Skip to content

Allow bcftools reheader --fai to read its input file from a stream #2129

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
pd3 merged 1 commit into from
Mar 21, 2024
+31 −45
Merged

Allow bcftools reheader --fai to read its input file from a stream #2129

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion doc/bcftools.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3142,7 +3142,7 @@ Modify header of VCF/BCF files, change sample names.
"Not\ a\ good\ sample\ name".

*-T, --temp-prefix* 'PATH'::
template for temporary file names, used with *-f*
this option is ignored, but left for compatibility with earlier versions of bcftools.

*--threads* 'INT'::
see *<<common_options,Common Options>>*
Expand Down
74 changes: 30 additions & 44 deletions reheader.c
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
/* reheader.c -- reheader subcommand.

Copyright (C) 2014-2022 Genome Research Ltd.
Copyright (C) 2014-2022,2024 Genome Research Ltd.

Author: Petr Danecek <pd3@sanger.ac.uk>

Expand Down Expand Up @@ -49,8 +49,9 @@ THE SOFTWARE. */
typedef struct _args_t
{
char **argv, *fname, *samples_fname, *header_fname, *output_fname;
char *fai_fname, *rm_tmpfile, *tmp_prefix;
char *fai_fname;
htsFile *fp;
faidx_t *fai;
htsFormat type;
htsThreadPool *threads;
int argc, n_threads;
Expand Down Expand Up @@ -168,33 +169,13 @@ char *init_tmp_prefix(const char *tmp_prefix)
kputs("/bcftools.XXXXXX", &prefix);
return prefix.s;
}
static void update_from_fai(args_t *args)
static void update_from_fai(faidx_t *fai, kstring_t *hdr_txt)
{
if ( !strcmp("-",args->fname) )
error("Cannot use the --fai option when reading from standard input.\n");

faidx_t *fai = fai_load3(args->fai_fname,args->fai_fname,NULL,FAI_FASTA);
if ( !fai ) error("Could not parse %s\n", args->fai_fname);
args->rm_tmpfile = init_tmp_prefix(args->tmp_prefix);
int fd = mkstemp(args->rm_tmpfile);
if ( fd<0 ) error("Could not open a temporary file for writing: %s\n", args->rm_tmpfile);

// get a template header: either from the original VCF or from --header
char *ori_hdr_fname = args->header_fname ? args->header_fname : args->fname;
htsFile *fp = hts_open(ori_hdr_fname,"r");
if ( !fp ) error("Failed to open: %s\n", ori_hdr_fname);
bcf_hdr_t *hdr = bcf_hdr_read(fp);
if ( !hdr ) error("Failed to read the header: %s\n", ori_hdr_fname);
hts_close(fp); // no need to check the return status here

// put the header in a text buffer
kstring_t hdr_txt_ori = {0,0,0}, hdr_txt_new = {0,0,0};
bcf_hdr_format(hdr, 0, &hdr_txt_ori);
bcf_hdr_destroy(hdr);
kstring_t hdr_txt_new = {0,0,0};

// update the existing contig lines and remove lines not present in the fai file
void *chr_seen = khash_str2int_init();
char *tmp, *beg = hdr_txt_ori.s;
char *tmp, *beg = hdr_txt->s;
while ( beg && *beg )
{
tmp = strstr(beg, "\n##contig=<");
Expand All @@ -216,13 +197,10 @@ static void update_from_fai(args_t *args)
}
kputs(tmp+1,&hdr_txt_new);

if ( write(fd, hdr_txt_new.s, hdr_txt_new.l)!=hdr_txt_new.l ) error("Failed to write %zu bytes to %s\n", hdr_txt_new.l,args->rm_tmpfile);
if ( close(fd)!=0 ) error("Failed to close %s\n", args->rm_tmpfile);
args->header_fname = args->rm_tmpfile;
// Switch the new header content for the old
free(hdr_txt->s);
memcpy(hdr_txt, &hdr_txt_new, sizeof(*hdr_txt));

free(hdr_txt_ori.s);
free(hdr_txt_new.s);
fai_destroy(fai);
khash_str2int_destroy_free(chr_seen);
}

Expand Down Expand Up @@ -420,6 +398,10 @@ static void reheader_vcf_gz(args_t *args)
free(hdr.s); hdr.s = NULL; hdr.l = hdr.m = 0;
read_header_file(args->header_fname, &hdr);
}

if ( args->fai )
update_from_fai(args->fai, &hdr);

if ( samples )
{
set_samples(samples, nsamples, &hdr);
Expand Down Expand Up @@ -479,6 +461,10 @@ static void reheader_vcf(args_t *args)
free(hdr.s); hdr.s = NULL; hdr.l = hdr.m = 0;
read_header_file(args->header_fname, &hdr);
}

if ( args->fai )
update_from_fai(args->fai, &hdr);

if ( samples )
{
set_samples(samples, nsamples, &hdr);
Expand Down Expand Up @@ -586,6 +572,10 @@ static void reheader_bcf(args_t *args, int is_compressed)
free(htxt.s); htxt.s = NULL; htxt.l = htxt.m = 0;
read_header_file(args->header_fname, &htxt);
}

if ( args->fai )
update_from_fai(args->fai, &htxt);

if ( samples )
{
set_samples(samples, nsamples, &htxt);
Expand Down Expand Up @@ -675,11 +665,7 @@ static void usage(args_t *args)
fprintf(stderr, " -h, --header FILE new header\n");
fprintf(stderr, " -o, --output FILE write output to a file [standard output]\n");
fprintf(stderr, " -s, --samples FILE new sample names\n");
#ifdef _WIN32
fprintf(stderr, " -T, --temp-prefix PATH template for temporary file name [/bcftools.XXXXXX]\n");
#else
fprintf(stderr, " -T, --temp-prefix PATH template for temporary file name [/tmp/bcftools.XXXXXX]\n");
#endif
fprintf(stderr, " -T, --temp-prefix PATH ignored; was template for temporary file name\n");
fprintf(stderr, " --threads INT use multithreading with <int> worker threads (BCF only) [0]\n");
fprintf(stderr, "\n");
fprintf(stderr, "Example:\n");
Expand Down Expand Up @@ -716,7 +702,7 @@ int main_reheader(int argc, char *argv[])
switch (c)
{
case 1 : args->n_threads = strtol(optarg, 0, 0); break;
case 'T': args->tmp_prefix = optarg; break;
case 'T': break; // unused - was temp file prefix
case 'f': args->fai_fname = optarg; break;
case 'o': args->output_fname = optarg; break;
case 's': args->samples_fname = optarg; break;
Expand All @@ -733,8 +719,11 @@ int main_reheader(int argc, char *argv[])
}
else args->fname = argv[optind];

if ( args->fai_fname ) update_from_fai(args);
if ( !args->samples_fname && !args->header_fname ) usage(args);
if ( args->fai_fname ) {
args->fai = fai_load3(args->fai_fname,args->fai_fname,NULL,FAI_FASTA);
if ( !args->fai ) error("Could not parse %s\n", args->fai_fname);
}
if ( !args->samples_fname && !args->header_fname && !args->fai) usage(args);
if ( !args->fname ) usage(args);

args->fp = hts_open(args->fname,"r");
Expand All @@ -755,11 +744,8 @@ int main_reheader(int argc, char *argv[])
else
reheader_bcf(args, args->type.compression==bgzf || args->type.compression==gzip);

if ( args->rm_tmpfile )
{
unlink(args->rm_tmpfile);
free(args->rm_tmpfile);
}
if (args->fai)
fai_destroy(args->fai);
free(args);
return 0;
}