bin/mandoc/dbm_map.c

   1 /*      $Id: dbm_map.c,v 1.8 2017/02/17 14:43:54 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2016 Ingo Schwarze <schwarze@openbsd.org>
   4  *
   5  * Permission to use, copy, modify, and distribute this software for any
   6  * purpose with or without fee is hereby granted, provided that the above
   7  * copyright notice and this permission notice appear in all copies.
   8  *
   9  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  10  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  11  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
  12  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  13  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  14  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  15  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  16  *
  17  * Low-level routines for the map-based version
  18  * of the mandoc database, for read-only access.
  19  * The interface is defined in "dbm_map.h".
  20  */
  21 #include "config.h"
  22
  23 #include <sys/mman.h>
  24 #include <sys/stat.h>
  25 #include <sys/types.h>
  26
  27 #if HAVE_ENDIAN
  28 #include <endian.h>
  29 #elif HAVE_SYS_ENDIAN
  30 #include <sys/endian.h>
  31 #elif HAVE_NTOHL
  32 #include <arpa/inet.h>
  33 #endif
  34 #if HAVE_ERR
  35 #include <err.h>
  36 #endif
  37 #include <errno.h>
  38 #include <fcntl.h>
  39 #include <regex.h>
  40 #include <stdint.h>
  41 #include <stdlib.h>
  42 #include <string.h>
  43 #include <unistd.h>
  44
  45 #include "mansearch.h"
  46 #include "dbm_map.h"
  47 #include "dbm.h"
  48
  49 static struct stat       st;
  50 static char             *dbm_base;
  51 static int               ifd;
  52 static int32_t           max_offset;
  53
  54 /*
  55  * Open a disk-based database for read-only access.
  56  * Validate the file format as far as it is not mandoc-specific.
  57  * Return 0 on success.  Return -1 and set errno on failure.
  58  */
  59 int
  60 dbm_map(const char *fname)
  61 {
  62         int              save_errno;
  63         const int32_t   *magic;
  64
  65         if ((ifd = open(fname, O_RDONLY)) == -1)
  66                 return -1;
  67         if (fstat(ifd, &st) == -1)
  68                 goto fail;
  69         if (st.st_size < 5) {
  70                 warnx("dbm_map(%s): File too short", fname);
  71                 errno = EFTYPE;
  72                 goto fail;
  73         }
  74         if (st.st_size > INT32_MAX) {
  75                 errno = EFBIG;
  76                 goto fail;
  77         }
  78         if ((dbm_base = mmap(NULL, st.st_size, PROT_READ, MAP_SHARED,
  79             ifd, 0)) == MAP_FAILED)
  80                 goto fail;
  81         magic = dbm_getint(0);
  82         if (be32toh(*magic) != MANDOCDB_MAGIC) {
  83                 if (strncmp(dbm_base, "SQLite format 3", 15))
  84                         warnx("dbm_map(%s): "
  85                             "Bad initial magic %x (expected %x)",
  86                             fname, be32toh(*magic), MANDOCDB_MAGIC);
  87                 else
  88                         warnx("dbm_map(%s): "
  89                             "Obsolete format based on SQLite 3",
  90                             fname);
  91                 errno = EFTYPE;
  92                 goto fail;
  93         }
  94         magic = dbm_getint(1);
  95         if (be32toh(*magic) != MANDOCDB_VERSION) {
  96                 warnx("dbm_map(%s): Bad version number %d (expected %d)",
  97                     fname, be32toh(*magic), MANDOCDB_VERSION);
  98                 errno = EFTYPE;
  99                 goto fail;
 100         }
 101         max_offset = be32toh(*dbm_getint(3)) + sizeof(int32_t);
 102         if (st.st_size != max_offset) {
 103                 warnx("dbm_map(%s): Inconsistent file size %lld (expected %d)",
 104                     fname, (long long)st.st_size, max_offset);
 105                 errno = EFTYPE;
 106                 goto fail;
 107         }
 108         if ((magic = dbm_get(*dbm_getint(3))) == NULL) {
 109                 errno = EFTYPE;
 110                 goto fail;
 111         }
 112         if (be32toh(*magic) != MANDOCDB_MAGIC) {
 113                 warnx("dbm_map(%s): Bad final magic %x (expected %x)",
 114                     fname, be32toh(*magic), MANDOCDB_MAGIC);
 115                 errno = EFTYPE;
 116                 goto fail;
 117         }
 118         return 0;
 119
 120 fail:
 121         save_errno = errno;
 122         close(ifd);
 123         errno = save_errno;
 124         return -1;
 125 }
 126
 127 void
 128 dbm_unmap(void)
 129 {
 130         if (munmap(dbm_base, st.st_size) == -1)
 131                 warn("dbm_unmap: munmap");
 132         if (close(ifd) == -1)
 133                 warn("dbm_unmap: close");
 134         dbm_base = (char *)-1;
 135 }
 136
 137 /*
 138  * Take a raw integer as it was read from the database.
 139  * Interpret it as an offset into the database file
 140  * and return a pointer to that place in the file.
 141  */
 142 void *
 143 dbm_get(int32_t offset)
 144 {
 145         offset = be32toh(offset);
 146         if (offset < 0) {
 147                 warnx("dbm_get: Database corrupt: offset %d", offset);
 148                 return NULL;
 149         }
 150         if (offset >= max_offset) {
 151                 warnx("dbm_get: Database corrupt: offset %d > %d",
 152                     offset, max_offset);
 153                 return NULL;
 154         }
 155         return dbm_base + offset;
 156 }
 157
 158 /*
 159  * Assume the database starts with some integers.
 160  * Assume they are numbered starting from 0, increasing.
 161  * Get a pointer to one with the number "offset".
 162  */
 163 int32_t *
 164 dbm_getint(int32_t offset)
 165 {
 166         return (int32_t *)dbm_base + offset;
 167 }
 168
 169 /*
 170  * The reverse of dbm_get().
 171  * Take pointer into the database file
 172  * and convert it to the raw integer
 173  * that would be used to refer to that place in the file.
 174  */
 175 int32_t
 176 dbm_addr(const void *p)
 177 {
 178         return htobe32((const char *)p - dbm_base);
 179 }
 180
 181 int
 182 dbm_match(const struct dbm_match *match, const char *str)
 183 {
 184         switch (match->type) {
 185         case DBM_EXACT:
 186                 return strcmp(str, match->str) == 0;
 187         case DBM_SUB:
 188                 return strcasestr(str, match->str) != NULL;
 189         case DBM_REGEX:
 190                 return regexec(match->re, str, 0, NULL, 0) == 0;
 191         default:
 192                 abort();
 193         }
 194 }