chars.c

   1 /*      $Id: chars.c,v 1.46 2011/05/24 21:31:23 kristaps Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2011 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 #ifdef HAVE_CONFIG_H
  19 #include "config.h"
  20 #endif
  21
  22 #include <assert.h>
  23 #include <ctype.h>
  24 #include <stdio.h>
  25 #include <stdlib.h>
  26 #include <string.h>
  27
  28 #include "mandoc.h"
  29 #include "libmandoc.h"
  30
  31 #define PRINT_HI         126
  32 #define PRINT_LO         32
  33
  34 struct  ln {
  35         struct ln        *next;
  36         const char       *code;
  37         const char       *ascii;
  38         int               unicode;
  39 };
  40
  41 #define LINES_MAX         325
  42
  43 #define CHAR(in, ch, code) \
  44         { NULL, (in), (ch), (code) },
  45
  46 #define CHAR_TBL_START    static struct ln lines[LINES_MAX] = {
  47 #define CHAR_TBL_END      };
  48
  49 #include "chars.in"
  50
  51 struct  mchars {
  52         struct ln       **htab;
  53 };
  54
  55 static  inline int        match(const struct ln *, const char *, size_t);
  56 static  const struct ln  *find(struct mchars *, const char *, size_t);
  57
  58 void
  59 mchars_free(struct mchars *arg)
  60 {
  61
  62         free(arg->htab);
  63         free(arg);
  64 }
  65
  66 struct mchars *
  67 mchars_alloc(void)
  68 {
  69         struct mchars    *tab;
  70         struct ln       **htab;
  71         struct ln        *pp;
  72         int               i, hash;
  73
  74         /*
  75          * Constructs a very basic chaining hashtable.  The hash routine
  76          * is simply the integral value of the first character.
  77          * Subsequent entries are chained in the order they're processed
  78          * (they're in-line re-ordered during lookup).
  79          */
  80
  81         tab = mandoc_malloc(sizeof(struct mchars));
  82         htab = mandoc_calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln **));
  83
  84         for (i = 0; i < LINES_MAX; i++) {
  85                 hash = (int)lines[i].code[0] - PRINT_LO;
  86
  87                 if (NULL == (pp = htab[hash])) {
  88                         htab[hash] = &lines[i];
  89                         continue;
  90                 }
  91
  92                 for ( ; pp->next; pp = pp->next)
  93                         /* Scan ahead. */ ;
  94                 pp->next = &lines[i];
  95         }
  96
  97         tab->htab = htab;
  98         return(tab);
  99 }
 100
 101
 102 /*
 103  * Special character to Unicode codepoint.
 104  */
 105 int
 106 mchars_spec2cp(struct mchars *arg, const char *p, size_t sz)
 107 {
 108         const struct ln *ln;
 109
 110         ln = find(arg, p, sz);
 111         if (NULL == ln)
 112                 return(-1);
 113         return(ln->unicode);
 114 }
 115
 116 /*
 117  * Numbered character string to ASCII codepoint.
 118  * This can only be a printable character (i.e., alnum, punct, space) so
 119  * prevent the character from ruining our state (backspace, newline, and
 120  * so on).
 121  * If the character is illegal, returns '\0'.
 122  */
 123 char
 124 mchars_num2char(const char *p, size_t sz)
 125 {
 126         int               i;
 127
 128         if ((i = mandoc_strntou(p, sz, 10)) < 0)
 129                 return('\0');
 130         return(isprint(i) ? i : '\0');
 131 }
 132
 133 /*
 134  * Hex character string to Unicode codepoint.
 135  * If the character is illegal, returns '\0'.
 136  */
 137 int
 138 mchars_num2uc(const char *p, size_t sz)
 139 {
 140         int               i;
 141
 142         if ((i = mandoc_strntou(p, sz, 16)) < 0)
 143                 return('\0');
 144         /* FIXME: make sure we're not in a bogus range. */
 145         return(i > 0x80 && i <= 0x10FFFF ? i : '\0');
 146 }
 147
 148 /*
 149  * Special character to string array.
 150  */
 151 const char *
 152 mchars_spec2str(struct mchars *arg, const char *p, size_t sz, size_t *rsz)
 153 {
 154         const struct ln *ln;
 155
 156         ln = find(arg, p, sz);
 157         if (NULL == ln)
 158                 return(NULL);
 159
 160         *rsz = strlen(ln->ascii);
 161         return(ln->ascii);
 162 }
 163
 164 static const struct ln *
 165 find(struct mchars *tab, const char *p, size_t sz)
 166 {
 167         struct ln        *pp, *prev;
 168         struct ln       **htab;
 169         int               hash;
 170
 171         assert(p);
 172         if (0 == sz)
 173                 return(NULL);
 174
 175         if (p[0] < PRINT_LO || p[0] > PRINT_HI)
 176                 return(NULL);
 177
 178         /*
 179          * Lookup the symbol in the symbol hash.  See ascii2htab for the
 180          * hashtable specs.  This dynamically re-orders the hash chain
 181          * to optimise for repeat hits.
 182          */
 183
 184         hash = (int)p[0] - PRINT_LO;
 185         htab = tab->htab;
 186
 187         if (NULL == (pp = htab[hash]))
 188                 return(NULL);
 189
 190         for (prev = NULL; pp; pp = pp->next) {
 191                 if ( ! match(pp, p, sz)) {
 192                         prev = pp;
 193                         continue;
 194                 }
 195
 196                 if (prev) {
 197                         prev->next = pp->next;
 198                         pp->next = htab[hash];
 199                         htab[hash] = pp;
 200                 }
 201
 202                 return(pp);
 203         }
 204
 205         return(NULL);
 206 }
 207
 208 static inline int
 209 match(const struct ln *ln, const char *p, size_t sz)
 210 {
 211
 212         if (strncmp(ln->code, p, sz))
 213                 return(0);
 214         return('\0' == ln->code[(int)sz]);
 215 }