]> git.cameronkatri.com Git - mandoc.git/blob - html.c
The bufcat() function in -T[x]html was eating one byte off the end of its
[mandoc.git] / html.c
1 /* $Id: html.c,v 1.148 2011/07/04 09:42:38 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 "libmandoc.h"
35 #include "out.h"
36 #include "html.h"
37 #include "main.h"
38
39 struct htmldata {
40 const char *name;
41 int flags;
42 #define HTML_CLRLINE (1 << 0)
43 #define HTML_NOSTACK (1 << 1)
44 #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
45 };
46
47 static const struct htmldata htmltags[TAG_MAX] = {
48 {"html", HTML_CLRLINE}, /* TAG_HTML */
49 {"head", HTML_CLRLINE}, /* TAG_HEAD */
50 {"body", HTML_CLRLINE}, /* TAG_BODY */
51 {"meta", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_META */
52 {"title", HTML_CLRLINE}, /* TAG_TITLE */
53 {"div", HTML_CLRLINE}, /* TAG_DIV */
54 {"h1", 0}, /* TAG_H1 */
55 {"h2", 0}, /* TAG_H2 */
56 {"span", 0}, /* TAG_SPAN */
57 {"link", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_LINK */
58 {"br", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_BR */
59 {"a", 0}, /* TAG_A */
60 {"table", HTML_CLRLINE}, /* TAG_TABLE */
61 {"tbody", HTML_CLRLINE}, /* TAG_TBODY */
62 {"col", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_COL */
63 {"tr", HTML_CLRLINE}, /* TAG_TR */
64 {"td", HTML_CLRLINE}, /* TAG_TD */
65 {"li", HTML_CLRLINE}, /* TAG_LI */
66 {"ul", HTML_CLRLINE}, /* TAG_UL */
67 {"ol", HTML_CLRLINE}, /* TAG_OL */
68 {"dl", HTML_CLRLINE}, /* TAG_DL */
69 {"dt", HTML_CLRLINE}, /* TAG_DT */
70 {"dd", HTML_CLRLINE}, /* TAG_DD */
71 {"blockquote", HTML_CLRLINE}, /* TAG_BLOCKQUOTE */
72 {"p", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_P */
73 {"pre", HTML_CLRLINE }, /* TAG_PRE */
74 {"b", 0 }, /* TAG_B */
75 {"i", 0 }, /* TAG_I */
76 {"code", 0 }, /* TAG_CODE */
77 {"small", 0 }, /* TAG_SMALL */
78 };
79
80 static const char *const htmlattrs[ATTR_MAX] = {
81 "http-equiv", /* ATTR_HTTPEQUIV */
82 "content", /* ATTR_CONTENT */
83 "name", /* ATTR_NAME */
84 "rel", /* ATTR_REL */
85 "href", /* ATTR_HREF */
86 "type", /* ATTR_TYPE */
87 "media", /* ATTR_MEDIA */
88 "class", /* ATTR_CLASS */
89 "style", /* ATTR_STYLE */
90 "width", /* ATTR_WIDTH */
91 "id", /* ATTR_ID */
92 "summary", /* ATTR_SUMMARY */
93 "align", /* ATTR_ALIGN */
94 "colspan", /* ATTR_COLSPAN */
95 };
96
97 static const char *const roffscales[SCALE_MAX] = {
98 "cm", /* SCALE_CM */
99 "in", /* SCALE_IN */
100 "pc", /* SCALE_PC */
101 "pt", /* SCALE_PT */
102 "em", /* SCALE_EM */
103 "em", /* SCALE_MM */
104 "ex", /* SCALE_EN */
105 "ex", /* SCALE_BU */
106 "em", /* SCALE_VS */
107 "ex", /* SCALE_FS */
108 };
109
110 static void bufncat(struct html *, const char *, size_t);
111 static void print_ctag(struct html *, enum htmltag);
112 static int print_encode(struct html *, const char *, int);
113 static void print_metaf(struct html *, enum mandoc_esc);
114 static void print_attr(struct html *, const char *, const char *);
115 static void *ml_alloc(char *, enum htmltype);
116
117 static void *
118 ml_alloc(char *outopts, enum htmltype type)
119 {
120 struct html *h;
121 const char *toks[4];
122 char *v;
123
124 toks[0] = "style";
125 toks[1] = "man";
126 toks[2] = "includes";
127 toks[3] = NULL;
128
129 h = mandoc_calloc(1, sizeof(struct html));
130
131 h->type = type;
132 h->tags.head = NULL;
133 h->symtab = mchars_alloc();
134
135 while (outopts && *outopts)
136 switch (getsubopt(&outopts, UNCONST(toks), &v)) {
137 case (0):
138 h->style = v;
139 break;
140 case (1):
141 h->base_man = v;
142 break;
143 case (2):
144 h->base_includes = v;
145 break;
146 default:
147 break;
148 }
149
150 return(h);
151 }
152
153 void *
154 html_alloc(char *outopts)
155 {
156
157 return(ml_alloc(outopts, HTML_HTML_4_01_STRICT));
158 }
159
160
161 void *
162 xhtml_alloc(char *outopts)
163 {
164
165 return(ml_alloc(outopts, HTML_XHTML_1_0_STRICT));
166 }
167
168
169 void
170 html_free(void *p)
171 {
172 struct tag *tag;
173 struct html *h;
174
175 h = (struct html *)p;
176
177 while ((tag = h->tags.head) != NULL) {
178 h->tags.head = tag->next;
179 free(tag);
180 }
181
182 if (h->symtab)
183 mchars_free(h->symtab);
184
185 free(h);
186 }
187
188
189 void
190 print_gen_head(struct html *h)
191 {
192 struct htmlpair tag[4];
193
194 tag[0].key = ATTR_HTTPEQUIV;
195 tag[0].val = "Content-Type";
196 tag[1].key = ATTR_CONTENT;
197 tag[1].val = "text/html; charset=utf-8";
198 print_otag(h, TAG_META, 2, tag);
199
200 tag[0].key = ATTR_NAME;
201 tag[0].val = "resource-type";
202 tag[1].key = ATTR_CONTENT;
203 tag[1].val = "document";
204 print_otag(h, TAG_META, 2, tag);
205
206 if (h->style) {
207 tag[0].key = ATTR_REL;
208 tag[0].val = "stylesheet";
209 tag[1].key = ATTR_HREF;
210 tag[1].val = h->style;
211 tag[2].key = ATTR_TYPE;
212 tag[2].val = "text/css";
213 tag[3].key = ATTR_MEDIA;
214 tag[3].val = "all";
215 print_otag(h, TAG_LINK, 4, tag);
216 }
217 }
218
219 static void
220 print_metaf(struct html *h, enum mandoc_esc deco)
221 {
222 enum htmlfont font;
223
224 switch (deco) {
225 case (ESCAPE_FONTPREV):
226 font = h->metal;
227 break;
228 case (ESCAPE_FONTITALIC):
229 font = HTMLFONT_ITALIC;
230 break;
231 case (ESCAPE_FONTBOLD):
232 font = HTMLFONT_BOLD;
233 break;
234 case (ESCAPE_FONT):
235 /* FALLTHROUGH */
236 case (ESCAPE_FONTROMAN):
237 font = HTMLFONT_NONE;
238 break;
239 default:
240 abort();
241 /* NOTREACHED */
242 }
243
244 if (h->metaf) {
245 print_tagq(h, h->metaf);
246 h->metaf = NULL;
247 }
248
249 h->metal = h->metac;
250 h->metac = font;
251
252 if (HTMLFONT_NONE != font)
253 h->metaf = HTMLFONT_BOLD == font ?
254 print_otag(h, TAG_B, 0, NULL) :
255 print_otag(h, TAG_I, 0, NULL);
256 }
257
258 int
259 html_strlen(const char *cp)
260 {
261 int ssz, sz;
262 const char *seq, *p;
263
264 /*
265 * Account for escaped sequences within string length
266 * calculations. This follows the logic in term_strlen() as we
267 * must calculate the width of produced strings.
268 * Assume that characters are always width of "1". This is
269 * hacky, but it gets the job done for approximation of widths.
270 */
271
272 sz = 0;
273 while (NULL != (p = strchr(cp, '\\'))) {
274 sz += (int)(p - cp);
275 ++cp;
276 switch (mandoc_escape(&cp, &seq, &ssz)) {
277 case (ESCAPE_ERROR):
278 return(sz);
279 case (ESCAPE_UNICODE):
280 /* FALLTHROUGH */
281 case (ESCAPE_NUMBERED):
282 /* FALLTHROUGH */
283 case (ESCAPE_SPECIAL):
284 sz++;
285 break;
286 default:
287 break;
288 }
289 }
290
291 assert(sz >= 0);
292 return(sz + strlen(cp));
293 }
294
295 static int
296 print_encode(struct html *h, const char *p, int norecurse)
297 {
298 size_t sz;
299 int c, len, nospace;
300 const char *seq;
301 enum mandoc_esc esc;
302 static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' };
303
304 nospace = 0;
305
306 while ('\0' != *p) {
307 sz = strcspn(p, rejs);
308
309 fwrite(p, 1, sz, stdout);
310 p += (int)sz;
311
312 if ('\0' == *p)
313 break;
314
315 switch (*p++) {
316 case ('<'):
317 printf("&lt;");
318 continue;
319 case ('>'):
320 printf("&gt;");
321 continue;
322 case ('&'):
323 printf("&amp;");
324 continue;
325 case (ASCII_HYPH):
326 putchar('-');
327 continue;
328 default:
329 break;
330 }
331
332 esc = mandoc_escape(&p, &seq, &len);
333 if (ESCAPE_ERROR == esc)
334 break;
335
336 switch (esc) {
337 case (ESCAPE_UNICODE):
338 /* Skip passed "u" header. */
339 c = mchars_num2uc(seq + 1, len - 1);
340 if ('\0' != c)
341 printf("&#x%x;", c);
342 break;
343 case (ESCAPE_NUMBERED):
344 c = mchars_num2char(seq, len);
345 if ('\0' != c)
346 putchar(c);
347 break;
348 case (ESCAPE_SPECIAL):
349 c = mchars_spec2cp(h->symtab, seq, len);
350 if (c > 0)
351 printf("&#%d;", c);
352 else if (-1 == c && 1 == len)
353 putchar((int)*seq);
354 break;
355 case (ESCAPE_FONT):
356 /* FALLTHROUGH */
357 case (ESCAPE_FONTPREV):
358 /* FALLTHROUGH */
359 case (ESCAPE_FONTBOLD):
360 /* FALLTHROUGH */
361 case (ESCAPE_FONTITALIC):
362 /* FALLTHROUGH */
363 case (ESCAPE_FONTROMAN):
364 if (norecurse)
365 break;
366 print_metaf(h, esc);
367 break;
368 case (ESCAPE_NOSPACE):
369 if ('\0' == *p)
370 nospace = 1;
371 break;
372 default:
373 break;
374 }
375 }
376
377 return(nospace);
378 }
379
380
381 static void
382 print_attr(struct html *h, const char *key, const char *val)
383 {
384 printf(" %s=\"", key);
385 (void)print_encode(h, val, 1);
386 putchar('\"');
387 }
388
389
390 struct tag *
391 print_otag(struct html *h, enum htmltag tag,
392 int sz, const struct htmlpair *p)
393 {
394 int i;
395 struct tag *t;
396
397 /* Push this tags onto the stack of open scopes. */
398
399 if ( ! (HTML_NOSTACK & htmltags[tag].flags)) {
400 t = mandoc_malloc(sizeof(struct tag));
401 t->tag = tag;
402 t->next = h->tags.head;
403 h->tags.head = t;
404 } else
405 t = NULL;
406
407 if ( ! (HTML_NOSPACE & h->flags))
408 if ( ! (HTML_CLRLINE & htmltags[tag].flags)) {
409 /* Manage keeps! */
410 if ( ! (HTML_KEEP & h->flags)) {
411 if (HTML_PREKEEP & h->flags)
412 h->flags |= HTML_KEEP;
413 putchar(' ');
414 } else
415 printf("&#160;");
416 }
417
418 if ( ! (h->flags & HTML_NONOSPACE))
419 h->flags &= ~HTML_NOSPACE;
420 else
421 h->flags |= HTML_NOSPACE;
422
423 /* Print out the tag name and attributes. */
424
425 printf("<%s", htmltags[tag].name);
426 for (i = 0; i < sz; i++)
427 print_attr(h, htmlattrs[p[i].key], p[i].val);
428
429 /* Add non-overridable attributes. */
430
431 if (TAG_HTML == tag && HTML_XHTML_1_0_STRICT == h->type) {
432 print_attr(h, "xmlns", "http://www.w3.org/1999/xhtml");
433 print_attr(h, "xml:lang", "en");
434 print_attr(h, "lang", "en");
435 }
436
437 /* Accommodate for XML "well-formed" singleton escaping. */
438
439 if (HTML_AUTOCLOSE & htmltags[tag].flags)
440 switch (h->type) {
441 case (HTML_XHTML_1_0_STRICT):
442 putchar('/');
443 break;
444 default:
445 break;
446 }
447
448 putchar('>');
449
450 h->flags |= HTML_NOSPACE;
451
452 if ((HTML_AUTOCLOSE | HTML_CLRLINE) & htmltags[tag].flags)
453 putchar('\n');
454
455 return(t);
456 }
457
458
459 static void
460 print_ctag(struct html *h, enum htmltag tag)
461 {
462
463 printf("</%s>", htmltags[tag].name);
464 if (HTML_CLRLINE & htmltags[tag].flags) {
465 h->flags |= HTML_NOSPACE;
466 putchar('\n');
467 }
468 }
469
470 void
471 print_gen_decls(struct html *h)
472 {
473 const char *doctype;
474 const char *dtd;
475 const char *name;
476
477 switch (h->type) {
478 case (HTML_HTML_4_01_STRICT):
479 name = "HTML";
480 doctype = "-//W3C//DTD HTML 4.01//EN";
481 dtd = "http://www.w3.org/TR/html4/strict.dtd";
482 break;
483 default:
484 puts("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
485 name = "html";
486 doctype = "-//W3C//DTD XHTML 1.0 Strict//EN";
487 dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
488 break;
489 }
490
491 printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
492 name, doctype, dtd);
493 }
494
495 void
496 print_text(struct html *h, const char *word)
497 {
498
499 if ( ! (HTML_NOSPACE & h->flags)) {
500 /* Manage keeps! */
501 if ( ! (HTML_KEEP & h->flags)) {
502 if (HTML_PREKEEP & h->flags)
503 h->flags |= HTML_KEEP;
504 putchar(' ');
505 } else
506 printf("&#160;");
507 }
508
509 assert(NULL == h->metaf);
510 if (HTMLFONT_NONE != h->metac)
511 h->metaf = HTMLFONT_BOLD == h->metac ?
512 print_otag(h, TAG_B, 0, NULL) :
513 print_otag(h, TAG_I, 0, NULL);
514
515 assert(word);
516 if ( ! print_encode(h, word, 0))
517 if ( ! (h->flags & HTML_NONOSPACE))
518 h->flags &= ~HTML_NOSPACE;
519
520 if (h->metaf) {
521 print_tagq(h, h->metaf);
522 h->metaf = NULL;
523 }
524
525 h->flags &= ~HTML_IGNDELIM;
526 }
527
528
529 void
530 print_tagq(struct html *h, const struct tag *until)
531 {
532 struct tag *tag;
533
534 while ((tag = h->tags.head) != NULL) {
535 /*
536 * Remember to close out and nullify the current
537 * meta-font and table, if applicable.
538 */
539 if (tag == h->metaf)
540 h->metaf = NULL;
541 if (tag == h->tblt)
542 h->tblt = NULL;
543 print_ctag(h, tag->tag);
544 h->tags.head = tag->next;
545 free(tag);
546 if (until && tag == until)
547 return;
548 }
549 }
550
551
552 void
553 print_stagq(struct html *h, const struct tag *suntil)
554 {
555 struct tag *tag;
556
557 while ((tag = h->tags.head) != NULL) {
558 if (suntil && tag == suntil)
559 return;
560 /*
561 * Remember to close out and nullify the current
562 * meta-font and table, if applicable.
563 */
564 if (tag == h->metaf)
565 h->metaf = NULL;
566 if (tag == h->tblt)
567 h->tblt = NULL;
568 print_ctag(h, tag->tag);
569 h->tags.head = tag->next;
570 free(tag);
571 }
572 }
573
574 void
575 bufinit(struct html *h)
576 {
577
578 h->buf[0] = '\0';
579 h->buflen = 0;
580 }
581
582 void
583 bufcat_style(struct html *h, const char *key, const char *val)
584 {
585
586 bufcat(h, key);
587 bufcat(h, ":");
588 bufcat(h, val);
589 bufcat(h, ";");
590 }
591
592 void
593 bufcat(struct html *h, const char *p)
594 {
595
596 h->buflen = strlcat(h->buf, p, BUFSIZ);
597 assert(h->buflen < BUFSIZ);
598 }
599
600 void
601 bufcat_fmt(struct html *h, const char *fmt, ...)
602 {
603 va_list ap;
604
605 va_start(ap, fmt);
606 (void)vsnprintf(h->buf + (int)h->buflen,
607 BUFSIZ - h->buflen - 1, fmt, ap);
608 va_end(ap);
609 h->buflen = strlen(h->buf);
610 }
611
612 static void
613 bufncat(struct html *h, const char *p, size_t sz)
614 {
615
616 assert(h->buflen + sz + 1 < BUFSIZ);
617 strncat(h->buf, p, sz);
618 h->buflen += sz;
619 }
620
621 void
622 buffmt_includes(struct html *h, const char *name)
623 {
624 const char *p, *pp;
625
626 pp = h->base_includes;
627
628 bufinit(h);
629 while (NULL != (p = strchr(pp, '%'))) {
630 bufncat(h, pp, (size_t)(p - pp));
631 switch (*(p + 1)) {
632 case('I'):
633 bufcat(h, name);
634 break;
635 default:
636 bufncat(h, p, 2);
637 break;
638 }
639 pp = p + 2;
640 }
641 if (pp)
642 bufcat(h, pp);
643 }
644
645 void
646 buffmt_man(struct html *h,
647 const char *name, const char *sec)
648 {
649 const char *p, *pp;
650
651 pp = h->base_man;
652
653 bufinit(h);
654 while (NULL != (p = strchr(pp, '%'))) {
655 bufncat(h, pp, (size_t)(p - pp));
656 switch (*(p + 1)) {
657 case('S'):
658 bufcat(h, sec ? sec : "1");
659 break;
660 case('N'):
661 bufcat_fmt(h, name);
662 break;
663 default:
664 bufncat(h, p, 2);
665 break;
666 }
667 pp = p + 2;
668 }
669 if (pp)
670 bufcat(h, pp);
671 }
672
673 void
674 bufcat_su(struct html *h, const char *p, const struct roffsu *su)
675 {
676 double v;
677
678 v = su->scale;
679 if (SCALE_MM == su->unit && 0.0 == (v /= 100.0))
680 v = 1.0;
681
682 bufcat_fmt(h, "%s: %.2f%s;", p, v, roffscales[su->unit]);
683 }
684
685 void
686 bufcat_id(struct html *h, const char *src)
687 {
688
689 /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
690
691 while ('\0' != *src)
692 bufcat_fmt(h, "%.2x", *src++);
693 }