mandoc.c

   1 /*      $Id: mandoc.c,v 1.25 2010/07/21 20:35:03 kristaps Exp $ */
   2 /*
   3  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
   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 #ifdef HAVE_CONFIG_H
  18 #include "config.h"
  19 #endif
  20
  21 #include <sys/types.h>
  22
  23 #include <assert.h>
  24 #include <ctype.h>
  25 #include <stdlib.h>
  26 #include <stdio.h>
  27 #include <string.h>
  28 #include <time.h>
  29
  30 #include "mandoc.h"
  31 #include "libmandoc.h"
  32
  33 static  int      a2time(time_t *, const char *, const char *);
  34
  35
  36 int
  37 mandoc_special(char *p)
  38 {
  39         int              len, i;
  40         char             term;
  41         char            *sv;
  42
  43         len = 0;
  44         term = '\0';
  45         sv = p;
  46
  47         assert('\\' == *p);
  48         p++;
  49
  50         switch (*p++) {
  51 #if 0
  52         case ('Z'):
  53                 /* FALLTHROUGH */
  54         case ('X'):
  55                 /* FALLTHROUGH */
  56         case ('x'):
  57                 /* FALLTHROUGH */
  58         case ('w'):
  59                 /* FALLTHROUGH */
  60         case ('v'):
  61                 /* FALLTHROUGH */
  62         case ('S'):
  63                 /* FALLTHROUGH */
  64         case ('R'):
  65                 /* FALLTHROUGH */
  66         case ('o'):
  67                 /* FALLTHROUGH */
  68         case ('N'):
  69                 /* FALLTHROUGH */
  70         case ('l'):
  71                 /* FALLTHROUGH */
  72         case ('L'):
  73                 /* FALLTHROUGH */
  74         case ('H'):
  75                 /* FALLTHROUGH */
  76         case ('h'):
  77                 /* FALLTHROUGH */
  78         case ('D'):
  79                 /* FALLTHROUGH */
  80         case ('C'):
  81                 /* FALLTHROUGH */
  82         case ('b'):
  83                 /* FALLTHROUGH */
  84         case ('B'):
  85                 /* FALLTHROUGH */
  86         case ('a'):
  87                 /* FALLTHROUGH */
  88         case ('A'):
  89                 if (*p++ != '\'')
  90                         return(0);
  91                 term = '\'';
  92                 break;
  93 #endif
  94         case ('s'):
  95                 if (ASCII_HYPH == *p)
  96                         *p = '-';
  97                 if ('+' == *p || '-' == *p)
  98                         p++;
  99
 100                 i = ('s' != *(p - 1));
 101
 102                 switch (*p++) {
 103                 case ('('):
 104                         len = 2;
 105                         break;
 106                 case ('['):
 107                         term = ']';
 108                         break;
 109                 case ('\''):
 110                         term = '\'';
 111                         break;
 112                 default:
 113                         len = 1;
 114                         p--;
 115                         break;
 116                 }
 117
 118                 if (ASCII_HYPH == *p)
 119                         *p = '-';
 120                 if ('+' == *p || '-' == *p) {
 121                         if (i++)
 122                                 return(0);
 123                         p++;
 124                 }
 125
 126                 if (0 == i)
 127                         return(0);
 128                 break;
 129 #if 0
 130         case ('Y'):
 131                 /* FALLTHROUGH */
 132         case ('V'):
 133                 /* FALLTHROUGH */
 134         case ('$'):
 135                 /* FALLTHROUGH */
 136         case ('n'):
 137                 /* FALLTHROUGH */
 138         case ('k'):
 139                 /* FALLTHROUGH */
 140 #endif
 141         case ('M'):
 142                 /* FALLTHROUGH */
 143         case ('m'):
 144                 /* FALLTHROUGH */
 145         case ('f'):
 146                 /* FALLTHROUGH */
 147         case ('F'):
 148                 /* FALLTHROUGH */
 149         case ('*'):
 150                 switch (*p++) {
 151                 case ('('):
 152                         len = 2;
 153                         break;
 154                 case ('['):
 155                         term = ']';
 156                         break;
 157                 default:
 158                         len = 1;
 159                         p--;
 160                         break;
 161                 }
 162                 break;
 163         case ('('):
 164                 len = 2;
 165                 break;
 166         case ('['):
 167                 term = ']';
 168                 break;
 169         default:
 170                 len = 1;
 171                 p--;
 172                 break;
 173         }
 174
 175         if (term) {
 176                 for ( ; *p && term != *p; p++)
 177                         if (ASCII_HYPH == *p)
 178                                 *p = '-';
 179                 return(*p ? (int)(p - sv) : 0);
 180         }
 181
 182         for (i = 0; *p && i < len; i++, p++)
 183                 if (ASCII_HYPH == *p)
 184                         *p = '-';
 185         return(i == len ? (int)(p - sv) : 0);
 186 }
 187
 188
 189 void *
 190 mandoc_calloc(size_t num, size_t size)
 191 {
 192         void            *ptr;
 193
 194         ptr = calloc(num, size);
 195         if (NULL == ptr) {
 196                 perror(NULL);
 197                 exit(EXIT_FAILURE);
 198         }
 199
 200         return(ptr);
 201 }
 202
 203
 204 void *
 205 mandoc_malloc(size_t size)
 206 {
 207         void            *ptr;
 208
 209         ptr = malloc(size);
 210         if (NULL == ptr) {
 211                 perror(NULL);
 212                 exit(EXIT_FAILURE);
 213         }
 214
 215         return(ptr);
 216 }
 217
 218
 219 void *
 220 mandoc_realloc(void *ptr, size_t size)
 221 {
 222
 223         ptr = realloc(ptr, size);
 224         if (NULL == ptr) {
 225                 perror(NULL);
 226                 exit(EXIT_FAILURE);
 227         }
 228
 229         return(ptr);
 230 }
 231
 232
 233 char *
 234 mandoc_strdup(const char *ptr)
 235 {
 236         char            *p;
 237
 238         p = strdup(ptr);
 239         if (NULL == p) {
 240                 perror(NULL);
 241                 exit(EXIT_FAILURE);
 242         }
 243
 244         return(p);
 245 }
 246
 247
 248 static int
 249 a2time(time_t *t, const char *fmt, const char *p)
 250 {
 251         struct tm        tm;
 252         char            *pp;
 253
 254         memset(&tm, 0, sizeof(struct tm));
 255
 256         pp = strptime(p, fmt, &tm);
 257         if (NULL != pp && '\0' == *pp) {
 258                 *t = mktime(&tm);
 259                 return(1);
 260         }
 261
 262         return(0);
 263 }
 264
 265
 266 /*
 267  * Convert from a manual date string (see mdoc(7) and man(7)) into a
 268  * date according to the stipulated date type.
 269  */
 270 time_t
 271 mandoc_a2time(int flags, const char *p)
 272 {
 273         time_t           t;
 274
 275         if (MTIME_MDOCDATE & flags) {
 276                 if (0 == strcmp(p, "$" "Mdocdate$"))
 277                         return(time(NULL));
 278                 if (a2time(&t, "$" "Mdocdate: %b %d %Y $", p))
 279                         return(t);
 280         }
 281
 282         if (MTIME_CANONICAL & flags || MTIME_REDUCED & flags)
 283                 if (a2time(&t, "%b %d, %Y", p))
 284                         return(t);
 285
 286         if (MTIME_ISO_8601 & flags)
 287                 if (a2time(&t, "%Y-%m-%d", p))
 288                         return(t);
 289
 290         if (MTIME_REDUCED & flags) {
 291                 if (a2time(&t, "%d, %Y", p))
 292                         return(t);
 293                 if (a2time(&t, "%Y", p))
 294                         return(t);
 295         }
 296
 297         return(0);
 298 }
 299
 300
 301 int
 302 mandoc_eos(const char *p, size_t sz, int enclosed)
 303 {
 304         const char *q;
 305         int found;
 306
 307         if (0 == sz)
 308                 return(0);
 309
 310         /*
 311          * End-of-sentence recognition must include situations where
 312          * some symbols, such as `)', allow prior EOS punctuation to
 313          * propogate outward.
 314          */
 315
 316         found = 0;
 317         for (q = p + (int)sz - 1; q >= p; q--) {
 318                 switch (*q) {
 319                 case ('\"'):
 320                         /* FALLTHROUGH */
 321                 case ('\''):
 322                         /* FALLTHROUGH */
 323                 case (']'):
 324                         /* FALLTHROUGH */
 325                 case (')'):
 326                         if (0 == found)
 327                                 enclosed = 1;
 328                         break;
 329                 case ('.'):
 330                         /* FALLTHROUGH */
 331                 case ('!'):
 332                         /* FALLTHROUGH */
 333                 case ('?'):
 334                         found = 1;
 335                         break;
 336                 default:
 337                         return(found && (!enclosed || isalnum(*q)));
 338                 }
 339         }
 340
 341         return(found && !enclosed);
 342 }
 343
 344
 345 int
 346 mandoc_hyph(const char *start, const char *c)
 347 {
 348
 349         /*
 350          * Choose whether to break at a hyphenated character.  We only
 351          * do this if it's free-standing within a word.
 352          */
 353
 354         /* Skip first/last character of buffer. */
 355         if (c == start || '\0' == *(c + 1))
 356                 return(0);
 357         /* Skip first/last character of word. */
 358         if ('\t' == *(c + 1) || '\t' == *(c - 1))
 359                 return(0);
 360         if (' ' == *(c + 1) || ' ' == *(c - 1))
 361                 return(0);
 362         /* Skip double invocations. */
 363         if ('-' == *(c + 1) || '-' == *(c - 1))
 364                 return(0);
 365         /* Skip escapes. */
 366         if ('\\' == *(c - 1))
 367                 return(0);
 368
 369         return(1);
 370 }