#include <boost/filesystem/fstream.hpp>
namespace fs = boost::filesystem;
-#include <boost/spirit/core.hpp>
-#include <boost/spirit/actor/push_back_actor.hpp>
-#include <boost/spirit/iterator/file_iterator.hpp>
-#include <boost/spirit/utility/chset.hpp>
-namespace spirit = boost::spirit;
-
#include <iostream>
#include <sstream>
#include "mussa_exceptions.hpp"
-#include "alg/mussa.hpp"
#include "alg/flp.hpp"
+#include "alg/mussa.hpp"
+#include "alg/motif_parser.hpp"
using namespace std;
update_sequences_motifs();
}
-// Helper functor to append created motifs to our Mussa analysis
-struct push_back_motif {
- Mussa::motif_set& motifs;
- boost::shared_ptr<AnnotationColors> color_mapper;
- std::string& seq_string;
- std::string& name;
- float& red;
- float& green;
- float& blue;
- float& alpha;
- int& parsed;
-
- push_back_motif(Mussa::motif_set& motifs_,
- boost::shared_ptr<AnnotationColors> color_mapper_,
- std::string& seq_,
- std::string& name_,
- float &red_, float &green_, float &blue_, float &alpha_,
- int &parsed_)
- : motifs(motifs_),
- color_mapper(color_mapper_),
- seq_string(seq_),
- name(name_),
- red(red_),
- green(green_),
- blue(blue_),
- alpha(alpha_),
- parsed(parsed_)
- {
- }
-
- void operator()(std::string::const_iterator,
- std::string::const_iterator) const
- {
- Sequence seq(seq_string, Sequence::nucleic_alphabet);
- // shouldn't we have a better field than "fasta header" and speices?
- seq.set_fasta_header(name);
- // we need to clear the name in case the next motif doesn't have one.
- name.clear();
- // be nice if glsequence was a subclass of sequence so we could
- // just attach colors directly to the motif.
- Color c(red, green, blue, alpha);
- color_mapper->appendInstanceColor("motif", seq.c_str(), c);
- alpha = 1.0;
- motifs.insert(seq);
- ++parsed;
- };
-};
-
void Mussa::load_motifs(fs::path filename)
{
fs::ifstream f;
f.open(filename, ifstream::in);
load_motifs(f);
}
-
-// I mostly split the ifstream out so I can use a stringstream to test it.
+
void Mussa::load_motifs(std::istream &in)
{
std::string data;
- const char *alphabet = Alphabet::nucleic_alphabet.c_str();
- string seq;
- string name;
- float red = 1.0;
- float green = 0.0;
- float blue = 0.0;
- float alpha = 1.0;
- int parsed = 1;
+ const char *alphabet = Alphabet::nucleic_cstr;
+ motif_parser::ParsedMotifs parsed_motifs(motif_sequences, color_mapper);
// slurp our data into a string
std::streamsize bytes_read = 1;
bytes_read = in.readsome(buf, bufsiz);
data.append(buf, buf+bytes_read);
}
- // parse our string
- bool ok = spirit::parse(data.begin(), data.end(),
- *(
- (
- (
- (+spirit::chset<>(alphabet))[spirit::assign_a(seq)] >>
- +spirit::space_p
- ) >>
- !(
- (
- // names can either be letter followed by non-space characters
- (spirit::alpha_p >> *spirit::graph_p)[spirit::assign_a(name)]
- |
- // or a quoted string
- (
- spirit::ch_p('"') >>
- (+(~spirit::ch_p('"')))[spirit::assign_a(name)] >>
- spirit::ch_p('"')
- )
- ) >> +spirit::space_p
- ) >>
- spirit::real_p[spirit::assign_a(red)] >> +spirit::space_p >>
- spirit::real_p[spirit::assign_a(green)] >> +spirit::space_p >>
- spirit::real_p[spirit::assign_a(blue)] >> +spirit::space_p >>
- !(spirit::real_p[spirit::assign_a(alpha)] >> +spirit::space_p)
- )[push_back_motif(motif_sequences, color_mapper, seq, name, red, green, blue, alpha, parsed)]
- )).full;
- if (not ok) {
- stringstream msg;
- msg << "Error parsing motif #" << parsed;
- // erase our potentially broken motif list
- motif_sequences.clear();
- throw motif_load_error(msg.str());
- }
+ parsed_motifs.parse(data);
update_sequences_motifs();
}