]>
git.cameronkatri.com Git - mandoc.git/blob - html.c
1 /* $Id: html.c,v 1.137 2011/04/30 22:24:31 kristaps Exp $ */
3 * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
4 * Copyright (c) 2011 Ingo Schwarze <schwarze@openbsd.org>
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.
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.
22 #include <sys/types.h>
41 #define HTML_CLRLINE (1 << 0)
42 #define HTML_NOSTACK (1 << 1)
43 #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
46 static const struct htmldata htmltags
[TAG_MAX
] = {
47 {"html", HTML_CLRLINE
}, /* TAG_HTML */
48 {"head", HTML_CLRLINE
}, /* TAG_HEAD */
49 {"body", HTML_CLRLINE
}, /* TAG_BODY */
50 {"meta", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_META */
51 {"title", HTML_CLRLINE
}, /* TAG_TITLE */
52 {"div", HTML_CLRLINE
}, /* TAG_DIV */
53 {"h1", 0}, /* TAG_H1 */
54 {"h2", 0}, /* TAG_H2 */
55 {"span", 0}, /* TAG_SPAN */
56 {"link", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_LINK */
57 {"br", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_BR */
59 {"table", HTML_CLRLINE
}, /* TAG_TABLE */
60 {"tbody", HTML_CLRLINE
}, /* TAG_TBODY */
61 {"col", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_COL */
62 {"tr", HTML_CLRLINE
}, /* TAG_TR */
63 {"td", HTML_CLRLINE
}, /* TAG_TD */
64 {"li", HTML_CLRLINE
}, /* TAG_LI */
65 {"ul", HTML_CLRLINE
}, /* TAG_UL */
66 {"ol", HTML_CLRLINE
}, /* TAG_OL */
67 {"dl", HTML_CLRLINE
}, /* TAG_DL */
68 {"dt", HTML_CLRLINE
}, /* TAG_DT */
69 {"dd", HTML_CLRLINE
}, /* TAG_DD */
70 {"blockquote", HTML_CLRLINE
}, /* TAG_BLOCKQUOTE */
71 {"p", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_P */
72 {"pre", HTML_CLRLINE
}, /* TAG_PRE */
73 {"b", 0 }, /* TAG_B */
74 {"i", 0 }, /* TAG_I */
75 {"code", 0 }, /* TAG_CODE */
76 {"small", 0 }, /* TAG_SMALL */
79 static const char *const htmlattrs
[ATTR_MAX
] = {
80 "http-equiv", /* ATTR_HTTPEQUIV */
81 "content", /* ATTR_CONTENT */
82 "name", /* ATTR_NAME */
84 "href", /* ATTR_HREF */
85 "type", /* ATTR_TYPE */
86 "media", /* ATTR_MEDIA */
87 "class", /* ATTR_CLASS */
88 "style", /* ATTR_STYLE */
89 "width", /* ATTR_WIDTH */
91 "summary", /* ATTR_SUMMARY */
92 "align", /* ATTR_ALIGN */
93 "colspan", /* ATTR_COLSPAN */
96 static void print_num(struct html
*, const char *, size_t);
97 static void print_spec(struct html
*, const char *, size_t);
98 static void print_res(struct html
*, const char *, size_t);
99 static void print_ctag(struct html
*, enum htmltag
);
100 static void print_doctype(struct html
*);
101 static void print_xmltype(struct html
*);
102 static int print_encode(struct html
*, const char *, int);
103 static void print_metaf(struct html
*, enum mandoc_esc
);
104 static void print_attr(struct html
*,
105 const char *, const char *);
106 static void *ml_alloc(char *, enum htmltype
);
110 ml_alloc(char *outopts
, enum htmltype type
)
118 toks
[2] = "includes";
121 h
= mandoc_calloc(1, sizeof(struct html
));
125 h
->symtab
= mchars_alloc();
127 while (outopts
&& *outopts
)
128 switch (getsubopt(&outopts
, UNCONST(toks
), &v
)) {
136 h
->base_includes
= v
;
146 html_alloc(char *outopts
)
149 return(ml_alloc(outopts
, HTML_HTML_4_01_STRICT
));
154 xhtml_alloc(char *outopts
)
157 return(ml_alloc(outopts
, HTML_XHTML_1_0_STRICT
));
167 h
= (struct html
*)p
;
169 while ((tag
= h
->tags
.head
) != NULL
) {
170 h
->tags
.head
= tag
->next
;
175 mchars_free(h
->symtab
);
182 print_gen_head(struct html
*h
)
184 struct htmlpair tag
[4];
186 tag
[0].key
= ATTR_HTTPEQUIV
;
187 tag
[0].val
= "Content-Type";
188 tag
[1].key
= ATTR_CONTENT
;
189 tag
[1].val
= "text/html; charset=utf-8";
190 print_otag(h
, TAG_META
, 2, tag
);
192 tag
[0].key
= ATTR_NAME
;
193 tag
[0].val
= "resource-type";
194 tag
[1].key
= ATTR_CONTENT
;
195 tag
[1].val
= "document";
196 print_otag(h
, TAG_META
, 2, tag
);
199 tag
[0].key
= ATTR_REL
;
200 tag
[0].val
= "stylesheet";
201 tag
[1].key
= ATTR_HREF
;
202 tag
[1].val
= h
->style
;
203 tag
[2].key
= ATTR_TYPE
;
204 tag
[2].val
= "text/css";
205 tag
[3].key
= ATTR_MEDIA
;
207 print_otag(h
, TAG_LINK
, 4, tag
);
213 print_num(struct html
*h
, const char *p
, size_t len
)
217 if ('\0' != (c
= mchars_num2char(p
, len
)))
222 print_spec(struct html
*h
, const char *p
, size_t len
)
228 if ((cp
= mchars_spec2cp(h
->symtab
, p
, len
)) > 0) {
231 } else if (-1 == cp
&& 1 == len
) {
232 fwrite(p
, 1, len
, stdout
);
237 if (NULL
!= (rhs
= mchars_spec2str(h
->symtab
, p
, len
, &sz
)))
238 fwrite(rhs
, 1, sz
, stdout
);
243 print_res(struct html
*h
, const char *p
, size_t len
)
249 if ((cp
= mchars_res2cp(h
->symtab
, p
, len
)) > 0) {
255 if (NULL
!= (rhs
= mchars_res2str(h
->symtab
, p
, len
, &sz
)))
256 fwrite(rhs
, 1, sz
, stdout
);
261 print_metaf(struct html
*h
, enum mandoc_esc deco
)
266 case (ESCAPE_FONTPREV
):
269 case (ESCAPE_FONTITALIC
):
270 font
= HTMLFONT_ITALIC
;
272 case (ESCAPE_FONTBOLD
):
273 font
= HTMLFONT_BOLD
;
275 case (ESCAPE_FONTROMAN
):
276 font
= HTMLFONT_NONE
;
284 print_tagq(h
, h
->metaf
);
291 if (HTMLFONT_NONE
!= font
)
292 h
->metaf
= HTMLFONT_BOLD
== font
?
293 print_otag(h
, TAG_B
, 0, NULL
) :
294 print_otag(h
, TAG_I
, 0, NULL
);
299 print_encode(struct html
*h
, const char *p
, int norecurse
)
305 static const char rejs
[6] = { '\\', '<', '>', '&', ASCII_HYPH
, '\0' };
310 sz
= strcspn(p
, rejs
);
312 fwrite(p
, 1, sz
, stdout
);
335 esc
= mandoc_escape(&p
, &seq
, &len
);
336 if (ESCAPE_ERROR
== esc
)
340 case (ESCAPE_NUMBERED
):
341 print_num(h
, seq
, len
);
343 case (ESCAPE_PREDEF
):
344 print_res(h
, seq
, len
);
346 case (ESCAPE_SPECIAL
):
347 print_spec(h
, seq
, len
);
349 case (ESCAPE_FONTPREV
):
351 case (ESCAPE_FONTBOLD
):
353 case (ESCAPE_FONTITALIC
):
355 case (ESCAPE_FONTROMAN
):
360 case (ESCAPE_NOSPACE
):
374 print_attr(struct html
*h
, const char *key
, const char *val
)
376 printf(" %s=\"", key
);
377 (void)print_encode(h
, val
, 1);
383 print_otag(struct html
*h
, enum htmltag tag
,
384 int sz
, const struct htmlpair
*p
)
389 /* Push this tags onto the stack of open scopes. */
391 if ( ! (HTML_NOSTACK
& htmltags
[tag
].flags
)) {
392 t
= mandoc_malloc(sizeof(struct tag
));
394 t
->next
= h
->tags
.head
;
399 if ( ! (HTML_NOSPACE
& h
->flags
))
400 if ( ! (HTML_CLRLINE
& htmltags
[tag
].flags
)) {
402 if ( ! (HTML_KEEP
& h
->flags
)) {
403 if (HTML_PREKEEP
& h
->flags
)
404 h
->flags
|= HTML_KEEP
;
410 if ( ! (h
->flags
& HTML_NONOSPACE
))
411 h
->flags
&= ~HTML_NOSPACE
;
413 h
->flags
|= HTML_NOSPACE
;
415 /* Print out the tag name and attributes. */
417 printf("<%s", htmltags
[tag
].name
);
418 for (i
= 0; i
< sz
; i
++)
419 print_attr(h
, htmlattrs
[p
[i
].key
], p
[i
].val
);
421 /* Add non-overridable attributes. */
423 if (TAG_HTML
== tag
&& HTML_XHTML_1_0_STRICT
== h
->type
) {
424 print_attr(h
, "xmlns", "http://www.w3.org/1999/xhtml");
425 print_attr(h
, "xml:lang", "en");
426 print_attr(h
, "lang", "en");
429 /* Accommodate for XML "well-formed" singleton escaping. */
431 if (HTML_AUTOCLOSE
& htmltags
[tag
].flags
)
433 case (HTML_XHTML_1_0_STRICT
):
442 h
->flags
|= HTML_NOSPACE
;
444 if ((HTML_AUTOCLOSE
| HTML_CLRLINE
) & htmltags
[tag
].flags
)
452 print_ctag(struct html
*h
, enum htmltag tag
)
455 printf("</%s>", htmltags
[tag
].name
);
456 if (HTML_CLRLINE
& htmltags
[tag
].flags
) {
457 h
->flags
|= HTML_NOSPACE
;
464 print_gen_decls(struct html
*h
)
473 print_xmltype(struct html
*h
)
476 if (HTML_XHTML_1_0_STRICT
== h
->type
)
477 puts("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
482 print_doctype(struct html
*h
)
489 case (HTML_HTML_4_01_STRICT
):
491 doctype
= "-//W3C//DTD HTML 4.01//EN";
492 dtd
= "http://www.w3.org/TR/html4/strict.dtd";
496 doctype
= "-//W3C//DTD XHTML 1.0 Strict//EN";
497 dtd
= "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
501 printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
506 print_text(struct html
*h
, const char *word
)
509 if ( ! (HTML_NOSPACE
& h
->flags
)) {
511 if ( ! (HTML_KEEP
& h
->flags
)) {
512 if (HTML_PREKEEP
& h
->flags
)
513 h
->flags
|= HTML_KEEP
;
519 assert(NULL
== h
->metaf
);
520 if (HTMLFONT_NONE
!= h
->metac
)
521 h
->metaf
= HTMLFONT_BOLD
== h
->metac
?
522 print_otag(h
, TAG_B
, 0, NULL
) :
523 print_otag(h
, TAG_I
, 0, NULL
);
526 if ( ! print_encode(h
, word
, 0))
527 if ( ! (h
->flags
& HTML_NONOSPACE
))
528 h
->flags
&= ~HTML_NOSPACE
;
531 print_tagq(h
, h
->metaf
);
535 h
->flags
&= ~HTML_IGNDELIM
;
540 print_tagq(struct html
*h
, const struct tag
*until
)
544 while ((tag
= h
->tags
.head
) != NULL
) {
546 * Remember to close out and nullify the current
547 * meta-font and table, if applicable.
553 print_ctag(h
, tag
->tag
);
554 h
->tags
.head
= tag
->next
;
556 if (until
&& tag
== until
)
563 print_stagq(struct html
*h
, const struct tag
*suntil
)
567 while ((tag
= h
->tags
.head
) != NULL
) {
568 if (suntil
&& tag
== suntil
)
571 * Remember to close out and nullify the current
572 * meta-font and table, if applicable.
578 print_ctag(h
, tag
->tag
);
579 h
->tags
.head
= tag
->next
;
586 bufinit(struct html
*h
)
595 bufcat_style(struct html
*h
, const char *key
, const char *val
)
606 bufcat(struct html
*h
, const char *p
)
609 bufncat(h
, p
, strlen(p
));
614 buffmt(struct html
*h
, const char *fmt
, ...)
619 (void)vsnprintf(h
->buf
+ (int)h
->buflen
,
620 BUFSIZ
- h
->buflen
- 1, fmt
, ap
);
622 h
->buflen
= strlen(h
->buf
);
627 bufncat(struct html
*h
, const char *p
, size_t sz
)
630 if (h
->buflen
+ sz
> BUFSIZ
- 1)
631 sz
= BUFSIZ
- 1 - h
->buflen
;
633 (void)strncat(h
->buf
, p
, sz
);
639 buffmt_includes(struct html
*h
, const char *name
)
643 pp
= h
->base_includes
;
645 while (NULL
!= (p
= strchr(pp
, '%'))) {
646 bufncat(h
, pp
, (size_t)(p
- pp
));
663 buffmt_man(struct html
*h
,
664 const char *name
, const char *sec
)
671 while (NULL
!= (p
= strchr(pp
, '%'))) {
672 bufncat(h
, pp
, (size_t)(p
- pp
));
675 bufcat(h
, sec
? sec
: "1");
692 bufcat_su(struct html
*h
, const char *p
, const struct roffsu
*su
)
735 * XXX: the CSS spec isn't clear as to which types accept
736 * integer or real numbers, so we just make them all decimals.
738 buffmt(h
, "%s: %.2f%s;", p
, v
, u
);
743 html_idcat(char *dst
, const char *src
, int sz
)
749 /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
751 /* We can't start with a number (bah). */
763 for ( ; *dst
!= '\0' && sz
; dst
++, sz
--)
766 for ( ; *src
!= '\0' && sz
> 1; src
++) {
767 ssz
= snprintf(dst
, (size_t)sz
, "%.2x", *src
);