c_cpp/lib/htslib/multipart.c

   1 /*  multipart.c -- GA4GH redirection and multipart backend for file streams.
   2
   3     Copyright (C) 2016 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 <stdio.h>
  28 #include <string.h>
  29 #include <errno.h>
  30
  31 #include "htslib/kstring.h"
  32
  33 #include "hts_internal.h"
  34 #include "hfile_internal.h"
  35
  36 #ifndef EPROTO
  37 #define EPROTO ENOEXEC
  38 #endif
  39
  40 typedef struct hfile_part {
  41     char *url;
  42     char **headers;
  43 } hfile_part;
  44
  45 typedef struct {
  46     hFILE base;
  47     hfile_part *parts;
  48     size_t nparts, maxparts, current;
  49     hFILE *currentfp;
  50 } hFILE_multipart;
  51
  52 static void free_part(hfile_part *p)
  53 {
  54     free(p->url);
  55     if (p->headers) {
  56         char **hdr;
  57         for (hdr = p->headers; *hdr; hdr++) free(*hdr);
  58         free(p->headers);
  59     }
  60
  61     p->url = NULL;
  62     p->headers = NULL;
  63 }
  64
  65 static void free_all_parts(hFILE_multipart *fp)
  66 {
  67     size_t i;
  68     for (i = 0; i < fp->nparts; i++) free_part(&fp->parts[i]);
  69     free(fp->parts);
  70 }
  71
  72 static ssize_t multipart_read(hFILE *fpv, void *buffer, size_t nbytes)
  73 {
  74     hFILE_multipart *fp = (hFILE_multipart *) fpv;
  75     size_t n;
  76
  77 open_next:
  78     if (fp->currentfp == NULL) {
  79         if (fp->current < fp->nparts) {
  80             const hfile_part *p = &fp->parts[fp->current];
  81             hts_log_debug("Opening part #%zu of %zu: \"%.120s%s\"",
  82                 fp->current+1, fp->nparts, p->url,
  83                 (strlen(p->url) > 120)? "..." : "");
  84
  85             fp->currentfp = p->headers?
  86                   hopen(p->url, "r:", "httphdr:v", p->headers, NULL)
  87                 : hopen(p->url, "r");
  88
  89             if (fp->currentfp == NULL) return -1;
  90         }
  91         else return 0;  // No more parts, so we're truly at EOF
  92     }
  93
  94     n = fp->currentfp->mobile?
  95           fp->currentfp->backend->read(fp->currentfp, buffer, nbytes)
  96         : hread(fp->currentfp, buffer, nbytes);
  97
  98     if (n == 0) {
  99         // We're at EOF on this part, so set up the next part
 100         hFILE *prevfp = fp->currentfp;
 101         free_part(&fp->parts[fp->current]);
 102         fp->current++;
 103         fp->currentfp = NULL;
 104         if (hclose(prevfp) < 0) return -1;
 105         goto open_next;
 106     }
 107
 108     return n;  // Number of bytes read by (or an error from) fp->currentfp
 109 }
 110
 111 static ssize_t multipart_write(hFILE *fpv, const void *buffer, size_t nbytes)
 112 {
 113     errno = EROFS;
 114     return -1;
 115 }
 116
 117 static off_t multipart_seek(hFILE *fpv, off_t offset, int whence)
 118 {
 119     errno = ESPIPE;
 120     return -1;
 121 }
 122
 123 static int multipart_close(hFILE *fpv)
 124 {
 125     hFILE_multipart *fp = (hFILE_multipart *) fpv;
 126
 127     free_all_parts(fp);
 128     if (fp->currentfp) {
 129         if (hclose(fp->currentfp) < 0) return -1;
 130     }
 131
 132     return 0;
 133 }
 134
 135 static const struct hFILE_backend multipart_backend =
 136 {
 137     multipart_read, multipart_write, multipart_seek, NULL, multipart_close
 138 };
 139
 140 // Returns 'v' (valid value), 'i' (invalid; required GA4GH field missing),
 141 // or upon encountering an unexpected token, that token's type.
 142 // Explicit `return '?'` means a JSON parsing error, typically a member key
 143 // that is not a string.  An unexpected token may be a valid token that was
 144 // not the type expected for a particular GA4GH field, or it may be '?' or
 145 // '\0' which should be propagated.
 146 static char
 147 parse_ga4gh_redirect_json(hFILE_multipart *fp, hFILE *json,
 148                           kstring_t *b, kstring_t *header)
 149 {
 150     hts_json_token t;
 151
 152     if (hts_json_fnext(json, &t, b) != '{') return t.type;
 153     while (hts_json_fnext(json, &t, b) != '}') {
 154         if (t.type != 's') return '?';
 155
 156         if (strcmp(t.str, "urls") == 0) {
 157             if (hts_json_fnext(json, &t, b) != '[') return t.type;
 158
 159             while (hts_json_fnext(json, &t, b) != ']') {
 160                 hfile_part *part;
 161                 size_t n = 0, max = 0;
 162
 163                 hts_expand(hfile_part, fp->nparts+1, fp->maxparts, fp->parts);
 164                 part = &fp->parts[fp->nparts++];
 165                 part->url = NULL;
 166                 part->headers = NULL;
 167
 168                 if (t.type != '{') return t.type;
 169                 while (hts_json_fnext(json, &t, b) != '}') {
 170                     if (t.type != 's') return '?';
 171
 172                     if (strcmp(t.str, "url") == 0) {
 173                         if (hts_json_fnext(json, &t, b) != 's') return t.type;
 174                         part->url = ks_release(b);
 175                     }
 176                     else if (strcmp(t.str, "headers") == 0) {
 177                         if (hts_json_fnext(json, &t, b) != '{') return t.type;
 178
 179                         while (hts_json_fnext(json, &t, header) != '}') {
 180                             if (t.type != 's') return '?';
 181
 182                             if (hts_json_fnext(json, &t, b) != 's')
 183                                 return t.type;
 184
 185                             kputs(": ", header);
 186                             kputs(t.str, header);
 187                             n++;
 188                             hts_expand(char *, n+1, max, part->headers);
 189                             part->headers[n-1] = ks_release(header);
 190                             part->headers[n] = NULL;
 191                         }
 192                     }
 193                     else if (hts_json_fskip_value(json, '\0') != 'v')
 194                         return '?';
 195                 }
 196
 197                 if (! part->url) return 'i';
 198             }
 199         }
 200         else if (strcmp(t.str, "format") == 0) {
 201             if (hts_json_fnext(json, &t, b) != 's') return t.type;
 202
 203             hts_log_debug("GA4GH JSON redirection to multipart %s data", t.str);
 204         }
 205         else if (hts_json_fskip_value(json, '\0') != 'v') return '?';
 206     }
 207
 208     if (hts_json_fnext(json, &t, b) != '\0') return '?';
 209
 210     return 'v';
 211 }
 212
 213 hFILE *hopen_json_redirect(hFILE *hfile, const char *mode)
 214 {
 215     hFILE_multipart *fp;
 216     kstring_t s1 = { 0, 0, NULL }, s2 = { 0, 0, NULL };
 217     char ret;
 218
 219     fp = (hFILE_multipart *) hfile_init(sizeof (hFILE_multipart), mode, 0);
 220     if (fp == NULL) return NULL;
 221
 222     fp->parts = NULL;
 223     fp->nparts = fp->maxparts = 0;
 224
 225     ret = parse_ga4gh_redirect_json(fp, hfile, &s1, &s2);
 226     free(s1.s);
 227     free(s2.s);
 228     if (ret != 'v') {
 229         free_all_parts(fp);
 230         hfile_destroy((hFILE *) fp);
 231         errno = (ret == '?' || ret == '\0')? EPROTO : EINVAL;
 232         return NULL;
 233     }
 234
 235     fp->current = 0;
 236     fp->currentfp = NULL;
 237     fp->base.backend = &multipart_backend;
 238     return &fp->base;
 239 }