1 /* $Id: html.c,v 1.96 2010/02/17 19:48:33 kristaps Exp $ */
3 * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
5 * Permission to use, copy, modify, and distribute this software for any
6 * purpose with or without fee is hereby granted, provided that the above
7 * copyright notice and this permission notice appear in all copies.
9 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
10 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
11 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
12 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
13 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
14 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
15 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
17 #include <sys/types.h>
33 #define UNCONST(a) ((void *)(uintptr_t)(const void *)(a))
38 #define HTML_CLRLINE (1 << 0)
39 #define HTML_NOSTACK (1 << 1)
40 #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
43 static const struct htmldata htmltags
[TAG_MAX
] = {
44 {"html", HTML_CLRLINE
}, /* TAG_HTML */
45 {"head", HTML_CLRLINE
}, /* TAG_HEAD */
46 {"body", HTML_CLRLINE
}, /* TAG_BODY */
47 {"meta", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_META */
48 {"title", HTML_CLRLINE
}, /* TAG_TITLE */
49 {"div", HTML_CLRLINE
}, /* TAG_DIV */
50 {"h1", 0}, /* TAG_H1 */
51 {"h2", 0}, /* TAG_H2 */
52 {"span", 0}, /* TAG_SPAN */
53 {"link", HTML_CLRLINE
| HTML_NOSTACK
}, /* TAG_LINK */
54 {"br", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_BR */
56 {"table", HTML_CLRLINE
}, /* TAG_TABLE */
57 {"col", HTML_CLRLINE
| HTML_NOSTACK
| HTML_AUTOCLOSE
}, /* TAG_COL */
58 {"tr", HTML_CLRLINE
}, /* TAG_TR */
59 {"td", HTML_CLRLINE
}, /* TAG_TD */
60 {"li", HTML_CLRLINE
}, /* TAG_LI */
61 {"ul", HTML_CLRLINE
}, /* TAG_UL */
62 {"ol", HTML_CLRLINE
}, /* TAG_OL */
65 static const char *const htmlfonts
[HTMLFONT_MAX
] = {
71 static const char *const htmlattrs
[ATTR_MAX
] = {
88 static void print_spec(struct html
*, const char *, size_t);
89 static void print_res(struct html
*, const char *, size_t);
90 static void print_ctag(struct html
*, enum htmltag
);
91 static void print_doctype(struct html
*);
92 static void print_xmltype(struct html
*);
93 static int print_encode(struct html
*, const char *, int);
94 static void print_metaf(struct html
*, enum roffdeco
);
95 static void print_attr(struct html
*,
96 const char *, const char *);
97 static void *ml_alloc(char *, enum htmltype
);
101 ml_alloc(char *outopts
, enum htmltype type
)
109 toks
[2] = "includes";
112 h
= calloc(1, sizeof(struct html
));
121 h
->symtab
= chars_init(CHARS_HTML
);
123 while (outopts
&& *outopts
)
124 switch (getsubopt(&outopts
, UNCONST(toks
), &v
)) {
132 h
->base_includes
= v
;
142 html_alloc(char *outopts
)
145 return(ml_alloc(outopts
, HTML_HTML_4_01_STRICT
));
150 xhtml_alloc(char *outopts
)
153 return(ml_alloc(outopts
, HTML_XHTML_1_0_STRICT
));
164 h
= (struct html
*)p
;
166 while ((ord
= h
->ords
.head
) != NULL
) {
167 h
->ords
.head
= ord
->next
;
171 while ((tag
= h
->tags
.head
) != NULL
) {
172 h
->tags
.head
= tag
->next
;
177 chars_free(h
->symtab
);
184 print_gen_head(struct html
*h
)
186 struct htmlpair tag
[4];
188 tag
[0].key
= ATTR_HTTPEQUIV
;
189 tag
[0].val
= "Content-Type";
190 tag
[1].key
= ATTR_CONTENT
;
191 tag
[1].val
= "text/html; charset=utf-8";
192 print_otag(h
, TAG_META
, 2, tag
);
194 tag
[0].key
= ATTR_NAME
;
195 tag
[0].val
= "resource-type";
196 tag
[1].key
= ATTR_CONTENT
;
197 tag
[1].val
= "document";
198 print_otag(h
, TAG_META
, 2, tag
);
201 tag
[0].key
= ATTR_REL
;
202 tag
[0].val
= "stylesheet";
203 tag
[1].key
= ATTR_HREF
;
204 tag
[1].val
= h
->style
;
205 tag
[2].key
= ATTR_TYPE
;
206 tag
[2].val
= "text/css";
207 tag
[3].key
= ATTR_MEDIA
;
209 print_otag(h
, TAG_LINK
, 4, tag
);
215 print_spec(struct html
*h
, const char *p
, size_t len
)
220 rhs
= chars_a2ascii(h
->symtab
, p
, len
, &sz
);
224 fwrite(rhs
, 1, sz
, stdout
);
229 print_res(struct html
*h
, const char *p
, size_t len
)
234 rhs
= chars_a2res(h
->symtab
, p
, len
, &sz
);
238 fwrite(rhs
, 1, sz
, stdout
);
243 print_ofont(struct html
*h
, enum htmlfont font
)
250 /* FIXME: DECO_ROMAN should just close out preexisting. */
252 if (h
->metaf
&& h
->tags
.head
== h
->metaf
)
253 print_tagq(h
, h
->metaf
);
255 PAIR_CLASS_INIT(&tag
, htmlfonts
[font
]);
256 h
->metaf
= print_otag(h
, TAG_SPAN
, 1, &tag
);
262 print_metaf(struct html
*h
, enum roffdeco deco
)
267 case (DECO_PREVIOUS
):
271 font
= HTMLFONT_ITALIC
;
274 font
= HTMLFONT_BOLD
;
277 font
= HTMLFONT_NONE
;
284 (void)print_ofont(h
, font
);
289 print_encode(struct html
*h
, const char *p
, int norecurse
)
299 sz
= strcspn(p
, "\\<>&");
301 fwrite(p
, 1, sz
, stdout
);
308 } else if ('>' == *p
) {
311 } else if ('&' == *p
) {
314 } else if ('\0' == *p
)
318 len
= a2roffdeco(&deco
, &seq
, &sz
);
321 case (DECO_RESERVED
):
322 print_res(h
, seq
, sz
);
325 print_spec(h
, seq
, sz
);
327 case (DECO_PREVIOUS
):
336 print_metaf(h
, deco
);
344 if (DECO_NOSPACE
== deco
&& '\0' == *(p
+ 1))
353 print_attr(struct html
*h
, const char *key
, const char *val
)
355 printf(" %s=\"", key
);
356 (void)print_encode(h
, val
, 1);
362 print_otag(struct html
*h
, enum htmltag tag
,
363 int sz
, const struct htmlpair
*p
)
368 /* Push this tags onto the stack of open scopes. */
370 if ( ! (HTML_NOSTACK
& htmltags
[tag
].flags
)) {
371 t
= malloc(sizeof(struct tag
));
377 t
->next
= h
->tags
.head
;
382 if ( ! (HTML_NOSPACE
& h
->flags
))
383 if ( ! (HTML_CLRLINE
& htmltags
[tag
].flags
))
386 /* Print out the tag name and attributes. */
388 printf("<%s", htmltags
[tag
].name
);
389 for (i
= 0; i
< sz
; i
++)
390 print_attr(h
, htmlattrs
[p
[i
].key
], p
[i
].val
);
392 /* Add non-overridable attributes. */
394 if (TAG_HTML
== tag
&& HTML_XHTML_1_0_STRICT
== h
->type
) {
395 print_attr(h
, "xmlns", "http://www.w3.org/1999/xhtml");
396 print_attr(h
, "xml:lang", "en");
397 print_attr(h
, "lang", "en");
400 /* Accomodate for XML "well-formed" singleton escaping. */
402 if (HTML_AUTOCLOSE
& htmltags
[tag
].flags
)
404 case (HTML_XHTML_1_0_STRICT
):
413 h
->flags
|= HTML_NOSPACE
;
419 print_ctag(struct html
*h
, enum htmltag tag
)
422 printf("</%s>", htmltags
[tag
].name
);
423 if (HTML_CLRLINE
& htmltags
[tag
].flags
) {
424 h
->flags
|= HTML_NOSPACE
;
431 print_gen_decls(struct html
*h
)
440 print_xmltype(struct html
*h
)
445 case (HTML_XHTML_1_0_STRICT
):
446 decl
= "<?xml version=\"1.0\" encoding=\"UTF-8\"?>";
456 printf("%s\n", decl
);
461 print_doctype(struct html
*h
)
468 case (HTML_HTML_4_01_STRICT
):
470 doctype
= "-//W3C//DTD HTML 4.01//EN";
471 dtd
= "http://www.w3.org/TR/html4/strict.dtd";
475 doctype
= "-//W3C//DTD XHTML 1.0 Strict//EN";
476 dtd
= "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
480 printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
486 print_text(struct html
*h
, const char *p
)
489 if (*p
&& 0 == *(p
+ 1))
508 if ( ! (HTML_IGNDELIM
& h
->flags
))
509 h
->flags
|= HTML_NOSPACE
;
515 if ( ! (h
->flags
& HTML_NOSPACE
))
519 if ( ! print_encode(h
, p
, 0))
520 h
->flags
&= ~HTML_NOSPACE
;
522 if (*p
&& 0 == *(p
+ 1))
529 h
->flags
|= HTML_NOSPACE
;
538 print_tagq(struct html
*h
, const struct tag
*until
)
542 while ((tag
= h
->tags
.head
) != NULL
) {
545 print_ctag(h
, tag
->tag
);
546 h
->tags
.head
= tag
->next
;
548 if (until
&& tag
== until
)
555 print_stagq(struct html
*h
, const struct tag
*suntil
)
559 while ((tag
= h
->tags
.head
) != NULL
) {
560 if (suntil
&& tag
== suntil
)
564 print_ctag(h
, tag
->tag
);
565 h
->tags
.head
= tag
->next
;
572 bufinit(struct html
*h
)
581 bufcat_style(struct html
*h
, const char *key
, const char *val
)
592 bufcat(struct html
*h
, const char *p
)
595 bufncat(h
, p
, strlen(p
));
600 buffmt(struct html
*h
, const char *fmt
, ...)
605 (void)vsnprintf(h
->buf
+ (int)h
->buflen
,
606 BUFSIZ
- h
->buflen
- 1, fmt
, ap
);
608 h
->buflen
= strlen(h
->buf
);
613 bufncat(struct html
*h
, const char *p
, size_t sz
)
616 if (h
->buflen
+ sz
> BUFSIZ
- 1)
617 sz
= BUFSIZ
- 1 - h
->buflen
;
619 (void)strncat(h
->buf
, p
, sz
);
625 buffmt_includes(struct html
*h
, const char *name
)
629 pp
= h
->base_includes
;
631 while (NULL
!= (p
= strchr(pp
, '%'))) {
632 bufncat(h
, pp
, (size_t)(p
- pp
));
649 buffmt_man(struct html
*h
,
650 const char *name
, const char *sec
)
657 while (NULL
!= (p
= strchr(pp
, '%'))) {
658 bufncat(h
, pp
, (size_t)(p
- pp
));
661 bufcat(h
, sec
? sec
: "1");
678 bufcat_su(struct html
*h
, const char *p
, const struct roffsu
*su
)
721 buffmt(h
, "%s: %f%s;", p
, v
, u
);
724 buffmt(h
, "%s: %d%s;", p
, (int)v
, u
);
729 html_idcat(char *dst
, const char *src
, int sz
)
735 /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
737 for ( ; *dst
!= '\0' && sz
; dst
++, sz
--)
742 /* We can't start with a number (bah). */
748 for ( ; *src
!= '\0' && sz
> 1; src
++) {
749 ssz
= snprintf(dst
, (size_t)sz
, "%.2x", *src
);