c_cpp/lib/htslib/htsfile.c

   1 /*  htsfile.c -- file identifier and minimal viewer.
   2
   3     Copyright (C) 2014-2017 Genome Research Ltd.
   4
   5     Author: John Marshall <jm18@sanger.ac.uk>
   6
   7 Permission is hereby granted, free of charge, to any person obtaining a copy
   8 of this software and associated documentation files (the "Software"), to deal
   9 in the Software without restriction, including without limitation the rights
  10 to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  11 copies of the Software, and to permit persons to whom the Software is
  12 furnished to do so, subject to the following conditions:
  13
  14 The above copyright notice and this permission notice shall be included in
  15 all copies or substantial portions of the Software.
  16
  17 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  18 IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  19 FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
  20 THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  21 LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  22 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
  23 DEALINGS IN THE SOFTWARE.  */
  24
  25 #include <config.h>
  26
  27 #include <ctype.h>
  28 #include <errno.h>
  29 #include <stdarg.h>
  30 #include <stdio.h>
  31 #include <stdlib.h>
  32 #include <string.h>
  33 #include <getopt.h>
  34 #include <unistd.h>
  35
  36 #include "htslib/hfile.h"
  37 #include "htslib/hts.h"
  38 #include "htslib/sam.h"
  39 #include "htslib/vcf.h"
  40
  41 enum { identify, view_headers, view_all } mode = identify;
  42 int show_headers = 1;
  43 int verbose = 0;
  44 int status = EXIT_SUCCESS;  /* Exit status from main */
  45
  46 void error(const char *format, ...)
  47 {
  48     int err = errno;
  49     va_list args;
  50     va_start(args, format);
  51     fflush(stdout);
  52     fprintf(stderr, "htsfile: ");
  53     vfprintf(stderr, format, args);
  54     if (err) fprintf(stderr, ": %s\n", strerror(err));
  55     else fprintf(stderr, "\n");
  56     fflush(stderr);
  57     va_end(args);
  58     status = EXIT_FAILURE;
  59 }
  60
  61 static htsFile *dup_stdout(const char *mode)
  62 {
  63     int fd = dup(STDOUT_FILENO);
  64     hFILE *hfp = (fd >= 0)? hdopen(fd, mode) : NULL;
  65     return hfp? hts_hopen(hfp, "-", mode) : NULL;
  66 }
  67
  68 static void view_sam(samFile *in, const char *filename)
  69 {
  70     bam1_t *b = NULL;
  71     bam_hdr_t *hdr = NULL;
  72     samFile *out = NULL;
  73
  74     hdr = sam_hdr_read(in);
  75     if (hdr == NULL) {
  76         errno = 0; error("reading headers from \"%s\" failed", filename);
  77         goto clean;
  78     }
  79
  80     out = dup_stdout("w");
  81     if (out == NULL) { error("reopening standard output failed"); goto clean; }
  82
  83     if (show_headers) {
  84         if (sam_hdr_write(out, hdr) != 0) {
  85             error("writing headers to standard output failed");
  86             goto clean;
  87         }
  88     }
  89
  90     if (mode == view_all) {
  91         int ret;
  92
  93         b = bam_init1();
  94         if (b == NULL) { error("can't create record"); goto clean; }
  95
  96         while ((ret = sam_read1(in, hdr, b)) >= 0) {
  97             if (sam_write1(out, hdr, b) < 0) {
  98                 error("writing to standard output failed");
  99                 goto clean;
 100             }
 101         }
 102
 103         if (ret < -1) { error("reading \"%s\" failed", filename); goto clean; }
 104     }
 105
 106  clean:
 107     bam_hdr_destroy(hdr);
 108     bam_destroy1(b);
 109     if (out) hts_close(out);
 110 }
 111
 112 static void view_vcf(vcfFile *in, const char *filename)
 113 {
 114     bcf1_t *rec = NULL;
 115     bcf_hdr_t *hdr = NULL;
 116     vcfFile *out = NULL;
 117
 118     hdr = bcf_hdr_read(in);
 119     if (hdr == NULL) {
 120         errno = 0; error("reading headers from \"%s\" failed", filename);
 121         goto clean;
 122     }
 123
 124     out = dup_stdout("w");
 125     if (out == NULL) { error("reopening standard output failed"); goto clean; }
 126
 127     if (show_headers) {
 128         if (bcf_hdr_write(out, hdr) != 0) {
 129             error("writing headers to standard output failed");
 130             goto clean;
 131         }
 132     }
 133
 134     if (mode == view_all) {
 135         int ret;
 136
 137         rec = bcf_init();
 138         if (rec == NULL) { error("can't create record"); goto clean; }
 139
 140         while ((ret = bcf_read(in, hdr, rec)) >= 0) {
 141             if (bcf_write(out, hdr, rec) < 0) {
 142                 error("writing to standard output failed");
 143                 goto clean;
 144             }
 145         }
 146
 147         if (ret < -1) { error("reading \"%s\" failed", filename); goto clean; }
 148     }
 149
 150  clean:
 151     if (hdr) bcf_hdr_destroy(hdr);
 152     if (rec) bcf_destroy(rec);
 153     if (out) hts_close(out);
 154 }
 155
 156 static void view_raw(hFILE *fp, const char *filename)
 157 {
 158     int c, prev;
 159     for (prev = '\n'; (c = hgetc(fp)) != EOF; prev = c)
 160         if (isprint(c) || c == '\n' || c == '\t') putchar(c);
 161         else if (c == '\r') fputs("\\r", stdout);
 162         else if (c == '\0') fputs("\\0", stdout);
 163         else printf("\\x%02x", c);
 164
 165     if (prev != '\n') putchar('\n');
 166
 167     if (herrno(fp)) {
 168         errno = herrno(fp);
 169         error("reading \"%s\" failed", filename);
 170     }
 171 }
 172
 173 static void usage(FILE *fp, int status)
 174 {
 175     fprintf(fp,
 176 "Usage: htsfile [-chHv] FILE...\n"
 177 "Options:\n"
 178 "  -c, --view         Write textual form of FILEs to standard output\n"
 179 "  -h, --header-only  Display only headers in view mode, not records\n"
 180 "  -H, --no-header    Suppress header display in view mode\n"
 181 "  -v, --verbose      Increase verbosity of warnings and diagnostics\n");
 182     exit(status);
 183 }
 184
 185 int main(int argc, char **argv)
 186 {
 187     static const struct option options[] = {
 188         { "header-only", no_argument, NULL, 'h' },
 189         { "no-header", no_argument, NULL, 'H' },
 190         { "view", no_argument, NULL, 'c' },
 191         { "verbose", no_argument, NULL, 'v' },
 192         { "help", no_argument, NULL, '?' },
 193         { "version", no_argument, NULL, 1 },
 194         { NULL, 0, NULL, 0 }
 195     };
 196
 197     int c, i;
 198
 199     status = EXIT_SUCCESS;
 200     while ((c = getopt_long(argc, argv, "chHv?", options, NULL)) >= 0)
 201         switch (c) {
 202         case 'c': mode = view_all; break;
 203         case 'h': mode = view_headers; show_headers = 1; break;
 204         case 'H': show_headers = 0; break;
 205         case 'v': hts_verbose++; verbose++; break;
 206         case 1:
 207             printf(
 208 "htsfile (htslib) %s\n"
 209 "Copyright (C) 2017 Genome Research Ltd.\n",
 210                    hts_version());
 211             exit(EXIT_SUCCESS);
 212             break;
 213         case '?': usage(stdout, EXIT_SUCCESS); break;
 214         default:  usage(stderr, EXIT_FAILURE); break;
 215         }
 216
 217     if (optind == argc) usage(stderr, EXIT_FAILURE);
 218
 219     for (i = optind; i < argc; i++) {
 220         hFILE *fp = hopen(argv[i], "r");
 221         if (fp == NULL) {
 222             error("can't open \"%s\"", argv[i]);
 223             continue;
 224         }
 225
 226         if (mode == identify) {
 227             htsFormat fmt;
 228             if (hts_detect_format(fp, &fmt) < 0) {
 229                 error("detecting \"%s\" format failed", argv[i]);
 230                 hclose_abruptly(fp);
 231                 continue;
 232             }
 233
 234             char *description = hts_format_description(&fmt);
 235             printf("%s:\t%s\n", argv[i], description);
 236             free(description);
 237         }
 238         else {
 239             htsFile *hts = hts_hopen(fp, argv[i], "r");
 240             if (hts) {
 241                 switch (hts_get_format(hts)->category) {
 242                 case sequence_data:
 243                     view_sam(hts, argv[i]);
 244                     break;
 245                 case variant_data:
 246                     view_vcf(hts, argv[i]);
 247                     break;
 248                 default:
 249                     if (verbose)
 250                         view_raw(fp, argv[i]);
 251                     else {
 252                         errno = 0;
 253                         error("can't view \"%s\": unknown format", argv[i]);
 254                     }
 255                     break;
 256                 }
 257
 258                 if (hts_close(hts) < 0) error("closing \"%s\" failed", argv[i]);
 259                 fp = NULL;
 260             }
 261             else if (errno == ENOEXEC && verbose)
 262                 view_raw(fp, argv[i]);
 263             else
 264                 error("can't view \"%s\"", argv[i]);
 265         }
 266
 267         if (fp && hclose(fp) < 0) error("closing \"%s\" failed", argv[i]);
 268     }
 269
 270     return status;
 271 }