]> git.cameronkatri.com Git - mandoc.git/blob - html.c
Remove TODO from prior commit of lifting warnings from `Sh', `Ss', `SH',
[mandoc.git] / html.c
1 /* $Id: html.c,v 1.132 2011/04/09 15:29:40 kristaps Exp $ */
2 /*
3 * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
4 * Copyright (c) 2011 Ingo Schwarze <schwarze@openbsd.org>
5 *
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.
9 *
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.
17 */
18 #ifdef HAVE_CONFIG_H
19 #include "config.h"
20 #endif
21
22 #include <sys/types.h>
23
24 #include <assert.h>
25 #include <ctype.h>
26 #include <stdarg.h>
27 #include <stdio.h>
28 #include <stdint.h>
29 #include <stdlib.h>
30 #include <string.h>
31 #include <unistd.h>
32
33 #include "mandoc.h"
34 #include "out.h"
35 #include "html.h"
36 #include "main.h"
37
38 struct htmldata {
39 const char *name;
40 int flags;
41 #define HTML_CLRLINE (1 << 0)
42 #define HTML_NOSTACK (1 << 1)
43 #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
44 };
45
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 */
58 {"a", 0}, /* TAG_A */
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 */
77 };
78
79 static const char *const htmlattrs[ATTR_MAX] = {
80 "http-equiv", /* ATTR_HTTPEQUIV */
81 "content", /* ATTR_CONTENT */
82 "name", /* ATTR_NAME */
83 "rel", /* ATTR_REL */
84 "href", /* ATTR_HREF */
85 "type", /* ATTR_TYPE */
86 "media", /* ATTR_MEDIA */
87 "class", /* ATTR_CLASS */
88 "style", /* ATTR_STYLE */
89 "width", /* ATTR_WIDTH */
90 "id", /* ATTR_ID */
91 "summary", /* ATTR_SUMMARY */
92 "align", /* ATTR_ALIGN */
93 "colspan", /* ATTR_COLSPAN */
94 };
95
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);
107
108
109 static void *
110 ml_alloc(char *outopts, enum htmltype type)
111 {
112 struct html *h;
113 const char *toks[4];
114 char *v;
115
116 toks[0] = "style";
117 toks[1] = "man";
118 toks[2] = "includes";
119 toks[3] = NULL;
120
121 h = mandoc_calloc(1, sizeof(struct html));
122
123 h->type = type;
124 h->tags.head = NULL;
125 h->symtab = chars_init(CHARS_HTML);
126
127 while (outopts && *outopts)
128 switch (getsubopt(&outopts, UNCONST(toks), &v)) {
129 case (0):
130 h->style = v;
131 break;
132 case (1):
133 h->base_man = v;
134 break;
135 case (2):
136 h->base_includes = v;
137 break;
138 default:
139 break;
140 }
141
142 return(h);
143 }
144
145 void *
146 html_alloc(char *outopts)
147 {
148
149 return(ml_alloc(outopts, HTML_HTML_4_01_STRICT));
150 }
151
152
153 void *
154 xhtml_alloc(char *outopts)
155 {
156
157 return(ml_alloc(outopts, HTML_XHTML_1_0_STRICT));
158 }
159
160
161 void
162 html_free(void *p)
163 {
164 struct tag *tag;
165 struct html *h;
166
167 h = (struct html *)p;
168
169 while ((tag = h->tags.head) != NULL) {
170 h->tags.head = tag->next;
171 free(tag);
172 }
173
174 if (h->symtab)
175 chars_free(h->symtab);
176
177 free(h);
178 }
179
180
181 void
182 print_gen_head(struct html *h)
183 {
184 struct htmlpair tag[4];
185
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);
191
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);
197
198 if (h->style) {
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;
206 tag[3].val = "all";
207 print_otag(h, TAG_LINK, 4, tag);
208 }
209 }
210
211 /* ARGSUSED */
212 static void
213 print_num(struct html *h, const char *p, size_t len)
214 {
215 const char *rhs;
216
217 rhs = chars_num2char(p, len);
218 if (rhs)
219 putchar((int)*rhs);
220 }
221
222 static void
223 print_spec(struct html *h, const char *p, size_t len)
224 {
225 int cp;
226 const char *rhs;
227 size_t sz;
228
229 if ((cp = chars_spec2cp(h->symtab, p, len)) > 0) {
230 printf("&#%d;", cp);
231 return;
232 } else if (-1 == cp && 1 == len) {
233 fwrite(p, 1, len, stdout);
234 return;
235 } else if (-1 == cp)
236 return;
237
238 if (NULL != (rhs = chars_spec2str(h->symtab, p, len, &sz)))
239 fwrite(rhs, 1, sz, stdout);
240 }
241
242
243 static void
244 print_res(struct html *h, const char *p, size_t len)
245 {
246 int cp;
247 const char *rhs;
248 size_t sz;
249
250 if ((cp = chars_res2cp(h->symtab, p, len)) > 0) {
251 printf("&#%d;", cp);
252 return;
253 } else if (-1 == cp)
254 return;
255
256 if (NULL != (rhs = chars_res2str(h->symtab, p, len, &sz)))
257 fwrite(rhs, 1, sz, stdout);
258 }
259
260
261 static void
262 print_metaf(struct html *h, enum mandoc_esc deco)
263 {
264 enum htmlfont font;
265
266 switch (deco) {
267 case (ESCAPE_FONTPREV):
268 font = h->metal;
269 break;
270 case (ESCAPE_FONTITALIC):
271 font = HTMLFONT_ITALIC;
272 break;
273 case (ESCAPE_FONTBOLD):
274 font = HTMLFONT_BOLD;
275 break;
276 case (ESCAPE_FONTROMAN):
277 font = HTMLFONT_NONE;
278 break;
279 default:
280 abort();
281 /* NOTREACHED */
282 }
283
284 if (h->metaf) {
285 print_tagq(h, h->metaf);
286 h->metaf = NULL;
287 }
288
289 h->metal = h->metac;
290 h->metac = font;
291
292 if (HTMLFONT_NONE != font)
293 h->metaf = HTMLFONT_BOLD == font ?
294 print_otag(h, TAG_B, 0, NULL) :
295 print_otag(h, TAG_I, 0, NULL);
296 }
297
298
299 static int
300 print_encode(struct html *h, const char *p, int norecurse)
301 {
302 size_t sz;
303 int len, nospace;
304 const char *seq;
305 enum mandoc_esc esc;
306 static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' };
307
308 nospace = 0;
309
310 while ('\0' != *p) {
311 sz = strcspn(p, rejs);
312
313 fwrite(p, 1, sz, stdout);
314 p += (int)sz;
315
316 if ('\0' == *p)
317 break;
318
319 switch (*p++) {
320 case ('<'):
321 printf("&lt;");
322 continue;
323 case ('>'):
324 printf("&gt;");
325 continue;
326 case ('&'):
327 printf("&amp;");
328 continue;
329 case (ASCII_HYPH):
330 putchar('-');
331 continue;
332 default:
333 break;
334 }
335
336 esc = mandoc_escape(&p, &seq, &len);
337 if (ESCAPE_ERROR == esc)
338 break;
339
340 switch (esc) {
341 case (ESCAPE_NUMBERED):
342 print_num(h, seq, len);
343 break;
344 case (ESCAPE_PREDEF):
345 print_res(h, seq, len);
346 break;
347 case (ESCAPE_SPECIAL):
348 print_spec(h, seq, len);
349 break;
350 case (ESCAPE_FONTPREV):
351 /* FALLTHROUGH */
352 case (ESCAPE_FONTBOLD):
353 /* FALLTHROUGH */
354 case (ESCAPE_FONTITALIC):
355 /* FALLTHROUGH */
356 case (ESCAPE_FONTROMAN):
357 if (norecurse)
358 break;
359 print_metaf(h, esc);
360 break;
361 case (ESCAPE_NOSPACE):
362 if ('\0' == *p)
363 nospace = 1;
364 break;
365 default:
366 break;
367 }
368 }
369
370 return(nospace);
371 }
372
373
374 static void
375 print_attr(struct html *h, const char *key, const char *val)
376 {
377 printf(" %s=\"", key);
378 (void)print_encode(h, val, 1);
379 putchar('\"');
380 }
381
382
383 struct tag *
384 print_otag(struct html *h, enum htmltag tag,
385 int sz, const struct htmlpair *p)
386 {
387 int i;
388 struct tag *t;
389
390 /* Push this tags onto the stack of open scopes. */
391
392 if ( ! (HTML_NOSTACK & htmltags[tag].flags)) {
393 t = mandoc_malloc(sizeof(struct tag));
394 t->tag = tag;
395 t->next = h->tags.head;
396 h->tags.head = t;
397 } else
398 t = NULL;
399
400 if ( ! (HTML_NOSPACE & h->flags))
401 if ( ! (HTML_CLRLINE & htmltags[tag].flags)) {
402 /* Manage keeps! */
403 if ( ! (HTML_KEEP & h->flags)) {
404 if (HTML_PREKEEP & h->flags)
405 h->flags |= HTML_KEEP;
406 putchar(' ');
407 } else
408 printf("&#160;");
409 }
410
411 if ( ! (h->flags & HTML_NONOSPACE))
412 h->flags &= ~HTML_NOSPACE;
413 else
414 h->flags |= HTML_NOSPACE;
415
416 /* Print out the tag name and attributes. */
417
418 printf("<%s", htmltags[tag].name);
419 for (i = 0; i < sz; i++)
420 print_attr(h, htmlattrs[p[i].key], p[i].val);
421
422 /* Add non-overridable attributes. */
423
424 if (TAG_HTML == tag && HTML_XHTML_1_0_STRICT == h->type) {
425 print_attr(h, "xmlns", "http://www.w3.org/1999/xhtml");
426 print_attr(h, "xml:lang", "en");
427 print_attr(h, "lang", "en");
428 }
429
430 /* Accomodate for XML "well-formed" singleton escaping. */
431
432 if (HTML_AUTOCLOSE & htmltags[tag].flags)
433 switch (h->type) {
434 case (HTML_XHTML_1_0_STRICT):
435 putchar('/');
436 break;
437 default:
438 break;
439 }
440
441 putchar('>');
442
443 h->flags |= HTML_NOSPACE;
444
445 if ((HTML_AUTOCLOSE | HTML_CLRLINE) & htmltags[tag].flags)
446 putchar('\n');
447
448 return(t);
449 }
450
451
452 static void
453 print_ctag(struct html *h, enum htmltag tag)
454 {
455
456 printf("</%s>", htmltags[tag].name);
457 if (HTML_CLRLINE & htmltags[tag].flags) {
458 h->flags |= HTML_NOSPACE;
459 putchar('\n');
460 }
461 }
462
463
464 void
465 print_gen_decls(struct html *h)
466 {
467
468 print_xmltype(h);
469 print_doctype(h);
470 }
471
472
473 static void
474 print_xmltype(struct html *h)
475 {
476
477 if (HTML_XHTML_1_0_STRICT == h->type)
478 puts("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
479 }
480
481
482 static void
483 print_doctype(struct html *h)
484 {
485 const char *doctype;
486 const char *dtd;
487 const char *name;
488
489 switch (h->type) {
490 case (HTML_HTML_4_01_STRICT):
491 name = "HTML";
492 doctype = "-//W3C//DTD HTML 4.01//EN";
493 dtd = "http://www.w3.org/TR/html4/strict.dtd";
494 break;
495 default:
496 name = "html";
497 doctype = "-//W3C//DTD XHTML 1.0 Strict//EN";
498 dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
499 break;
500 }
501
502 printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
503 name, doctype, dtd);
504 }
505
506 void
507 print_text(struct html *h, const char *word)
508 {
509
510 if ( ! (HTML_NOSPACE & h->flags)) {
511 /* Manage keeps! */
512 if ( ! (HTML_KEEP & h->flags)) {
513 if (HTML_PREKEEP & h->flags)
514 h->flags |= HTML_KEEP;
515 putchar(' ');
516 } else
517 printf("&#160;");
518 }
519
520 assert(NULL == h->metaf);
521 if (HTMLFONT_NONE != h->metac)
522 h->metaf = HTMLFONT_BOLD == h->metac ?
523 print_otag(h, TAG_B, 0, NULL) :
524 print_otag(h, TAG_I, 0, NULL);
525
526 assert(word);
527 if ( ! print_encode(h, word, 0))
528 if ( ! (h->flags & HTML_NONOSPACE))
529 h->flags &= ~HTML_NOSPACE;
530
531 if (h->metaf) {
532 print_tagq(h, h->metaf);
533 h->metaf = NULL;
534 }
535
536 h->flags &= ~HTML_IGNDELIM;
537 }
538
539
540 void
541 print_tagq(struct html *h, const struct tag *until)
542 {
543 struct tag *tag;
544
545 while ((tag = h->tags.head) != NULL) {
546 /*
547 * Remember to close out and nullify the current
548 * meta-font and table, if applicable.
549 */
550 if (tag == h->metaf)
551 h->metaf = NULL;
552 if (tag == h->tblt)
553 h->tblt = NULL;
554 print_ctag(h, tag->tag);
555 h->tags.head = tag->next;
556 free(tag);
557 if (until && tag == until)
558 return;
559 }
560 }
561
562
563 void
564 print_stagq(struct html *h, const struct tag *suntil)
565 {
566 struct tag *tag;
567
568 while ((tag = h->tags.head) != NULL) {
569 if (suntil && tag == suntil)
570 return;
571 /*
572 * Remember to close out and nullify the current
573 * meta-font and table, if applicable.
574 */
575 if (tag == h->metaf)
576 h->metaf = NULL;
577 if (tag == h->tblt)
578 h->tblt = NULL;
579 print_ctag(h, tag->tag);
580 h->tags.head = tag->next;
581 free(tag);
582 }
583 }
584
585
586 void
587 bufinit(struct html *h)
588 {
589
590 h->buf[0] = '\0';
591 h->buflen = 0;
592 }
593
594
595 void
596 bufcat_style(struct html *h, const char *key, const char *val)
597 {
598
599 bufcat(h, key);
600 bufncat(h, ":", 1);
601 bufcat(h, val);
602 bufncat(h, ";", 1);
603 }
604
605
606 void
607 bufcat(struct html *h, const char *p)
608 {
609
610 bufncat(h, p, strlen(p));
611 }
612
613
614 void
615 buffmt(struct html *h, const char *fmt, ...)
616 {
617 va_list ap;
618
619 va_start(ap, fmt);
620 (void)vsnprintf(h->buf + (int)h->buflen,
621 BUFSIZ - h->buflen - 1, fmt, ap);
622 va_end(ap);
623 h->buflen = strlen(h->buf);
624 }
625
626
627 void
628 bufncat(struct html *h, const char *p, size_t sz)
629 {
630
631 if (h->buflen + sz > BUFSIZ - 1)
632 sz = BUFSIZ - 1 - h->buflen;
633
634 (void)strncat(h->buf, p, sz);
635 h->buflen += sz;
636 }
637
638
639 void
640 buffmt_includes(struct html *h, const char *name)
641 {
642 const char *p, *pp;
643
644 pp = h->base_includes;
645
646 while (NULL != (p = strchr(pp, '%'))) {
647 bufncat(h, pp, (size_t)(p - pp));
648 switch (*(p + 1)) {
649 case('I'):
650 bufcat(h, name);
651 break;
652 default:
653 bufncat(h, p, 2);
654 break;
655 }
656 pp = p + 2;
657 }
658 if (pp)
659 bufcat(h, pp);
660 }
661
662
663 void
664 buffmt_man(struct html *h,
665 const char *name, const char *sec)
666 {
667 const char *p, *pp;
668
669 pp = h->base_man;
670
671 /* LINTED */
672 while (NULL != (p = strchr(pp, '%'))) {
673 bufncat(h, pp, (size_t)(p - pp));
674 switch (*(p + 1)) {
675 case('S'):
676 bufcat(h, sec ? sec : "1");
677 break;
678 case('N'):
679 buffmt(h, name);
680 break;
681 default:
682 bufncat(h, p, 2);
683 break;
684 }
685 pp = p + 2;
686 }
687 if (pp)
688 bufcat(h, pp);
689 }
690
691
692 void
693 bufcat_su(struct html *h, const char *p, const struct roffsu *su)
694 {
695 double v;
696 const char *u;
697
698 v = su->scale;
699
700 switch (su->unit) {
701 case (SCALE_CM):
702 u = "cm";
703 break;
704 case (SCALE_IN):
705 u = "in";
706 break;
707 case (SCALE_PC):
708 u = "pc";
709 break;
710 case (SCALE_PT):
711 u = "pt";
712 break;
713 case (SCALE_EM):
714 u = "em";
715 break;
716 case (SCALE_MM):
717 if (0 == (v /= 100))
718 v = 1;
719 u = "em";
720 break;
721 case (SCALE_EN):
722 u = "ex";
723 break;
724 case (SCALE_BU):
725 u = "ex";
726 break;
727 case (SCALE_VS):
728 u = "em";
729 break;
730 default:
731 u = "ex";
732 break;
733 }
734
735 /*
736 * XXX: the CSS spec isn't clear as to which types accept
737 * integer or real numbers, so we just make them all decimals.
738 */
739 buffmt(h, "%s: %.2f%s;", p, v, u);
740 }
741
742
743 void
744 html_idcat(char *dst, const char *src, int sz)
745 {
746 int ssz;
747
748 assert(sz > 2);
749
750 /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
751
752 /* We can't start with a number (bah). */
753
754 if ('#' == *dst) {
755 dst++;
756 sz--;
757 }
758 if ('\0' == *dst) {
759 *dst++ = 'x';
760 *dst = '\0';
761 sz--;
762 }
763
764 for ( ; *dst != '\0' && sz; dst++, sz--)
765 /* Jump to end. */ ;
766
767 for ( ; *src != '\0' && sz > 1; src++) {
768 ssz = snprintf(dst, (size_t)sz, "%.2x", *src);
769 sz -= ssz;
770 dst += ssz;
771 }
772 }