxons.find_first_of('('); - if (hasConPos != string::npos) { - taxonsHasConfidence = true; - searchTaxons = getTaxons(taxons); - noConfidenceTaxons = removeConfidences(taxons); + vector taxonsHasConfidence; taxonsHasConfidence.resize(listOfTaxons.size(), false); + vector< vector< map > > searchTaxons; searchTaxons.resize(listOfTaxons.size()); + vector noConfidenceTaxons; noConfidenceTaxons.resize(listOfTaxons.size(), ""); + + for (int i = 0; i < listOfTaxons.size(); i++) { + noConfidenceTaxons[i] = listOfTaxons[i]; + int hasConPos = listOfTaxons[i].find_first_of('('); + if (hasConPos != string::npos) { + taxonsHasConfidence[i] = true; + searchTaxons[i] = getTaxons(listOfTaxons[i]); + noConfidenceTaxons[i] = removeConfidences(listOfTaxons[i]); + } } @@ -559,102 +563,107 @@ int RemoveLineageCommand::readTax(){ in >> name; //read from first column in >> tax; //read from second column - string newtax = tax; + bool remove = false; - //if the users file contains confidence scores we want to ignore them when searching for the taxons, unless the taxon has them - if (!taxonsHasConfidence) { - - int hasConfidences = tax.find_first_of('('); - if (hasConfidences != string::npos) { - newtax = removeConfidences(tax); - } - - int pos = newtax.find(taxons); + for (int j = 0; j < listOfTaxons.size(); j++) { + string newtax = tax; - if (pos == string::npos) { - wroteSomething = true; - out << name << '\t' << tax << endl; - }else{ //this sequence contains the taxon the user wants to remove - names.insert(name); - } - - }else{//if taxons has them and you don't them remove taxons - int hasConfidences = tax.find_first_of('('); - if (hasConfidences == string::npos) { - - int pos = newtax.find(noConfidenceTaxons); + //if the users file contains confidence scores we want to ignore them when searching for the taxons, unless the taxon has them + if (!taxonsHasConfidence[j]) { - if (pos == string::npos) { - wroteSomething = true; - out << name << '\t' << tax << endl; - }else{ //this sequence contains the taxon the user wants to remove - names.insert(name); - } - }else { //both have confidences so we want to make sure the users confidences are greater then or equal to the taxons - //first remove confidences from both and see if the taxonomy exists - - string noNewTax = tax; int hasConfidences = tax.find_first_of('('); if (hasConfidences != string::npos) { - noNewTax = removeConfidences(tax); + newtax = removeConfidences(tax); } - int pos = noNewTax.find(noConfidenceTaxons); + int pos = newtax.find(noConfidenceTaxons[j]); - if (pos != string::npos) { //if yes, then are the confidences okay + if (pos == string::npos) { + //wroteSomething = true; + //out << name << '\t' << tax << endl; + }else{ //this sequence contains the taxon the user wants to remove + names.insert(name); + remove=true; break; + } + + }else{//if taxons has them and you don't them remove taxons + int hasConfidences = tax.find_first_of('('); + if (hasConfidences == string::npos) { - bool remove = false; - vector< map > usersTaxon = getTaxons(newtax); + int pos = newtax.find(noConfidenceTaxons[j]); - //the usersTaxon is most likely longer than the searchTaxons, and searchTaxon[0] may relate to userTaxon[4] - //we want to "line them up", so we will find the the index where the searchstring starts - int index = 0; - for (int i = 0; i < usersTaxon.size(); i++) { + if (pos == string::npos) { + //wroteSomething = true; + //out << name << '\t' << tax << endl; + }else{ //this sequence contains the taxon the user wants to remove + names.insert(name); + remove=true; break; + } + }else { //both have confidences so we want to make sure the users confidences are greater then or equal to the taxons + //first remove confidences from both and see if the taxonomy exists + + string noNewTax = tax; + int hasConfidences = tax.find_first_of('('); + if (hasConfidences != string::npos) { + noNewTax = removeConfidences(tax); + } + + int pos = noNewTax.find(noConfidenceTaxons[j]); + + if (pos != string::npos) { //if yes, then are the confidences okay - if (usersTaxon[i].begin()->first == searchTaxons[0].begin()->first) { - index = i; - int spot = 0; - bool goodspot = true; - //is this really the start, or are we dealing with a taxon of the same name? - while ((spot < searchTaxons.size()) && ((i+spot) < usersTaxon.size())) { - if (usersTaxon[i+spot].begin()->first != searchTaxons[spot].begin()->first) { goodspot = false; break; } - else { spot++; } - } + vector< map > usersTaxon = getTaxons(newtax); + + //the usersTaxon is most likely longer than the searchTaxons, and searchTaxon[0] may relate to userTaxon[4] + //we want to "line them up", so we will find the the index where the searchstring starts + int index = 0; + for (int i = 0; i < usersTaxon.size(); i++) { - if (goodspot) { break; } + if (usersTaxon[i].begin()->first == searchTaxons[j][0].begin()->first) { + index = i; + int spot = 0; + bool goodspot = true; + //is this really the start, or are we dealing with a taxon of the same name? + while ((spot < searchTaxons[j].size()) && ((i+spot) < usersTaxon.size())) { + if (usersTaxon[i+spot].begin()->first != searchTaxons[j][spot].begin()->first) { goodspot = false; break; } + else { spot++; } + } + + if (goodspot) { break; } + } } - } - - for (int i = 0; i < searchTaxons.size(); i++) { - if ((i+index) < usersTaxon.size()) { //just in case, should never be false - if (usersTaxon[i+index].begin()->second < searchTaxons[i].begin()->second) { //is the users cutoff less than the search taxons + for (int i = 0; i < searchTaxons[j].size(); i++) { + + if ((i+index) < usersTaxon.size()) { //just in case, should never be false + if (usersTaxon[i+index].begin()->second < searchTaxons[j][i].begin()->second) { //is the users cutoff less than the search taxons + remove = true; + break; + } + }else { remove = true; break; } + } + + //passed the test so remove you + if (remove) { + names.insert(name); + remove=true; break; }else { - remove = true; - break; + //wroteSomething = true; + //out << name << '\t' << tax << endl; } - } - - //passed the test so remove you - if (remove) { - names.insert(name); }else { - wroteSomething = true; - out << name << '\t' << tax << endl; + //wroteSomething = true; + //out << name << '\t' << tax << endl; } - }else { - wroteSomething = true; - out << name << '\t' << tax << endl; } } + } - - - + if (!remove) { wroteSomething = true; out << name << '\t' << tax << endl; } m->gobble(in); } in.close();