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;
}
static int query_regions(args_t *args, char *fname, char **regs, int nregs)
{
int i;
htsFile *fp = hts_open(fname,"r");
if ( !fp ) error("Could not read %s\n", fname);
enum htsExactFormat format = hts_get_format(fp)->format;
regidx_t *reg_idx = NULL;
if ( args->targets_fname )
{
reg_idx = regidx_init(args->targets_fname, NULL, NULL, 0, NULL);
if ( !reg_idx ) error("Could not read %s\n", args->targets_fname);
}
if ( format == bcf )
{
htsFile *out = hts_open("-","w");
if ( !out ) error("Could not open stdout\n", fname);
hts_idx_t *idx = bcf_index_load(fname);
if ( !idx ) error("Could not load .csi index of %s\n", fname);
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) error("Could not read the header: %s\n", fname);
if ( args->print_header )
bcf_hdr_write(out,hdr);
if ( !args->header_only )
{
bcf1_t *rec = bcf_init();
for (i=0; i<nregs; i++)
{
hts_itr_t *itr = bcf_itr_querys(idx,hdr,regs[i]);
while ( bcf_itr_next(fp, itr, rec) >=0 )
{
if ( reg_idx && !regidx_overlap(reg_idx, bcf_seqname(hdr,rec),rec->pos,rec->pos+rec->rlen-1, NULL) ) continue;
bcf_write(out,hdr,rec);
}
tbx_itr_destroy(itr);
}
bcf_destroy(rec);
}
if ( hts_close(out) ) error("hts_close returned non-zero status for stdout\n");
bcf_hdr_destroy(hdr);
hts_idx_destroy(idx);
}
else if ( format==vcf || format==sam || format==unknown_format )
{
tbx_t *tbx = tbx_index_load(fname);
if ( !tbx ) error("Could not load .tbi/.csi index of %s\n", fname);
kstring_t str = {0,0,0};
if ( args->print_header )
{
while ( hts_getline(fp, KS_SEP_LINE, &str) >= 0 )
{
if ( !str.l || str.s[0]!=tbx->conf.meta_char ) break;
puts(str.s);
}
}
if ( !args->header_only )
{
int nseq;
const char **seq = NULL;
if ( reg_idx ) seq = tbx_seqnames(tbx, &nseq);
for (i=0; i<nregs; i++)
{
hts_itr_t *itr = tbx_itr_querys(tbx, regs[i]);
if ( !itr ) continue;
while (tbx_itr_next(fp, tbx, itr, &str) >= 0)
{
if ( reg_idx && !regidx_overlap(reg_idx,seq[itr->curr_tid],itr->curr_beg,itr->curr_end, NULL) ) continue;
puts(str.s);
}
tbx_itr_destroy(itr);
}
free(seq);
}
free(str.s);
tbx_destroy(tbx);
}
else if ( format==bam )
error("Please use \"samtools view\" for querying BAM files.\n");
if ( reg_idx ) regidx_destroy(reg_idx);
if ( hts_close(fp) ) error("hts_close returned non-zero status: %s\n", fname);
for (i=0; i<nregs; i++) free(regs[i]);
free(regs);
return 0;
}
int convert(int argc, char **argv)
{
if (argc < 2) return convert_help();
int c;
char *in=NULL, *out=NULL, *bim=NULL, *vid=NULL, *tmp_dir=NULL, *ped=NULL;
uint32_t num_fields, num_records, col = 2;
int i_is_set = 0,
o_is_set = 0,
f_is_set = 0,
b_is_set = 0,
v_is_set = 0,
t_is_set = 0,
p_is_set = 0,
r_is_set = 0;
while((c = getopt (argc, argv, "hi:o:f:r:b:v:t:p:c:")) != -1) {
switch (c) {
case 'c':
col = atoi(optarg);
break;
case 'p':
p_is_set = 1;
ped = optarg;
break;
case 't':
t_is_set = 1;
tmp_dir = optarg;
break;
case 'v':
v_is_set = 1;
vid = optarg;
break;
case 'b':
b_is_set = 1;
bim = optarg;
break;
case 'i':
i_is_set = 1;
in = optarg;
break;
case 'o':
o_is_set = 1;
out = optarg;
break;
case 'f':
f_is_set = 1;
num_fields = atoi(optarg);
break;
case 'r':
r_is_set = 1;
num_records = atoi(optarg);
break;
case 'h':
convert_help();
return 1;
case '?':
if ( (optopt == 'i') ||
(optopt == 'f') ||
(optopt == 'r') ||
(optopt == 't') ||
(optopt == 's') ||
(optopt == 'p') ||
(optopt == 'c') ||
(optopt == 'o') )
fprintf (stderr, "Option -%c requires an argument.\n",
optopt);
else if (isprint (optopt))
fprintf (stderr, "Unknown option `-%c'.\n", optopt);
else
fprintf (stderr, "Unknown option character `\\x%x'.\n", optopt);
default:
convert_help();
return 1;
}
}
char *type = argv[0];
if (i_is_set == 0) {
printf("Input file is not set\n");
return convert_help();
}
if (strcmp(type, "bcf") == 0) {
if ( (f_is_set == 0) || (r_is_set == 0) ) {
fprintf(stderr,"Attempting to autodetect num of records "
"and fields from %s\n", in);
//Try and auto detect the sizes, need the index
tbx_t *tbx = NULL;
hts_idx_t *idx = NULL;
htsFile *fp = hts_open(in,"rb");
if ( !fp ) {
fprintf(stderr,"Could not read %s\n", in);
return 1;
}
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) {
fprintf(stderr,"Could not read the header: %s\n", in);
return 1;
}
if (hts_get_format(fp)->format==vcf) {
tbx = tbx_index_load(in);
if ( !tbx ) {
fprintf(stderr,"Could not load TBI index: %s\n", in);
return 1;
}
} else if ( hts_get_format(fp)->format==bcf ) {
idx = bcf_index_load(in);
if ( !idx ) {
fprintf(stderr,"Could not load CSI index: %s\n", in);
return 1;
}
} else {
fprintf(stderr,
"Could not detect the file type as VCF or BCF: %s\n",
in);
return 1;
}
num_fields = hdr->n[BCF_DT_SAMPLE];
num_records = 0;
const char **seq;
int nseq;
seq = tbx ? tbx_seqnames(tbx, &nseq) :
bcf_index_seqnames(idx, hdr, &nseq);
int i;
uint32_t sum = 0;
for (i = 0; i < nseq; ++i) {
uint64_t records, v;
hts_idx_get_stat(tbx ? tbx->idx: idx, i, &records, &v);
num_records += records;
}
fprintf(stderr, "Number of records:%u\tNumber of fields:%u\n",
num_records, num_fields);
free(seq);
hts_close(fp);
bcf_hdr_destroy(hdr);
if (idx)
hts_idx_destroy(idx);
if (tbx)
tbx_destroy(tbx);
}
if (o_is_set == 0) {
out = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(out,in);
strcat(out, ".gqt");
}
if (b_is_set == 0) {
bim = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(bim,in);
strcat(bim, ".bim");
}
if (v_is_set == 0) {
vid = (char*)malloc(strlen(in) + 5); // 5 for ext and \0
strcpy(vid,in);
strcat(vid, ".vid");
}
if (t_is_set == 0) {
tmp_dir = (char*)malloc(3*sizeof(char)); // "./\0"
strcpy(tmp_dir,"./");
}
int r = bcf_wahbm(in, out, bim, vid, tmp_dir, num_fields, num_records);
return r;
}
if (strcmp(type, "ped") == 0) {
if (o_is_set == 0) {
if (p_is_set == 1) {
out = (char*)malloc(strlen(ped) + 4); // 4 for ext and \0
strcpy(out,ped);
strcat(out, ".db");
} else {
out = (char*)malloc(strlen(in) + 4); // 4 for ext and \0
strcpy(out,in);
strcat(out, ".db");
}
}
fprintf(stderr, "Creating sample database %s\n", out);
return ped_ped(in, ped, col, out);
}
return convert_help();
}
