Annotation of mandoc/chars.c, Revision 1.59
1.59 ! schwarze 1: /* $Id: chars.c,v 1.58 2014/07/23 15:00:08 schwarze Exp $ */
1.1 kristaps 2: /*
1.51 schwarze 3: * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
1.32 schwarze 4: * Copyright (c) 2011 Ingo Schwarze <schwarze@openbsd.org>
1.1 kristaps 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: */
1.14 kristaps 18: #include "config.h"
1.59 ! schwarze 19:
! 20: #include <sys/types.h>
1.14 kristaps 21:
1.1 kristaps 22: #include <assert.h>
1.40 kristaps 23: #include <ctype.h>
1.1 kristaps 24: #include <stdlib.h>
25: #include <string.h>
26:
1.18 kristaps 27: #include "mandoc.h"
1.56 schwarze 28: #include "mandoc_aux.h"
1.43 kristaps 29: #include "libmandoc.h"
1.1 kristaps 30:
1.2 kristaps 31: #define PRINT_HI 126
32: #define PRINT_LO 32
1.1 kristaps 33:
34: struct ln {
35: struct ln *next;
36: const char *code;
1.2 kristaps 37: const char *ascii;
1.21 kristaps 38: int unicode;
1.1 kristaps 39: };
40:
1.55 schwarze 41: #define LINES_MAX 330
1.1 kristaps 42:
1.24 kristaps 43: #define CHAR(in, ch, code) \
1.45 kristaps 44: { NULL, (in), (ch), (code) },
1.1 kristaps 45:
1.13 kristaps 46: #define CHAR_TBL_START static struct ln lines[LINES_MAX] = {
47: #define CHAR_TBL_END };
48:
1.1 kristaps 49: #include "chars.in"
50:
1.36 kristaps 51: struct mchars {
1.1 kristaps 52: struct ln **htab;
53: };
54:
1.57 schwarze 55: static const struct ln *find(const struct mchars *,
1.52 kristaps 56: const char *, size_t);
1.1 kristaps 57:
1.57 schwarze 58:
1.1 kristaps 59: void
1.36 kristaps 60: mchars_free(struct mchars *arg)
1.1 kristaps 61: {
62:
1.36 kristaps 63: free(arg->htab);
64: free(arg);
1.1 kristaps 65: }
66:
1.36 kristaps 67: struct mchars *
1.38 kristaps 68: mchars_alloc(void)
1.1 kristaps 69: {
1.36 kristaps 70: struct mchars *tab;
1.1 kristaps 71: struct ln **htab;
72: struct ln *pp;
73: int i, hash;
74:
75: /*
76: * Constructs a very basic chaining hashtable. The hash routine
77: * is simply the integral value of the first character.
1.47 kristaps 78: * Subsequent entries are chained in the order they're processed.
1.1 kristaps 79: */
80:
1.36 kristaps 81: tab = mandoc_malloc(sizeof(struct mchars));
1.53 schwarze 82: htab = mandoc_calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln *));
1.1 kristaps 83:
84: for (i = 0; i < LINES_MAX; i++) {
1.2 kristaps 85: hash = (int)lines[i].code[0] - PRINT_LO;
1.1 kristaps 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:
1.21 kristaps 101: int
1.52 kristaps 102: mchars_spec2cp(const struct mchars *arg, const char *p, size_t sz)
1.21 kristaps 103: {
104: const struct ln *ln;
105:
1.45 kristaps 106: ln = find(arg, p, sz);
1.21 kristaps 107: if (NULL == ln)
108: return(-1);
109: return(ln->unicode);
110: }
111:
1.39 kristaps 112: char
1.36 kristaps 113: mchars_num2char(const char *p, size_t sz)
1.32 schwarze 114: {
1.57 schwarze 115: int i;
1.32 schwarze 116:
1.48 kristaps 117: if ((i = mandoc_strntoi(p, sz, 10)) < 0)
1.39 kristaps 118: return('\0');
1.57 schwarze 119:
120: return(i > 0 && i < 256 && isprint(i) ? i : '\0');
1.44 kristaps 121: }
122:
123: int
124: mchars_num2uc(const char *p, size_t sz)
125: {
1.57 schwarze 126: int i;
1.39 kristaps 127:
1.48 kristaps 128: if ((i = mandoc_strntoi(p, sz, 16)) < 0)
1.44 kristaps 129: return('\0');
1.58 schwarze 130:
131: /*
132: * Security warning:
133: * Never extend the range of accepted characters
134: * to overlap with the ASCII range, 0x00-0x7F
135: * without re-auditing the callers of this function.
136: * Some callers might relay on the fact that we never
137: * return ASCII characters for their escaping decisions.
138: *
139: * XXX Code is missing here to exclude bogus ranges.
140: */
141:
1.44 kristaps 142: return(i > 0x80 && i <= 0x10FFFF ? i : '\0');
1.21 kristaps 143: }
144:
1.1 kristaps 145: const char *
1.57 schwarze 146: mchars_spec2str(const struct mchars *arg,
1.52 kristaps 147: const char *p, size_t sz, size_t *rsz)
1.1 kristaps 148: {
1.21 kristaps 149: const struct ln *ln;
1.1 kristaps 150:
1.45 kristaps 151: ln = find(arg, p, sz);
1.50 schwarze 152: if (NULL == ln) {
153: *rsz = 1;
1.21 kristaps 154: return(NULL);
1.50 schwarze 155: }
1.21 kristaps 156:
1.24 kristaps 157: *rsz = strlen(ln->ascii);
1.21 kristaps 158: return(ln->ascii);
1.1 kristaps 159: }
160:
1.21 kristaps 161: static const struct ln *
1.52 kristaps 162: find(const struct mchars *tab, const char *p, size_t sz)
1.1 kristaps 163: {
1.52 kristaps 164: const struct ln *pp;
1.1 kristaps 165: int hash;
166:
167: assert(p);
168:
1.47 kristaps 169: if (0 == sz || p[0] < PRINT_LO || p[0] > PRINT_HI)
1.1 kristaps 170: return(NULL);
171:
1.2 kristaps 172: hash = (int)p[0] - PRINT_LO;
1.1 kristaps 173:
1.47 kristaps 174: for (pp = tab->htab[hash]; pp; pp = pp->next)
1.57 schwarze 175: if (0 == strncmp(pp->code, p, sz) &&
176: '\0' == pp->code[(int)sz])
1.47 kristaps 177: return(pp);
1.1 kristaps 178:
179: return(NULL);
180: }
CVSweb