// module.names
// -o output_path
// -s skip_module.names
int parseTask(std::string& content, Task* task, int left, int right) {
int space = findSymbol(content, ' ', left, right);
splitNames(task->moduleNames, content, left, space);
if (space == right) return ERROR_OK; // no more data
// parse task options
int cursor = space + 1;
int hyphen;
while (cursor < right) {
hyphen = findSymbol(content, '-', cursor, right);
if (hyphen == right) break; // no options
space = findSymbol(content, ' ', hyphen + 3, right);
switch (content[hyphen + 1]) {
case 'o':
task->output = content.substr(hyphen + 3, space - (hyphen + 3));
break;
case 's':
splitNames(task->skipModuleNames, content, hyphen + 3, space);
break;
default:
return PROJECT_UNKNOWN_TASK_OPTION_ERROR;
}
cursor = space + 1;
}
return ERROR_OK;
}
static void
collapse(char *path)
{
char *names = (path[0] == '/') ? path + 1 : path; /* Preserve first '/' */
int nc;
char **ix;
int i, j;
char *p, *q;
nc = collapsible(names);
if (nc < 2) return; /* Nothing to do */
ix = (char **)alloca(nc * sizeof(char *));
splitNames(names, ix);
for (i = 0; i < nc; i++) {
int dots = 0;
/* Find next occurrence of "." or ".." */
do {
char *p = ix[i];
if (p[0] == '.') {
if (p[1] == '\0') {
dots = 1;
break;
}
if ((p[1] == '.') && (p[2] == '\0')) {
dots = 2;
break;
}
}
i++;
} while (i < nc);
if (i >= nc) break;
/* At this point i is the index of either a "." or a "..", so take the
appropriate action and then continue the outer loop */
if (dots == 1) {
/* Remove this instance of "." */
ix[i] = 0;
}
else {
/* If there is a preceding name, remove both that name and this
instance of ".."; otherwise, leave the ".." as is */
for (j = i - 1; j >= 0; j--) {
if (ix[j]) break;
}
if (j < 0) continue;
ix[j] = 0;
ix[i] = 0;
}
/* i will be incremented at the top of the loop */
}
joinNames(names, nc, ix);
}
NAMELIST *edit_list(NAMELIST *list, size_t *list_size_p){
long list_size=MAXLETTER, i, j;
NAMELIST *list2, tmp;
char *name, *periodName;
long from, to, lengthLimit;
fpos_t fposition;
char *periodPosition;
list2=(NAMELIST *)malloc(list_size*sizeof(NAMELIST));
/* delimit by space */
j=0;
for(i=0;list[i].name!=NULL;i++){
tmp=seqnameDelimitBySpace(list[i]);
/*fprintf(stderr, "after delimit by space: [%s]\n", tmp.name);*/
if(tmp.name[0]!='\0'){
if(j+10 >= list_size){
list_size+=MAXLETTER;
list2=(NAMELIST *)realloc(list2,list_size*sizeof(NAMELIST));
}
list2[j]=tmp;
j++;
if(isspace(tmp.name[0])){
list2[j]=nameWithSpace(&tmp);
j++;
}
name=(char *)malloc((strlen(tmp.name)+1)*sizeof(char));
strcpy(name, tmp.name);
lengthLimit=strlen(name)-1;
/*printf(" input: %s, length: %ld\n",name, lengthLimit); */
for(from=tmp.from-1, to=tmp.from-1, fposition=tmp.fposition;
name!=NULL && from < lengthLimit;
j++){
if(j>=list_size-2){
list_size+=MAXLETTER;
list2=(NAMELIST *)realloc(list2,list_size*sizeof(NAMELIST));
}
/* printf("Input: %s %ld-%ld::%lld\n", name, from, to, fposition); *//*debug */
tmp=splitNames(&from, &to, name, fposition);
/* printf("result; %s %ld-%ld::%lld / %ld\n", tmp.name, from, to, fposition, lengthLimit); */ /* debug */
if(tmp.name!=NULL && tmp.name[0]!='\0'){
list2[j]=tmp;
/* name+=list2[j].to+1;*/
}else{
break;
}
if(strstr(tmp.name, ".")!=NULL){
j++;
periodName=(char *)malloc((strlen(tmp.name)+1)*sizeof(char));
strcpy(periodName, tmp.name);
periodPosition=strstr(periodName, ".");
*periodPosition='\0';
list2[j].name=periodName;
list2[j].from =list2[j-1].from;
list2[j].to =list2[j-1].to-(strlen(tmp.name)-strlen(periodName));
list2[j].fposition=list2[j-1].fposition;
/* printf("found period: %s %d-%d::%lld\n", list2[j].name, list2[j].from,
list2[j].to, list2[j].fposition);*/
}
}
}
}
/* add line to line2 */
for(i=0;list[i].name!=NULL;j++, i++){
if(j>=list_size){
list_size+=MAXLETTER;
list2=(NAMELIST *)realloc(list2,list_size*sizeof(NAMELIST));
}
list2[j]=list[i];
}
*list_size_p=j;
return list2;
}
//********************************************************************************************************************
int SplitMatrix::splitDistanceRAM() {
try {
vector<set<string> > groups;
vector<string> outputs;
int numGroups = 0;
ifstream dFile;
m->openInputFile(distFile, dFile);
while(dFile) {
string seqA, seqB;
float dist;
dFile >> seqA >> seqB >> dist;
if (m->control_pressed) {
dFile.close();
for(int i=0; i<numGroups; i++) {
if(groups[i].size() > 0) {
m->mothurRemove((distFile + "." + toString(i) + ".temp"));
}
}
return 0;
}
if(dist < cutoff) {
//cout << "in cutoff: " << dist << endl;
int groupIDA = -1;
int groupIDB = -1;
int groupID = -1;
for(int i=0; i<numGroups; i++) {
set<string>::iterator aIt = groups[i].find(seqA);
set<string>::iterator bIt = groups[i].find(seqB);
if(groupIDA == -1 && aIt != groups[i].end()) { //seqA is not already assigned to a group and is in group[i], so assign seqB to group[i]
groups[i].insert(seqB);
groupIDA = i;
groupID = groupIDA;
//cout << "in aIt: " << groupID << endl;
// break;
}
else if(groupIDB == -1 && bIt != groups[i].end()) { //seqB is not already assigned to a group and is in group[i], so assign seqA to group[i]
groups[i].insert(seqA);
groupIDB = i;
groupID = groupIDB;
// cout << "in bIt: " << groupID << endl;
// break;
}
if(groupIDA != -1 && groupIDB != -1) { //both ifs above have been executed, so we need to decide who to assign them to
if(groupIDA < groupIDB) {
// cout << "A: " << groupIDA << "\t" << groupIDB << endl;
groups[groupIDA].insert(groups[groupIDB].begin(), groups[groupIDB].end()); //merge two groups into groupIDA
groups[groupIDB].clear();
groupID = groupIDA;
}
else {
// cout << "B: " << groupIDA << "\t" << groupIDB << endl;
groups[groupIDB].insert(groups[groupIDA].begin(), groups[groupIDA].end()); //merge two groups into groupIDB
groups[groupIDA].clear();
groupID = groupIDB;
}
break;
}
}
//windows is gonna gag on the reuse of outFile, will need to make it local...
if(groupIDA == -1 && groupIDB == -1) { //we need a new group
set<string> newGroup;
newGroup.insert(seqA);
newGroup.insert(seqB);
groups.push_back(newGroup);
string tempOut = seqA + '\t' + seqB + '\t' + toString(dist) + '\n';
outputs.push_back(tempOut);
numGroups++;
}
else {
outputs[groupID] += seqA + '\t' + seqB + '\t' + toString(dist) + '\n';
if(groupIDA != -1 && groupIDB != -1) { //merge distance files of two groups you merged above
string row, column, distance;
if(groupIDA<groupIDB) {
//merge memory
outputs[groupID] += outputs[groupIDB];
outputs[groupIDB] = "";
} else {
outputs[groupID] += outputs[groupIDA];
outputs[groupIDA] = "";
}
}
}
}
m->gobble(dFile);
}
dFile.close();
vector<string> tempDistFiles;
for (int i = 0; i < numGroups; i++) {
string fileName = distFile + "." + toString(i) + ".temp";
tempDistFiles.push_back(fileName);
if (outputs[i] != "") {
ofstream outFile;
outFile.open(fileName.c_str(), ios::ate);
outFile << outputs[i];
outFile.close();
}
}
map<string, int> seqGroup;
for (int i = 0; i < groups.size(); i++) {
for (set<string>::iterator itNames = groups[i].begin(); itNames != groups[i].end();) {
seqGroup[*itNames] = i;
groups[i].erase(itNames++);
}
}
splitNames(seqGroup, numGroups, tempDistFiles);
return 0;
}
catch(exception& e) {
m->errorOut(e, "SplitMatrix", "splitDistanceRAM");
exit(1);
}
}
int SplitMatrix::splitDistanceLarge() {
try {
vector<set<string> > groups;
//for buffering the io to improve speed
//allow for 30 dists to be stored, then output.
vector<string> outputs;
vector<int> numOutputs;
vector<bool> wroteOutPut;
int numGroups = 0;
//ofstream outFile;
ifstream dFile;
m->openInputFile(distFile, dFile);
while(dFile) {
string seqA, seqB;
float dist;
dFile >> seqA >> seqB >> dist;
if (m->control_pressed) {
dFile.close();
for(int i=0; i<numGroups; i++) {
if(groups[i].size() > 0) {
m->mothurRemove((distFile + "." + toString(i) + ".temp"));
}
}
return 0;
}
if(dist < cutoff) {
//cout << "in cutoff: " << dist << endl;
int groupIDA = -1;
int groupIDB = -1;
int groupID = -1;
for(int i=0; i<numGroups; i++) {
set<string>::iterator aIt = groups[i].find(seqA);
set<string>::iterator bIt = groups[i].find(seqB);
if(groupIDA == -1 && aIt != groups[i].end()) { //seqA is not already assigned to a group and is in group[i], so assign seqB to group[i]
groups[i].insert(seqB);
groupIDA = i;
groupID = groupIDA;
//cout << "in aIt: " << groupID << endl;
// break;
}
else if(groupIDB == -1 && bIt != groups[i].end()) { //seqB is not already assigned to a group and is in group[i], so assign seqA to group[i]
groups[i].insert(seqA);
groupIDB = i;
groupID = groupIDB;
// cout << "in bIt: " << groupID << endl;
// break;
}
if(groupIDA != -1 && groupIDB != -1) { //both ifs above have been executed, so we need to decide who to assign them to
if(groupIDA < groupIDB) {
// cout << "A: " << groupIDA << "\t" << groupIDB << endl;
groups[groupIDA].insert(groups[groupIDB].begin(), groups[groupIDB].end()); //merge two groups into groupIDA
groups[groupIDB].clear();
groupID = groupIDA;
}
else {
// cout << "B: " << groupIDA << "\t" << groupIDB << endl;
groups[groupIDB].insert(groups[groupIDA].begin(), groups[groupIDA].end()); //merge two groups into groupIDB
groups[groupIDA].clear();
groupID = groupIDB;
}
break;
}
}
//windows is gonna gag on the reuse of outFile, will need to make it local...
if(groupIDA == -1 && groupIDB == -1) { //we need a new group
set<string> newGroup;
newGroup.insert(seqA);
newGroup.insert(seqB);
groups.push_back(newGroup);
string tempOut = seqA + '\t' + seqB + '\t' + toString(dist) + '\n';
outputs.push_back(tempOut);
numOutputs.push_back(1);
wroteOutPut.push_back(false);
numGroups++;
}
else {
string fileName = distFile + "." + toString(groupID) + ".temp";
//have we reached the max buffer size
if (numOutputs[groupID] > 60) { //write out sequence
ofstream outFile;
outFile.open(fileName.c_str(), ios::app);
outFile << outputs[groupID] << seqA << '\t' << seqB << '\t' << dist << endl;
outFile.close();
outputs[groupID] = "";
numOutputs[groupID] = 0;
wroteOutPut[groupID] = true;
} else {
outputs[groupID] += seqA + '\t' + seqB + '\t' + toString(dist) + '\n';
numOutputs[groupID]++;
}
if(groupIDA != -1 && groupIDB != -1) { //merge distance files of two groups you merged above
string row, column, distance;
if(groupIDA<groupIDB) {
//merge memory
numOutputs[groupID] += numOutputs[groupIDB];
outputs[groupID] += outputs[groupIDB];
outputs[groupIDB] = "";
numOutputs[groupIDB] = 0;
//if groupB is written to file it is above buffer size so read and write to new merged file
if (wroteOutPut[groupIDB]) {
string fileName2 = distFile + "." + toString(groupIDB) + ".temp";
m->appendFiles(fileName2, fileName);
m->mothurRemove(fileName2);
//write out the merged memory
if (numOutputs[groupID] > 60) {
ofstream tempOut;
m->openOutputFile(fileName, tempOut);
tempOut << outputs[groupID];
outputs[groupID] = "";
numOutputs[groupID] = 0;
tempOut.close();
}
//outFile.close();
wroteOutPut[groupID] = true;
wroteOutPut[groupIDB] = false;
} else { } //just merge b's memory with a's memory
}
else {
numOutputs[groupID] += numOutputs[groupIDA];
outputs[groupID] += outputs[groupIDA];
outputs[groupIDA] = "";
numOutputs[groupIDA] = 0;
if (wroteOutPut[groupIDA]) {
string fileName2 = distFile + "." + toString(groupIDA) + ".temp";
m->appendFiles(fileName2, fileName);
m->mothurRemove(fileName2);
//write out the merged memory
if (numOutputs[groupID] > 60) {
ofstream tempOut;
m->openOutputFile(fileName, tempOut);
tempOut << outputs[groupID];
outputs[groupID] = "";
numOutputs[groupID] = 0;
tempOut.close();
}
//outFile.close();
wroteOutPut[groupID] = true;
wroteOutPut[groupIDA] = false;
} else { } //just merge memory
}
}
}
}
m->gobble(dFile);
}
dFile.close();
vector<string> tempDistFiles;
for (int i = 0; i < numGroups; i++) {
string fileName = distFile + "." + toString(i) + ".temp";
tempDistFiles.push_back(fileName);
//remove old names files just in case
if (numOutputs[i] > 0) {
ofstream outFile;
outFile.open(fileName.c_str(), ios::app);
outFile << outputs[i];
outFile.close();
}
}
map<string, int> seqGroup;
for (int i = 0; i < groups.size(); i++) {
for (set<string>::iterator itNames = groups[i].begin(); itNames != groups[i].end();) {
seqGroup[*itNames] = i;
groups[i].erase(itNames++);
}
}
splitNames(seqGroup, numGroups, tempDistFiles);
return 0;
}
catch(exception& e) {
m->errorOut(e, "SplitMatrix", "splitDistanceLarge");
exit(1);
}
}
int SplitMatrix::splitDistanceFileByTax(map<string, int>& seqGroup, int numGroups) {
try {
map<string, int>::iterator it;
map<string, int>::iterator it2;
ofstream outFile;
ifstream dFile;
m->openInputFile(distFile, dFile);
for (int i = 0; i < numGroups; i++) { //remove old temp files, just in case
m->mothurRemove((distFile + "." + toString(i) + ".temp"));
}
//for buffering the io to improve speed
//allow for 10 dists to be stored, then output.
vector<string> outputs;
outputs.resize(numGroups, "");
vector<int> numOutputs;
numOutputs.resize(numGroups, 0);
//you can have a group made, but their may be no distances in the file for this group if the taxonomy file and distance file don't match
//this can occur if we have converted the phylip to column, since we reduce the size at that step by using the cutoff value
vector<bool> validDistances;
validDistances.resize(numGroups, false);
//for each distance
while(dFile) {
string seqA, seqB;
float dist;
if (m->control_pressed) {
dFile.close();
for (int i = 0; i < numGroups; i++) {
m->mothurRemove((distFile + "." + toString(i) + ".temp"));
}
}
dFile >> seqA >> seqB >> dist;
m->gobble(dFile);
//if both sequences are in the same group then they are within the cutoff
it = seqGroup.find(seqA);
it2 = seqGroup.find(seqB);
if ((it != seqGroup.end()) && (it2 != seqGroup.end())) { //they are both not singletons
if (it->second == it2->second) { //they are from the same group so add the distance
if (numOutputs[it->second] > 30) {
m->openOutputFileAppend((distFile + "." + toString(it->second) + ".temp"), outFile);
outFile << outputs[it->second] << seqA << '\t' << seqB << '\t' << dist << endl;
outFile.close();
outputs[it->second] = "";
numOutputs[it->second] = 0;
validDistances[it->second] = true;
} else {
outputs[it->second] += seqA + '\t' + seqB + '\t' + toString(dist) + '\n';
numOutputs[it->second]++;
}
}
}
}
dFile.close();
string inputFile = namefile;
if (countfile != "") {
inputFile = countfile;
}
vector<string> tempDistFiles;
for (int i = 0; i < numGroups; i++) { //remove old temp files, just in case
string tempDistFile = distFile + "." + toString(i) + ".temp";
tempDistFiles.push_back(tempDistFile);
m->mothurRemove((inputFile + "." + toString(i) + ".temp"));
//write out any remaining buffers
if (numOutputs[i] > 0) {
m->openOutputFileAppend((distFile + "." + toString(i) + ".temp"), outFile);
outFile << outputs[i];
outFile.close();
outputs[i] = "";
numOutputs[i] = 0;
validDistances[i] = true;
}
}
splitNames(seqGroup, numGroups, tempDistFiles);
if (m->control_pressed) {
for (int i = 0; i < dists.size(); i++) {
m->mothurRemove((dists[i].begin()->first));
m->mothurRemove((dists[i].begin()->second));
}
dists.clear();
}
return 0;
}
catch(exception& e) {
m->errorOut(e, "SplitMatrix", "splitDistanceFileByTax");
exit(1);
}
}
int SplitMatrix::createDistanceFilesFromTax(map<string, int>& seqGroup, int numGroups) {
try {
map<string, int> copyGroups = seqGroup;
map<string, int>::iterator it;
set<string> names;
ifstream in;
m->openInputFile(fastafile, in);
for (int i = 0; i < numGroups; i++) {
m->mothurRemove((fastafile + "." + toString(i) + ".temp"));
}
//parse fastafile
while (!in.eof()) {
Sequence query(in);
m->gobble(in);
if (query.getName() != "") {
it = seqGroup.find(query.getName());
//save names in case no namefile is given
if ((namefile == "") && (countfile == "")) {
names.insert(query.getName());
}
if (it != seqGroup.end()) { //not singleton
ofstream outFile;
m->openOutputFileAppend((fastafile + "." + toString(it->second) + ".temp"), outFile);
query.printSequence(outFile);
outFile.close();
copyGroups.erase(query.getName());
}
}
}
in.close();
bool error = false;
//warn about sequence in groups that are not in fasta file
for(it = copyGroups.begin(); it != copyGroups.end(); it++) {
m->mothurOut("ERROR: " + it->first + " is missing from your fastafile. This could happen if your taxonomy file is not unique and your fastafile is, or it could indicate an error.");
m->mothurOutEndLine();
error = true;
}
copyGroups.clear();
if (error) {
exit(1);
}
if (outputType == "distance") { //create distance matrices for each fasta file
//process each distance file
for (int i = 0; i < numGroups; i++) {
string options = "";
if (classic) {
options = "fasta=" + (fastafile + "." + toString(i) + ".temp") + ", processors=" + toString(processors) + ", output=lt";
}
else {
options = "fasta=" + (fastafile + "." + toString(i) + ".temp") + ", processors=" + toString(processors) + ", cutoff=" + toString(distCutoff);
}
if (outputDir != "") {
options += ", outputdir=" + outputDir;
}
m->mothurCalling = true;
m->mothurOut("/******************************************/");
m->mothurOutEndLine();
m->mothurOut("Running command: dist.seqs(" + options + ")");
m->mothurOutEndLine();
m->mothurCalling = true;
Command* command = new DistanceCommand(options);
m->mothurOut("/******************************************/");
m->mothurOutEndLine();
command->execute();
delete command;
m->mothurCalling = false;
m->mothurRemove((fastafile + "." + toString(i) + ".temp"));
//remove old names files just in case
if (namefile != "") {
m->mothurRemove((namefile + "." + toString(i) + ".temp"));
}
else {
m->mothurRemove((countfile + "." + toString(i) + ".temp"));
}
}
}
//restore old fasta file name since dist.seqs overwrites it with the temp files
m->setFastaFile(fastafile);
vector<string> tempDistFiles;
for(int i=0; i<numGroups; i++) {
if (outputDir == "") {
outputDir = m->hasPath(fastafile);
}
string tempDistFile = (fastafile + "." + toString(i) + ".temp");
if (outputType == "distance") {
if (classic) {
tempDistFile = outputDir + m->getRootName(m->getSimpleName((fastafile + "." + toString(i) + ".temp"))) + "phylip.dist";
}
else {
tempDistFile = outputDir + m->getRootName(m->getSimpleName((fastafile + "." + toString(i) + ".temp"))) + "dist";
}
}
tempDistFiles.push_back(tempDistFile);
}
if (method == "vsearch") {
splitNamesVsearch(seqGroup, numGroups, tempDistFiles);
}
else {
splitNames(seqGroup, numGroups, tempDistFiles);
}
if (m->control_pressed) {
for (int i = 0; i < dists.size(); i++) {
m->mothurRemove((dists[i].begin()->first));
m->mothurRemove((dists[i].begin()->second));
}
dists.clear();
}
return 0;
}
catch(exception& e) {
m->errorOut(e, "SplitMatrix", "createDistanceFilesFromTax");
exit(1);
}
}
