Annotation of src/usr.bin/mandoc/html.c, Revision 1.9
1.9 ! schwarze 1: /* $Id: html.c,v 1.8 2010/05/14 01:54:37 schwarze Exp $ */
1.1 schwarze 2: /*
3: * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
4: *
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.
8: *
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.
16: */
17: #include <sys/types.h>
18:
19: #include <assert.h>
1.3 schwarze 20: #include <ctype.h>
1.4 schwarze 21: #include <stdarg.h>
1.1 schwarze 22: #include <stdio.h>
23: #include <stdint.h>
24: #include <stdlib.h>
25: #include <string.h>
26: #include <unistd.h>
27:
1.9 ! schwarze 28: #include "mandoc.h"
1.1 schwarze 29: #include "out.h"
30: #include "chars.h"
31: #include "html.h"
32: #include "main.h"
33:
34: #define UNCONST(a) ((void *)(uintptr_t)(const void *)(a))
35:
36: struct htmldata {
37: const char *name;
38: int flags;
39: #define HTML_CLRLINE (1 << 0)
40: #define HTML_NOSTACK (1 << 1)
1.6 schwarze 41: #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
1.1 schwarze 42: };
43:
44: static const struct htmldata htmltags[TAG_MAX] = {
45: {"html", HTML_CLRLINE}, /* TAG_HTML */
46: {"head", HTML_CLRLINE}, /* TAG_HEAD */
47: {"body", HTML_CLRLINE}, /* TAG_BODY */
1.6 schwarze 48: {"meta", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_META */
1.1 schwarze 49: {"title", HTML_CLRLINE}, /* TAG_TITLE */
50: {"div", HTML_CLRLINE}, /* TAG_DIV */
51: {"h1", 0}, /* TAG_H1 */
52: {"h2", 0}, /* TAG_H2 */
53: {"span", 0}, /* TAG_SPAN */
1.8 schwarze 54: {"link", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_LINK */
1.6 schwarze 55: {"br", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_BR */
1.1 schwarze 56: {"a", 0}, /* TAG_A */
57: {"table", HTML_CLRLINE}, /* TAG_TABLE */
1.6 schwarze 58: {"col", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_COL */
1.1 schwarze 59: {"tr", HTML_CLRLINE}, /* TAG_TR */
60: {"td", HTML_CLRLINE}, /* TAG_TD */
61: {"li", HTML_CLRLINE}, /* TAG_LI */
62: {"ul", HTML_CLRLINE}, /* TAG_UL */
63: {"ol", HTML_CLRLINE}, /* TAG_OL */
64: };
65:
1.5 schwarze 66: static const char *const htmlfonts[HTMLFONT_MAX] = {
67: "roman",
68: "bold",
69: "italic"
70: };
71:
72: static const char *const htmlattrs[ATTR_MAX] = {
1.1 schwarze 73: "http-equiv",
74: "content",
75: "name",
76: "rel",
77: "href",
78: "type",
79: "media",
80: "class",
81: "style",
82: "width",
83: "valign",
84: "target",
85: "id",
1.3 schwarze 86: "summary",
1.1 schwarze 87: };
88:
1.5 schwarze 89: static void print_spec(struct html *, const char *, size_t);
90: static void print_res(struct html *, const char *, size_t);
91: static void print_ctag(struct html *, enum htmltag);
1.6 schwarze 92: static void print_doctype(struct html *);
93: static void print_xmltype(struct html *);
1.5 schwarze 94: static int print_encode(struct html *, const char *, int);
95: static void print_metaf(struct html *, enum roffdeco);
1.6 schwarze 96: static void print_attr(struct html *,
97: const char *, const char *);
98: static void *ml_alloc(char *, enum htmltype);
1.5 schwarze 99:
100:
1.6 schwarze 101: static void *
102: ml_alloc(char *outopts, enum htmltype type)
1.1 schwarze 103: {
104: struct html *h;
105: const char *toks[4];
106: char *v;
107:
108: toks[0] = "style";
109: toks[1] = "man";
110: toks[2] = "includes";
111: toks[3] = NULL;
112:
1.3 schwarze 113: h = calloc(1, sizeof(struct html));
114: if (NULL == h) {
115: perror(NULL);
116: exit(EXIT_FAILURE);
117: }
1.1 schwarze 118:
1.6 schwarze 119: h->type = type;
1.2 schwarze 120: h->tags.head = NULL;
121: h->ords.head = NULL;
1.3 schwarze 122: h->symtab = chars_init(CHARS_HTML);
1.1 schwarze 123:
124: while (outopts && *outopts)
125: switch (getsubopt(&outopts, UNCONST(toks), &v)) {
126: case (0):
127: h->style = v;
128: break;
129: case (1):
130: h->base_man = v;
131: break;
132: case (2):
133: h->base_includes = v;
134: break;
135: default:
136: break;
137: }
138:
139: return(h);
140: }
141:
1.6 schwarze 142: void *
143: html_alloc(char *outopts)
144: {
145:
146: return(ml_alloc(outopts, HTML_HTML_4_01_STRICT));
147: }
148:
149:
150: void *
151: xhtml_alloc(char *outopts)
152: {
153:
154: return(ml_alloc(outopts, HTML_XHTML_1_0_STRICT));
155: }
156:
1.1 schwarze 157:
158: void
159: html_free(void *p)
160: {
161: struct tag *tag;
162: struct ord *ord;
163: struct html *h;
164:
165: h = (struct html *)p;
166:
1.2 schwarze 167: while ((ord = h->ords.head) != NULL) {
168: h->ords.head = ord->next;
1.1 schwarze 169: free(ord);
170: }
171:
1.2 schwarze 172: while ((tag = h->tags.head) != NULL) {
173: h->tags.head = tag->next;
1.1 schwarze 174: free(tag);
175: }
176:
177: if (h->symtab)
178: chars_free(h->symtab);
179:
180: free(h);
181: }
182:
183:
184: void
185: print_gen_head(struct html *h)
186: {
187: struct htmlpair tag[4];
188:
189: tag[0].key = ATTR_HTTPEQUIV;
190: tag[0].val = "Content-Type";
191: tag[1].key = ATTR_CONTENT;
192: tag[1].val = "text/html; charset=utf-8";
193: print_otag(h, TAG_META, 2, tag);
194:
195: tag[0].key = ATTR_NAME;
196: tag[0].val = "resource-type";
197: tag[1].key = ATTR_CONTENT;
198: tag[1].val = "document";
199: print_otag(h, TAG_META, 2, tag);
200:
201: if (h->style) {
202: tag[0].key = ATTR_REL;
203: tag[0].val = "stylesheet";
204: tag[1].key = ATTR_HREF;
205: tag[1].val = h->style;
206: tag[2].key = ATTR_TYPE;
207: tag[2].val = "text/css";
208: tag[3].key = ATTR_MEDIA;
209: tag[3].val = "all";
210: print_otag(h, TAG_LINK, 4, tag);
211: }
212: }
213:
214:
215: static void
1.5 schwarze 216: print_spec(struct html *h, const char *p, size_t len)
1.1 schwarze 217: {
218: const char *rhs;
219: size_t sz;
220:
1.5 schwarze 221: rhs = chars_a2ascii(h->symtab, p, len, &sz);
1.1 schwarze 222:
223: if (NULL == rhs)
224: return;
1.4 schwarze 225: fwrite(rhs, 1, sz, stdout);
1.1 schwarze 226: }
227:
228:
229: static void
1.5 schwarze 230: print_res(struct html *h, const char *p, size_t len)
1.1 schwarze 231: {
232: const char *rhs;
233: size_t sz;
234:
1.5 schwarze 235: rhs = chars_a2res(h->symtab, p, len, &sz);
1.1 schwarze 236:
237: if (NULL == rhs)
238: return;
1.4 schwarze 239: fwrite(rhs, 1, sz, stdout);
1.1 schwarze 240: }
241:
242:
1.5 schwarze 243: struct tag *
244: print_ofont(struct html *h, enum htmlfont font)
1.1 schwarze 245: {
1.5 schwarze 246: struct htmlpair tag;
1.1 schwarze 247:
1.5 schwarze 248: h->metal = h->metac;
249: h->metac = font;
1.1 schwarze 250:
1.5 schwarze 251: /* FIXME: DECO_ROMAN should just close out preexisting. */
1.1 schwarze 252:
1.5 schwarze 253: if (h->metaf && h->tags.head == h->metaf)
254: print_tagq(h, h->metaf);
1.1 schwarze 255:
1.5 schwarze 256: PAIR_CLASS_INIT(&tag, htmlfonts[font]);
257: h->metaf = print_otag(h, TAG_SPAN, 1, &tag);
258: return(h->metaf);
259: }
1.1 schwarze 260:
261:
1.5 schwarze 262: static void
263: print_metaf(struct html *h, enum roffdeco deco)
264: {
265: enum htmlfont font;
1.1 schwarze 266:
1.5 schwarze 267: switch (deco) {
268: case (DECO_PREVIOUS):
269: font = h->metal;
270: break;
271: case (DECO_ITALIC):
272: font = HTMLFONT_ITALIC;
273: break;
274: case (DECO_BOLD):
275: font = HTMLFONT_BOLD;
276: break;
277: case (DECO_ROMAN):
278: font = HTMLFONT_NONE;
279: break;
280: default:
281: abort();
282: /* NOTREACHED */
1.1 schwarze 283: }
284:
1.5 schwarze 285: (void)print_ofont(h, font);
1.1 schwarze 286: }
287:
288:
1.5 schwarze 289: static int
290: print_encode(struct html *h, const char *p, int norecurse)
1.1 schwarze 291: {
1.4 schwarze 292: size_t sz;
1.5 schwarze 293: int len, nospace;
294: const char *seq;
295: enum roffdeco deco;
1.9 ! schwarze 296: static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' };
1.5 schwarze 297:
298: nospace = 0;
1.1 schwarze 299:
300: for (; *p; p++) {
1.9 ! schwarze 301: sz = strcspn(p, rejs);
1.4 schwarze 302:
303: fwrite(p, 1, sz, stdout);
304: p += /* LINTED */
305: sz;
306:
1.5 schwarze 307: if ('<' == *p) {
308: printf("<");
309: continue;
310: } else if ('>' == *p) {
311: printf(">");
312: continue;
313: } else if ('&' == *p) {
314: printf("&");
1.1 schwarze 315: continue;
1.9 ! schwarze 316: } else if (ASCII_HYPH == *p) {
! 317: /*
! 318: * Note: "soft hyphens" aren't graphically
! 319: * displayed when not breaking the text; we want
! 320: * them to be displayed.
! 321: */
! 322: /*printf("­");*/
! 323: putchar('-');
! 324: continue;
1.4 schwarze 325: } else if ('\0' == *p)
326: break;
327:
1.5 schwarze 328: seq = ++p;
329: len = a2roffdeco(&deco, &seq, &sz);
330:
331: switch (deco) {
332: case (DECO_RESERVED):
333: print_res(h, seq, sz);
334: break;
335: case (DECO_SPECIAL):
336: print_spec(h, seq, sz);
337: break;
338: case (DECO_PREVIOUS):
339: /* FALLTHROUGH */
340: case (DECO_BOLD):
341: /* FALLTHROUGH */
342: case (DECO_ITALIC):
343: /* FALLTHROUGH */
344: case (DECO_ROMAN):
345: if (norecurse)
346: break;
347: print_metaf(h, deco);
348: break;
349: default:
350: break;
351: }
352:
353: p += len - 1;
354:
355: if (DECO_NOSPACE == deco && '\0' == *(p + 1))
356: nospace = 1;
1.1 schwarze 357: }
1.5 schwarze 358:
359: return(nospace);
1.1 schwarze 360: }
361:
362:
1.6 schwarze 363: static void
364: print_attr(struct html *h, const char *key, const char *val)
365: {
366: printf(" %s=\"", key);
367: (void)print_encode(h, val, 1);
368: putchar('\"');
369: }
370:
371:
1.1 schwarze 372: struct tag *
373: print_otag(struct html *h, enum htmltag tag,
374: int sz, const struct htmlpair *p)
375: {
376: int i;
377: struct tag *t;
378:
1.6 schwarze 379: /* Push this tags onto the stack of open scopes. */
380:
1.1 schwarze 381: if ( ! (HTML_NOSTACK & htmltags[tag].flags)) {
1.3 schwarze 382: t = malloc(sizeof(struct tag));
383: if (NULL == t) {
384: perror(NULL);
385: exit(EXIT_FAILURE);
386: }
1.1 schwarze 387: t->tag = tag;
1.2 schwarze 388: t->next = h->tags.head;
389: h->tags.head = t;
1.1 schwarze 390: } else
391: t = NULL;
392:
393: if ( ! (HTML_NOSPACE & h->flags))
394: if ( ! (HTML_CLRLINE & htmltags[tag].flags))
1.4 schwarze 395: putchar(' ');
1.1 schwarze 396:
1.6 schwarze 397: /* Print out the tag name and attributes. */
398:
1.1 schwarze 399: printf("<%s", htmltags[tag].name);
1.6 schwarze 400: for (i = 0; i < sz; i++)
401: print_attr(h, htmlattrs[p[i].key], p[i].val);
402:
403: /* Add non-overridable attributes. */
404:
405: if (TAG_HTML == tag && HTML_XHTML_1_0_STRICT == h->type) {
406: print_attr(h, "xmlns", "http://www.w3.org/1999/xhtml");
407: print_attr(h, "xml:lang", "en");
408: print_attr(h, "lang", "en");
1.1 schwarze 409: }
1.6 schwarze 410:
411: /* Accomodate for XML "well-formed" singleton escaping. */
412:
413: if (HTML_AUTOCLOSE & htmltags[tag].flags)
414: switch (h->type) {
415: case (HTML_XHTML_1_0_STRICT):
416: putchar('/');
417: break;
418: default:
419: break;
420: }
421:
1.4 schwarze 422: putchar('>');
1.1 schwarze 423:
424: h->flags |= HTML_NOSPACE;
425: return(t);
426: }
427:
428:
429: static void
430: print_ctag(struct html *h, enum htmltag tag)
431: {
432:
433: printf("</%s>", htmltags[tag].name);
1.3 schwarze 434: if (HTML_CLRLINE & htmltags[tag].flags) {
1.1 schwarze 435: h->flags |= HTML_NOSPACE;
1.4 schwarze 436: putchar('\n');
1.5 schwarze 437: }
1.1 schwarze 438: }
439:
440:
441: void
1.6 schwarze 442: print_gen_decls(struct html *h)
443: {
444:
445: print_xmltype(h);
446: print_doctype(h);
447: }
448:
449:
450: static void
451: print_xmltype(struct html *h)
452: {
453:
1.9 ! schwarze 454: if (HTML_XHTML_1_0_STRICT == h->type)
! 455: printf("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
1.6 schwarze 456: }
457:
458:
459: static void
460: print_doctype(struct html *h)
1.1 schwarze 461: {
1.6 schwarze 462: const char *doctype;
463: const char *dtd;
464: const char *name;
465:
466: switch (h->type) {
467: case (HTML_HTML_4_01_STRICT):
468: name = "HTML";
469: doctype = "-//W3C//DTD HTML 4.01//EN";
470: dtd = "http://www.w3.org/TR/html4/strict.dtd";
471: break;
472: default:
473: name = "html";
474: doctype = "-//W3C//DTD XHTML 1.0 Strict//EN";
475: dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
476: break;
477: }
478:
479: printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
480: name, doctype, dtd);
1.1 schwarze 481: }
482:
483:
484: void
485: print_text(struct html *h, const char *p)
486: {
487:
488: if (*p && 0 == *(p + 1))
489: switch (*p) {
490: case('.'):
491: /* FALLTHROUGH */
492: case(','):
493: /* FALLTHROUGH */
494: case(';'):
495: /* FALLTHROUGH */
496: case(':'):
497: /* FALLTHROUGH */
498: case('?'):
499: /* FALLTHROUGH */
500: case('!'):
501: /* FALLTHROUGH */
502: case(')'):
503: /* FALLTHROUGH */
504: case(']'):
505: if ( ! (HTML_IGNDELIM & h->flags))
506: h->flags |= HTML_NOSPACE;
507: break;
508: default:
509: break;
510: }
511:
512: if ( ! (h->flags & HTML_NOSPACE))
1.4 schwarze 513: putchar(' ');
1.1 schwarze 514:
1.5 schwarze 515: assert(p);
516: if ( ! print_encode(h, p, 0))
517: h->flags &= ~HTML_NOSPACE;
1.1 schwarze 518:
1.8 schwarze 519: /*
520: * Note that we don't process the pipe: the parser sees it as
521: * punctuation, but we don't in terms of typography.
522: */
1.1 schwarze 523: if (*p && 0 == *(p + 1))
524: switch (*p) {
525: case('('):
526: /* FALLTHROUGH */
527: case('['):
528: h->flags |= HTML_NOSPACE;
529: break;
530: default:
531: break;
532: }
533: }
534:
535:
536: void
537: print_tagq(struct html *h, const struct tag *until)
538: {
539: struct tag *tag;
540:
1.2 schwarze 541: while ((tag = h->tags.head) != NULL) {
1.5 schwarze 542: if (tag == h->metaf)
543: h->metaf = NULL;
1.1 schwarze 544: print_ctag(h, tag->tag);
1.2 schwarze 545: h->tags.head = tag->next;
1.1 schwarze 546: free(tag);
547: if (until && tag == until)
548: return;
549: }
550: }
551:
552:
553: void
554: print_stagq(struct html *h, const struct tag *suntil)
555: {
556: struct tag *tag;
557:
1.2 schwarze 558: while ((tag = h->tags.head) != NULL) {
1.1 schwarze 559: if (suntil && tag == suntil)
560: return;
1.5 schwarze 561: if (tag == h->metaf)
562: h->metaf = NULL;
1.1 schwarze 563: print_ctag(h, tag->tag);
1.2 schwarze 564: h->tags.head = tag->next;
1.1 schwarze 565: free(tag);
566: }
567: }
568:
569:
570: void
571: bufinit(struct html *h)
572: {
573:
574: h->buf[0] = '\0';
575: h->buflen = 0;
576: }
577:
578:
579: void
580: bufcat_style(struct html *h, const char *key, const char *val)
581: {
582:
583: bufcat(h, key);
584: bufncat(h, ":", 1);
585: bufcat(h, val);
586: bufncat(h, ";", 1);
587: }
588:
589:
590: void
591: bufcat(struct html *h, const char *p)
592: {
593:
594: bufncat(h, p, strlen(p));
595: }
596:
597:
598: void
599: buffmt(struct html *h, const char *fmt, ...)
600: {
601: va_list ap;
602:
603: va_start(ap, fmt);
604: (void)vsnprintf(h->buf + (int)h->buflen,
605: BUFSIZ - h->buflen - 1, fmt, ap);
606: va_end(ap);
607: h->buflen = strlen(h->buf);
608: }
609:
610:
611: void
612: bufncat(struct html *h, const char *p, size_t sz)
613: {
614:
615: if (h->buflen + sz > BUFSIZ - 1)
616: sz = BUFSIZ - 1 - h->buflen;
617:
618: (void)strncat(h->buf, p, sz);
619: h->buflen += sz;
620: }
621:
622:
623: void
624: buffmt_includes(struct html *h, const char *name)
625: {
626: const char *p, *pp;
627:
628: pp = h->base_includes;
629:
630: while (NULL != (p = strchr(pp, '%'))) {
631: bufncat(h, pp, (size_t)(p - pp));
632: switch (*(p + 1)) {
633: case('I'):
634: bufcat(h, name);
635: break;
636: default:
637: bufncat(h, p, 2);
638: break;
639: }
640: pp = p + 2;
641: }
642: if (pp)
643: bufcat(h, pp);
644: }
645:
646:
647: void
648: buffmt_man(struct html *h,
649: const char *name, const char *sec)
650: {
651: const char *p, *pp;
652:
653: pp = h->base_man;
654:
655: /* LINTED */
656: while (NULL != (p = strchr(pp, '%'))) {
657: bufncat(h, pp, (size_t)(p - pp));
658: switch (*(p + 1)) {
659: case('S'):
660: bufcat(h, sec ? sec : "1");
661: break;
662: case('N'):
663: buffmt(h, name);
664: break;
665: default:
666: bufncat(h, p, 2);
667: break;
668: }
669: pp = p + 2;
670: }
671: if (pp)
672: bufcat(h, pp);
673: }
674:
675:
676: void
677: bufcat_su(struct html *h, const char *p, const struct roffsu *su)
678: {
679: double v;
680: const char *u;
681:
682: v = su->scale;
683:
684: switch (su->unit) {
685: case (SCALE_CM):
686: u = "cm";
687: break;
688: case (SCALE_IN):
689: u = "in";
690: break;
691: case (SCALE_PC):
692: u = "pc";
693: break;
694: case (SCALE_PT):
695: u = "pt";
696: break;
697: case (SCALE_EM):
698: u = "em";
699: break;
700: case (SCALE_MM):
701: if (0 == (v /= 100))
702: v = 1;
703: u = "em";
704: break;
705: case (SCALE_EN):
706: u = "ex";
707: break;
708: case (SCALE_BU):
709: u = "ex";
710: break;
711: case (SCALE_VS):
712: u = "em";
713: break;
714: default:
715: u = "ex";
716: break;
717: }
718:
719: if (su->pt)
720: buffmt(h, "%s: %f%s;", p, v, u);
721: else
722: /* LINTED */
723: buffmt(h, "%s: %d%s;", p, (int)v, u);
724: }
725:
1.3 schwarze 726:
727: void
728: html_idcat(char *dst, const char *src, int sz)
729: {
730: int ssz;
731:
732: assert(sz);
733:
734: /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
735:
736: for ( ; *dst != '\0' && sz; dst++, sz--)
737: /* Jump to end. */ ;
738:
739: assert(sz > 2);
740:
741: /* We can't start with a number (bah). */
742:
743: *dst++ = 'x';
744: *dst = '\0';
745: sz--;
746:
747: for ( ; *src != '\0' && sz > 1; src++) {
748: ssz = snprintf(dst, (size_t)sz, "%.2x", *src);
749: sz -= ssz;
750: dst += ssz;
751: }
752: }