BCFOrderedReader::BCFOrderedReader(std::string file_name, std::vector<GenomeInterval>& intervals)
{
this->file_name = (file_name=="+")? "-" : file_name;
file = NULL;
hdr = NULL;
idx = NULL;
tbx = NULL;
itr = NULL;
this->intervals = intervals;
interval_index = 0;
index_loaded = false;
file = hts_open(this->file_name.c_str(), "r");
if (!file)
{
fprintf(stderr, "[%s:%d %s] Cannot open %s\n", __FILE__, __LINE__, __FUNCTION__, file_name.c_str());
exit(1);
}
ftype = file->format;
if (ftype.format!=vcf && ftype.format!=bcf)
{
fprintf(stderr, "[%s:%d %s] Not a VCF/BCF file: %s\n", __FILE__, __LINE__, __FUNCTION__, file_name.c_str());
exit(1);
}
s = {0, 0, 0};
if (file==NULL) exit(1);
hdr = bcf_alt_hdr_read(file);
if (!hdr) exit(1);
intervals_present = intervals.size()!=0;
if (ftype.format==bcf)
{
if ((idx = bcf_index_load(file_name.c_str())))
{
index_loaded = true;
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] index cannot be loaded for %s for random access, ignoring specified intervals and reading from start.\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
else if (ftype.format==vcf)
{
if (ftype.compression==bgzf)
{
if ((tbx = tbx_index_load(file_name.c_str())))
{
index_loaded = true;
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] index cannot be loaded for %s for random access, ignoring specified intervals and reading from start.\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
else
{
if (intervals_present)
{
fprintf(stderr, "[E:%s] no random access support for VCF file: %s\n", __FUNCTION__, file_name.c_str());
// exit(1);
}
}
}
random_access_enabled = intervals_present && index_loaded;
};
/**
* Constructor.
*
* @intervals - if empty, will add the contigs found in the header files
*/
BCFSyncedStreamReader::BCFSyncedStreamReader(std::vector<std::string>& vcf_files, std::vector<GenomeInterval>& intervals, bool sync_by_pos)
:vcf_files(vcf_files), intervals(intervals), sync_by_pos(sync_by_pos)
{
nfiles = vcf_files.size();
vcfs.resize(nfiles, 0);
hdrs.resize(nfiles, 0);
idxs.resize(nfiles, 0);
tbxs.resize(nfiles, 0);
itrs.resize(nfiles, 0);
ftypes.resize(nfiles, -1);
current_interval = "";
current_pos1 = 0;
buffer.resize(nfiles);
s = {0, 0, 0};
exists_selected_intervals = (intervals.size()!=0);
for (uint32_t i=0; i<intervals.size(); ++i)
{
intervals_map[intervals[i].to_string()] = i;
}
intervals_index = 0;
//1. check file type validity
//2. loads indices
//3. adds sequences found in all indexed files, this allows us to iterate through all sequences.
for (int32_t i = 0; i<nfiles; ++i)
{
ftypes[i] = hts_file_type(vcf_files[i].c_str());
vcfs[i] = bcf_open(vcf_files[i].c_str(), "r");
if (vcfs[i]==NULL) exit(1);
hdrs[i] = bcf_alt_hdr_read(vcfs[i]);
if (i==0)
{
if (!(ftypes[i] & (FT_VCF|FT_BCF|FT_STDIN)))
{
fprintf(stderr, "[E:%s:%d %s] %s not a VCF or BCF file\n", __FILE__, __LINE__, __FUNCTION__, vcf_files[i].c_str());
exit(1);
}
if (!load_index(i))
{
fprintf(stderr, "[I:%s:%d %s] index cannot be loaded for %s\n", __FILE__, __LINE__, __FUNCTION__, vcf_files[i].c_str());
exit(1);
}
if (!exists_selected_intervals)
{
//add sequences from file i
add_interval(i);
}
}
else
{
if (!(ftypes[i] & (FT_VCF_GZ|FT_BCF_GZ)))
{
fprintf(stderr, "[E:%s:%d %s] %s not a VCF_GZ or BCF file\n", __FILE__, __LINE__, __FUNCTION__, vcf_files[i].c_str());
exit(1);
}
if (!load_index(i))
{
fprintf(stderr, "[E:%s:%d %s] index cannot be loaded for %s\n", __FILE__, __LINE__, __FUNCTION__, vcf_files[i].c_str());
exit(1);
}
if (!exists_selected_intervals)
{
//add sequences from file i
add_interval(i);
}
}
}
}
/**
* Constructor.
*
* @intervals - if empty, will add the contigs found in the header files
*/
BCFSyncedReader::BCFSyncedReader(std::vector<std::string>& file_names, std::vector<GenomeInterval>& intervals, bool sync_by_pos)
:file_names(file_names), intervals(intervals), sync_by_pos(sync_by_pos)
{
nfiles = file_names.size();
files.resize(nfiles, 0);
hdrs.resize(nfiles, 0);
idxs.resize(nfiles, 0);
tbxs.resize(nfiles, 0);
itrs.resize(nfiles, 0);
ftypes.resize(nfiles);
current_interval = "";
current_pos1 = 0;
buffer.resize(nfiles);
s = {0, 0, 0};
random_access = (intervals.size()!=0);
for (size_t i=0; i<intervals.size(); ++i)
{
intervals_map[intervals[i].to_string()] = i;
}
intervals_index = 0;
uint32_t no_stdins = 0;
for (size_t i = 0; i<nfiles; ++i)
{
if (file_names[0]=="+")
{
file_names[0]="-";
++no_stdins;
}
else if (file_names[0]=="-")
{
file_names[0]="-";
++no_stdins;
}
if (no_stdins>1)
{
fprintf(stderr, "[E:%s:%d %s] BCFSyncedReader does not support reading from more than one STDIN stream\n", __FILE__, __LINE__, __FUNCTION__);
exit(1);
}
files[i] = hts_open(file_names[i].c_str(), "r");
if (files[i]==NULL)
{
fprintf(stderr, "[%s:%d %s] Cannot open %s\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
ftypes[i] = files[i]->format;
//check format
if (ftypes[i].format!=vcf && ftypes[i].format!=bcf)
{
fprintf(stderr, "[E:%s:%d %s] %s not a VCF or BCF file\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
//read header
hdrs[i] = bcf_alt_hdr_read(files[i]);
if (!hdrs[i])
{
fprintf(stderr, "[E:%s:%d %s] header cannot be read for %s\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
//load index if intervals are specified
if (random_access && !load_index(i))
{
fprintf(stderr, "[E:%s:%d %s] index cannot be loaded for %s for random access\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
//check contigs consistency
if (i)
{
int32_t nseqs0;
const char ** seqnames0 = bcf_hdr_seqnames(hdrs[i], &nseqs0);
int32_t nseqs;
const char ** seqnames = bcf_hdr_seqnames(hdrs[i], &nseqs);
if (nseqs0==0 || nseqs==0 || nseqs0!=nseqs)
{
fprintf(stderr, "[E:%s:%d %s] contigs in header not consistent with first file for %s\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
for (size_t j=0; j<nseqs; ++j)
{
if (strcmp(seqnames0[j], seqnames[j]))
{
fprintf(stderr, "[E:%s:%d %s] contigs in header not consistent with first file for %s\n", __FILE__, __LINE__, __FUNCTION__, file_names[i].c_str());
exit(1);
}
}
free(seqnames0);
free(seqnames);
}
}
}
