preconv.c

   1 /*      $Id: preconv.c,v 1.10 2014/10/26 18:22:51 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2014 Ingo Schwarze <schwarze@openbsd.org>
   5  *
   6  * Permission to use, copy, modify, and distribute this software for any
   7  * purpose with or without fee is hereby granted, provided that the above
   8  * copyright notice and this permission notice appear in all copies.
   9  *
  10  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  11  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  12  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
  13  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  14  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  15  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  16  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  17  */
  18 #include "config.h"
  19
  20 #include <sys/types.h>
  21
  22 #include <stdio.h>
  23 #include <string.h>
  24 #include "mandoc.h"
  25 #include "libmandoc.h"
  26
  27 int
  28 preconv_encode(struct buf *ib, struct buf *ob, int *filenc)
  29 {
  30         size_t           i;
  31         const long       one = 1L;
  32         int              state, be;
  33         unsigned int     accum;
  34         unsigned char    cu;
  35
  36         if ( ! (*filenc & MPARSE_UTF8))
  37                 goto latin;
  38
  39         state = 0;
  40         accum = 0U;
  41         be = 0;
  42
  43         /* Quick test for big-endian value. */
  44
  45         if ( ! (*((const char *)(&one))))
  46                 be = 1;
  47
  48         for (i = ib->offs; i < ib->sz; i++) {
  49                 cu = ib->buf[i];
  50                 if (state) {
  51                         if ( ! (cu & 128) || (cu & 64)) {
  52                                 /* Bad sequence header. */
  53                                 break;
  54                         }
  55
  56                         /* Accept only legitimate bit patterns. */
  57
  58                         if (cu > 191 || cu < 128) {
  59                                 /* Bad in-sequence bits. */
  60                                 break;
  61                         }
  62
  63                         accum |= (cu & 63) << --state * 6;
  64
  65                         if (state)
  66                                 continue;
  67
  68                         /*
  69                          * Accum is held in little-endian order as
  70                          * stipulated by the UTF-8 sequence coding.  We
  71                          * need to convert to a native big-endian if our
  72                          * architecture requires it.
  73                          */
  74
  75                         if (be)
  76                                 accum = (accum >> 24) |
  77                                         ((accum << 8) & 0x00FF0000) |
  78                                         ((accum >> 8) & 0x0000FF00) |
  79                                         (accum << 24);
  80
  81                         if (accum < 0x80)
  82                                 ob->buf[ob->offs++] = accum;
  83                         else
  84                                 ob->offs += snprintf(ob->buf + ob->offs,
  85                                     11, "\\[u%.4X]", accum);
  86                         ib->offs = i + 1;
  87                         *filenc &= ~MPARSE_LATIN1;
  88                         return(1);
  89                 } else {
  90                         /*
  91                          * Entering a UTF-8 state:  if we encounter a
  92                          * UTF-8 bitmask, calculate the expected UTF-8
  93                          * state from it.
  94                          */
  95                         for (state = 0; state < 7; state++)
  96                                 if ( ! (cu & (1 << (7 - state))))
  97                                         break;
  98
  99                         /* Accept only legitimate bit patterns. */
 100
 101                         switch (state--) {
 102                         case (4):
 103                                 if (cu <= 244 && cu >= 240) {
 104                                         accum = (cu & 7) << 18;
 105                                         continue;
 106                                 }
 107                                 /* Bad 4-sequence start bits. */
 108                                 break;
 109                         case (3):
 110                                 if (cu <= 239 && cu >= 224) {
 111                                         accum = (cu & 15) << 12;
 112                                         continue;
 113                                 }
 114                                 /* Bad 3-sequence start bits. */
 115                                 break;
 116                         case (2):
 117                                 if (cu <= 223 && cu >= 194) {
 118                                         accum = (cu & 31) << 6;
 119                                         continue;
 120                                 }
 121                                 /* Bad 2-sequence start bits. */
 122                                 break;
 123                         default:
 124                                 /* Bad sequence bit mask. */
 125                                 break;
 126                         }
 127                         break;
 128                 }
 129         }
 130
 131         /* FALLTHROUGH: Invalid or incomplete UTF-8 sequence. */
 132
 133 latin:
 134         if ( ! (*filenc & MPARSE_LATIN1))
 135                 return(0);
 136
 137         ob->offs += snprintf(ob->buf + ob->offs, 11,
 138             "\\[u%.4X]", (unsigned char)ib->buf[ib->offs++]);
 139
 140         *filenc &= ~MPARSE_UTF8;
 141         return(1);
 142 }
 143
 144 int
 145 preconv_cue(const struct buf *b)
 146 {
 147         const char      *ln, *eoln, *eoph;
 148         size_t           sz, phsz;
 149
 150         ln = b->buf + b->offs;
 151         sz = b->sz - b->offs;
 152
 153         /* Look for the end-of-line. */
 154
 155         if (NULL == (eoln = memchr(ln, '\n', sz)))
 156                 eoln = ln + sz;
 157
 158         /* Check if we have the correct header/trailer. */
 159
 160         if ((sz = (size_t)(eoln - ln)) < 10 ||
 161             memcmp(ln, ".\\\" -*-", 7) || memcmp(eoln - 3, "-*-", 3))
 162                 return(MPARSE_UTF8 | MPARSE_LATIN1);
 163
 164         /* Move after the header and adjust for the trailer. */
 165
 166         ln += 7;
 167         sz -= 10;
 168
 169         while (sz > 0) {
 170                 while (sz > 0 && ' ' == *ln) {
 171                         ln++;
 172                         sz--;
 173                 }
 174                 if (0 == sz)
 175                         break;
 176
 177                 /* Find the end-of-phrase marker (or eoln). */
 178
 179                 if (NULL == (eoph = memchr(ln, ';', sz)))
 180                         eoph = eoln - 3;
 181                 else
 182                         eoph++;
 183
 184                 /* Only account for the "coding" phrase. */
 185
 186                 if ((phsz = eoph - ln) < 7 ||
 187                     strncasecmp(ln, "coding:", 7)) {
 188                         sz -= phsz;
 189                         ln += phsz;
 190                         continue;
 191                 }
 192
 193                 sz -= 7;
 194                 ln += 7;
 195
 196                 while (sz > 0 && ' ' == *ln) {
 197                         ln++;
 198                         sz--;
 199                 }
 200                 if (0 == sz)
 201                         return(0);
 202
 203                 /* Check us against known encodings. */
 204
 205                 if (phsz > 4 && !strncasecmp(ln, "utf-8", 5))
 206                         return(MPARSE_UTF8);
 207                 if (phsz > 10 && !strncasecmp(ln, "iso-latin-1", 11))
 208                         return(MPARSE_LATIN1);
 209                 return(0);
 210         }
 211         return(MPARSE_UTF8 | MPARSE_LATIN1);
 212 }