=================================================================== RCS file: /cvs/mandoc/chars.c,v retrieving revision 1.19 retrieving revision 1.45 diff -u -p -r1.19 -r1.45 --- mandoc/chars.c 2010/06/01 11:47:28 1.19 +++ mandoc/chars.c 2011/05/24 21:24:16 1.45 @@ -1,6 +1,7 @@ -/* $Id: chars.c,v 1.19 2010/06/01 11:47:28 kristaps Exp $ */ +/* $Id: chars.c,v 1.45 2011/05/24 21:24:16 kristaps Exp $ */ /* - * Copyright (c) 2009 Kristaps Dzonsons + * Copyright (c) 2009, 2010 Kristaps Dzonsons + * Copyright (c) 2011 Ingo Schwarze * * Permission to use, copy, modify, and distribute this software for any * purpose with or without fee is hereby granted, provided that the above @@ -19,12 +20,13 @@ #endif #include +#include #include #include #include #include "mandoc.h" -#include "chars.h" +#include "libmandoc.h" #define PRINT_HI 126 #define PRINT_LO 32 @@ -33,57 +35,38 @@ struct ln { struct ln *next; const char *code; const char *ascii; - const char *html; - size_t codesz; - size_t asciisz; - size_t htmlsz; - int type; -#define CHARS_CHAR (1 << 0) -#define CHARS_STRING (1 << 1) -#define CHARS_BOTH (CHARS_CHAR | CHARS_STRING) + int unicode; }; -#define LINES_MAX 370 +#define LINES_MAX 325 -#define CHAR(w, x, y, z, a, b) \ - { NULL, (w), (y), (a), (x), (z), (b), CHARS_CHAR }, -#define STRING(w, x, y, z, a, b) \ - { NULL, (w), (y), (a), (x), (z), (b), CHARS_STRING }, -#define BOTH(w, x, y, z, a, b) \ - { NULL, (w), (y), (a), (x), (z), (b), CHARS_BOTH }, +#define CHAR(in, ch, code) \ + { NULL, (in), (ch), (code) }, #define CHAR_TBL_START static struct ln lines[LINES_MAX] = { #define CHAR_TBL_END }; #include "chars.in" -struct tbl { - enum chars type; +struct mchars { struct ln **htab; }; -static inline int match(const struct ln *, - const char *, size_t, int); -static const char *find(struct tbl *, const char *, - size_t, size_t *, int); +static inline int match(const struct ln *, const char *, size_t); +static const struct ln *find(struct mchars *, const char *, size_t); - void -chars_free(void *arg) +mchars_free(struct mchars *arg) { - struct tbl *tab; - tab = (struct tbl *)arg; - - free(tab->htab); - free(tab); + free(arg->htab); + free(arg); } - -void * -chars_init(enum chars type) +struct mchars * +mchars_alloc(void) { - struct tbl *tab; + struct mchars *tab; struct ln **htab; struct ln *pp; int i, hash; @@ -95,18 +78,9 @@ chars_init(enum chars type) * (they're in-line re-ordered during lookup). */ - tab = malloc(sizeof(struct tbl)); - if (NULL == tab) { - perror(NULL); - exit(EXIT_FAILURE); - } + tab = mandoc_malloc(sizeof(struct mchars)); + htab = mandoc_calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln **)); - htab = calloc(PRINT_HI - PRINT_LO + 1, sizeof(struct ln **)); - if (NULL == htab) { - perror(NULL); - exit(EXIT_FAILURE); - } - for (i = 0; i < LINES_MAX; i++) { hash = (int)lines[i].code[0] - PRINT_LO; @@ -121,36 +95,113 @@ chars_init(enum chars type) } tab->htab = htab; - tab->type = type; return(tab); } -const char * -chars_a2ascii(void *arg, const char *p, size_t sz, size_t *rsz) +/* + * Special character to Unicode codepoint. + */ +int +mchars_spec2cp(struct mchars *arg, const char *p, size_t sz) { + const struct ln *ln; - return(find((struct tbl *)arg, p, sz, rsz, 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); + if (NULL == ln) + return(-1); + return(ln->unicode); +} + +/* + * Numbered character string to ASCII codepoint. + * 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; + + if ((i = mandoc_strntou(p, sz, 10)) < 0) + return('\0'); + return(isprint(i) ? i : '\0'); +} + +/* + * Hex character string to Unicode codepoint. + * If the character is illegal, returns '\0'. + */ +int +mchars_num2uc(const char *p, size_t sz) +{ + int i; + + if ((i = mandoc_strntou(p, sz, 16)) < 0) + return('\0'); + /* FIXME: make sure we're not in a bogus range. */ + return(i > 0x80 && i <= 0x10FFFF ? i : '\0'); +} + +/* + * Special character to string array. + */ const char * -chars_a2res(void *arg, const char *p, size_t sz, size_t *rsz) +mchars_spec2str(struct mchars *arg, const char *p, size_t sz, size_t *rsz) { + const struct ln *ln; - return(find((struct tbl *)arg, p, sz, rsz, CHARS_STRING)); + ln = find(arg, p, sz); + if (NULL == ln) + return(NULL); + + *rsz = strlen(ln->ascii); + return(ln->ascii); } +/* + * Reserved word to string array. + */ +const char * +mchars_res2str(struct mchars *arg, const char *p, size_t sz, size_t *rsz) +{ + const struct ln *ln; -static const char * -find(struct tbl *tab, const char *p, size_t sz, size_t *rsz, int type) + ln = find(arg, p, sz); + if (NULL == ln) + return(NULL); + + *rsz = strlen(ln->ascii); + return(ln->ascii); +} + +static const struct ln * +find(struct mchars *tab, const char *p, size_t sz) { struct ln *pp, *prev; struct ln **htab; int hash; assert(p); - assert(sz > 0); + if (0 == sz) + return(NULL); if (p[0] < PRINT_LO || p[0] > PRINT_HI) return(NULL); @@ -168,7 +219,7 @@ find(struct tbl *tab, const char *p, size_t sz, size_t return(NULL); for (prev = NULL; pp; pp = pp->next) { - if ( ! match(pp, p, sz, type)) { + if ( ! match(pp, p, sz)) { prev = pp; continue; } @@ -179,25 +230,17 @@ find(struct tbl *tab, const char *p, size_t sz, size_t htab[hash] = pp; } - if (CHARS_HTML == tab->type) { - *rsz = pp->htmlsz; - return(pp->html); - } - *rsz = pp->asciisz; - return(pp->ascii); + return(pp); } return(NULL); } - static inline int -match(const struct ln *ln, const char *p, size_t sz, int type) +match(const struct ln *ln, const char *p, size_t sz) { - if ( ! (ln->type & type)) + if (strncmp(ln->code, p, sz)) return(0); - if (ln->codesz != sz) - return(0); - return(0 == strncmp(ln->code, p, sz)); + return('\0' == ln->code[(int)sz]); }