void parse_rg_sample_map(char* hts_header, map<string, string>& rg_sample) {
string header(hts_header);
vector<string> header_lines = split_delims(header, "\n");
for (auto& line : header_lines) {
// get next line from header, skip if empty
if ( line.empty() ) { continue; }
// lines of the header look like:
// "@RG ID:- SM:NA11832 CN:BCM PL:454"
// ^^^^^^^\ is our sample name
if (line.find("@RG") == 0) {
vector<string> rg_parts = split_delims(line, "\t ");
string name;
string rg_id;
for (auto& part : rg_parts) {
size_t colpos = part.find(":");
if (colpos != string::npos) {
string fieldname = part.substr(0, colpos);
if (fieldname == "SM") {
name = part.substr(colpos+1);
} else if (fieldname == "ID") {
rg_id = part.substr(colpos+1);
}
}
}
if (name.empty()) {
cerr << "[vg::alignment] Error: could not find 'SM' in @RG line " << endl << line << endl;
exit(1);
}
if (rg_id.empty()) {
cerr << "[vg::alignment] Error: could not find 'ID' in @RG line " << endl << line << endl;
exit(1);
}
map<string, string>::iterator s = rg_sample.find(rg_id);
if (s != rg_sample.end()) {
if (s->second != name) {
cerr << "[vg::alignment] Error: multiple samples (SM) map to the same read group (RG)" << endl
<< endl
<< "samples " << name << " and " << s->second << " map to " << rg_id << endl
<< endl
<< "It will not be possible to determine what sample an alignment belongs to" << endl
<< "at runtime." << endl
<< endl
<< "To resolve the issue, ensure that RG ids are unique to one sample" << endl
<< "across all the input files to freebayes." << endl
<< endl
<< "See bamaddrg (https://github.com/ekg/bamaddrg) for a method which can" << endl
<< "add RG tags to alignments." << endl;
exit(1);
}
}
// if it's the same sample name and RG combo, no worries
rg_sample[rg_id] = name;
}
}
}
std::vector<std::string> split_delims(const std::string &s, const std::string& delims) {
std::vector<std::string> elems;
return split_delims(s, delims, elems);
}