/*
operates on normalized and localized sequences and patterns only
alphabet can be modified
*/
list<string> extendSequence(const list<string>& seq,
string pat) {
list<string> results;
std::string result;
boost::regex_constants::syntax_option_type flags =
boost::regex_constants::perl;
boost::regex re;
boost::smatch what;
pat = "^" + pat;
re.assign(pat, flags);
for (list<string>::const_iterator it = seq.begin(); it!=seq.end(); it++) {
string activeSeq = *it;
Normalizer n;
n.norm(activeSeq);
string input = n.getInput() + "z";
for (int i=0; i<input.length(); i++) {
string ref = activeSeq+input[i]+"~~~~~~~~~~~~~~~~~~";
DBG cout << "pattern " << pat << " against " << ref << endl;
try {
if (regex_search(ref,what,re)) {
string m = what[0];
if (m.length()>activeSeq.length()) {
string bit;
bit += input[i];
string ext = activeSeq + n.norm(bit);
DBG cout << " got something " << ext << endl;
results.push_back(string(ext));
}
}
} catch(boost::regex_error& regErr) {
cerr << "regular expression failed: " << regErr.what() << endl;
}
}
}
return results;
}
std::string sequenceToPattern(std::string seq) {
checkSequence(seq);
Normalizer n;
std::string sub = "";
if (seq.length()>10) {
sub = seq.substr(0,seq.length()-10);
seq = seq.substr(seq.length()-10,seq.length());
}
string pat = n.denorm(submain(n.norm(seq)));
pat = sub + pat;
return pat;
}
std::string extendSequence(std::string seq, std::string pattern, int len) {
srand (std::time(0));
checkSequence(seq);
Normalizer n;
int pruneLen = 100;
seq = n.norm(seq);
pattern = n.norm(pattern);
string refPattern = pattern;
pattern = localizePattern(pattern);
DBG cout << "extend " << seq << " with " << pattern << endl;
n.add("ABCDEFGHIJKLMNOPQRSTUVWXYZ");
list<string> lst;
lst.push_back(seq);
for (int k=0; k<len; k++) {
lst = extendSequence(lst,pattern);
list<string> lst2;
for (list<string>::const_iterator it = lst.begin(); it!=lst.end();
it++) {
string nextSeq = *it;
string nextPat = refPattern;
if (nextSeq.length()<=10) {
nextPat = sequenceToPattern(nextSeq);
}
if (nextPat==refPattern) {
lst2.push_back(nextSeq);
} else {
DBG cout << nextSeq << ": " << "mismatch " << nextPat << " versus " << refPattern << endl;
}
}
if (lst2.size()>0) {
lst = lst2;
}
if (lst.size()>pruneLen) {
DBG cout << "NEED TO PRUNE" << endl;
vector<string> v(lst.begin(),lst.end());
random_shuffle(v.begin(),v.end());
lst.clear();
lst = list<string>(v.begin(),v.begin()+pruneLen);
//lst.erase((++(++(lst.begin()))),lst.end());
}
DBG {
cout << "possibilities: " << endl;
for (list<string>::const_iterator it = lst.begin(); it!=lst.end();
it++) {
cout << " -- " << n.denorm(*it) << endl;
}
}
}
for (list<string>::const_iterator it = lst.begin(); it!=lst.end();
it++) {
DBG cout << " final possibility " << n.denorm(*it) << endl;
}
vector<string> v(lst.begin(),lst.end());
random_shuffle(v.begin(),v.end());
if (v.size()>0) {
string result = n.denorm(v[0]);
result = result.substr(seq.length(),result.length());
return result;
}
return "";
}