#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
#include <sys/wait.h>
+ #include <sys/time.h>
+ #include <sys/resource.h>
#include <unistd.h>
#ifdef USE_READLINE
#else
#include <conio.h> //allows unbuffered screen capture from stdin
#include <direct.h> //get cwd
+ #include <windows.h>
+ #include <psapi.h>
+
#endif
using namespace std;
}
/***********************************************************************/
-inline string getline(ifstream& fileHandle) {
+inline void gobble(istringstream& f){
+
+ char d;
+ while(isspace(d=f.get())) {;}
+ f.putback(d);
+
+}
+
+/***********************************************************************/
+
+inline string getline(istringstream& fileHandle) {
try {
string line = "";
exit(1);
}
}
+/***********************************************************************/
+inline string getline(ifstream& fileHandle) {
+ try {
+
+ string line = "";
+
+ while (!fileHandle.eof()) {
+ //get next character
+ char c = fileHandle.get();
+
+ //are you at the end of the line
+ if ((c == '\n') || (c == '\r') || (c == '\f')){ break; }
+ else { line += c; }
+ }
+
+ return line;
+
+ }
+ catch(exception& e) {
+ cout << "Error in mothur function getline" << endl;
+ exit(1);
+ }
+}
/***********************************************************************/
inline bool isTrue(string f){
return int(dist * precision + 0.5)/float(precision);
+}
+/***********************************************************************/
+
+inline float ceilDist(float dist, int precision){
+
+ return int(ceil(dist * precision))/float(precision);
+
}
/***********************************************************************/
string path = "";
size_t found;
- found=longName.find_last_of("/\\");
+ found=longName.find_last_of("~/\\");
if(found != longName.npos){
path = longName.substr(0, found+1);
string cwd;
//get current working directory
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- if (path.rfind("./") == -1) { return fileName; } //already complete name
- else { newFileName = fileName.substr(fileName.rfind("./")+2); } //save the complete part of the name
- char* cwdpath = new char[1024];
+ if (path.find("~") != -1) { //go to home directory
+ string homeDir = getenv ("HOME");
+ newFileName = homeDir + fileName.substr(fileName.find("~")+1);
+ return newFileName;
+ }else { //find path
+ if (path.rfind("./") == -1) { return fileName; } //already complete name
+ else { newFileName = fileName.substr(fileName.rfind("./")+2); } //save the complete part of the name
+
+ char* cwdpath = new char[1024];
- size_t size;
- cwdpath=getcwd(cwdpath,size);
-
- cwd = cwdpath;
+ size_t size;
+ cwdpath=getcwd(cwdpath,size);
- //rip off first '/'
- string simpleCWD;
- if (cwd.length() > 0) { simpleCWD = cwd.substr(1); }
-
- //break apart the current working directory
- vector<string> dirs;
- while (simpleCWD.find_first_of('/') != -1) {
- string dir = simpleCWD.substr(0,simpleCWD.find_first_of('/'));
- simpleCWD = simpleCWD.substr(simpleCWD.find_first_of('/')+1, simpleCWD.length());
- dirs.push_back(dir);
- }
- //get last one // ex. ../../../filename = /user/work/desktop/filename
- dirs.push_back(simpleCWD); //ex. dirs[0] = user, dirs[1] = work, dirs[2] = desktop
+ cwd = cwdpath;
+
+ //rip off first '/'
+ string simpleCWD;
+ if (cwd.length() > 0) { simpleCWD = cwd.substr(1); }
+
+ //break apart the current working directory
+ vector<string> dirs;
+ while (simpleCWD.find_first_of('/') != -1) {
+ string dir = simpleCWD.substr(0,simpleCWD.find_first_of('/'));
+ simpleCWD = simpleCWD.substr(simpleCWD.find_first_of('/')+1, simpleCWD.length());
+ dirs.push_back(dir);
+ }
+ //get last one // ex. ../../../filename = /user/work/desktop/filename
+ dirs.push_back(simpleCWD); //ex. dirs[0] = user, dirs[1] = work, dirs[2] = desktop
+
+ int index = dirs.size()-1;
- int index = dirs.size()-1;
-
- while((pos = path.rfind("./")) != -1) { //while you don't have a complete path
- if (path[(pos-1)] == '.') { //you want your parent directory ../
- path = path.substr(0, pos-1);
- index--;
- if (index == 0) { break; }
- }else if (path[(pos-1)] == '/') { //you want the current working dir ./
- path = path.substr(0, pos);
- }else if (pos == 1) { break;
- }else { cout << "cannot resolve path for " << fileName << endl; return fileName; }
- }
-
- for (int i = index; i >= 0; i--) {
- newFileName = dirs[i] + "/" + newFileName;
- }
+ while((pos = path.rfind("./")) != -1) { //while you don't have a complete path
+ if (pos == 0) { break; //you are at the end
+ }else if (path[(pos-1)] == '.') { //you want your parent directory ../
+ path = path.substr(0, pos-1);
+ index--;
+ if (index == 0) { break; }
+ }else if (path[(pos-1)] == '/') { //you want the current working dir ./
+ path = path.substr(0, pos);
+ }else if (pos == 1) { break; //you are at the end
+ }else { cout << "cannot resolve path for " << fileName << endl; return fileName; }
+ }
- newFileName = "/" + newFileName;
- return newFileName;
+ for (int i = index; i >= 0; i--) {
+ newFileName = dirs[i] + "/" + newFileName;
+ }
+ newFileName = "/" + newFileName;
+ return newFileName;
+ }
#else
- if (path.rfind(".\\") == -1) { return fileName; } //already complete name
- else { newFileName = fileName.substr(fileName.rfind(".\\")+2); } //save the complete part of the name
-
- char *cwdpath = NULL;
- cwdpath = getcwd(NULL, 0); // or _getcwd
- if ( cwdpath != NULL) { cwd = cwdpath; }
- else { cwd = ""; }
-
- //break apart the current working directory
- vector<string> dirs;
- while (cwd.find_first_of('\\') != -1) {
- string dir = cwd.substr(0,cwd.find_first_of('\\'));
- cwd = cwd.substr(cwd.find_first_of('\\')+1, cwd.length());
- dirs.push_back(dir);
-
- }
- //get last one
- dirs.push_back(cwd); //ex. dirs[0] = user, dirs[1] = work, dirs[2] = desktop
-
- int index = dirs.size()-1;
+ if (path.find("~") != -1) { //go to home directory
+ string homeDir = getenv ("HOMEPATH");
+ newFileName = homeDir + fileName.substr(fileName.find("~")+1);
+ return newFileName;
+ }else { //find path
+ if (path.rfind(".\\") == -1) { return fileName; } //already complete name
+ else { newFileName = fileName.substr(fileName.rfind(".\\")+2); } //save the complete part of the name
+
+ char *cwdpath = NULL;
+ cwdpath = getcwd(NULL, 0); // or _getcwd
+ if ( cwdpath != NULL) { cwd = cwdpath; }
+ else { cwd = ""; }
- while((pos = path.rfind(".\\")) != -1) { //while you don't have a complete path
- if (path[(pos-1)] == '.') { //you want your parent directory ../
- path = path.substr(0, pos-1);
- index--;
- if (index == 0) { break; }
- }else if (path[(pos-1)] == '\\') { //you want the current working dir ./
- path = path.substr(0, pos);
- }else if (pos == 1) { break;
- }else { cout << "cannot resolve path for " << fileName << endl; return fileName; }
- }
+ //break apart the current working directory
+ vector<string> dirs;
+ while (cwd.find_first_of('\\') != -1) {
+ string dir = cwd.substr(0,cwd.find_first_of('\\'));
+ cwd = cwd.substr(cwd.find_first_of('\\')+1, cwd.length());
+ dirs.push_back(dir);
- for (int i = index; i >= 0; i--) {
- newFileName = dirs[i] + "\\" + newFileName;
- }
+ }
+ //get last one
+ dirs.push_back(cwd); //ex. dirs[0] = user, dirs[1] = work, dirs[2] = desktop
+
+ int index = dirs.size()-1;
+
+ while((pos = path.rfind(".\\")) != -1) { //while you don't have a complete path
+ if (pos == 0) { break; //you are at the end
+ }else if (path[(pos-1)] == '.') { //you want your parent directory ../
+ path = path.substr(0, pos-1);
+ index--;
+ if (index == 0) { break; }
+ }else if (path[(pos-1)] == '\\') { //you want the current working dir ./
+ path = path.substr(0, pos);
+ }else if (pos == 1) { break; //you are at the end
+ }else { cout << "cannot resolve path for " << fileName << endl; return fileName; }
+ }
- return newFileName;
+ for (int i = index; i >= 0; i--) {
+ newFileName = dirs[i] + "\\" + newFileName;
+ }
+
+ return newFileName;
+ }
#endif
}
else {
//check for blank file
gobble(fileHandle);
- if (fileHandle.eof()) { cout << completeFileName << " is blank. Please correct." << endl; return 1; }
+ if (fileHandle.eof()) { cout << completeFileName << " is blank. Please correct." << endl; }
return 0;
}
return numSeqs;
}
+/***********************************************************************/
+inline void getNumSeqs(ifstream& file, int& numSeqs){
+
+ string input;
+ numSeqs = 0;
+ while(!file.eof()){
+ input = getline(file);
+ if (input.length() != 0) {
+ if(input[0] == '>'){ numSeqs++; }
+ }
+ }
+}
+
/***********************************************************************/
inline bool inVector(string member, vector<string> group){
}
/***********************************************************************/
+//This function parses the estimator options and puts them in a vector
+inline void splitAtChar(string& estim, vector<string>& container, char symbol) {
+ try {
+ string individual;
+
+ while (estim.find_first_of(symbol) != -1) {
+ individual = estim.substr(0,estim.find_first_of(symbol));
+ if ((estim.find_first_of(symbol)+1) <= estim.length()) { //checks to make sure you don't have dash at end of string
+ estim = estim.substr(estim.find_first_of(symbol)+1, estim.length());
+ container.push_back(individual);
+ }
+ }
+ //get last one
+ container.push_back(estim);
+ }
+ catch(exception& e) {
+ cout << "Standard Error: " << e.what() << " has occurred in the mothur.h function splitAtDash. Please contact Pat Schloss at mothur.bugs@gmail.com." << "\n";
+ exit(1);
+ }
+}
+
+/***********************************************************************/
+
//This function parses the estimator options and puts them in a vector
inline void splitAtDash(string& estim, vector<string>& container) {
try {
exit(1);
}
}
+
/**************************************************************************************************/
inline bool inUsersGroups(string groupname, vector<string> Groups) {
//unique is the smallest line
if (label == "unique") { return false; }
- else { convert(label, labelFloat); }
+ else {
+ if (convertTestFloat(label, labelFloat)) {
+ convert(label, labelFloat);
+ }else { //cant convert
+ return false;
+ }
+ }
//go through users set and make them floats
for(it = userLabels.begin(); it != userLabels.end(); ++it) {
//open output file in append mode
openOutputFileAppend(filename, output);
- openInputFile(temp, input);
+ int ableToOpen = openInputFile(temp, input, "no error");
- while(char c = input.get()){
- if(input.eof()) { break; }
- else { output << c; }
+ if (ableToOpen == 0) { //you opened it
+ while(char c = input.get()){
+ if(input.eof()) { break; }
+ else { output << c; }
+ }
+ input.close();
}
- input.close();
output.close();
}
catch(exception& e) {
}
/**************************************************************************************************/
-inline string sortFile(string distFile){
+inline string sortFile(string distFile, string outputDir){
try {
+
+ //if (outputDir == "") { outputDir += hasPath(distFile); }
string outfile = getRootName(distFile) + "sorted.dist";
+
//if you can, use the unix sort since its been optimized for years
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
}
}
/**************************************************************************************************/
-inline vector<long> setFilePosFasta(string filename, int& num) {
+inline vector<unsigned long int> setFilePosFasta(string filename, int& num) {
- vector<long> positions;
+ vector<unsigned long int> positions;
ifstream inFASTA;
openInputFile(filename, inFASTA);
-
+
string input;
while(!inFASTA.eof()){
- input = getline(inFASTA); gobble(inFASTA);
+ input = getline(inFASTA);
if (input.length() != 0) {
- if(input[0] == '>'){ long pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
+ if(input[0] == '>'){ unsigned long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
}
+ gobble(inFASTA); //has to be here since windows line endings are 2 characters and mess up the positions
}
inFASTA.close();
num = positions.size();
- FILE * pFile;
+ /*FILE * pFile;
long size;
//get num bytes in file
fseek (pFile, 0, SEEK_END);
size=ftell (pFile);
fclose (pFile);
+ }*/
+
+ unsigned long int size = positions[(positions.size()-1)];
+ ifstream in;
+ openInputFile(filename, in);
+
+ in.seekg(size);
+
+ while(char c = in.get()){
+ if(in.eof()) { break; }
+ else { size++; }
}
+ in.close();
positions.push_back(size);
return positions;
}
/**************************************************************************************************/
-inline vector<long> setFilePosEachLine(string filename, int& num) {
+inline vector<unsigned long int> setFilePosEachLine(string filename, int& num) {
- vector<long> positions;
+ vector<unsigned long int> positions;
ifstream in;
openInputFile(filename, in);
string input;
while(!in.eof()){
- long lastpos = in.tellg();
- input = getline(in); gobble(in);
+ unsigned long int lastpos = in.tellg();
+ input = getline(in);
if (input.length() != 0) {
- long pos = in.tellg();
+ unsigned long int pos = in.tellg();
if (pos != -1) { positions.push_back(pos - input.length() - 1); }
else { positions.push_back(lastpos); }
}
+ gobble(in); //has to be here since windows line endings are 2 characters and mess up the positions
}
in.close();
num = positions.size();
FILE * pFile;
- long size;
+ unsigned long int size;
//get num bytes in file
pFile = fopen (filename.c_str(),"rb");
return positions;
}
+/**************************************************************************************************/
+
+inline vector<unsigned long int> divideFile(string filename, int& proc) {
+ try{
+
+ vector<unsigned long int> filePos;
+ filePos.push_back(0);
+
+ FILE * pFile;
+ unsigned long int 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);
+ }
+
+ //estimate file breaks
+ unsigned long int chunkSize = 0;
+ chunkSize = size / proc;
+
+ //file to small to divide by processors
+ if (chunkSize == 0) { proc = 1; filePos.push_back(size); return filePos; }
+
+ //for each process seekg to closest file break and search for next '>' char. make that the filebreak
+ for (int i = 0; i < proc; i++) {
+ unsigned long int spot = (i+1) * chunkSize;
+
+ ifstream in;
+ openInputFile(filename, in);
+ in.seekg(spot);
+
+ //look for next '>'
+ unsigned long int newSpot = spot;
+ while (!in.eof()) {
+ char c = in.get();
+ if (c == '>') { in.putback(c); newSpot = in.tellg(); break; }
+ }
+
+ //there was not another sequence before the end of the file
+ if (newSpot == spot) { break; }
+ else { filePos.push_back(newSpot); }
+
+ in.close();
+ }
+
+ //save end pos
+ filePos.push_back(size);
+
+ //sanity check filePos
+ for (int i = 0; i < (filePos.size()-1); i++) {
+ if (filePos[(i+1)] <= filePos[i]) { filePos.erase(filePos.begin()+(i+1)); i--; }
+ }
+
+ proc = (filePos.size() - 1);
+
+ return filePos;
+ }
+ catch(exception& e) {
+ cout << "Standard Error: " << e.what() << " has occurred in the mothur.h function divideFile. Please contact Pat Schloss at mothur.bugs@gmail.com." << "\n";
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+inline bool checkReleaseVersion(ifstream& file, string version) {
+ try {
+
+ bool good = true;
+
+ string line = getline(file);
+ //before we added this check
+ if (line[0] != '#') { good = false; }
+ else {
+ //rip off #
+ line = line.substr(1);
+
+ vector<string> versionVector;
+ splitAtChar(version, versionVector, '.');
+
+ //check file version
+ vector<string> linesVector;
+ splitAtChar(line, linesVector, '.');
+
+ if (versionVector.size() != linesVector.size()) { good = false; }
+ else {
+ for (int j = 0; j < versionVector.size(); j++) {
+ int num1, num2;
+ convert(versionVector[j], num1);
+ convert(linesVector[j], num2);
+
+ //if mothurs version is newer than this files version, then we want to remake it
+ if (num1 > num2) { good = false; break; }
+ }
+ }
+
+ }
+
+ if (!good) { file.close(); }
+ else { file.seekg(0); }
+
+ return good;
+ }
+ catch(exception& e) {
+ cout << "Standard Error: " << e.what() << " has occurred in the mothur.h function checkReleaseVersion. Please contact Pat Schloss at mothur.bugs@gmail.com." << "\n";
+ exit(1);
+ }
+}
/**************************************************************************************************/
#endif