Annotation of src/usr.bin/mandoc/html.c, Revision 1.12
1.12 ! schwarze 1: /* $Id: html.c,v 1.11 2010/06/27 20:28:56 schwarze Exp $ */
1.1 schwarze 2: /*
1.12 ! schwarze 3: * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
1.1 schwarze 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: struct htmldata {
35: const char *name;
36: int flags;
37: #define HTML_CLRLINE (1 << 0)
38: #define HTML_NOSTACK (1 << 1)
1.6 schwarze 39: #define HTML_AUTOCLOSE (1 << 2) /* Tag has auto-closure. */
1.1 schwarze 40: };
41:
42: static const struct htmldata htmltags[TAG_MAX] = {
43: {"html", HTML_CLRLINE}, /* TAG_HTML */
44: {"head", HTML_CLRLINE}, /* TAG_HEAD */
45: {"body", HTML_CLRLINE}, /* TAG_BODY */
1.6 schwarze 46: {"meta", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_META */
1.1 schwarze 47: {"title", HTML_CLRLINE}, /* TAG_TITLE */
48: {"div", HTML_CLRLINE}, /* TAG_DIV */
49: {"h1", 0}, /* TAG_H1 */
50: {"h2", 0}, /* TAG_H2 */
51: {"span", 0}, /* TAG_SPAN */
1.8 schwarze 52: {"link", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_LINK */
1.6 schwarze 53: {"br", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_BR */
1.1 schwarze 54: {"a", 0}, /* TAG_A */
55: {"table", HTML_CLRLINE}, /* TAG_TABLE */
1.6 schwarze 56: {"col", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_COL */
1.1 schwarze 57: {"tr", HTML_CLRLINE}, /* TAG_TR */
58: {"td", HTML_CLRLINE}, /* TAG_TD */
59: {"li", HTML_CLRLINE}, /* TAG_LI */
60: {"ul", HTML_CLRLINE}, /* TAG_UL */
61: {"ol", HTML_CLRLINE}, /* TAG_OL */
62: };
63:
1.5 schwarze 64: static const char *const htmlfonts[HTMLFONT_MAX] = {
65: "roman",
66: "bold",
67: "italic"
68: };
69:
70: static const char *const htmlattrs[ATTR_MAX] = {
1.1 schwarze 71: "http-equiv",
72: "content",
73: "name",
74: "rel",
75: "href",
76: "type",
77: "media",
78: "class",
79: "style",
80: "width",
81: "valign",
82: "target",
83: "id",
1.3 schwarze 84: "summary",
1.1 schwarze 85: };
86:
1.5 schwarze 87: static void print_spec(struct html *, const char *, size_t);
88: static void print_res(struct html *, const char *, size_t);
89: static void print_ctag(struct html *, enum htmltag);
1.6 schwarze 90: static void print_doctype(struct html *);
91: static void print_xmltype(struct html *);
1.5 schwarze 92: static int print_encode(struct html *, const char *, int);
93: static void print_metaf(struct html *, enum roffdeco);
1.6 schwarze 94: static void print_attr(struct html *,
95: const char *, const char *);
96: static void *ml_alloc(char *, enum htmltype);
1.5 schwarze 97:
98:
1.6 schwarze 99: static void *
100: ml_alloc(char *outopts, enum htmltype type)
1.1 schwarze 101: {
102: struct html *h;
103: const char *toks[4];
104: char *v;
105:
106: toks[0] = "style";
107: toks[1] = "man";
108: toks[2] = "includes";
109: toks[3] = NULL;
110:
1.3 schwarze 111: h = calloc(1, sizeof(struct html));
112: if (NULL == h) {
113: perror(NULL);
114: exit(EXIT_FAILURE);
115: }
1.1 schwarze 116:
1.6 schwarze 117: h->type = type;
1.2 schwarze 118: h->tags.head = NULL;
119: h->ords.head = NULL;
1.3 schwarze 120: h->symtab = chars_init(CHARS_HTML);
1.1 schwarze 121:
122: while (outopts && *outopts)
123: switch (getsubopt(&outopts, UNCONST(toks), &v)) {
124: case (0):
125: h->style = v;
126: break;
127: case (1):
128: h->base_man = v;
129: break;
130: case (2):
131: h->base_includes = v;
132: break;
133: default:
134: break;
135: }
136:
137: return(h);
138: }
139:
1.6 schwarze 140: void *
141: html_alloc(char *outopts)
142: {
143:
144: return(ml_alloc(outopts, HTML_HTML_4_01_STRICT));
145: }
146:
147:
148: void *
149: xhtml_alloc(char *outopts)
150: {
151:
152: return(ml_alloc(outopts, HTML_XHTML_1_0_STRICT));
153: }
154:
1.1 schwarze 155:
156: void
157: html_free(void *p)
158: {
159: struct tag *tag;
160: struct ord *ord;
161: struct html *h;
162:
163: h = (struct html *)p;
164:
1.2 schwarze 165: while ((ord = h->ords.head) != NULL) {
166: h->ords.head = ord->next;
1.1 schwarze 167: free(ord);
168: }
169:
1.2 schwarze 170: while ((tag = h->tags.head) != NULL) {
171: h->tags.head = tag->next;
1.1 schwarze 172: free(tag);
173: }
174:
175: if (h->symtab)
176: chars_free(h->symtab);
177:
178: free(h);
179: }
180:
181:
182: void
183: print_gen_head(struct html *h)
184: {
185: struct htmlpair tag[4];
186:
187: tag[0].key = ATTR_HTTPEQUIV;
188: tag[0].val = "Content-Type";
189: tag[1].key = ATTR_CONTENT;
190: tag[1].val = "text/html; charset=utf-8";
191: print_otag(h, TAG_META, 2, tag);
192:
193: tag[0].key = ATTR_NAME;
194: tag[0].val = "resource-type";
195: tag[1].key = ATTR_CONTENT;
196: tag[1].val = "document";
197: print_otag(h, TAG_META, 2, tag);
198:
199: if (h->style) {
200: tag[0].key = ATTR_REL;
201: tag[0].val = "stylesheet";
202: tag[1].key = ATTR_HREF;
203: tag[1].val = h->style;
204: tag[2].key = ATTR_TYPE;
205: tag[2].val = "text/css";
206: tag[3].key = ATTR_MEDIA;
207: tag[3].val = "all";
208: print_otag(h, TAG_LINK, 4, tag);
209: }
210: }
211:
212:
213: static void
1.5 schwarze 214: print_spec(struct html *h, const char *p, size_t len)
1.1 schwarze 215: {
216: const char *rhs;
217: size_t sz;
218:
1.5 schwarze 219: rhs = chars_a2ascii(h->symtab, p, len, &sz);
1.1 schwarze 220:
221: if (NULL == rhs)
222: return;
1.4 schwarze 223: fwrite(rhs, 1, sz, stdout);
1.1 schwarze 224: }
225:
226:
227: static void
1.5 schwarze 228: print_res(struct html *h, const char *p, size_t len)
1.1 schwarze 229: {
230: const char *rhs;
231: size_t sz;
232:
1.5 schwarze 233: rhs = chars_a2res(h->symtab, p, len, &sz);
1.1 schwarze 234:
235: if (NULL == rhs)
236: return;
1.4 schwarze 237: fwrite(rhs, 1, sz, stdout);
1.1 schwarze 238: }
239:
240:
1.5 schwarze 241: struct tag *
242: print_ofont(struct html *h, enum htmlfont font)
1.1 schwarze 243: {
1.5 schwarze 244: struct htmlpair tag;
1.1 schwarze 245:
1.5 schwarze 246: h->metal = h->metac;
247: h->metac = font;
1.1 schwarze 248:
1.5 schwarze 249: /* FIXME: DECO_ROMAN should just close out preexisting. */
1.1 schwarze 250:
1.5 schwarze 251: if (h->metaf && h->tags.head == h->metaf)
252: print_tagq(h, h->metaf);
1.1 schwarze 253:
1.5 schwarze 254: PAIR_CLASS_INIT(&tag, htmlfonts[font]);
255: h->metaf = print_otag(h, TAG_SPAN, 1, &tag);
256: return(h->metaf);
257: }
1.1 schwarze 258:
259:
1.5 schwarze 260: static void
261: print_metaf(struct html *h, enum roffdeco deco)
262: {
263: enum htmlfont font;
1.1 schwarze 264:
1.5 schwarze 265: switch (deco) {
266: case (DECO_PREVIOUS):
267: font = h->metal;
268: break;
269: case (DECO_ITALIC):
270: font = HTMLFONT_ITALIC;
271: break;
272: case (DECO_BOLD):
273: font = HTMLFONT_BOLD;
274: break;
275: case (DECO_ROMAN):
276: font = HTMLFONT_NONE;
277: break;
278: default:
279: abort();
280: /* NOTREACHED */
1.1 schwarze 281: }
282:
1.5 schwarze 283: (void)print_ofont(h, font);
1.1 schwarze 284: }
285:
286:
1.5 schwarze 287: static int
288: print_encode(struct html *h, const char *p, int norecurse)
1.1 schwarze 289: {
1.4 schwarze 290: size_t sz;
1.5 schwarze 291: int len, nospace;
292: const char *seq;
293: enum roffdeco deco;
1.9 schwarze 294: static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' };
1.5 schwarze 295:
296: nospace = 0;
1.1 schwarze 297:
298: for (; *p; p++) {
1.9 schwarze 299: sz = strcspn(p, rejs);
1.4 schwarze 300:
301: fwrite(p, 1, sz, stdout);
302: p += /* LINTED */
303: sz;
304:
1.5 schwarze 305: if ('<' == *p) {
306: printf("<");
307: continue;
308: } else if ('>' == *p) {
309: printf(">");
310: continue;
311: } else if ('&' == *p) {
312: printf("&");
1.1 schwarze 313: continue;
1.9 schwarze 314: } else if (ASCII_HYPH == *p) {
315: /*
316: * Note: "soft hyphens" aren't graphically
317: * displayed when not breaking the text; we want
318: * them to be displayed.
319: */
320: /*printf("­");*/
321: putchar('-');
322: continue;
1.4 schwarze 323: } else if ('\0' == *p)
324: break;
325:
1.5 schwarze 326: seq = ++p;
327: len = a2roffdeco(&deco, &seq, &sz);
328:
329: switch (deco) {
330: case (DECO_RESERVED):
331: print_res(h, seq, sz);
332: break;
333: case (DECO_SPECIAL):
334: print_spec(h, seq, sz);
335: break;
336: case (DECO_PREVIOUS):
337: /* FALLTHROUGH */
338: case (DECO_BOLD):
339: /* FALLTHROUGH */
340: case (DECO_ITALIC):
341: /* FALLTHROUGH */
342: case (DECO_ROMAN):
343: if (norecurse)
344: break;
345: print_metaf(h, deco);
346: break;
347: default:
348: break;
349: }
350:
351: p += len - 1;
352:
353: if (DECO_NOSPACE == deco && '\0' == *(p + 1))
354: nospace = 1;
1.1 schwarze 355: }
1.5 schwarze 356:
357: return(nospace);
1.1 schwarze 358: }
359:
360:
1.6 schwarze 361: static void
362: print_attr(struct html *h, const char *key, const char *val)
363: {
364: printf(" %s=\"", key);
365: (void)print_encode(h, val, 1);
366: putchar('\"');
367: }
368:
369:
1.1 schwarze 370: struct tag *
371: print_otag(struct html *h, enum htmltag tag,
372: int sz, const struct htmlpair *p)
373: {
374: int i;
375: struct tag *t;
376:
1.6 schwarze 377: /* Push this tags onto the stack of open scopes. */
378:
1.1 schwarze 379: if ( ! (HTML_NOSTACK & htmltags[tag].flags)) {
1.3 schwarze 380: t = malloc(sizeof(struct tag));
381: if (NULL == t) {
382: perror(NULL);
383: exit(EXIT_FAILURE);
384: }
1.1 schwarze 385: t->tag = tag;
1.2 schwarze 386: t->next = h->tags.head;
387: h->tags.head = t;
1.1 schwarze 388: } else
389: t = NULL;
390:
391: if ( ! (HTML_NOSPACE & h->flags))
1.12 ! schwarze 392: if ( ! (HTML_CLRLINE & htmltags[tag].flags)) {
! 393: /* Manage keeps! */
! 394: if ( ! (HTML_KEEP & h->flags)) {
! 395: if (HTML_PREKEEP & h->flags)
! 396: h->flags |= HTML_KEEP;
! 397: putchar(' ');
! 398: } else
! 399: printf(" ");
! 400: }
1.1 schwarze 401:
1.6 schwarze 402: /* Print out the tag name and attributes. */
403:
1.1 schwarze 404: printf("<%s", htmltags[tag].name);
1.6 schwarze 405: for (i = 0; i < sz; i++)
406: print_attr(h, htmlattrs[p[i].key], p[i].val);
407:
408: /* Add non-overridable attributes. */
409:
410: if (TAG_HTML == tag && HTML_XHTML_1_0_STRICT == h->type) {
411: print_attr(h, "xmlns", "http://www.w3.org/1999/xhtml");
412: print_attr(h, "xml:lang", "en");
413: print_attr(h, "lang", "en");
1.1 schwarze 414: }
1.6 schwarze 415:
416: /* Accomodate for XML "well-formed" singleton escaping. */
417:
418: if (HTML_AUTOCLOSE & htmltags[tag].flags)
419: switch (h->type) {
420: case (HTML_XHTML_1_0_STRICT):
421: putchar('/');
422: break;
423: default:
424: break;
425: }
426:
1.4 schwarze 427: putchar('>');
1.1 schwarze 428:
429: h->flags |= HTML_NOSPACE;
430: return(t);
431: }
432:
433:
434: static void
435: print_ctag(struct html *h, enum htmltag tag)
436: {
437:
438: printf("</%s>", htmltags[tag].name);
1.3 schwarze 439: if (HTML_CLRLINE & htmltags[tag].flags) {
1.1 schwarze 440: h->flags |= HTML_NOSPACE;
1.4 schwarze 441: putchar('\n');
1.5 schwarze 442: }
1.1 schwarze 443: }
444:
445:
446: void
1.6 schwarze 447: print_gen_decls(struct html *h)
448: {
449:
450: print_xmltype(h);
451: print_doctype(h);
452: }
453:
454:
455: static void
456: print_xmltype(struct html *h)
457: {
458:
1.9 schwarze 459: if (HTML_XHTML_1_0_STRICT == h->type)
460: printf("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
1.6 schwarze 461: }
462:
463:
464: static void
465: print_doctype(struct html *h)
1.1 schwarze 466: {
1.6 schwarze 467: const char *doctype;
468: const char *dtd;
469: const char *name;
470:
471: switch (h->type) {
472: case (HTML_HTML_4_01_STRICT):
473: name = "HTML";
474: doctype = "-//W3C//DTD HTML 4.01//EN";
475: dtd = "http://www.w3.org/TR/html4/strict.dtd";
476: break;
477: default:
478: name = "html";
479: doctype = "-//W3C//DTD XHTML 1.0 Strict//EN";
480: dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
481: break;
482: }
483:
484: printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
485: name, doctype, dtd);
1.1 schwarze 486: }
487:
488:
489: void
1.12 ! schwarze 490: print_text(struct html *h, const char *word)
1.1 schwarze 491: {
492:
1.12 ! schwarze 493: if (word[0] && '\0' == word[1])
! 494: switch (word[0]) {
1.1 schwarze 495: case('.'):
496: /* FALLTHROUGH */
497: case(','):
498: /* FALLTHROUGH */
499: case(';'):
500: /* FALLTHROUGH */
501: case(':'):
502: /* FALLTHROUGH */
503: case('?'):
504: /* FALLTHROUGH */
505: case('!'):
506: /* FALLTHROUGH */
507: case(')'):
508: /* FALLTHROUGH */
509: case(']'):
510: if ( ! (HTML_IGNDELIM & h->flags))
511: h->flags |= HTML_NOSPACE;
512: break;
513: default:
514: break;
515: }
516:
1.12 ! schwarze 517: if ( ! (HTML_NOSPACE & h->flags)) {
! 518: /* Manage keeps! */
! 519: if ( ! (HTML_KEEP & h->flags)) {
! 520: if (HTML_PREKEEP & h->flags)
! 521: h->flags |= HTML_KEEP;
! 522: putchar(' ');
! 523: } else
! 524: printf(" ");
! 525: }
1.1 schwarze 526:
1.12 ! schwarze 527: assert(word);
! 528: if ( ! print_encode(h, word, 0))
1.5 schwarze 529: h->flags &= ~HTML_NOSPACE;
1.1 schwarze 530:
1.8 schwarze 531: /*
532: * Note that we don't process the pipe: the parser sees it as
533: * punctuation, but we don't in terms of typography.
534: */
1.12 ! schwarze 535: if (word[0] && '\0' == word[1])
! 536: switch (word[0]) {
1.1 schwarze 537: case('('):
538: /* FALLTHROUGH */
539: case('['):
540: h->flags |= HTML_NOSPACE;
541: break;
542: default:
543: break;
544: }
545: }
546:
547:
548: void
549: print_tagq(struct html *h, const struct tag *until)
550: {
551: struct tag *tag;
552:
1.2 schwarze 553: while ((tag = h->tags.head) != NULL) {
1.5 schwarze 554: if (tag == h->metaf)
555: h->metaf = NULL;
1.1 schwarze 556: print_ctag(h, tag->tag);
1.2 schwarze 557: h->tags.head = tag->next;
1.1 schwarze 558: free(tag);
559: if (until && tag == until)
560: return;
561: }
562: }
563:
564:
565: void
566: print_stagq(struct html *h, const struct tag *suntil)
567: {
568: struct tag *tag;
569:
1.2 schwarze 570: while ((tag = h->tags.head) != NULL) {
1.1 schwarze 571: if (suntil && tag == suntil)
572: return;
1.5 schwarze 573: if (tag == h->metaf)
574: h->metaf = NULL;
1.1 schwarze 575: print_ctag(h, tag->tag);
1.2 schwarze 576: h->tags.head = tag->next;
1.1 schwarze 577: free(tag);
578: }
579: }
580:
581:
582: void
583: bufinit(struct html *h)
584: {
585:
586: h->buf[0] = '\0';
587: h->buflen = 0;
588: }
589:
590:
591: void
592: bufcat_style(struct html *h, const char *key, const char *val)
593: {
594:
595: bufcat(h, key);
596: bufncat(h, ":", 1);
597: bufcat(h, val);
598: bufncat(h, ";", 1);
599: }
600:
601:
602: void
603: bufcat(struct html *h, const char *p)
604: {
605:
606: bufncat(h, p, strlen(p));
607: }
608:
609:
610: void
611: buffmt(struct html *h, const char *fmt, ...)
612: {
613: va_list ap;
614:
615: va_start(ap, fmt);
616: (void)vsnprintf(h->buf + (int)h->buflen,
617: BUFSIZ - h->buflen - 1, fmt, ap);
618: va_end(ap);
619: h->buflen = strlen(h->buf);
620: }
621:
622:
623: void
624: bufncat(struct html *h, const char *p, size_t sz)
625: {
626:
627: if (h->buflen + sz > BUFSIZ - 1)
628: sz = BUFSIZ - 1 - h->buflen;
629:
630: (void)strncat(h->buf, p, sz);
631: h->buflen += sz;
632: }
633:
634:
635: void
636: buffmt_includes(struct html *h, const char *name)
637: {
638: const char *p, *pp;
639:
640: pp = h->base_includes;
641:
642: while (NULL != (p = strchr(pp, '%'))) {
643: bufncat(h, pp, (size_t)(p - pp));
644: switch (*(p + 1)) {
645: case('I'):
646: bufcat(h, name);
647: break;
648: default:
649: bufncat(h, p, 2);
650: break;
651: }
652: pp = p + 2;
653: }
654: if (pp)
655: bufcat(h, pp);
656: }
657:
658:
659: void
660: buffmt_man(struct html *h,
661: const char *name, const char *sec)
662: {
663: const char *p, *pp;
664:
665: pp = h->base_man;
666:
667: /* LINTED */
668: while (NULL != (p = strchr(pp, '%'))) {
669: bufncat(h, pp, (size_t)(p - pp));
670: switch (*(p + 1)) {
671: case('S'):
672: bufcat(h, sec ? sec : "1");
673: break;
674: case('N'):
675: buffmt(h, name);
676: break;
677: default:
678: bufncat(h, p, 2);
679: break;
680: }
681: pp = p + 2;
682: }
683: if (pp)
684: bufcat(h, pp);
685: }
686:
687:
688: void
689: bufcat_su(struct html *h, const char *p, const struct roffsu *su)
690: {
691: double v;
692: const char *u;
693:
694: v = su->scale;
695:
696: switch (su->unit) {
697: case (SCALE_CM):
698: u = "cm";
699: break;
700: case (SCALE_IN):
701: u = "in";
702: break;
703: case (SCALE_PC):
704: u = "pc";
705: break;
706: case (SCALE_PT):
707: u = "pt";
708: break;
709: case (SCALE_EM):
710: u = "em";
711: break;
712: case (SCALE_MM):
713: if (0 == (v /= 100))
714: v = 1;
715: u = "em";
716: break;
717: case (SCALE_EN):
718: u = "ex";
719: break;
720: case (SCALE_BU):
721: u = "ex";
722: break;
723: case (SCALE_VS):
724: u = "em";
725: break;
726: default:
727: u = "ex";
728: break;
729: }
730:
1.11 schwarze 731: /*
732: * XXX: the CSS spec isn't clear as to which types accept
733: * integer or real numbers, so we just make them all decimals.
734: */
735: buffmt(h, "%s: %.2f%s;", p, v, u);
1.1 schwarze 736: }
737:
1.3 schwarze 738:
739: void
740: html_idcat(char *dst, const char *src, int sz)
741: {
742: int ssz;
743:
744: assert(sz);
745:
746: /* Cf. <http://www.w3.org/TR/html4/types.html#h-6.2>. */
747:
748: for ( ; *dst != '\0' && sz; dst++, sz--)
749: /* Jump to end. */ ;
750:
751: assert(sz > 2);
752:
753: /* We can't start with a number (bah). */
754:
755: *dst++ = 'x';
756: *dst = '\0';
757: sz--;
758:
759: for ( ; *src != '\0' && sz > 1; src++) {
760: ssz = snprintf(dst, (size_t)sz, "%.2x", *src);
761: sz -= ssz;
762: dst += ssz;
763: }
764: }