static void list_columns(args_t *args)
{
void *has_sample = NULL;
if ( args->sample_list )
{
has_sample = khash_str2int_init();
int i, nsmpl;
char **smpl = hts_readlist(args->sample_list, args->sample_is_file, &nsmpl);
for (i=0; i<nsmpl; i++) khash_str2int_inc(has_sample, smpl[i]);
free(smpl);
}
int i;
bcf_sr_t *reader = &args->files->readers[0];
for (i=0; i<bcf_hdr_nsamples(reader->header); i++)
{
if ( has_sample && !khash_str2int_has_key(has_sample, reader->header->samples[i]) ) continue;
printf("%s\n", reader->header->samples[i]);
}
if ( has_sample )
khash_str2int_destroy_free(has_sample);
}
static void bsmpl_add_readgroup(bam_smpl_t *bsmpl, file_t *file, const char *rg_id, const char *smpl_name)
{
int ismpl = -1;
if ( smpl_name )
{
if ( khash_str2int_get(bsmpl->name2idx,smpl_name,&ismpl) < 0 )
{
// new sample
bsmpl->nsmpl++;
bsmpl->smpl = (char**) realloc(bsmpl->smpl,sizeof(char*)*bsmpl->nsmpl);
bsmpl->smpl[bsmpl->nsmpl-1] = strdup(smpl_name);
ismpl = khash_str2int_inc(bsmpl->name2idx,bsmpl->smpl[bsmpl->nsmpl-1]);
}
}
if ( !strcmp("*",rg_id) )
{
// all read groups in the bam treated as the same sample
file->default_idx = ismpl;
return;
}
if ( !file->rg2idx ) file->rg2idx = khash_str2int_init();
if ( khash_str2int_has_key(file->rg2idx,rg_id) ) return; // duplicate @RG:ID
khash_str2int_set(file->rg2idx, strdup(rg_id), ismpl);
}
int bcf_sr_set_samples(bcf_srs_t *files, const char *fname, int is_file)
{
int i, j, nsmpl, free_smpl = 0;
char **smpl = NULL;
void *exclude = (fname[0]=='^') ? khash_str2int_init() : NULL;
if ( exclude || strcmp("-",fname) ) // "-" stands for all samples
{
smpl = hts_readlist(fname, is_file, &nsmpl);
if ( !smpl )
{
fprintf(stderr,"Could not read the file: \"%s\"\n", fname);
return 0;
}
if ( exclude )
{
for (i=0; i<nsmpl; i++)
khash_str2int_inc(exclude, smpl[i]);
}
free_smpl = 1;
}
if ( !smpl )
{
smpl = files->readers[0].header->samples; // intersection of all samples
nsmpl = bcf_hdr_nsamples(files->readers[0].header);
}
files->samples = NULL;
files->n_smpl = 0;
for (i=0; i<nsmpl; i++)
{
if ( exclude && khash_str2int_has_key(exclude,smpl[i]) ) continue;
int n_isec = 0;
for (j=0; j<files->nreaders; j++)
{
if ( bcf_hdr_id2int(files->readers[j].header, BCF_DT_SAMPLE, smpl[i])<0 ) break;
n_isec++;
}
if ( n_isec!=files->nreaders )
{
fprintf(stderr,"Warning: The sample \"%s\" was not found in %s, skipping\n", smpl[i], files->readers[n_isec].fname);
continue;
}
files->samples = (char**) realloc(files->samples, (files->n_smpl+1)*sizeof(const char*));
files->samples[files->n_smpl++] = strdup(smpl[i]);
}
if ( exclude ) khash_str2int_destroy(exclude);
if ( free_smpl )
{
for (i=0; i<nsmpl; i++) free(smpl[i]);
free(smpl);
}
if ( !files->n_smpl )
{
if ( files->nreaders>1 )
fprintf(stderr,"No samples in common.\n");
return 0;
}
for (i=0; i<files->nreaders; i++)
{
bcf_sr_t *reader = &files->readers[i];
reader->samples = (int*) malloc(sizeof(int)*files->n_smpl);
reader->n_smpl = files->n_smpl;
for (j=0; j<files->n_smpl; j++)
reader->samples[j] = bcf_hdr_id2int(reader->header, BCF_DT_SAMPLE, files->samples[j]);
}
return 1;
}
static void init_data(args_t *args)
{
int i;
args->hdr = args->files->readers[0].header;
if (args->calc_ac && args->update_info)
{
bcf_hdr_append(args->hdr,"##INFO=<ID=AC,Number=A,Type=Integer,Description=\"Allele count in genotypes\">");
bcf_hdr_append(args->hdr,"##INFO=<ID=AN,Number=1,Type=Integer,Description=\"Total number of alleles in called genotypes\">");
}
bcf_hdr_append_version(args->hdr, args->argc, args->argv, "bcftools_view");
// setup sample data
if (args->sample_names)
{
void *hdr_samples = khash_str2int_init();
for (i=0; i<bcf_hdr_nsamples(args->hdr); i++)
khash_str2int_inc(hdr_samples, bcf_hdr_int2id(args->hdr,BCF_DT_SAMPLE,i));
void *exclude = (args->sample_names[0]=='^') ? khash_str2int_init() : NULL;
int nsmpl;
char **smpl = NULL;
args->samples = NULL; args->n_samples = 0;
smpl = hts_readlist(exclude ? &args->sample_names[1] : args->sample_names, args->sample_is_file, &nsmpl);
if ( !smpl )
{
error("Could not read the list: \"%s\"\n", exclude ? &args->sample_names[1] : args->sample_names);
}
if ( exclude )
{
for (i=0; i<nsmpl; i++) {
if (!khash_str2int_has_key(hdr_samples,smpl[i])) {
if (args->force_samples) {
fprintf(stderr, "Warn: exclude called for sample that does not exist in header: \"%s\"... skipping\n", smpl[i]);
} else {
error("Error: exclude called for sample that does not exist in header: \"%s\". Use \"--force-samples\" to ignore this error.\n", smpl[i]);
}
}
khash_str2int_inc(exclude, smpl[i]);
}
for (i=0; i<bcf_hdr_nsamples(args->hdr); i++)
{
if ( exclude && khash_str2int_has_key(exclude,bcf_hdr_int2id(args->hdr,BCF_DT_SAMPLE,i)) ) continue;
args->samples = (char**) realloc(args->samples, (args->n_samples+1)*sizeof(const char*));
args->samples[args->n_samples++] = strdup(bcf_hdr_int2id(args->hdr,BCF_DT_SAMPLE,i));
}
khash_str2int_destroy(exclude);
}
else
{
for (i=0; i<nsmpl; i++) {
if (!khash_str2int_has_key(hdr_samples,smpl[i])) {
if (args->force_samples) {
fprintf(stderr, "Warn: subset called for sample that does not exist in header: \"%s\"... skipping\n", smpl[i]);
continue;
} else {
error("Error: subset called for sample that does not exist in header: \"%s\". Use \"--force-samples\" to ignore this error.\n", smpl[i]);
}
}
args->samples = (char**) realloc(args->samples, (args->n_samples+1)*sizeof(const char*));
args->samples[args->n_samples++] = strdup(smpl[i]);
}
}
for (i=0; i<nsmpl; i++) free(smpl[i]);
free(smpl);
khash_str2int_destroy(hdr_samples);
if (args->n_samples == 0) {
fprintf(stderr, "Warn: subsetting has removed all samples\n");
args->sites_only = 1;
}
}
if (args->n_samples)
args->imap = (int*)malloc(args->n_samples * sizeof(int));
// determine variant types to include/exclude
if (args->include_types || args->exclude_types) {
if (args->include_types && args->exclude_types) {
fprintf(stderr, "Error: only supply one of --include-types, --exclude-types options\n");
exit(1);
}
char **type_list = 0;
int m = 0, n = 0;
const char *q, *p;
for (q = p = args->include_types ? args->include_types : args->exclude_types;; ++p) {
if (*p == ',' || *p == 0) {
if (m == n) {
m = m? m<<1 : 16;
type_list = (char**)realloc(type_list, m * sizeof(char*));
}
type_list[n] = (char*)calloc(p - q + 1, 1);
strncpy(type_list[n++], q, p - q);
q = p + 1;
if (*p == 0) break;
}
}
type_list = (char**)realloc(type_list, n * sizeof(char*));
if (args->include_types) {
args->include = 0;
for (i = 0; i < n; ++i) {
if (strcmp(type_list[i], "snps") == 0) args->include |= VCF_SNP;
else if (strcmp(type_list[i], "indels") == 0) args->include |= VCF_INDEL;
else if (strcmp(type_list[i], "mnps") == 0) args->include |= VCF_MNP;
else if (strcmp(type_list[i], "other") == 0) args->include |= VCF_OTHER;
else {
fprintf(stderr, "[E::%s] unknown type\n", type_list[i]);
fprintf(stderr, "Accepted types are snps, indels, mnps, other\n");
exit(1);
}
}
}
if (args->exclude_types) {
args->exclude = 0;
for (i = 0; i < n; ++i) {
if (strcmp(type_list[i], "snps") == 0) args->exclude |= VCF_SNP;
else if (strcmp(type_list[i], "indels") == 0) args->exclude |= VCF_INDEL;
else if (strcmp(type_list[i], "mnps") == 0) args->exclude |= VCF_MNP;
else if (strcmp(type_list[i], "other") == 0) args->exclude |= VCF_OTHER;
else {
fprintf(stderr, "[E::%s] unknown type\n", type_list[i]);
fprintf(stderr, "Accepted types are snps, indels, mnps, other\n");
exit(1);
}
}
}
for (i = 0; i < n; ++i)
free(type_list[i]);
free(type_list);
}
// setup output
char modew[8];
strcpy(modew, "w");
if (args->clevel >= 0 && args->clevel <= 9) sprintf(modew + 1, "%d", args->clevel);
if (args->output_type==FT_BCF) strcat(modew, "bu"); // uncompressed BCF
else if (args->output_type & FT_BCF) strcat(modew, "b"); // compressed BCF
else if (args->output_type & FT_GZ) strcat(modew,"z"); // compressed VCF
args->out = hts_open(args->fn_out ? args->fn_out : "-", modew);
if ( !args->out ) error("%s: %s\n", args->fn_out,strerror(errno));
// headers: hdr=full header, hsub=subset header, hnull=sites only header
if (args->sites_only){
args->hnull = bcf_hdr_subset(args->hdr, 0, 0, 0);
bcf_hdr_remove(args->hnull, BCF_HL_FMT, NULL);
}
if (args->n_samples > 0)
{
args->hsub = bcf_hdr_subset(args->hdr, args->n_samples, args->samples, args->imap);
if ( !args->hsub ) error("Error occurred while subsetting samples\n");
if ( args->n_samples != bcf_hdr_nsamples(args->hsub) )
{
int i;
for (i=0; i<args->n_samples; i++)
if ( args->imap[i]<0 ) error("Error: No such sample: \"%s\"\n", args->samples[i]);
}
}
if ( args->filter_str )
args->filter = filter_init(args->hdr, args->filter_str);
}
/*
The logic of this function is a bit complicated because we want to work
also with broken bams containing read groups that are not listed in the
header. The desired behavior is as follows:
- when -G is given, read groups which are not listed in the header must
be given explicitly using the "?" symbol in -G.
Otherwise:
- if the bam has no header, all reads in the file are assigned to a
single sample named after the file
- if there is at least one sample defined in the header, reads with no
read group id or with a read group id not listed in the header are
assigned to the first sample encountered in the header
*/
int bam_smpl_add_bam(bam_smpl_t *bsmpl, char *bam_hdr, const char *fname)
{
bsmpl->nfiles++;
bsmpl->files = (file_t*) realloc(bsmpl->files,bsmpl->nfiles*sizeof(file_t));
file_t *file = &bsmpl->files[bsmpl->nfiles-1];
memset(file,0,sizeof(file_t));
file->fname = strdup(fname);
file->default_idx = -1;
if ( bsmpl->ignore_rg || !bam_hdr )
{
// The option --ignore-RG is set or there is no BAM header: use the file name as the sample name
bsmpl_add_readgroup(bsmpl,file,"*",file->fname);
return bsmpl->nfiles-1;
}
void *bam_smpls = khash_str2int_init();
int first_smpl = -1, nskipped = 0;
const char *p = bam_hdr, *q, *r;
while (p != NULL && (q = strstr(p, "@RG")) != 0)
{
char *eol = strchr(q + 3, '\n');
if (q > bam_hdr && *(q - 1) != '\n') { // @RG must be at start of line
p = eol;
continue;
}
p = q + 3;
if ((q = strstr(p, "\tID:")) != 0) q += 4;
if ((r = strstr(p, "\tSM:")) != 0) r += 4;
if (r && q)
{
char *u, *v;
int ioq, ior;
for (u = (char*)q; *u && *u != '\t' && *u != '\n'; ++u);
for (v = (char*)r; *v && *v != '\t' && *v != '\n'; ++v);
ioq = *u; ior = *v; *u = *v = '\0';
// q now points to a null terminated read group id
// r points to a null terminated sample name
if ( !strcmp("*",q) || !strcmp("?",q) )
error("Error: the read group IDs \"*\" and \"?\" have a special meaning in the mpileup code. Please fix the code or the bam: %s\n", fname);
int accept_rg = 1;
if ( bsmpl->sample_list )
{
// restrict samples based on the -s/-S options
char *name = khash_str2str_get(bsmpl->sample_list,r);
if ( bsmpl->sample_logic==0 )
accept_rg = name ? 0 : 1;
else if ( !name )
accept_rg = 0;
else
r = name;
}
if ( accept_rg && bsmpl->rg_list )
{
// restrict readgroups based on the -G option, possibly renaming the sample
accept_rg = bsmpl_keep_readgroup(bsmpl,file,q,&r);
}
if ( accept_rg )
bsmpl_add_readgroup(bsmpl,file,q,r);
else
{
bsmpl_add_readgroup(bsmpl,file,q,NULL); // ignore this RG but note that it was seen in the header
nskipped++;
}
if ( first_smpl<0 )
khash_str2int_get(bsmpl->name2idx,r,&first_smpl);
if ( !khash_str2int_has_key(bam_smpls,r) )
khash_str2int_inc(bam_smpls,strdup(r));
*u = ioq; *v = ior;
}
else
break;
p = eol;
}
int nsmpls = khash_str2int_size(bam_smpls);
khash_str2int_destroy_free(bam_smpls);
const char *smpl_name = NULL;
int accept_null_rg = 1;
if ( bsmpl->rg_list && !bsmpl_keep_readgroup(bsmpl,file,"?",&smpl_name) ) accept_null_rg = 0;
if ( bsmpl->sample_list && first_smpl==-1 ) accept_null_rg = 0;
if ( !accept_null_rg && first_smpl==-1 )
{
// no suitable read group is available in this bam: ignore the whole file.
free(file->fname);
if ( file->rg2idx ) khash_str2int_destroy_free(file->rg2idx);
bsmpl->nfiles--;
return -1;
}
if ( !accept_null_rg ) return bsmpl->nfiles-1;
if ( nsmpls==1 && !nskipped )
{
file->default_idx = first_smpl;
return bsmpl->nfiles-1;
}
if ( !smpl_name ) smpl_name = first_smpl==-1 ? file->fname : bsmpl->smpl[first_smpl];
bsmpl_add_readgroup(bsmpl,file,"?",smpl_name);
return bsmpl->nfiles-1;
}