// File name or a list of genomic locations. If file name, NULL is returned.
static bcf_sr_regions_t *_regions_init_string(const char *str)
{
bcf_sr_regions_t *reg = (bcf_sr_regions_t *) calloc(1, sizeof(bcf_sr_regions_t));
reg->start = reg->end = -1;
reg->prev_start = reg->prev_seq = -1;
kstring_t tmp = {0,0,0};
const char *sp = str, *ep = str;
int from, to;
while ( 1 )
{
while ( *ep && *ep!=',' && *ep!=':' ) ep++;
tmp.l = 0;
kputsn(sp,ep-sp,&tmp);
if ( *ep==':' )
{
sp = ep+1;
from = hts_parse_decimal(sp,(char**)&ep);
if ( sp==ep )
{
fprintf(stderr,"[%s:%d %s] Could not parse the region(s): %s\n", __FILE__,__LINE__,__FUNCTION__,str);
free(reg); free(tmp.s); return NULL;
}
if ( !*ep || *ep==',' )
{
_regions_add(reg, tmp.s, from, from);
sp = ep;
continue;
}
if ( *ep!='-' )
{
fprintf(stderr,"[%s:%d %s] Could not parse the region(s): %s\n", __FILE__,__LINE__,__FUNCTION__,str);
free(reg); free(tmp.s); return NULL;
}
ep++;
sp = ep;
to = hts_parse_decimal(sp,(char**)&ep);
if ( *ep && *ep!=',' )
{
fprintf(stderr,"[%s:%d %s] Could not parse the region(s): %s\n", __FILE__,__LINE__,__FUNCTION__,str);
free(reg); free(tmp.s); return NULL;
}
if ( sp==ep ) to = MAX_CSI_COOR-1;
_regions_add(reg, tmp.s, from, to);
if ( !*ep ) break;
sp = ep;
}
else
{
if ( tmp.l ) _regions_add(reg, tmp.s, -1, -1);
if ( !*ep ) break;
sp = ++ep;
}
}
free(tmp.s);
return reg;
}
bcf_sr_regions_t *bcf_sr_regions_init(const char *regions, int is_file, int ichr, int ifrom, int ito)
{
bcf_sr_regions_t *reg;
if ( !is_file ) return _regions_init_string(regions);
reg = (bcf_sr_regions_t *) calloc(1, sizeof(bcf_sr_regions_t));
reg->start = reg->end = -1;
reg->prev_start = reg->prev_seq = -1;
reg->file = hts_open(regions, "rb");
if ( !reg->file )
{
fprintf(stderr,"[%s:%d %s] Could not open file: %s\n", __FILE__,__LINE__,__FUNCTION__,regions);
free(reg);
return NULL;
}
reg->tbx = tbx_index_load(regions);
if ( !reg->tbx )
{
int len = strlen(regions);
int is_bed = strcasecmp(".bed",regions+len-4) ? 0 : 1;
if ( !is_bed && !strcasecmp(".bed.gz",regions+len-7) ) is_bed = 1;
if ( reg->file->format.format==vcf ) ito = 1;
// read the whole file, tabix index is not present
while ( hts_getline(reg->file, KS_SEP_LINE, ®->line) > 0 )
{
char *chr, *chr_end;
int from, to, ret;
ret = _regions_parse_line(reg->line.s, ichr,ifrom,abs(ito), &chr,&chr_end,&from,&to);
if ( ret < 0 )
{
if ( ito<0 )
ret = _regions_parse_line(reg->line.s, ichr,ifrom,ifrom, &chr,&chr_end,&from,&to);
if ( ret<0 )
{
fprintf(stderr,"[%s:%d] Could not parse the file %s, using the columns %d,%d[,%d]\n", __FILE__,__LINE__,regions,ichr+1,ifrom+1,ito+1);
hts_close(reg->file); reg->file = NULL; free(reg);
return NULL;
}
}
if ( !ret ) continue;
if ( is_bed ) from++;
*chr_end = 0;
_regions_add(reg, chr, from, to);
*chr_end = '\t';
}
hts_close(reg->file); reg->file = NULL;
if ( !reg->nseqs ) { free(reg); return NULL; }
return reg;
}
reg->seq_names = (char**) tbx_seqnames(reg->tbx, ®->nseqs);
if ( !reg->seq_hash )
reg->seq_hash = khash_str2int_init();
int i;
for (i=0; i<reg->nseqs; i++)
{
khash_str2int_set(reg->seq_hash,reg->seq_names[i],i);
}
reg->fname = strdup(regions);
reg->is_bin = 1;
return reg;
}
int bcf_sr_add_reader(bcf_srs_t *files, const char *fname)
{
htsFile* file_ptr = hts_open(fname, "r");
if ( ! file_ptr ) {
files->errnum = open_failed;
return 0;
}
files->has_line = (int*) realloc(files->has_line, sizeof(int)*(files->nreaders+1));
files->has_line[files->nreaders] = 0;
files->readers = (bcf_sr_t*) realloc(files->readers, sizeof(bcf_sr_t)*(files->nreaders+1));
bcf_sr_t *reader = &files->readers[files->nreaders++];
memset(reader,0,sizeof(bcf_sr_t));
reader->file = file_ptr;
files->errnum = 0;
if ( files->require_index )
{
if ( reader->file->format.format==vcf )
{
if ( reader->file->format.compression!=bgzf )
{
files->errnum = not_bgzf;
return 0;
}
reader->tbx_idx = tbx_index_load(fname);
if ( !reader->tbx_idx )
{
files->errnum = idx_load_failed;
return 0;
}
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->file->format.format==bcf )
{
if ( reader->file->format.compression!=bgzf )
{
files->errnum = not_bgzf;
return 0;
}
reader->header = bcf_hdr_read(reader->file);
reader->bcf_idx = bcf_index_load(fname);
if ( !reader->bcf_idx )
{
files->errnum = idx_load_failed;
return 0;
}
}
else
{
files->errnum = file_type_error;
return 0;
}
}
else
{
if ( reader->file->format.format==bcf || reader->file->format.format==vcf )
{
reader->header = bcf_hdr_read(reader->file);
}
else
{
files->errnum = file_type_error;
return 0;
}
files->streaming = 1;
}
if ( files->streaming && files->nreaders>1 )
{
files->errnum = api_usage_error;
fprintf(stderr,"[%s:%d %s] Error: %d readers, yet require_index not set\n", __FILE__,__LINE__,__FUNCTION__,files->nreaders);
return 0;
}
if ( files->streaming && files->regions )
{
files->errnum = api_usage_error;
fprintf(stderr,"[%s:%d %s] Error: cannot tabix-jump in streaming mode\n", __FILE__,__LINE__,__FUNCTION__);
return 0;
}
if ( !reader->header )
{
files->errnum = header_error;
return 0;
}
reader->fname = fname;
if ( files->apply_filters )
reader->filter_ids = init_filters(reader->header, files->apply_filters, &reader->nfilter_ids);
// Update list of chromosomes
if ( !files->explicit_regs && !files->streaming )
{
int n,i;
const char **names = reader->tbx_idx ? tbx_seqnames(reader->tbx_idx, &n) : bcf_hdr_seqnames(reader->header, &n);
for (i=0; i<n; i++)
{
if ( !files->regions )
files->regions = _regions_init_string(names[i]);
else
_regions_add(files->regions, names[i], -1, -1);
}
free(names);
}
return 1;
}
int bcf_sr_add_reader(bcf_srs_t *files, const char *fname)
{
files->has_line = (int*) realloc(files->has_line, sizeof(int)*(files->nreaders+1));
files->has_line[files->nreaders] = 0;
files->readers = (bcf_sr_t*) realloc(files->readers, sizeof(bcf_sr_t)*(files->nreaders+1));
bcf_sr_t *reader = &files->readers[files->nreaders++];
memset(reader,0,sizeof(bcf_sr_t));
reader->file = hts_open(fname, "r");
if ( !reader->file ) return 0;
reader->type = reader->file->is_bin? FT_BCF : FT_VCF;
if (reader->file->is_compressed) reader->type |= FT_GZ;
if ( files->require_index )
{
if ( reader->type==FT_VCF_GZ )
{
reader->tbx_idx = tbx_index_load(fname);
if ( !reader->tbx_idx )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0;
}
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->type==FT_BCF_GZ )
{
reader->header = bcf_hdr_read(reader->file);
reader->bcf_idx = bcf_index_load(fname);
if ( !reader->bcf_idx )
{
fprintf(stderr,"[add_reader] Could not load the index of %s\n", fname);
return 0; // not indexed..?
}
}
else
{
fprintf(stderr,"Index required, expected .vcf.gz or .bcf file: %s\n", fname);
return 0;
}
}
else
{
if ( reader->type & FT_BCF )
{
reader->header = bcf_hdr_read(reader->file);
}
else if ( reader->type & FT_VCF )
{
reader->header = bcf_hdr_read(reader->file);
}
else
{
fprintf(stderr,"File type not recognised: %s\n", fname);
return 0;
}
files->streaming = 1;
}
if ( files->streaming && files->nreaders>1 )
{
fprintf(stderr,"[%s:%d %s] Error: %d readers, yet require_index not set\n", __FILE__,__LINE__,__FUNCTION__,files->nreaders);
return 0;
}
if ( files->streaming && files->regions )
{
fprintf(stderr,"[%s:%d %s] Error: cannot tabix-jump in streaming mode\n", __FILE__,__LINE__,__FUNCTION__);
return 0;
}
if ( !reader->header ) return 0;
reader->fname = fname;
if ( files->apply_filters )
reader->filter_ids = init_filters(reader->header, files->apply_filters, &reader->nfilter_ids);
// Update list of chromosomes
if ( !files->explicit_regs && !files->streaming )
{
int n,i;
const char **names = reader->tbx_idx ? tbx_seqnames(reader->tbx_idx, &n) : bcf_hdr_seqnames(reader->header, &n);
for (i=0; i<n; i++)
{
if ( !files->regions )
files->regions = _regions_init_string(names[i]);
else
_regions_add(files->regions, names[i], -1, -1);
}
free(names);
}
return 1;
}