1 // This file is part of the Mussa source distribution.
2 // http://mussa.caltech.edu/
3 // Contact author: Tristan De Buysscher, tristan@caltech.edu
5 // This program and all associated source code files are Copyright (C) 2005
6 // the California Institute of Technology, Pasadena, CA, 91125 USA. It is
7 // under the GNU Public License; please see the included LICENSE.txt
8 // file for more information, or contact Tristan directly.
11 // This file is part of the Mussa source distribution.
12 // http://mussa.caltech.edu/
13 // Contact author: Tristan De Buysscher, tristan@caltech.edu
15 // This program and all associated source code files are Copyright (C) 2005
16 // the California Institute of Technology, Pasadena, CA, 91125 USA. It is
17 // under the GNU Public License; please see the included LICENSE.txt
18 // file for more information, or contact Tristan directly.
21 // ----------------------------------------
22 // ---------- sequence.cc -----------
23 // ----------------------------------------
24 #include <boost/filesystem/fstream.hpp>
25 namespace fs = boost::filesystem;
27 #include <boost/spirit/core.hpp>
28 #include <boost/spirit/actor/push_back_actor.hpp>
29 #include <boost/spirit/iterator/file_iterator.hpp>
30 #include <boost/spirit/utility/chset.hpp>
31 namespace spirit = boost::spirit;
33 #include "alg/sequence.hpp"
34 #include "mussa_exceptions.hpp"
48 annot::annot(int begin, int end, std::string type, std::string name)
60 bool operator==(const annot& left, const annot& right)
62 return ((left.begin== right.begin) and
63 (left.end == right.end) and
64 (left.type == right.type) and
65 (left.name == right.name));
68 motif::motif(int begin, std::string motif)
69 : annot(begin, begin+motif.size(), "motif", motif),
78 const std::string Sequence::dna_alphabet("AaCcGgTtNn\012\015");
79 const std::string Sequence::rna_alphabet("AaCcGgNnUu\012\015");
80 //! this is the general iupac alphabet for nucleotides
81 const std::string Sequence::nucleic_iupac_alphabet("AaCcGgTtUuRrYyMmKkSsWwBbDdHhVvNn\012\015");
82 //! the protein alphabet
83 const std::string Sequence::protein_alphabet("AaCcDdEeFfGgHhIiKkLlMmNnPpQqRrSsTtVvWwYy\012\015");
95 Sequence::Sequence(const char *seq)
101 set_filtered_sequence(seq);
104 Sequence::Sequence(const std::string& seq)
110 set_filtered_sequence(seq);
113 Sequence::Sequence(const Sequence& o)
115 seq_start(o.seq_start),
116 seq_count(o.seq_count),
120 motif_list(o.motif_list)
124 Sequence &Sequence::operator=(const Sequence& s)
128 seq_start = s.seq_start;
129 seq_count = s.seq_count;
133 motif_list = s.motif_list;
138 static void multiplatform_getline(std::istream& in, std::string& line)
143 while(in.good() and !(c == '\012' or c == '\015') ) {
147 // if we have cr-lf eat it
149 if (c=='\012' or c == '\015') {
154 //! load a fasta file into a sequence
156 * \param file_path the location of the fasta file in the filesystem
157 * \param seq_num which sequence in the file to load
158 * \param start_index starting position in the fasta sequence, 0 for beginning
159 * \param end_index ending position in the fasta sequence, 0 for end
160 * \return error message, empty string if no error. (gag!)
162 void Sequence::load_fasta(fs::path file_path, int seq_num,
163 int start_index, int end_index)
165 fs::fstream data_file;
166 data_file.open(file_path, std::ios::in);
168 if (!data_file.good())
170 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
173 load_fasta(data_file, seq_num, start_index, end_index);
174 } catch(sequence_empty_error e) {
175 // there doesn't appear to be any sequence
176 // catch and rethrow to include the filename
177 std::stringstream msg;
178 msg << "The selected sequence in "
179 << file_path.native_file_string()
180 << " appears to be empty";
181 throw sequence_empty_error(msg.str());
182 } catch(sequence_empty_file_error e) {
183 std::stringstream errormsg;
184 errormsg << file_path.native_file_string()
185 << " did not have any fasta sequences" << std::endl;
186 throw sequence_empty_file_error(errormsg.str());
192 Sequence::load_fasta(std::iostream& data_file, int seq_num,
193 int start_index, int end_index)
195 std::string file_data_line;
196 int header_counter = 0;
197 bool read_seq = true;
198 std::string rev_comp;
199 std::string sequence_raw;
200 std::string seq_tmp; // holds sequence during basic filtering
203 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
206 // search for the header of the fasta sequence we want
207 while ( (!data_file.eof()) && (header_counter < seq_num) )
209 multiplatform_getline(data_file, file_data_line);
210 if (file_data_line.substr(0,1) == ">")
214 if (header_counter > 0) {
215 header = file_data_line.substr(1);
219 while ( !data_file.eof() && read_seq ) {
220 multiplatform_getline(data_file,file_data_line);
221 if (file_data_line.substr(0,1) == ">")
223 else sequence_raw += file_data_line;
226 // Lastly, if subselection of the sequence was specified we keep cut out
227 // and only keep that part
228 // end_index = 0 means no end was specified, so cut to the end
230 end_index = sequence_raw.size();
232 // sequence filtering for upcasing agctn and convert non AGCTN to N
233 if (end_index-start_index <= 0) {
234 std::string msg("The selected sequence appears to be empty");
235 throw sequence_empty_error(msg);
237 set_filtered_sequence(sequence_raw, start_index, end_index-start_index);
239 std::string errormsg("There were no fasta sequences");
240 throw sequence_empty_file_error(errormsg);
244 void Sequence::set_filtered_sequence(const std::string &old_seq,
245 std::string::size_type start,
246 std::string::size_type count)
248 char conversionTable[257];
251 count = old_seq.size() - start;
252 boost::shared_ptr<seq_string> new_seq(new seq_string);
253 new_seq->reserve(count);
255 // Make a conversion table
257 // everything we don't specify below will become 'N'
258 for(int table_i=0; table_i < 256; table_i++)
260 conversionTable[table_i] = 'N';
262 // add end of string character for printing out table for testing purposes
263 conversionTable[256] = '\0';
265 // we want these to map to themselves - ie not to change
266 conversionTable[(int)'A'] = 'A';
267 conversionTable[(int)'T'] = 'T';
268 conversionTable[(int)'G'] = 'G';
269 conversionTable[(int)'C'] = 'C';
271 conversionTable[(int)'a'] = 'A';
272 conversionTable[(int)'t'] = 'T';
273 conversionTable[(int)'g'] = 'G';
274 conversionTable[(int)'c'] = 'C';
276 // finally, the actual conversion loop
277 for(std::string::size_type seq_index = 0; seq_index < count; seq_index++)
279 new_seq->append(1, conversionTable[ (int)old_seq[seq_index+start]]);
287 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
289 fs::fstream data_stream(file_path, std::ios::in);
292 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
295 // so i should probably be passing the parse function some iterators
296 // but the annotations files are (currently) small, so i think i can
297 // get away with loading the whole file into memory
300 while(data_stream.good()) {
306 parse_annot(data, start_index, end_index);
309 /* If this works, yikes, this is some brain hurting code.
311 * what's going on is that when pb_annot is instantiated it stores references
312 * to begin, end, name, type, declared in the parse function, then
313 * when operator() is called it grabs values from those references
314 * and uses that to instantiate an annot object and append that to our
317 * This weirdness is because the spirit library requires that actions
318 * conform to a specific prototype operator()(IteratorT, IteratorT)
319 * which doesn't provide any useful opportunity for me to actually
320 * grab the results of our parsing.
322 * so I instantiate this structure in order to have a place to grab
326 struct push_back_annot {
327 std::list<annot>& annot_list;
333 push_back_annot(std::list<annot>& annot_list_,
338 : annot_list(annot_list_),
346 void operator()(std::string::const_iterator,
347 std::string::const_iterator) const
349 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
350 annot_list.push_back(annot(begin, end, name, type));
354 struct push_back_seq {
355 std::list<Sequence>& seq_list;
359 push_back_seq(std::list<Sequence>& seq_list_,
362 : seq_list(seq_list_),
368 void operator()(std::string::const_iterator,
369 std::string::const_iterator) const
371 // filter out newlines from our sequence
373 for(std::string::const_iterator seq_i = seq.begin();
377 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
379 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
382 s.set_fasta_header(name);
383 seq_list.push_back(s);
388 Sequence::parse_annot(std::string data, int start_index, int end_index)
395 std::list<Sequence> query_seqs;
397 bool status = spirit::parse(data.begin(), data.end(),
404 )[spirit::assign_a(species)] >>
408 ( // ignore html tags
409 *(spirit::space_p) >>
411 +(~spirit::ch_p('>')) >>
416 ( // parse an absolute location name
417 (spirit::uint_p[spirit::assign_a(start)] >>
419 spirit::uint_p[spirit::assign_a(end)] >>
424 )[spirit::assign_a(name)] >>
431 )[spirit::assign_a(type)]
433 // to understand how this group gets set
434 // read the comment above struct push_back_annot
435 )[push_back_annot(annots, start, end, type, name)]
437 ((spirit::ch_p('>')|spirit::str_p(">")) >>
438 (*(spirit::print_p))[spirit::assign_a(name)] >>
440 (+(spirit::chset<>(nucleic_iupac_alphabet.c_str())))[spirit::assign_a(seq)]
441 )[push_back_seq(query_seqs, name, seq)]
448 // go seearch for query sequences
449 find_sequences(query_seqs.begin(), query_seqs.end());
453 void Sequence::add_annotation(const annot& a)
458 const std::list<annot>& Sequence::annotations() const
464 Sequence::subseq(int start, int count) const
471 // there might be an off by one error with start+count > size()
472 if ( count == npos || start+count > size()) {
473 count = size()-start;
475 Sequence new_seq(*this);
476 new_seq.seq_start = start;
477 new_seq.seq_count = count;
479 new_seq.motif_list = motif_list;
480 new_seq.annots.clear();
481 // attempt to copy & reannotate position based annotations
482 int end = start+count;
484 for(std::list<annot>::const_iterator annot_i = annots.begin();
485 annot_i != annots.end();
488 int annot_begin= annot_i->begin;
489 int annot_end = annot_i->end;
491 if (annot_begin < end) {
492 if (annot_begin >= start) {
493 annot_begin -= start;
498 if (annot_end < end) {
504 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
505 new_seq.annots.push_back(new_annot);
513 Sequence::rev_comp() const
515 std::string rev_comp;
516 char conversionTable[257];
517 int seq_i, table_i, len;
520 rev_comp.reserve(len);
521 // make a conversion table
522 // init all parts of conversion table to '~' character
523 // '~' I doubt will ever appear in a sequence file (jeez, I hope)
524 // and may the fleas of 1000 camels infest the genitals of any biologist (and
525 // seven generations of their progeny) who decides to make it mean
526 // something special!!!
527 // PS - double the curse for any smartass non-biologist who tries it as well
528 for(table_i=0; table_i < 256; table_i++)
530 conversionTable[table_i] = '~';
532 // add end of string character for printing out table for testing purposes
533 conversionTable[256] = '\0';
535 // add in the characters for the bases we want to convert
536 conversionTable[(int)'A'] = 'T';
537 conversionTable[(int)'T'] = 'A';
538 conversionTable[(int)'G'] = 'C';
539 conversionTable[(int)'C'] = 'G';
540 conversionTable[(int)'N'] = 'N';
542 // finally, the actual conversion loop
543 for(seq_i = len - 1; seq_i >= 0; seq_i--)
545 table_i = (int) at(seq_i);
546 rev_comp += conversionTable[table_i];
552 void Sequence::set_fasta_header(std::string header_)
557 void Sequence::set_species(const std::string& name)
562 std::string Sequence::get_species() const
569 Sequence::get_fasta_header() const
575 Sequence::get_name() const
577 if (header.size() > 0)
579 else if (species.size() > 0)
585 void Sequence::set_sequence(const std::string& s)
587 set_filtered_sequence(s);
590 std::string Sequence::get_sequence() const
595 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
600 Sequence::const_reference Sequence::at(Sequence::size_type i) const
602 if (!seq) throw std::out_of_range("empty sequence");
603 return seq->at(i+seq_start);
618 const char *Sequence::c_str() const
621 return seq->c_str()+seq_start;
626 Sequence::const_iterator Sequence::begin() const
628 if (seq and seq_count != 0)
629 return seq->begin()+seq_start;
631 return Sequence::const_iterator(0);
634 Sequence::const_iterator Sequence::end() const
636 if (seq and seq_count != 0) {
637 return seq->begin() + seq_start + seq_count;
639 return Sequence::const_iterator(0);
643 bool Sequence::empty() const
645 return (seq_count == 0) ? true : false;
648 Sequence::size_type Sequence::size() const
653 Sequence::size_type Sequence::length() const
658 Sequence::save(fs::fstream &save_file)
659 //std::string save_file_path)
662 std::list<annot>::iterator annots_i;
664 // not sure why, or if i'm doing something wrong, but can't seem to pass
665 // file pointers down to this method from the mussa control class
666 // so each call to save a sequence appends to the file started by mussa_class
667 //save_file.open(save_file_path.c_str(), std::ios::app);
669 save_file << "<Sequence>" << std::endl;
670 save_file << *this << std::endl;
671 save_file << "</Sequence>" << std::endl;
673 save_file << "<Annotations>" << std::endl;
674 save_file << species << std::endl;
675 for (annots_i = annots.begin(); annots_i != annots.end(); ++annots_i)
677 save_file << annots_i->begin << " " << annots_i->end << " " ;
678 save_file << annots_i->name << " " << annots_i->type << std::endl;
680 save_file << "</Annotations>" << std::endl;
685 Sequence::load_museq(fs::path load_file_path, int seq_num)
687 fs::fstream load_file;
688 std::string file_data_line;
691 std::string::size_type space_split_i;
692 std::string annot_value;
695 load_file.open(load_file_path, std::ios::in);
698 // search for the seq_num-th sequence
699 while ( (!load_file.eof()) && (seq_counter < seq_num) )
701 getline(load_file,file_data_line);
702 if (file_data_line == "<Sequence>")
705 getline(load_file, file_data_line);
706 // looks like the sequence is written as a single line
707 set_filtered_sequence(file_data_line);
708 getline(load_file, file_data_line);
709 getline(load_file, file_data_line);
710 if (file_data_line == "<Annotations>")
712 getline(load_file, file_data_line);
713 species = file_data_line;
714 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
716 getline(load_file,file_data_line);
717 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
719 // need to get 4 values...almost same code 4 times...
720 // get annot start index
721 space_split_i = file_data_line.find(" ");
722 annot_value = file_data_line.substr(0,space_split_i);
723 an_annot.begin = atoi (annot_value.c_str());
724 file_data_line = file_data_line.substr(space_split_i+1);
725 // get annot end index
726 space_split_i = file_data_line.find(" ");
727 annot_value = file_data_line.substr(0,space_split_i);
728 an_annot.end = atoi (annot_value.c_str());
730 if (space_split_i == std::string::npos) // no entry for type or name
732 std::cout << "seq, annots - no type or name\n";
736 else // else get annot type
738 file_data_line = file_data_line.substr(space_split_i+1);
739 space_split_i = file_data_line.find(" ");
740 annot_value = file_data_line.substr(0,space_split_i);
741 an_annot.type = annot_value;
742 if (space_split_i == std::string::npos) // no entry for name
744 std::cout << "seq, annots - no name\n";
747 else // get annot name
749 file_data_line = file_data_line.substr(space_split_i+1);
750 space_split_i = file_data_line.find(" ");
751 annot_value = file_data_line.substr(0,space_split_i);
752 an_annot.type = annot_value;
755 annots.push_back(an_annot); // don't forget to actually add the annot
757 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
758 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
765 Sequence::rc_motif(std::string a_motif) const
767 std::string rev_comp;
768 char conversionTable[257];
769 int seq_i, table_i, len;
771 len = a_motif.length();
772 rev_comp.reserve(len);
774 for(table_i=0; table_i < 256; table_i++)
776 conversionTable[table_i] = '~';
778 // add end of std::string character for printing out table for testing purposes
779 conversionTable[256] = '\0';
781 // add in the characters for the bases we want to convert (IUPAC)
782 conversionTable[(int)'A'] = 'T';
783 conversionTable[(int)'T'] = 'A';
784 conversionTable[(int)'G'] = 'C';
785 conversionTable[(int)'C'] = 'G';
786 conversionTable[(int)'N'] = 'N';
787 conversionTable[(int)'M'] = 'K';
788 conversionTable[(int)'R'] = 'Y';
789 conversionTable[(int)'W'] = 'W';
790 conversionTable[(int)'S'] = 'S';
791 conversionTable[(int)'Y'] = 'R';
792 conversionTable[(int)'K'] = 'M';
793 conversionTable[(int)'V'] = 'B';
794 conversionTable[(int)'H'] = 'D';
795 conversionTable[(int)'D'] = 'H';
796 conversionTable[(int)'B'] = 'V';
798 // finally, the actual conversion loop
799 for(seq_i = len - 1; seq_i >= 0; seq_i--)
801 //std::cout << "** i = " << seq_i << " bp = " <<
802 table_i = (int) a_motif[seq_i];
803 rev_comp += conversionTable[table_i];
806 //std::cout << "seq: " << a_motif << std::endl;
807 //std::cout << "rc: " << rev_comp << std::endl;
813 Sequence::motif_normalize(const std::string& a_motif)
815 std::string valid_motif;
818 len = a_motif.length();
819 valid_motif.reserve(len);
821 // this just upcases IUPAC symbols. Eventually should return an error if non IUPAC is present.
822 // current nonIUPAC symbols are omitted, which is not reported atm
823 for(seq_i = 0; seq_i < len; seq_i++)
825 if ((a_motif[seq_i] == 'a') || (a_motif[seq_i] == 'A'))
827 else if ((a_motif[seq_i] == 't') || (a_motif[seq_i] == 'T'))
829 else if ((a_motif[seq_i] == 'g') || (a_motif[seq_i] == 'G'))
831 else if ((a_motif[seq_i] == 'c') || (a_motif[seq_i] == 'C'))
833 else if ((a_motif[seq_i] == 'n') || (a_motif[seq_i] == 'N'))
835 else if ((a_motif[seq_i] == 'm') || (a_motif[seq_i] == 'M'))
837 else if ((a_motif[seq_i] == 'r') || (a_motif[seq_i] == 'R'))
839 else if ((a_motif[seq_i] == 'w') || (a_motif[seq_i] == 'W'))
841 else if ((a_motif[seq_i] == 's') || (a_motif[seq_i] == 'S'))
843 else if ((a_motif[seq_i] == 'y') || (a_motif[seq_i] == 'Y'))
845 else if ((a_motif[seq_i] == 'k') || (a_motif[seq_i] == 'K'))
847 else if ((a_motif[seq_i] == 'v') || (a_motif[seq_i] == 'V'))
849 else if ((a_motif[seq_i] == 'h') || (a_motif[seq_i] == 'H'))
851 else if ((a_motif[seq_i] == 'd') || (a_motif[seq_i] == 'D'))
853 else if ((a_motif[seq_i] == 'b') || (a_motif[seq_i] == 'B'))
856 std::string msg = "Letter ";
857 msg += a_motif[seq_i];
858 msg += " is not a valid IUPAC symbol";
859 throw motif_normalize_error(msg);
862 //std::cout << "valid_motif is: " << valid_motif << std::endl;
866 void Sequence::add_motif(const Sequence& a_motif)
868 std::vector<int> motif_starts = find_motif(a_motif);
870 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
871 motif_start_i != motif_starts.end();
874 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
878 void Sequence::clear_motifs()
883 const std::list<motif>& Sequence::motifs() const
889 Sequence::find_motif(const std::string& a_motif) const
891 std::vector<int> motif_match_starts;
892 std::string norm_motif_rc;
894 motif_match_starts.clear();
896 //std::cout << "motif is: " << a_motif << std::endl;
897 std::string norm_motif = motif_normalize(a_motif);
898 //std::cout << "motif is: " << a_motif << std::endl;
900 if (norm_motif.size() > 0)
902 //std::cout << "Sequence: none blank motif\n";
903 motif_scan(norm_motif, &motif_match_starts);
905 norm_motif_rc = rc_motif(a_motif);
906 // make sure not to do search again if it is a palindrome
907 if (norm_motif_rc != norm_motif) {
908 motif_scan(norm_motif_rc, &motif_match_starts);
911 return motif_match_starts;
915 Sequence::find_motif(const Sequence& a_motif) const
917 return find_motif(a_motif.get_sequence());
921 Sequence::motif_scan(std::string a_motif, std::vector<int> * motif_match_starts) const
923 // if there's no sequence we can't scan for it?
924 // should this throw an exception?
927 std::string::const_iterator seq_c = seq->begin();
928 std::string::size_type seq_i;
929 int motif_i, motif_len;
931 //std::cout << "Sequence: motif, seq len = " << sequence.length() << std::endl;
932 motif_len = a_motif.length();
934 //std::cout << "motif_length: " << motif_len << std::endl;
935 //std::cout << "RAAARRRRR\n";
939 //std::cout << "motif: " << a_motif << std::endl;
941 //std::cout << "Sequence: motif, length= " << length << std::endl;
943 while (seq_i < length())
945 //std::cout << seq_c[seq_i];
946 //std::cout << seq_c[seq_i] << "?" << a_motif[motif_i] << ":" << motif_i << " ";
947 // this is pretty much a straight translation of Nora's python code
948 // to match iupac letter codes
949 if (a_motif[motif_i] =='N')
951 else if (a_motif[motif_i] == seq_c[seq_i])
953 else if ((a_motif[motif_i] =='M') &&
954 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C')))
956 else if ((a_motif[motif_i] =='R') &&
957 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G')))
959 else if ((a_motif[motif_i] =='W') &&
960 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='T')))
962 else if ((a_motif[motif_i] =='S') &&
963 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G')))
965 else if ((a_motif[motif_i] =='Y') &&
966 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='T')))
968 else if ((a_motif[motif_i] =='K') &&
969 ((seq_c[seq_i]=='G') || (seq_c[seq_i]=='T')))
971 else if ((a_motif[motif_i] =='V') &&
972 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
973 (seq_c[seq_i]=='G')))
975 else if ((a_motif[seq_i] =='H') &&
976 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
977 (seq_c[seq_i]=='T')))
979 else if ((a_motif[motif_i] =='D') &&
980 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G') ||
981 (seq_c[seq_i]=='T')))
983 else if ((a_motif[motif_i] =='B') &&
984 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G') ||
985 (seq_c[seq_i]=='T')))
994 // end Nora stuff, now we see if a match is found this pass
995 if (motif_i == motif_len)
999 motif_match_starts->push_back(seq_i - motif_len + 1);
1005 //std::cout << std::endl;
1008 void Sequence::add_string_annotation(std::string a_seq,
1011 std::vector<int> seq_starts = find_motif(a_seq);
1013 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1015 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1016 seq_start_i != seq_starts.end();
1019 annots.push_back(annot(*seq_start_i,
1020 *seq_start_i+a_seq.size(),
1026 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1027 std::list<Sequence>::iterator end)
1029 while (start != end) {
1030 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1036 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1038 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1044 bool operator<(const Sequence& x, const Sequence& y)
1046 Sequence::const_iterator x_i = x.begin();
1047 Sequence::const_iterator y_i = y.begin();
1048 // for sequences there's some computation associated with computing .end
1049 // so lets cache it.
1050 Sequence::const_iterator xend = x.end();
1051 Sequence::const_iterator yend = y.end();
1053 if( x_i == xend and y_i == yend ) {
1055 } else if ( x_i == xend ) {
1057 } else if ( y_i == yend ) {
1059 } else if ( (*x_i) < (*y_i)) {
1061 } else if ( (*x_i) > (*y_i) ) {
1070 bool operator==(const Sequence& x, const Sequence& y)
1072 if (x.empty() and y.empty()) {
1073 // if there's no sequence in either sequence structure, they're equal
1075 } else if (x.empty() or y.empty()) {
1076 // if we fail the first test, and we discover one is empty,
1077 // we know they can't be equal. (and we need to do this
1078 // to prevent dereferencing an empty pointer)
1080 } else if (x.seq_count != y.seq_count) {
1081 // if they're of different lenghts, they're not equal
1084 Sequence::const_iterator xseq_i = x.begin();
1085 Sequence::const_iterator yseq_i = y.begin();
1086 // since the length of the two sequences is equal, we only need to
1088 for(; xseq_i != x.end(); ++xseq_i, ++yseq_i) {
1089 if (*xseq_i != *yseq_i) {