bool ribo_profile::assign_reads(fp_list_t& fp_base_list, const unordered_set<int>& type_set)
{
for (auto& r: fp_base_list) {
// round 1: get candidate alignment loci
if (r.used) continue;
vector<position> loci;
for (auto loc: r.al_loci) {
rid_t refID(loc.refID); //transcript index
rid_t t(get_transcript_index(refID));
if (loc.start<0 or loc.stop>len(t)) {
cout<<"profile index out of bound! readID:"<<*r.seqs.begin()<<" ";
cout<<r.al_loci.size()<<" "<<refID<<" "<<loc.start<<"-"<<loc.stop<<" "<<len(t)<<endl;
continue;
}
if (type_set.find(loc.type)!=type_set.end())
loci.emplace_back(loc);
} // for r.al_loci
// no mappings of the desired type, move on to the next read
if (loci.size() == 0) continue;
// read processed mark as used
r.used = true;
// round 2: get transcript abundance
vector<double> prob(loci.size(), 0);
double tot_prob(0);
for (size_t i=0; i!=prob.size(); ++i) {
const position& loc(loci[i]);
rid_t refID(loc.refID);
rid_t t(get_transcript_index(refID));
prob[i] = get_tot_abundance(t);
tot_prob += prob[i];
}
// round 3: asign reads to loci proportional to loci prob
for (size_t i=0; i!=prob.size(); ++i) {
const position& loc(loci[i]);
rid_t refID(loc.refID);
rid_t t(get_transcript_index(refID));
double count(r.count);
if (prob.size()!=1)
count *= prob[i] / tot_prob;
for (rid_t base = loc.start; base != loc.stop; ++base)
add_read_count(t,base, count);
add_tot_count(t, count);
}
}// for r: fp_base_list
return check_tot_count(fp_base_list);
}
size_t fread(void *ptr, size_t size, size_t nmemb,FILE *stream)
{
int realsize=size*nmemb;
int retsize;
orig_fwrite_f_type orig_fread;
//int gettimeofday(struct timeval *tv, struct timezone *tz);
struct timeval tv0;
struct timeval tv1;
struct timezone tz;
int fd;
//mtx.lock();
fd=fileno(stream);
orig_fread = (orig_fread_f_type)dlsym(RTLD_NEXT,"fread");
gettimeofday(&tv0,&tz);
retsize=orig_fread(ptr,size,nmemb,stream);
gettimeofday(&tv1,&tz);
std::string str;
std::ostringstream oss;
oss << "fread(" << ptr << "(" << fd << ")" << "," << size << "," << nmemb << "," << stream << ")="<< retsize << "\n";
log.add(oss.str());
if (size>=0)
{
if (myiio.existFd(fd)==-1)
{
Ifile ifi;
ifi.setFd(fd);
//ifi.setName(std::string("UNKNOWN-FWRITE"));
ifi.setName(getStrFDInfo(fd));
ifi.setState(IOBYFILE_WRITE);
myiio.iiof.push_back(ifi);
}
add_read_count(fd,retsize);
add_read_time(fd,tv0,tv1);
}
//mtx.unlock();
return retsize;
}
ssize_t read(int fd, void *buf, size_t count)
{
int size;
int size_t_count=count;
orig_read_f_type orig_read;
//int gettimeofday(struct timeval *tv, struct timezone *tz);
struct timeval tv0;
struct timeval tv1;
struct timezone tz;
//mtx.lock();
orig_read = (orig_read_f_type)dlsym(RTLD_NEXT,"read");
gettimeofday(&tv0,&tz);
size=orig_read(fd,buf,count);
gettimeofday(&tv1,&tz);
std::string str;
std::ostringstream oss;
oss << "read(" << fd << "," << buf << "," << count << ")\n";
log.add(oss.str());
if (size>=0)
{
if (myiio.existFd(fd)==-1)
{
Ifile ifi;
ifi.setFd(fd);
ifi.setName(getStrFDInfo(fd));
//ifi.setName(std::string("UNKNOWN-WRITE"));
ifi.setState(IOBYFILE_READ);
myiio.iiof.push_back(ifi);
}
add_read_count(fd,size);
add_read_time(fd,tv0,tv1);
}
//mtx.unlock();
return size;
}