=================================================================== RCS file: /cvs/mandoc/chars.c,v retrieving revision 1.42 retrieving revision 1.59 diff -u -p -r1.42 -r1.59 --- mandoc/chars.c 2011/05/15 15:47:46 1.42 +++ mandoc/chars.c 2014/08/10 23:54:41 1.59 @@ -1,6 +1,6 @@ -/* $Id: chars.c,v 1.42 2011/05/15 15:47:46 kristaps Exp $ */ +/* $Id: chars.c,v 1.59 2014/08/10 23:54:41 schwarze Exp $ */ /* - * Copyright (c) 2009, 2010 Kristaps Dzonsons + * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons * Copyright (c) 2011 Ingo Schwarze * * Permission to use, copy, modify, and distribute this software for any @@ -15,17 +15,18 @@ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ -#ifdef HAVE_CONFIG_H #include "config.h" -#endif +#include + #include #include -#include #include #include #include "mandoc.h" +#include "mandoc_aux.h" +#include "libmandoc.h" #define PRINT_HI 126 #define PRINT_LO 32 @@ -35,20 +36,12 @@ struct ln { const char *code; const char *ascii; int unicode; - int type; -#define CHARS_CHAR (1 << 0) -#define CHARS_STRING (1 << 1) -#define CHARS_BOTH (CHARS_CHAR | CHARS_STRING) }; -#define LINES_MAX 353 +#define LINES_MAX 330 #define CHAR(in, ch, code) \ - { NULL, (in), (ch), (code), CHARS_CHAR }, -#define STRING(in, ch, code) \ - { NULL, (in), (ch), (code), CHARS_STRING }, -#define BOTH(in, ch, code) \ - { NULL, (in), (ch), (code), CHARS_BOTH }, + { NULL, (in), (ch), (code) }, #define CHAR_TBL_START static struct ln lines[LINES_MAX] = { #define CHAR_TBL_END }; @@ -59,10 +52,10 @@ struct mchars { struct ln **htab; }; -static inline int match(const struct ln *, - const char *, size_t, int); -static const struct ln *find(struct mchars *, const char *, size_t, int); +static const struct ln *find(const struct mchars *, + const char *, size_t); + void mchars_free(struct mchars *arg) { @@ -82,12 +75,11 @@ mchars_alloc(void) /* * Constructs a very basic chaining hashtable. The hash routine * is simply the integral value of the first character. - * Subsequent entries are chained in the order they're processed - * (they're in-line re-ordered during lookup). + * Subsequent entries are chained in the order they're processed. */ tab = mandoc_malloc(sizeof(struct mchars)); - htab = mandoc_calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln **)); + htab = mandoc_calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln *)); for (i = 0; i < LINES_MAX; i++) { hash = (int)lines[i].code[0] - PRINT_LO; @@ -106,139 +98,83 @@ mchars_alloc(void) return(tab); } - -/* - * Special character to Unicode codepoint. - */ int -mchars_spec2cp(struct mchars *arg, const char *p, size_t sz) +mchars_spec2cp(const struct mchars *arg, const char *p, size_t sz) { const struct ln *ln; - ln = find(arg, p, sz, CHARS_CHAR); + ln = find(arg, p, sz); if (NULL == ln) return(-1); return(ln->unicode); } - -/* - * Reserved word to Unicode codepoint. - */ -int -mchars_res2cp(struct mchars *arg, const char *p, size_t sz) -{ - const struct ln *ln; - - ln = find(arg, p, sz, CHARS_STRING); - if (NULL == ln) - return(-1); - return(ln->unicode); -} - -/* - * Numbered character to literal character. - * This can only be a printable character (i.e., alnum, punct, space) so - * prevent the character from ruining our state (backspace, newline, and - * so on). - * If the character is illegal, returns '\0'. - */ char mchars_num2char(const char *p, size_t sz) { - int i; + int i; - if (sz > 3) + if ((i = mandoc_strntoi(p, sz, 10)) < 0) return('\0'); - i = atoi(p); - /* LINTED */ - return(isprint(i) ? i : '\0'); + return(i > 0 && i < 256 && isprint(i) ? i : '\0'); } -/* - * Special character to string array. - */ -const char * -mchars_spec2str(struct mchars *arg, const char *p, size_t sz, size_t *rsz) +int +mchars_num2uc(const char *p, size_t sz) { - const struct ln *ln; + int i; - ln = find(arg, p, sz, CHARS_CHAR); - if (NULL == ln) - return(NULL); + if ((i = mandoc_strntoi(p, sz, 16)) < 0) + return('\0'); - *rsz = strlen(ln->ascii); - return(ln->ascii); + /* + * Security warning: + * Never extend the range of accepted characters + * to overlap with the ASCII range, 0x00-0x7F + * without re-auditing the callers of this function. + * Some callers might relay on the fact that we never + * return ASCII characters for their escaping decisions. + * + * XXX Code is missing here to exclude bogus ranges. + */ + + return(i > 0x80 && i <= 0x10FFFF ? i : '\0'); } -/* - * Reserved word to string array. - */ const char * -mchars_res2str(struct mchars *arg, const char *p, size_t sz, size_t *rsz) +mchars_spec2str(const struct mchars *arg, + const char *p, size_t sz, size_t *rsz) { const struct ln *ln; - ln = find(arg, p, sz, CHARS_STRING); - if (NULL == ln) + ln = find(arg, p, sz); + if (NULL == ln) { + *rsz = 1; return(NULL); + } *rsz = strlen(ln->ascii); return(ln->ascii); } static const struct ln * -find(struct mchars *tab, const char *p, size_t sz, int type) +find(const struct mchars *tab, const char *p, size_t sz) { - struct ln *pp, *prev; - struct ln **htab; + const struct ln *pp; int hash; assert(p); - if (0 == sz) - return(NULL); - if (p[0] < PRINT_LO || p[0] > PRINT_HI) + if (0 == sz || p[0] < PRINT_LO || p[0] > PRINT_HI) return(NULL); - /* - * Lookup the symbol in the symbol hash. See ascii2htab for the - * hashtable specs. This dynamically re-orders the hash chain - * to optimise for repeat hits. - */ - hash = (int)p[0] - PRINT_LO; - htab = tab->htab; - if (NULL == (pp = htab[hash])) - return(NULL); + for (pp = tab->htab[hash]; pp; pp = pp->next) + if (0 == strncmp(pp->code, p, sz) && + '\0' == pp->code[(int)sz]) + return(pp); - for (prev = NULL; pp; pp = pp->next) { - if ( ! match(pp, p, sz, type)) { - prev = pp; - continue; - } - - if (prev) { - prev->next = pp->next; - pp->next = htab[hash]; - htab[hash] = pp; - } - - return(pp); - } - return(NULL); -} - -static inline int -match(const struct ln *ln, const char *p, size_t sz, int type) -{ - - if ( ! (ln->type & type)) - return(0); - if (strncmp(ln->code, p, sz)) - return(0); - return('\0' == ln->code[(int)sz]); }