=================================================================== RCS file: /cvs/mandoc/html.c,v retrieving revision 1.138 retrieving revision 1.163 diff -u -p -r1.138 -r1.163 --- mandoc/html.c 2011/05/14 16:28:23 1.138 +++ mandoc/html.c 2014/08/14 00:31:43 1.163 @@ -1,7 +1,7 @@ -/* $Id: html.c,v 1.138 2011/05/14 16:28:23 kristaps Exp $ */ +/* $Id: html.c,v 1.163 2014/08/14 00:31:43 schwarze Exp $ */ /* * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons - * Copyright (c) 2011 Ingo Schwarze + * Copyright (c) 2011, 2012, 2013, 2014 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 @@ -15,9 +15,7 @@ * 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 @@ -31,6 +29,7 @@ #include #include "mandoc.h" +#include "mandoc_aux.h" #include "libmandoc.h" #include "out.h" #include "html.h" @@ -94,30 +93,40 @@ static const char *const htmlattrs[ATTR_MAX] = { "colspan", /* ATTR_COLSPAN */ }; -static void print_num(struct html *, const char *, size_t); -static void print_spec(struct html *, const char *, size_t); -static void print_res(struct html *, const char *, size_t); -static void print_ctag(struct html *, enum htmltag); -static void print_doctype(struct html *); -static void print_xmltype(struct html *); -static int print_encode(struct html *, const char *, int); -static void print_metaf(struct html *, enum mandoc_esc); -static void print_attr(struct html *, - const char *, const char *); -static void *ml_alloc(char *, enum htmltype); +static const char *const roffscales[SCALE_MAX] = { + "cm", /* SCALE_CM */ + "in", /* SCALE_IN */ + "pc", /* SCALE_PC */ + "pt", /* SCALE_PT */ + "em", /* SCALE_EM */ + "em", /* SCALE_MM */ + "ex", /* SCALE_EN */ + "ex", /* SCALE_BU */ + "em", /* SCALE_VS */ + "ex", /* SCALE_FS */ +}; +static void bufncat(struct html *, const char *, size_t); +static void print_ctag(struct html *, enum htmltag); +static int print_escape(char); +static int print_encode(struct html *, const char *, int); +static void print_metaf(struct html *, enum mandoc_esc); +static void print_attr(struct html *, const char *, const char *); +static void *ml_alloc(char *, enum htmltype); + static void * ml_alloc(char *outopts, enum htmltype type) { struct html *h; - const char *toks[4]; + const char *toks[5]; char *v; toks[0] = "style"; toks[1] = "man"; toks[2] = "includes"; - toks[3] = NULL; + toks[3] = "fragment"; + toks[4] = NULL; h = mandoc_calloc(1, sizeof(struct html)); @@ -127,15 +136,18 @@ ml_alloc(char *outopts, enum htmltype type) while (outopts && *outopts) switch (getsubopt(&outopts, UNCONST(toks), &v)) { - case (0): + case 0: h->style = v; break; - case (1): + case 1: h->base_man = v; break; - case (2): + case 2: h->base_includes = v; break; + case 3: + h->oflags |= HTML_FRAGMENT; + break; default: break; } @@ -150,7 +162,6 @@ html_alloc(char *outopts) return(ml_alloc(outopts, HTML_HTML_4_01_STRICT)); } - void * xhtml_alloc(char *outopts) { @@ -158,7 +169,6 @@ xhtml_alloc(char *outopts) return(ml_alloc(outopts, HTML_XHTML_1_0_STRICT)); } - void html_free(void *p) { @@ -168,17 +178,16 @@ html_free(void *p) h = (struct html *)p; while ((tag = h->tags.head) != NULL) { - h->tags.head = tag->next; + h->tags.head = tag->next; free(tag); } - + if (h->symtab) mchars_free(h->symtab); free(h); } - void print_gen_head(struct html *h) { @@ -209,71 +218,27 @@ print_gen_head(struct html *h) } } -/* ARGSUSED */ static void -print_num(struct html *h, const char *p, size_t len) -{ - char c; - - if ('\0' != (c = mchars_num2char(p, len))) - putchar((int)c); -} - -static void -print_spec(struct html *h, const char *p, size_t len) -{ - int cp; - const char *rhs; - size_t sz; - - if ((cp = mchars_spec2cp(h->symtab, p, len)) > 0) { - printf("&#%d;", cp); - return; - } else if (-1 == cp && 1 == len) { - fwrite(p, 1, len, stdout); - return; - } else if (-1 == cp) - return; - - if (NULL != (rhs = mchars_spec2str(h->symtab, p, len, &sz))) - fwrite(rhs, 1, sz, stdout); -} - - -static void -print_res(struct html *h, const char *p, size_t len) -{ - int cp; - const char *rhs; - size_t sz; - - if ((cp = mchars_res2cp(h->symtab, p, len)) > 0) { - printf("&#%d;", cp); - return; - } else if (-1 == cp) - return; - - if (NULL != (rhs = mchars_res2str(h->symtab, p, len, &sz))) - fwrite(rhs, 1, sz, stdout); -} - - -static void print_metaf(struct html *h, enum mandoc_esc deco) { enum htmlfont font; switch (deco) { - case (ESCAPE_FONTPREV): + case ESCAPE_FONTPREV: font = h->metal; break; - case (ESCAPE_FONTITALIC): + case ESCAPE_FONTITALIC: font = HTMLFONT_ITALIC; break; - case (ESCAPE_FONTBOLD): + case ESCAPE_FONTBOLD: font = HTMLFONT_BOLD; break; - case (ESCAPE_FONTROMAN): + case ESCAPE_FONTBI: + font = HTMLFONT_BI; + break; + case ESCAPE_FONT: + /* FALLTHROUGH */ + case ESCAPE_FONTROMAN: font = HTMLFONT_NONE; break; default: @@ -289,17 +254,27 @@ print_metaf(struct html *h, enum mandoc_esc deco) h->metal = h->metac; h->metac = font; - if (HTMLFONT_NONE != font) - h->metaf = HTMLFONT_BOLD == font ? - print_otag(h, TAG_B, 0, NULL) : - print_otag(h, TAG_I, 0, NULL); + switch (font) { + case HTMLFONT_ITALIC: + h->metaf = print_otag(h, TAG_I, 0, NULL); + break; + case HTMLFONT_BOLD: + h->metaf = print_otag(h, TAG_B, 0, NULL); + break; + case HTMLFONT_BI: + h->metaf = print_otag(h, TAG_B, 0, NULL); + print_otag(h, TAG_I, 0, NULL); + break; + default: + break; + } } int html_strlen(const char *cp) { - int ssz, sz; - const char *seq, *p; + size_t rsz; + int skip, sz; /* * Account for escaped sequences within string length @@ -310,39 +285,93 @@ html_strlen(const char *cp) */ sz = 0; - while (NULL != (p = strchr(cp, '\\'))) { - sz += (int)(p - cp); - ++cp; - switch (mandoc_escape(&cp, &seq, &ssz)) { - case (ESCAPE_ERROR): + skip = 0; + while (1) { + rsz = strcspn(cp, "\\"); + if (rsz) { + cp += rsz; + if (skip) { + skip = 0; + rsz--; + } + sz += rsz; + } + if ('\0' == *cp) + break; + cp++; + switch (mandoc_escape(&cp, NULL, NULL)) { + case ESCAPE_ERROR: return(sz); - case (ESCAPE_PREDEF): - sz++; + case ESCAPE_UNICODE: + /* FALLTHROUGH */ + case ESCAPE_NUMBERED: + /* FALLTHROUGH */ + case ESCAPE_SPECIAL: + if (skip) + skip = 0; + else + sz++; break; - case (ESCAPE_SPECIAL): - sz++; + case ESCAPE_SKIPCHAR: + skip = 1; break; default: break; } } + return(sz); +} - assert(sz >= 0); - return(sz + strlen(cp)); +static int +print_escape(char c) +{ + + switch (c) { + case '<': + printf("<"); + break; + case '>': + printf(">"); + break; + case '&': + printf("&"); + break; + case '"': + printf("""); + break; + case ASCII_NBRSP: + putchar('-'); + break; + case ASCII_HYPH: + putchar('-'); + /* FALLTHROUGH */ + case ASCII_BREAK: + break; + default: + return(0); + } + return(1); } static int print_encode(struct html *h, const char *p, int norecurse) { size_t sz; - int len, nospace; + int c, len, nospace; const char *seq; enum mandoc_esc esc; - static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' }; + static const char rejs[9] = { '\\', '<', '>', '&', '"', + ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' }; nospace = 0; while ('\0' != *p) { + if (HTML_SKIPCHAR & h->flags && '\\' != *p) { + h->flags &= ~HTML_SKIPCHAR; + p++; + continue; + } + sz = strcspn(p, rejs); fwrite(p, 1, sz, stdout); @@ -351,49 +380,61 @@ print_encode(struct html *h, const char *p, int norecu if ('\0' == *p) break; - switch (*p++) { - case ('<'): - printf("<"); + if (print_escape(*p++)) continue; - case ('>'): - printf(">"); + + esc = mandoc_escape(&p, &seq, &len); + if (ESCAPE_ERROR == esc) + break; + + switch (esc) { + case ESCAPE_FONT: + /* FALLTHROUGH */ + case ESCAPE_FONTPREV: + /* FALLTHROUGH */ + case ESCAPE_FONTBOLD: + /* FALLTHROUGH */ + case ESCAPE_FONTITALIC: + /* FALLTHROUGH */ + case ESCAPE_FONTBI: + /* FALLTHROUGH */ + case ESCAPE_FONTROMAN: + if (0 == norecurse) + print_metaf(h, esc); continue; - case ('&'): - printf("&"); + case ESCAPE_SKIPCHAR: + h->flags |= HTML_SKIPCHAR; continue; - case (ASCII_HYPH): - putchar('-'); - continue; default: break; } - esc = mandoc_escape(&p, &seq, &len); - if (ESCAPE_ERROR == esc) - break; + if (h->flags & HTML_SKIPCHAR) { + h->flags &= ~HTML_SKIPCHAR; + continue; + } switch (esc) { - case (ESCAPE_NUMBERED): - print_num(h, seq, len); + case ESCAPE_UNICODE: + /* Skip past "u" header. */ + c = mchars_num2uc(seq + 1, len - 1); + if ('\0' != c) + printf("&#x%x;", c); break; - case (ESCAPE_PREDEF): - print_res(h, seq, len); + case ESCAPE_NUMBERED: + c = mchars_num2char(seq, len); + if ( ! ('\0' == c || print_escape(c))) + putchar(c); break; - case (ESCAPE_SPECIAL): - print_spec(h, seq, len); + case ESCAPE_SPECIAL: + c = mchars_spec2cp(h->symtab, seq, len); + if (c > 0) + printf("&#%d;", c); + else if (-1 == c && 1 == len && + !print_escape(*seq)) + putchar((int)*seq); break; - case (ESCAPE_FONTPREV): - /* FALLTHROUGH */ - case (ESCAPE_FONTBOLD): - /* FALLTHROUGH */ - case (ESCAPE_FONTITALIC): - /* FALLTHROUGH */ - case (ESCAPE_FONTROMAN): - if (norecurse) - break; - print_metaf(h, esc); - break; - case (ESCAPE_NOSPACE): + case ESCAPE_NOSPACE: if ('\0' == *p) nospace = 1; break; @@ -405,7 +446,6 @@ print_encode(struct html *h, const char *p, int norecu return(nospace); } - static void print_attr(struct html *h, const char *key, const char *val) { @@ -414,9 +454,8 @@ print_attr(struct html *h, const char *key, const char putchar('\"'); } - struct tag * -print_otag(struct html *h, enum htmltag tag, +print_otag(struct html *h, enum htmltag tag, int sz, const struct htmlpair *p) { int i; @@ -466,7 +505,7 @@ print_otag(struct html *h, enum htmltag tag, if (HTML_AUTOCLOSE & htmltags[tag].flags) switch (h->type) { - case (HTML_XHTML_1_0_STRICT): + case HTML_XHTML_1_0_STRICT: putchar('/'); break; default: @@ -483,59 +522,40 @@ print_otag(struct html *h, enum htmltag tag, return(t); } - static void print_ctag(struct html *h, enum htmltag tag) { - + printf("", htmltags[tag].name); if (HTML_CLRLINE & htmltags[tag].flags) { h->flags |= HTML_NOSPACE; putchar('\n'); - } + } } - void print_gen_decls(struct html *h) { - - print_xmltype(h); - print_doctype(h); -} - - -static void -print_xmltype(struct html *h) -{ - - if (HTML_XHTML_1_0_STRICT == h->type) - puts(""); -} - - -static void -print_doctype(struct html *h) -{ const char *doctype; const char *dtd; const char *name; switch (h->type) { - case (HTML_HTML_4_01_STRICT): + case HTML_HTML_4_01_STRICT: name = "HTML"; doctype = "-//W3C//DTD HTML 4.01//EN"; dtd = "http://www.w3.org/TR/html4/strict.dtd"; break; default: + puts(""); name = "html"; doctype = "-//W3C//DTD XHTML 1.0 Strict//EN"; dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"; break; } - printf("\n", - name, doctype, dtd); + printf("\n", + name, doctype, dtd); } void @@ -553,15 +573,27 @@ print_text(struct html *h, const char *word) } assert(NULL == h->metaf); - if (HTMLFONT_NONE != h->metac) - h->metaf = HTMLFONT_BOLD == h->metac ? - print_otag(h, TAG_B, 0, NULL) : - print_otag(h, TAG_I, 0, NULL); + switch (h->metac) { + case HTMLFONT_ITALIC: + h->metaf = print_otag(h, TAG_I, 0, NULL); + break; + case HTMLFONT_BOLD: + h->metaf = print_otag(h, TAG_B, 0, NULL); + break; + case HTMLFONT_BI: + h->metaf = print_otag(h, TAG_B, 0, NULL); + print_otag(h, TAG_I, 0, NULL); + break; + default: + break; + } assert(word); - if ( ! print_encode(h, word, 0)) + if ( ! print_encode(h, word, 0)) { if ( ! (h->flags & HTML_NONOSPACE)) h->flags &= ~HTML_NOSPACE; + } else + h->flags |= HTML_NOSPACE; if (h->metaf) { print_tagq(h, h->metaf); @@ -571,14 +603,13 @@ print_text(struct html *h, const char *word) h->flags &= ~HTML_IGNDELIM; } - void print_tagq(struct html *h, const struct tag *until) { struct tag *tag; while ((tag = h->tags.head) != NULL) { - /* + /* * Remember to close out and nullify the current * meta-font and table, if applicable. */ @@ -594,7 +625,6 @@ print_tagq(struct html *h, const struct tag *until) } } - void print_stagq(struct html *h, const struct tag *suntil) { @@ -603,7 +633,7 @@ print_stagq(struct html *h, const struct tag *suntil) while ((tag = h->tags.head) != NULL) { if (suntil && tag == suntil) return; - /* + /* * Remember to close out and nullify the current * meta-font and table, if applicable. */ @@ -617,7 +647,6 @@ print_stagq(struct html *h, const struct tag *suntil) } } - void bufinit(struct html *h) { @@ -626,62 +655,63 @@ bufinit(struct html *h) h->buflen = 0; } - void bufcat_style(struct html *h, const char *key, const char *val) { bufcat(h, key); - bufncat(h, ":", 1); + bufcat(h, ":"); bufcat(h, val); - bufncat(h, ";", 1); + bufcat(h, ";"); } - void bufcat(struct html *h, const char *p) { - bufncat(h, p, strlen(p)); + /* + * XXX This is broken and not easy to fix. + * When using the -Oincludes option, buffmt_includes() + * may pass in strings overrunning BUFSIZ, causing a crash. + */ + + h->buflen = strlcat(h->buf, p, BUFSIZ); + assert(h->buflen < BUFSIZ); } - void -buffmt(struct html *h, const char *fmt, ...) +bufcat_fmt(struct html *h, const char *fmt, ...) { va_list ap; va_start(ap, fmt); - (void)vsnprintf(h->buf + (int)h->buflen, - BUFSIZ - h->buflen - 1, fmt, ap); + (void)vsnprintf(h->buf + (int)h->buflen, + BUFSIZ - h->buflen - 1, fmt, ap); va_end(ap); h->buflen = strlen(h->buf); } - -void +static void bufncat(struct html *h, const char *p, size_t sz) { - if (h->buflen + sz > BUFSIZ - 1) - sz = BUFSIZ - 1 - h->buflen; - - (void)strncat(h->buf, p, sz); + assert(h->buflen + sz + 1 < BUFSIZ); + strncat(h->buf, p, sz); h->buflen += sz; } - void buffmt_includes(struct html *h, const char *name) { const char *p, *pp; pp = h->base_includes; - + + bufinit(h); while (NULL != (p = strchr(pp, '%'))) { bufncat(h, pp, (size_t)(p - pp)); switch (*(p + 1)) { - case('I'): + case'I': bufcat(h, name); break; default: @@ -694,24 +724,22 @@ buffmt_includes(struct html *h, const char *name) bufcat(h, pp); } - void -buffmt_man(struct html *h, - const char *name, const char *sec) +buffmt_man(struct html *h, const char *name, const char *sec) { const char *p, *pp; pp = h->base_man; - - /* LINTED */ + + bufinit(h); while (NULL != (p = strchr(pp, '%'))) { bufncat(h, pp, (size_t)(p - pp)); switch (*(p + 1)) { - case('S'): + case 'S': bufcat(h, sec ? sec : "1"); break; - case('N'): - buffmt(h, name); + case 'N': + bufcat_fmt(h, "%s", name); break; default: bufncat(h, p, 2); @@ -723,85 +751,26 @@ buffmt_man(struct html *h, bufcat(h, pp); } - void bufcat_su(struct html *h, const char *p, const struct roffsu *su) { double v; - const char *u; v = su->scale; + if (SCALE_MM == su->unit && 0.0 == (v /= 100.0)) + v = 1.0; + else if (SCALE_BU == su->unit) + v /= 24.0; - switch (su->unit) { - case (SCALE_CM): - u = "cm"; - break; - case (SCALE_IN): - u = "in"; - break; - case (SCALE_PC): - u = "pc"; - break; - case (SCALE_PT): - u = "pt"; - break; - case (SCALE_EM): - u = "em"; - break; - case (SCALE_MM): - if (0 == (v /= 100)) - v = 1; - u = "em"; - break; - case (SCALE_EN): - u = "ex"; - break; - case (SCALE_BU): - u = "ex"; - break; - case (SCALE_VS): - u = "em"; - break; - default: - u = "ex"; - break; - } - - /* - * XXX: the CSS spec isn't clear as to which types accept - * integer or real numbers, so we just make them all decimals. - */ - buffmt(h, "%s: %.2f%s;", p, v, u); + bufcat_fmt(h, "%s: %.2f%s;", p, v, roffscales[su->unit]); } - void -html_idcat(char *dst, const char *src, int sz) +bufcat_id(struct html *h, const char *src) { - int ssz; - assert(sz > 2); - /* Cf. . */ - /* We can't start with a number (bah). */ - - if ('#' == *dst) { - dst++; - sz--; - } - if ('\0' == *dst) { - *dst++ = 'x'; - *dst = '\0'; - sz--; - } - - for ( ; *dst != '\0' && sz; dst++, sz--) - /* Jump to end. */ ; - - for ( ; *src != '\0' && sz > 1; src++) { - ssz = snprintf(dst, (size_t)sz, "%.2x", *src); - sz -= ssz; - dst += ssz; - } + while ('\0' != *src) + bufcat_fmt(h, "%.2x", *src++); }