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");
96 Sequence::Sequence(const char *seq)
99 strand(UnknownStrand),
103 set_filtered_sequence(seq);
106 Sequence::Sequence(const std::string& seq)
109 strand(UnknownStrand),
113 set_filtered_sequence(seq);
116 Sequence::Sequence(const Sequence& o)
118 seq_start(o.seq_start),
119 seq_count(o.seq_count),
124 motif_list(o.motif_list)
128 Sequence &Sequence::operator=(const Sequence& s)
132 seq_start = s.seq_start;
133 seq_count = s.seq_count;
138 motif_list = s.motif_list;
143 static void multiplatform_getline(std::istream& in, std::string& line)
148 while(in.good() and !(c == '\012' or c == '\015') ) {
152 // if we have cr-lf eat it
154 if (c=='\012' or c == '\015') {
159 //! load a fasta file into a sequence
161 * \param file_path the location of the fasta file in the filesystem
162 * \param seq_num which sequence in the file to load
163 * \param start_index starting position in the fasta sequence, 0 for beginning
164 * \param end_index ending position in the fasta sequence, 0 for end
165 * \return error message, empty string if no error. (gag!)
167 void Sequence::load_fasta(fs::path file_path, int seq_num,
168 int start_index, int end_index)
170 fs::fstream data_file;
171 data_file.open(file_path, std::ios::in);
173 if (!data_file.good())
175 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
178 load_fasta(data_file, seq_num, start_index, end_index);
179 } catch(sequence_empty_error e) {
180 // there doesn't appear to be any sequence
181 // catch and rethrow to include the filename
182 std::stringstream msg;
183 msg << "The selected sequence in "
184 << file_path.native_file_string()
185 << " appears to be empty";
186 throw sequence_empty_error(msg.str());
187 } catch(sequence_empty_file_error e) {
188 std::stringstream errormsg;
189 errormsg << file_path.native_file_string()
190 << " did not have any fasta sequences" << std::endl;
191 throw sequence_empty_file_error(errormsg.str());
197 Sequence::load_fasta(std::iostream& data_file, int seq_num,
198 int start_index, int end_index)
200 std::string file_data_line;
201 int header_counter = 0;
202 bool read_seq = true;
203 std::string rev_comp;
204 std::string sequence_raw;
205 std::string seq_tmp; // holds sequence during basic filtering
208 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
211 // search for the header of the fasta sequence we want
212 while ( (!data_file.eof()) && (header_counter < seq_num) )
214 multiplatform_getline(data_file, file_data_line);
215 if (file_data_line.substr(0,1) == ">")
219 if (header_counter > 0) {
220 header = file_data_line.substr(1);
224 while ( !data_file.eof() && read_seq ) {
225 multiplatform_getline(data_file,file_data_line);
226 if (file_data_line.substr(0,1) == ">")
228 else sequence_raw += file_data_line;
231 // Lastly, if subselection of the sequence was specified we keep cut out
232 // and only keep that part
233 // end_index = 0 means no end was specified, so cut to the end
235 end_index = sequence_raw.size();
237 // sequence filtering for upcasing agctn and convert non AGCTN to N
238 if (end_index-start_index <= 0) {
239 std::string msg("The selected sequence appears to be empty");
240 throw sequence_empty_error(msg);
242 set_filtered_sequence(sequence_raw, start_index, end_index-start_index);
244 std::string errormsg("There were no fasta sequences");
245 throw sequence_empty_file_error(errormsg);
249 void Sequence::set_filtered_sequence(const std::string &old_seq,
254 char conversionTable[257];
257 count = old_seq.size() - start;
258 boost::shared_ptr<seq_string> new_seq(new seq_string);
259 new_seq->reserve(count);
261 // Make a conversion table
263 // everything we don't specify below will become 'N'
264 for(int table_i=0; table_i < 256; table_i++)
266 conversionTable[table_i] = 'N';
268 // add end of string character for printing out table for testing purposes
269 conversionTable[256] = '\0';
271 // we want these to map to themselves - ie not to change
272 conversionTable[(int)'A'] = 'A';
273 conversionTable[(int)'T'] = 'T';
274 conversionTable[(int)'G'] = 'G';
275 conversionTable[(int)'C'] = 'C';
277 conversionTable[(int)'a'] = 'A';
278 conversionTable[(int)'t'] = 'T';
279 conversionTable[(int)'g'] = 'G';
280 conversionTable[(int)'c'] = 'C';
282 // finally, the actual conversion loop
283 for(std::string::size_type seq_index = 0; seq_index < count; seq_index++)
285 new_seq->append(1, conversionTable[ (int)old_seq[seq_index+start]]);
294 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
296 fs::fstream data_stream(file_path, std::ios::in);
299 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
302 // so i should probably be passing the parse function some iterators
303 // but the annotations files are (currently) small, so i think i can
304 // get away with loading the whole file into memory
307 while(data_stream.good()) {
313 parse_annot(data, start_index, end_index);
316 /* If this works, yikes, this is some brain hurting code.
318 * what's going on is that when pb_annot is instantiated it stores references
319 * to begin, end, name, type, declared in the parse function, then
320 * when operator() is called it grabs values from those references
321 * and uses that to instantiate an annot object and append that to our
324 * This weirdness is because the spirit library requires that actions
325 * conform to a specific prototype operator()(IteratorT, IteratorT)
326 * which doesn't provide any useful opportunity for me to actually
327 * grab the results of our parsing.
329 * so I instantiate this structure in order to have a place to grab
333 struct push_back_annot {
334 std::list<annot>& annot_list;
340 push_back_annot(std::list<annot>& annot_list_,
345 : annot_list(annot_list_),
353 void operator()(std::string::const_iterator,
354 std::string::const_iterator) const
356 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
357 annot_list.push_back(annot(begin, end, name, type));
361 struct push_back_seq {
362 std::list<Sequence>& seq_list;
366 push_back_seq(std::list<Sequence>& seq_list_,
369 : seq_list(seq_list_),
375 void operator()(std::string::const_iterator,
376 std::string::const_iterator) const
378 // filter out newlines from our sequence
380 for(std::string::const_iterator seq_i = seq.begin();
384 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
386 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
389 s.set_fasta_header(name);
390 seq_list.push_back(s);
395 Sequence::parse_annot(std::string data, int start_index, int end_index)
402 std::list<Sequence> query_seqs;
404 bool status = spirit::parse(data.begin(), data.end(),
411 )[spirit::assign_a(species)] >>
415 ( // ignore html tags
416 *(spirit::space_p) >>
418 +(~spirit::ch_p('>')) >>
423 ( // parse an absolute location name
424 (spirit::uint_p[spirit::assign_a(start)] >>
426 spirit::uint_p[spirit::assign_a(end)] >>
431 )[spirit::assign_a(name)] >>
438 )[spirit::assign_a(type)]
440 // to understand how this group gets set
441 // read the comment above struct push_back_annot
442 )[push_back_annot(annots, start, end, type, name)]
444 ((spirit::ch_p('>')|spirit::str_p(">")) >>
445 (*(spirit::print_p))[spirit::assign_a(name)] >>
447 (+(spirit::chset<>(nucleic_iupac_alphabet.c_str())))[spirit::assign_a(seq)]
448 )[push_back_seq(query_seqs, name, seq)]
455 // go seearch for query sequences
456 find_sequences(query_seqs.begin(), query_seqs.end());
460 void Sequence::add_annotation(const annot& a)
465 const std::list<annot>& Sequence::annotations() const
471 Sequence::subseq(int start, int count) const
478 // there might be an off by one error with start+count > size()
479 if ( count == npos || start+count > size()) {
480 count = size()-start;
482 Sequence new_seq(*this);
483 new_seq.seq_start = start;
484 new_seq.seq_count = count;
486 new_seq.motif_list = motif_list;
487 new_seq.annots.clear();
488 // attempt to copy & reannotate position based annotations
489 int end = start+count;
491 for(std::list<annot>::const_iterator annot_i = annots.begin();
492 annot_i != annots.end();
495 int annot_begin= annot_i->begin;
496 int annot_end = annot_i->end;
498 if (annot_begin < end) {
499 if (annot_begin >= start) {
500 annot_begin -= start;
505 if (annot_end < end) {
511 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
512 new_seq.annots.push_back(new_annot);
520 Sequence::rev_comp() const
522 std::string rev_comp;
523 char conversionTable[257];
524 int seq_i, table_i, len;
527 rev_comp.reserve(len);
528 // make a conversion table
529 // init all parts of conversion table to '~' character
530 // '~' I doubt will ever appear in a sequence file (jeez, I hope)
531 // and may the fleas of 1000 camels infest the genitals of any biologist (and
532 // seven generations of their progeny) who decides to make it mean
533 // something special!!!
534 // PS - double the curse for any smartass non-biologist who tries it as well
535 for(table_i=0; table_i < 256; table_i++)
537 conversionTable[table_i] = '~';
539 // add end of string character for printing out table for testing purposes
540 conversionTable[256] = '\0';
542 // add in the characters for the bases we want to convert
543 conversionTable[(int)'A'] = 'T';
544 conversionTable[(int)'T'] = 'A';
545 conversionTable[(int)'G'] = 'C';
546 conversionTable[(int)'C'] = 'G';
547 conversionTable[(int)'N'] = 'N';
549 // finally, the actual conversion loop
550 for(seq_i = len - 1; seq_i >= 0; seq_i--)
552 table_i = (int) at(seq_i);
553 rev_comp += conversionTable[table_i];
559 void Sequence::set_fasta_header(std::string header_)
564 void Sequence::set_species(const std::string& name)
569 std::string Sequence::get_species() const
576 Sequence::get_fasta_header() const
582 Sequence::get_name() const
584 if (header.size() > 0)
586 else if (species.size() > 0)
592 void Sequence::set_sequence(const std::string& s)
594 set_filtered_sequence(s);
597 std::string Sequence::get_sequence() const
602 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
607 Sequence::const_reference Sequence::at(Sequence::size_type i) const
609 if (!seq) throw std::out_of_range("empty sequence");
610 return seq->at(i+seq_start);
619 strand = UnknownStrand;
626 const char *Sequence::c_str() const
629 return seq->c_str()+seq_start;
634 Sequence::const_iterator Sequence::begin() const
636 if (seq and seq_count != 0)
637 return seq->begin()+seq_start;
639 return Sequence::const_iterator(0);
642 Sequence::const_iterator Sequence::end() const
644 if (seq and seq_count != 0) {
645 return seq->begin() + seq_start + seq_count;
647 return Sequence::const_iterator(0);
651 bool Sequence::empty() const
653 return (seq_count == 0) ? true : false;
656 Sequence::size_type Sequence::start() const
661 Sequence::size_type Sequence::stop() const
663 return seq_start + seq_count;
666 Sequence::size_type Sequence::size() const
671 Sequence::size_type Sequence::length() const
677 Sequence::save(fs::fstream &save_file)
680 std::list<annot>::iterator annots_i;
682 // not sure why, or if i'm doing something wrong, but can't seem to pass
683 // file pointers down to this method from the mussa control class
684 // so each call to save a sequence appends to the file started by mussa_class
685 //save_file.open(save_file_path.c_str(), std::ios::app);
687 save_file << "<Sequence>" << std::endl;
688 save_file << *this << std::endl;
689 save_file << "</Sequence>" << std::endl;
691 save_file << "<Annotations>" << std::endl;
692 save_file << species << std::endl;
693 for (annots_i = annots.begin(); annots_i != annots.end(); ++annots_i)
695 save_file << annots_i->begin << " " << annots_i->end << " " ;
696 save_file << annots_i->name << " " << annots_i->type << std::endl;
698 save_file << "</Annotations>" << std::endl;
703 Sequence::load_museq(fs::path load_file_path, int seq_num)
705 fs::fstream load_file;
706 std::string file_data_line;
709 std::string::size_type space_split_i;
710 std::string annot_value;
713 load_file.open(load_file_path, std::ios::in);
716 // search for the seq_num-th sequence
717 while ( (!load_file.eof()) && (seq_counter < seq_num) )
719 getline(load_file,file_data_line);
720 if (file_data_line == "<Sequence>")
723 getline(load_file, file_data_line);
724 // looks like the sequence is written as a single line
725 set_filtered_sequence(file_data_line);
726 getline(load_file, file_data_line);
727 getline(load_file, file_data_line);
728 if (file_data_line == "<Annotations>")
730 getline(load_file, file_data_line);
731 species = file_data_line;
732 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
734 getline(load_file,file_data_line);
735 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
737 // need to get 4 values...almost same code 4 times...
738 // get annot start index
739 space_split_i = file_data_line.find(" ");
740 annot_value = file_data_line.substr(0,space_split_i);
741 an_annot.begin = atoi (annot_value.c_str());
742 file_data_line = file_data_line.substr(space_split_i+1);
743 // get annot end index
744 space_split_i = file_data_line.find(" ");
745 annot_value = file_data_line.substr(0,space_split_i);
746 an_annot.end = atoi (annot_value.c_str());
748 if (space_split_i == std::string::npos) // no entry for type or name
750 std::cout << "seq, annots - no type or name\n";
754 else // else get annot type
756 file_data_line = file_data_line.substr(space_split_i+1);
757 space_split_i = file_data_line.find(" ");
758 annot_value = file_data_line.substr(0,space_split_i);
759 an_annot.type = annot_value;
760 if (space_split_i == std::string::npos) // no entry for name
762 std::cout << "seq, annots - no name\n";
765 else // get annot name
767 file_data_line = file_data_line.substr(space_split_i+1);
768 space_split_i = file_data_line.find(" ");
769 annot_value = file_data_line.substr(0,space_split_i);
770 an_annot.type = annot_value;
773 annots.push_back(an_annot); // don't forget to actually add the annot
775 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
776 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
783 Sequence::rc_motif(std::string a_motif) const
785 std::string rev_comp;
786 char conversionTable[257];
787 int seq_i, table_i, len;
789 len = a_motif.length();
790 rev_comp.reserve(len);
792 for(table_i=0; table_i < 256; table_i++)
794 conversionTable[table_i] = '~';
796 // add end of std::string character for printing out table for testing purposes
797 conversionTable[256] = '\0';
799 // add in the characters for the bases we want to convert (IUPAC)
800 conversionTable[(int)'A'] = 'T';
801 conversionTable[(int)'T'] = 'A';
802 conversionTable[(int)'G'] = 'C';
803 conversionTable[(int)'C'] = 'G';
804 conversionTable[(int)'N'] = 'N';
805 conversionTable[(int)'M'] = 'K';
806 conversionTable[(int)'R'] = 'Y';
807 conversionTable[(int)'W'] = 'W';
808 conversionTable[(int)'S'] = 'S';
809 conversionTable[(int)'Y'] = 'R';
810 conversionTable[(int)'K'] = 'M';
811 conversionTable[(int)'V'] = 'B';
812 conversionTable[(int)'H'] = 'D';
813 conversionTable[(int)'D'] = 'H';
814 conversionTable[(int)'B'] = 'V';
816 // finally, the actual conversion loop
817 for(seq_i = len - 1; seq_i >= 0; seq_i--)
819 //std::cout << "** i = " << seq_i << " bp = " <<
820 table_i = (int) a_motif[seq_i];
821 rev_comp += conversionTable[table_i];
824 //std::cout << "seq: " << a_motif << std::endl;
825 //std::cout << "rc: " << rev_comp << std::endl;
831 Sequence::motif_normalize(const std::string& a_motif)
833 std::string valid_motif;
836 len = a_motif.length();
837 valid_motif.reserve(len);
839 // this just upcases IUPAC symbols. Eventually should return an error if non IUPAC is present.
840 // current nonIUPAC symbols are omitted, which is not reported atm
841 for(seq_i = 0; seq_i < len; seq_i++)
843 if ((a_motif[seq_i] == 'a') || (a_motif[seq_i] == 'A'))
845 else if ((a_motif[seq_i] == 't') || (a_motif[seq_i] == 'T'))
847 else if ((a_motif[seq_i] == 'g') || (a_motif[seq_i] == 'G'))
849 else if ((a_motif[seq_i] == 'c') || (a_motif[seq_i] == 'C'))
851 else if ((a_motif[seq_i] == 'n') || (a_motif[seq_i] == 'N'))
853 else if ((a_motif[seq_i] == 'm') || (a_motif[seq_i] == 'M'))
855 else if ((a_motif[seq_i] == 'r') || (a_motif[seq_i] == 'R'))
857 else if ((a_motif[seq_i] == 'w') || (a_motif[seq_i] == 'W'))
859 else if ((a_motif[seq_i] == 's') || (a_motif[seq_i] == 'S'))
861 else if ((a_motif[seq_i] == 'y') || (a_motif[seq_i] == 'Y'))
863 else if ((a_motif[seq_i] == 'k') || (a_motif[seq_i] == 'K'))
865 else if ((a_motif[seq_i] == 'v') || (a_motif[seq_i] == 'V'))
867 else if ((a_motif[seq_i] == 'h') || (a_motif[seq_i] == 'H'))
869 else if ((a_motif[seq_i] == 'd') || (a_motif[seq_i] == 'D'))
871 else if ((a_motif[seq_i] == 'b') || (a_motif[seq_i] == 'B'))
874 std::string msg = "Letter ";
875 msg += a_motif[seq_i];
876 msg += " is not a valid IUPAC symbol";
877 throw motif_normalize_error(msg);
880 //std::cout << "valid_motif is: " << valid_motif << std::endl;
884 void Sequence::add_motif(const Sequence& a_motif)
886 std::vector<int> motif_starts = find_motif(a_motif);
888 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
889 motif_start_i != motif_starts.end();
892 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
896 void Sequence::clear_motifs()
901 const std::list<motif>& Sequence::motifs() const
907 Sequence::find_motif(const std::string& a_motif) const
909 std::vector<int> motif_match_starts;
910 std::string norm_motif_rc;
912 motif_match_starts.clear();
914 //std::cout << "motif is: " << a_motif << std::endl;
915 std::string norm_motif = motif_normalize(a_motif);
916 //std::cout << "motif is: " << a_motif << std::endl;
918 if (norm_motif.size() > 0)
920 //std::cout << "Sequence: none blank motif\n";
921 motif_scan(norm_motif, &motif_match_starts);
923 norm_motif_rc = rc_motif(a_motif);
924 // make sure not to do search again if it is a palindrome
925 if (norm_motif_rc != norm_motif) {
926 motif_scan(norm_motif_rc, &motif_match_starts);
929 return motif_match_starts;
933 Sequence::find_motif(const Sequence& a_motif) const
935 return find_motif(a_motif.get_sequence());
939 Sequence::motif_scan(std::string a_motif, std::vector<int> * motif_match_starts) const
941 // if there's no sequence we can't scan for it?
942 // should this throw an exception?
945 std::string::const_iterator seq_c = seq->begin();
946 std::string::size_type seq_i;
947 int motif_i, motif_len;
949 //std::cout << "Sequence: motif, seq len = " << sequence.length() << std::endl;
950 motif_len = a_motif.length();
952 //std::cout << "motif_length: " << motif_len << std::endl;
953 //std::cout << "RAAARRRRR\n";
957 //std::cout << "motif: " << a_motif << std::endl;
959 //std::cout << "Sequence: motif, length= " << length << std::endl;
961 while (seq_i < length())
963 //std::cout << seq_c[seq_i];
964 //std::cout << seq_c[seq_i] << "?" << a_motif[motif_i] << ":" << motif_i << " ";
965 // this is pretty much a straight translation of Nora's python code
966 // to match iupac letter codes
967 if (a_motif[motif_i] =='N')
969 else if (a_motif[motif_i] == seq_c[seq_i])
971 else if ((a_motif[motif_i] =='M') &&
972 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C')))
974 else if ((a_motif[motif_i] =='R') &&
975 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G')))
977 else if ((a_motif[motif_i] =='W') &&
978 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='T')))
980 else if ((a_motif[motif_i] =='S') &&
981 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G')))
983 else if ((a_motif[motif_i] =='Y') &&
984 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='T')))
986 else if ((a_motif[motif_i] =='K') &&
987 ((seq_c[seq_i]=='G') || (seq_c[seq_i]=='T')))
989 else if ((a_motif[motif_i] =='V') &&
990 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
991 (seq_c[seq_i]=='G')))
993 else if ((a_motif[seq_i] =='H') &&
994 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
995 (seq_c[seq_i]=='T')))
997 else if ((a_motif[motif_i] =='D') &&
998 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G') ||
999 (seq_c[seq_i]=='T')))
1001 else if ((a_motif[motif_i] =='B') &&
1002 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G') ||
1003 (seq_c[seq_i]=='T')))
1012 // end Nora stuff, now we see if a match is found this pass
1013 if (motif_i == motif_len)
1015 //std::cout << "!!";
1017 motif_match_starts->push_back(seq_i - motif_len + 1);
1023 //std::cout << std::endl;
1026 void Sequence::add_string_annotation(std::string a_seq,
1029 std::vector<int> seq_starts = find_motif(a_seq);
1031 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1033 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1034 seq_start_i != seq_starts.end();
1037 annots.push_back(annot(*seq_start_i,
1038 *seq_start_i+a_seq.size(),
1044 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1045 std::list<Sequence>::iterator end)
1047 while (start != end) {
1048 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1054 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1056 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1062 bool operator<(const Sequence& x, const Sequence& y)
1064 Sequence::const_iterator x_i = x.begin();
1065 Sequence::const_iterator y_i = y.begin();
1066 // for sequences there's some computation associated with computing .end
1067 // so lets cache it.
1068 Sequence::const_iterator xend = x.end();
1069 Sequence::const_iterator yend = y.end();
1071 if( x_i == xend and y_i == yend ) {
1073 } else if ( x_i == xend ) {
1075 } else if ( y_i == yend ) {
1077 } else if ( (*x_i) < (*y_i)) {
1079 } else if ( (*x_i) > (*y_i) ) {
1088 bool operator==(const Sequence& x, const Sequence& y)
1090 if (x.empty() and y.empty()) {
1091 // if there's no sequence in either sequence structure, they're equal
1093 } else if (x.empty() or y.empty()) {
1094 // if we fail the first test, and we discover one is empty,
1095 // we know they can't be equal. (and we need to do this
1096 // to prevent dereferencing an empty pointer)
1098 } else if (x.seq_count != y.seq_count) {
1099 // if they're of different lenghts, they're not equal
1102 Sequence::const_iterator xseq_i = x.begin();
1103 Sequence::const_iterator yseq_i = y.begin();
1104 // since the length of the two sequences is equal, we only need to
1106 for(; xseq_i != x.end(); ++xseq_i, ++yseq_i) {
1107 if (*xseq_i != *yseq_i) {