Annotation of src/usr.bin/mandoc/html.c, Revision 1.114
1.114 ! schwarze 1: /* $OpenBSD: html.c,v 1.113 2018/11/23 19:15:32 schwarze Exp $ */
1.1 schwarze 2: /*
1.42 schwarze 3: * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.91 schwarze 4: * Copyright (c) 2011-2015, 2017, 2018 Ingo Schwarze <schwarze@openbsd.org>
1.1 schwarze 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: *
1.56 schwarze 10: * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.1 schwarze 11: * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.56 schwarze 12: * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.1 schwarze 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: #include <sys/types.h>
1.110 schwarze 19: #include <sys/stat.h>
1.1 schwarze 20:
21: #include <assert.h>
1.3 schwarze 22: #include <ctype.h>
1.4 schwarze 23: #include <stdarg.h>
1.99 schwarze 24: #include <stddef.h>
1.1 schwarze 25: #include <stdio.h>
26: #include <stdint.h>
27: #include <stdlib.h>
28: #include <string.h>
29: #include <unistd.h>
30:
1.80 schwarze 31: #include "mandoc_aux.h"
1.99 schwarze 32: #include "mandoc_ohash.h"
1.9 schwarze 33: #include "mandoc.h"
1.80 schwarze 34: #include "roff.h"
1.1 schwarze 35: #include "out.h"
36: #include "html.h"
1.56 schwarze 37: #include "manconf.h"
1.1 schwarze 38: #include "main.h"
39:
40: struct htmldata {
41: const char *name;
42: int flags;
1.66 schwarze 43: #define HTML_NOSTACK (1 << 0)
44: #define HTML_AUTOCLOSE (1 << 1)
45: #define HTML_NLBEFORE (1 << 2)
46: #define HTML_NLBEGIN (1 << 3)
47: #define HTML_NLEND (1 << 4)
48: #define HTML_NLAFTER (1 << 5)
49: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
50: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
51: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
52: #define HTML_INDENT (1 << 6)
53: #define HTML_NOINDENT (1 << 7)
1.1 schwarze 54: };
55:
56: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 57: {"html", HTML_NLALL},
58: {"head", HTML_NLALL | HTML_INDENT},
59: {"body", HTML_NLALL},
60: {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
61: {"title", HTML_NLAROUND},
62: {"div", HTML_NLAROUND},
1.95 schwarze 63: {"div", 0},
1.66 schwarze 64: {"h1", HTML_NLAROUND},
65: {"h2", HTML_NLAROUND},
66: {"span", 0},
67: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
68: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
69: {"a", 0},
70: {"table", HTML_NLALL | HTML_INDENT},
71: {"tr", HTML_NLALL | HTML_INDENT},
72: {"td", HTML_NLAROUND},
73: {"li", HTML_NLAROUND | HTML_INDENT},
74: {"ul", HTML_NLALL | HTML_INDENT},
75: {"ol", HTML_NLALL | HTML_INDENT},
76: {"dl", HTML_NLALL | HTML_INDENT},
77: {"dt", HTML_NLAROUND},
78: {"dd", HTML_NLAROUND | HTML_INDENT},
79: {"pre", HTML_NLALL | HTML_NOINDENT},
1.77 schwarze 80: {"var", 0},
1.76 schwarze 81: {"cite", 0},
1.66 schwarze 82: {"b", 0},
83: {"i", 0},
84: {"code", 0},
85: {"small", 0},
86: {"style", HTML_NLALL | HTML_INDENT},
87: {"math", HTML_NLALL | HTML_INDENT},
88: {"mrow", 0},
89: {"mi", 0},
1.85 schwarze 90: {"mn", 0},
1.66 schwarze 91: {"mo", 0},
92: {"msup", 0},
93: {"msub", 0},
94: {"msubsup", 0},
95: {"mfrac", 0},
96: {"msqrt", 0},
97: {"mfenced", 0},
98: {"mtable", 0},
99: {"mtr", 0},
100: {"mtd", 0},
101: {"munderover", 0},
102: {"munder", 0},
103: {"mover", 0},
1.5 schwarze 104: };
105:
1.99 schwarze 106: /* Avoid duplicate HTML id= attributes. */
107: static struct ohash id_unique;
108:
1.67 schwarze 109: static void print_byte(struct html *, char);
110: static void print_endword(struct html *);
111: static void print_indent(struct html *);
112: static void print_word(struct html *, const char *);
113:
1.54 schwarze 114: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 115: static int print_escape(struct html *, char);
1.65 schwarze 116: static int print_encode(struct html *, const char *, const char *, int);
117: static void print_href(struct html *, const char *, const char *, int);
1.26 schwarze 118: static void print_metaf(struct html *, enum mandoc_esc);
1.5 schwarze 119:
1.35 schwarze 120:
1.50 schwarze 121: void *
1.61 schwarze 122: html_alloc(const struct manoutput *outopts)
1.1 schwarze 123: {
124: struct html *h;
125:
1.24 schwarze 126: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 127:
1.74 schwarze 128: h->tag = NULL;
1.56 schwarze 129: h->style = outopts->style;
1.110 schwarze 130: if ((h->base_man1 = outopts->man) == NULL)
131: h->base_man2 = NULL;
132: else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL)
133: *h->base_man2++ = '\0';
1.56 schwarze 134: h->base_includes = outopts->includes;
135: if (outopts->fragment)
136: h->oflags |= HTML_FRAGMENT;
1.111 schwarze 137: if (outopts->toc)
138: h->oflags |= HTML_TOC;
1.1 schwarze 139:
1.99 schwarze 140: mandoc_ohash_init(&id_unique, 4, 0);
141:
1.58 schwarze 142: return h;
1.1 schwarze 143: }
144:
145: void
146: html_free(void *p)
147: {
148: struct tag *tag;
149: struct html *h;
1.99 schwarze 150: char *cp;
151: unsigned int slot;
1.1 schwarze 152:
153: h = (struct html *)p;
1.74 schwarze 154: while ((tag = h->tag) != NULL) {
155: h->tag = tag->next;
1.1 schwarze 156: free(tag);
157: }
1.99 schwarze 158: free(h);
1.1 schwarze 159:
1.99 schwarze 160: cp = ohash_first(&id_unique, &slot);
161: while (cp != NULL) {
162: free(cp);
163: cp = ohash_next(&id_unique, &slot);
164: }
165: ohash_delete(&id_unique);
1.1 schwarze 166: }
167:
168: void
169: print_gen_head(struct html *h)
170: {
1.42 schwarze 171: struct tag *t;
172:
1.64 schwarze 173: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 174: if (h->style != NULL) {
175: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
176: h->style, "type", "text/css", "media", "all");
177: return;
178: }
1.1 schwarze 179:
1.42 schwarze 180: /*
1.92 schwarze 181: * Print a minimal embedded style sheet.
1.42 schwarze 182: */
1.66 schwarze 183:
1.64 schwarze 184: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 185: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 186: print_endline(h);
1.66 schwarze 187: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 188: print_endline(h);
1.66 schwarze 189: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 190: print_endline(h);
1.68 schwarze 191: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.95 schwarze 192: print_endline(h);
193: print_text(h, "div.Nd, div.Bf, div.Op { display: inline; }");
1.96 schwarze 194: print_endline(h);
1.97 schwarze 195: print_text(h, "span.Pa, span.Ad { font-style: italic; }");
1.98 schwarze 196: print_endline(h);
197: print_text(h, "span.Ms { font-weight: bold; }");
1.94 schwarze 198: print_endline(h);
199: print_text(h, "dl.Bl-diag ");
200: print_byte(h, '>');
201: print_text(h, " dt { font-weight: bold; }");
1.93 schwarze 202: print_endline(h);
203: print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, "
204: "code.In, code.Fd, code.Fn,");
205: print_endline(h);
206: print_text(h, "code.Cd { font-weight: bold; "
207: "font-family: inherit; }");
1.42 schwarze 208: print_tagq(h, t);
1.1 schwarze 209: }
210:
1.5 schwarze 211: static void
1.26 schwarze 212: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 213: {
214: enum htmlfont font;
1.1 schwarze 215:
1.5 schwarze 216: switch (deco) {
1.35 schwarze 217: case ESCAPE_FONTPREV:
1.5 schwarze 218: font = h->metal;
219: break;
1.35 schwarze 220: case ESCAPE_FONTITALIC:
1.5 schwarze 221: font = HTMLFONT_ITALIC;
222: break;
1.35 schwarze 223: case ESCAPE_FONTBOLD:
1.5 schwarze 224: font = HTMLFONT_BOLD;
225: break;
1.35 schwarze 226: case ESCAPE_FONTBI:
1.31 schwarze 227: font = HTMLFONT_BI;
228: break;
1.112 schwarze 229: case ESCAPE_FONTCW:
230: font = HTMLFONT_CW;
231: break;
1.35 schwarze 232: case ESCAPE_FONT:
233: case ESCAPE_FONTROMAN:
1.5 schwarze 234: font = HTMLFONT_NONE;
235: break;
236: default:
237: abort();
1.1 schwarze 238: }
239:
1.20 schwarze 240: if (h->metaf) {
241: print_tagq(h, h->metaf);
242: h->metaf = NULL;
243: }
244:
245: h->metal = h->metac;
246: h->metac = font;
247:
1.31 schwarze 248: switch (font) {
1.35 schwarze 249: case HTMLFONT_ITALIC:
1.64 schwarze 250: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 251: break;
1.35 schwarze 252: case HTMLFONT_BOLD:
1.64 schwarze 253: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 254: break;
1.35 schwarze 255: case HTMLFONT_BI:
1.64 schwarze 256: h->metaf = print_otag(h, TAG_B, "");
257: print_otag(h, TAG_I, "");
1.31 schwarze 258: break;
1.112 schwarze 259: case HTMLFONT_CW:
260: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
261: break;
1.31 schwarze 262: default:
263: break;
264: }
1.80 schwarze 265: }
266:
267: char *
1.99 schwarze 268: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 269: {
270: const struct roff_node *nch;
1.99 schwarze 271: char *buf, *bufs, *cp;
272: unsigned int slot;
273: int suffix;
1.80 schwarze 274:
275: for (nch = n->child; nch != NULL; nch = nch->next)
276: if (nch->type != ROFFT_TEXT)
277: return NULL;
278:
279: buf = NULL;
280: deroff(&buf, n);
1.90 schwarze 281: if (buf == NULL)
282: return NULL;
1.80 schwarze 283:
1.100 schwarze 284: /*
285: * In ID attributes, only use ASCII characters that are
286: * permitted in URL-fragment strings according to the
287: * explicit list at:
288: * https://url.spec.whatwg.org/#url-fragment-string
289: */
1.80 schwarze 290:
291: for (cp = buf; *cp != '\0'; cp++)
1.100 schwarze 292: if (isalnum((unsigned char)*cp) == 0 &&
293: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.80 schwarze 294: *cp = '_';
295:
1.99 schwarze 296: if (unique == 0)
297: return buf;
298:
299: /* Avoid duplicate HTML id= attributes. */
300:
301: bufs = NULL;
302: suffix = 1;
303: slot = ohash_qlookup(&id_unique, buf);
304: cp = ohash_find(&id_unique, slot);
305: if (cp != NULL) {
306: while (cp != NULL) {
307: free(bufs);
308: if (++suffix > 127) {
309: free(buf);
310: return NULL;
311: }
312: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
313: slot = ohash_qlookup(&id_unique, bufs);
314: cp = ohash_find(&id_unique, slot);
315: }
316: free(buf);
317: buf = bufs;
318: }
319: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 320: return buf;
1.1 schwarze 321: }
322:
1.5 schwarze 323: static int
1.67 schwarze 324: print_escape(struct html *h, char c)
1.38 schwarze 325: {
326:
327: switch (c) {
328: case '<':
1.67 schwarze 329: print_word(h, "<");
1.38 schwarze 330: break;
331: case '>':
1.67 schwarze 332: print_word(h, ">");
1.38 schwarze 333: break;
334: case '&':
1.67 schwarze 335: print_word(h, "&");
1.38 schwarze 336: break;
337: case '"':
1.67 schwarze 338: print_word(h, """);
1.38 schwarze 339: break;
340: case ASCII_NBRSP:
1.67 schwarze 341: print_word(h, " ");
1.38 schwarze 342: break;
343: case ASCII_HYPH:
1.67 schwarze 344: print_byte(h, '-');
1.59 schwarze 345: break;
1.38 schwarze 346: case ASCII_BREAK:
347: break;
348: default:
1.58 schwarze 349: return 0;
1.38 schwarze 350: }
1.58 schwarze 351: return 1;
1.38 schwarze 352: }
353:
354: static int
1.65 schwarze 355: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 356: {
1.67 schwarze 357: char numbuf[16];
1.84 schwarze 358: struct tag *t;
359: const char *seq;
1.4 schwarze 360: size_t sz;
1.84 schwarze 361: int c, len, breakline, nospace;
1.26 schwarze 362: enum mandoc_esc esc;
1.84 schwarze 363: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 364: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 365:
1.65 schwarze 366: if (pend == NULL)
367: pend = strchr(p, '\0');
368:
1.84 schwarze 369: breakline = 0;
1.5 schwarze 370: nospace = 0;
1.1 schwarze 371:
1.65 schwarze 372: while (p < pend) {
1.30 schwarze 373: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
374: h->flags &= ~HTML_SKIPCHAR;
375: p++;
376: continue;
377: }
378:
1.67 schwarze 379: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 380: print_byte(h, *p);
381:
382: if (breakline &&
383: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
384: t = print_otag(h, TAG_DIV, "");
385: print_text(h, "\\~");
386: print_tagq(h, t);
387: breakline = 0;
388: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
389: p++;
390: continue;
391: }
1.4 schwarze 392:
1.65 schwarze 393: if (p >= pend)
1.26 schwarze 394: break;
395:
1.84 schwarze 396: if (*p == ' ') {
397: print_endword(h);
398: p++;
399: continue;
400: }
401:
1.67 schwarze 402: if (print_escape(h, *p++))
1.33 schwarze 403: continue;
1.4 schwarze 404:
1.26 schwarze 405: esc = mandoc_escape(&p, &seq, &len);
406: if (ESCAPE_ERROR == esc)
407: break;
1.5 schwarze 408:
1.26 schwarze 409: switch (esc) {
1.35 schwarze 410: case ESCAPE_FONT:
411: case ESCAPE_FONTPREV:
412: case ESCAPE_FONTBOLD:
413: case ESCAPE_FONTITALIC:
414: case ESCAPE_FONTBI:
1.112 schwarze 415: case ESCAPE_FONTCW:
1.35 schwarze 416: case ESCAPE_FONTROMAN:
1.113 schwarze 417: if (0 == norecurse) {
418: h->flags |= HTML_NOSPACE;
1.30 schwarze 419: print_metaf(h, esc);
1.113 schwarze 420: h->flags &= ~HTML_NOSPACE;
421: }
1.30 schwarze 422: continue;
1.35 schwarze 423: case ESCAPE_SKIPCHAR:
1.30 schwarze 424: h->flags |= HTML_SKIPCHAR;
425: continue;
426: default:
427: break;
428: }
429:
430: if (h->flags & HTML_SKIPCHAR) {
431: h->flags &= ~HTML_SKIPCHAR;
432: continue;
433: }
434:
435: switch (esc) {
1.35 schwarze 436: case ESCAPE_UNICODE:
1.38 schwarze 437: /* Skip past "u" header. */
1.26 schwarze 438: c = mchars_num2uc(seq + 1, len - 1);
439: break;
1.35 schwarze 440: case ESCAPE_NUMBERED:
1.26 schwarze 441: c = mchars_num2char(seq, len);
1.51 schwarze 442: if (c < 0)
443: continue;
1.26 schwarze 444: break;
1.35 schwarze 445: case ESCAPE_SPECIAL:
1.61 schwarze 446: c = mchars_spec2cp(seq, len);
1.51 schwarze 447: if (c <= 0)
448: continue;
1.26 schwarze 449: break;
1.109 schwarze 450: case ESCAPE_DEVICE:
451: print_word(h, "html");
452: continue;
1.84 schwarze 453: case ESCAPE_BREAK:
454: breakline = 1;
455: continue;
1.35 schwarze 456: case ESCAPE_NOSPACE:
1.26 schwarze 457: if ('\0' == *p)
458: nospace = 1;
1.49 schwarze 459: continue;
1.55 schwarze 460: case ESCAPE_OVERSTRIKE:
461: if (len == 0)
462: continue;
463: c = seq[len - 1];
464: break;
1.5 schwarze 465: default:
1.49 schwarze 466: continue;
1.5 schwarze 467: }
1.51 schwarze 468: if ((c < 0x20 && c != 0x09) ||
469: (c > 0x7E && c < 0xA0))
1.49 schwarze 470: c = 0xFFFD;
1.67 schwarze 471: if (c > 0x7E) {
1.86 bentley 472: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 473: print_word(h, numbuf);
474: } else if (print_escape(h, c) == 0)
475: print_byte(h, c);
1.1 schwarze 476: }
1.5 schwarze 477:
1.58 schwarze 478: return nospace;
1.1 schwarze 479: }
480:
1.6 schwarze 481: static void
1.65 schwarze 482: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 483: {
1.110 schwarze 484: struct stat sb;
1.65 schwarze 485: const char *p, *pp;
1.110 schwarze 486: char *filename;
487:
488: if (man) {
489: pp = h->base_man1;
490: if (h->base_man2 != NULL) {
491: mandoc_asprintf(&filename, "%s.%s", name, sec);
492: if (stat(filename, &sb) == -1)
493: pp = h->base_man2;
494: free(filename);
495: }
496: } else
497: pp = h->base_includes;
1.65 schwarze 498:
499: while ((p = strchr(pp, '%')) != NULL) {
500: print_encode(h, pp, p, 1);
501: if (man && p[1] == 'S') {
502: if (sec == NULL)
1.67 schwarze 503: print_byte(h, '1');
1.65 schwarze 504: else
505: print_encode(h, sec, NULL, 1);
506: } else if ((man && p[1] == 'N') ||
507: (man == 0 && p[1] == 'I'))
508: print_encode(h, name, NULL, 1);
509: else
510: print_encode(h, p, p + 2, 1);
511: pp = p + 2;
512: }
513: if (*pp != '\0')
514: print_encode(h, pp, NULL, 1);
1.6 schwarze 515: }
516:
1.1 schwarze 517: struct tag *
1.64 schwarze 518: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 519: {
1.64 schwarze 520: va_list ap;
1.1 schwarze 521: struct tag *t;
1.65 schwarze 522: const char *attr;
1.73 schwarze 523: char *arg1, *arg2;
1.114 ! schwarze 524: int style_written, tflags;
1.66 schwarze 525:
526: tflags = htmltags[tag].flags;
1.1 schwarze 527:
1.74 schwarze 528: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 529:
1.66 schwarze 530: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 531: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 532: t->tag = tag;
1.74 schwarze 533: t->next = h->tag;
534: h->tag = t;
1.1 schwarze 535: } else
536: t = NULL;
537:
1.66 schwarze 538: if (tflags & HTML_NLBEFORE)
1.67 schwarze 539: print_endline(h);
540: if (h->col == 0)
541: print_indent(h);
1.66 schwarze 542: else if ((h->flags & HTML_NOSPACE) == 0) {
543: if (h->flags & HTML_KEEP)
1.86 bentley 544: print_word(h, " ");
1.66 schwarze 545: else {
546: if (h->flags & HTML_PREKEEP)
547: h->flags |= HTML_KEEP;
1.67 schwarze 548: print_endword(h);
1.12 schwarze 549: }
1.66 schwarze 550: }
1.1 schwarze 551:
1.13 schwarze 552: if ( ! (h->flags & HTML_NONOSPACE))
553: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 554: else
555: h->flags |= HTML_NOSPACE;
1.13 schwarze 556:
1.6 schwarze 557: /* Print out the tag name and attributes. */
558:
1.67 schwarze 559: print_byte(h, '<');
560: print_word(h, htmltags[tag].name);
1.64 schwarze 561:
562: va_start(ap, fmt);
563:
1.114 ! schwarze 564: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 565:
1.108 schwarze 566: /* Parse attributes and arguments. */
1.73 schwarze 567:
568: arg1 = va_arg(ap, char *);
1.108 schwarze 569: arg2 = NULL;
1.64 schwarze 570: switch (*fmt++) {
571: case 'c':
1.65 schwarze 572: attr = "class";
1.64 schwarze 573: break;
574: case 'h':
1.65 schwarze 575: attr = "href";
1.64 schwarze 576: break;
577: case 'i':
1.65 schwarze 578: attr = "id";
1.64 schwarze 579: break;
580: case '?':
1.73 schwarze 581: attr = arg1;
582: arg1 = va_arg(ap, char *);
1.64 schwarze 583: break;
584: default:
585: abort();
586: }
1.73 schwarze 587: if (*fmt == 'M')
588: arg2 = va_arg(ap, char *);
589: if (arg1 == NULL)
590: continue;
591:
1.108 schwarze 592: /* Print the attributes. */
1.73 schwarze 593:
1.67 schwarze 594: print_byte(h, ' ');
595: print_word(h, attr);
596: print_byte(h, '=');
597: print_byte(h, '"');
1.65 schwarze 598: switch (*fmt) {
1.78 schwarze 599: case 'I':
600: print_href(h, arg1, NULL, 0);
601: fmt++;
602: break;
1.65 schwarze 603: case 'M':
1.73 schwarze 604: print_href(h, arg1, arg2, 1);
1.65 schwarze 605: fmt++;
606: break;
1.78 schwarze 607: case 'R':
608: print_byte(h, '#');
609: print_encode(h, arg1, NULL, 1);
1.65 schwarze 610: fmt++;
611: break;
1.78 schwarze 612: case 'T':
613: print_encode(h, arg1, NULL, 1);
614: print_word(h, "\" title=\"");
615: print_encode(h, arg1, NULL, 1);
1.65 schwarze 616: fmt++;
1.78 schwarze 617: break;
1.65 schwarze 618: default:
1.114 ! schwarze 619: print_encode(h, arg1, NULL, 1);
1.65 schwarze 620: break;
621: }
1.67 schwarze 622: print_byte(h, '"');
1.64 schwarze 623: }
1.114 ! schwarze 624:
! 625: style_written = 0;
! 626: while (*fmt++ == 's') {
! 627: arg1 = va_arg(ap, char *);
! 628: arg2 = va_arg(ap, char *);
! 629: if (arg2 == NULL)
! 630: continue;
! 631: print_byte(h, ' ');
! 632: if (style_written == 0) {
! 633: print_word(h, "style=\"");
! 634: style_written = 1;
! 635: }
! 636: print_word(h, arg1);
! 637: print_byte(h, ':');
! 638: print_byte(h, ' ');
! 639: print_word(h, arg2);
! 640: print_byte(h, ';');
! 641: }
! 642: if (style_written)
! 643: print_byte(h, '"');
! 644:
1.64 schwarze 645: va_end(ap);
1.6 schwarze 646:
1.42 schwarze 647: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 648:
649: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 650: print_byte(h, '/');
1.6 schwarze 651:
1.67 schwarze 652: print_byte(h, '>');
1.1 schwarze 653:
1.66 schwarze 654: if (tflags & HTML_NLBEGIN)
1.67 schwarze 655: print_endline(h);
1.66 schwarze 656: else
657: h->flags |= HTML_NOSPACE;
1.18 schwarze 658:
1.66 schwarze 659: if (tflags & HTML_INDENT)
660: h->indent++;
661: if (tflags & HTML_NOINDENT)
662: h->noindent++;
1.18 schwarze 663:
1.58 schwarze 664: return t;
1.1 schwarze 665: }
666:
667: static void
1.54 schwarze 668: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 669: {
1.66 schwarze 670: int tflags;
1.35 schwarze 671:
1.54 schwarze 672: /*
673: * Remember to close out and nullify the current
674: * meta-font and table, if applicable.
675: */
676: if (tag == h->metaf)
677: h->metaf = NULL;
678: if (tag == h->tblt)
679: h->tblt = NULL;
680:
1.66 schwarze 681: tflags = htmltags[tag->tag].flags;
682:
683: if (tflags & HTML_INDENT)
684: h->indent--;
685: if (tflags & HTML_NOINDENT)
686: h->noindent--;
687: if (tflags & HTML_NLEND)
1.67 schwarze 688: print_endline(h);
689: print_indent(h);
690: print_byte(h, '<');
691: print_byte(h, '/');
692: print_word(h, htmltags[tag->tag].name);
693: print_byte(h, '>');
1.66 schwarze 694: if (tflags & HTML_NLAFTER)
1.67 schwarze 695: print_endline(h);
1.54 schwarze 696:
1.74 schwarze 697: h->tag = tag->next;
1.54 schwarze 698: free(tag);
1.1 schwarze 699: }
700:
701: void
1.6 schwarze 702: print_gen_decls(struct html *h)
703: {
1.67 schwarze 704: print_word(h, "<!DOCTYPE html>");
705: print_endline(h);
1.91 schwarze 706: }
707:
708: void
709: print_gen_comment(struct html *h, struct roff_node *n)
710: {
711: int wantblank;
712:
713: print_word(h, "<!-- This is an automatically generated file."
714: " Do not edit.");
715: h->indent = 1;
716: wantblank = 0;
717: while (n != NULL && n->type == ROFFT_COMMENT) {
718: if (strstr(n->string, "-->") == NULL &&
719: (wantblank || *n->string != '\0')) {
720: print_endline(h);
721: print_indent(h);
722: print_word(h, n->string);
723: wantblank = *n->string != '\0';
724: }
725: n = n->next;
726: }
727: if (wantblank)
728: print_endline(h);
729: print_word(h, " -->");
730: print_endline(h);
731: h->indent = 0;
1.1 schwarze 732: }
733:
734: void
1.12 schwarze 735: print_text(struct html *h, const char *word)
1.1 schwarze 736: {
1.67 schwarze 737: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 738: if ( ! (HTML_KEEP & h->flags)) {
739: if (HTML_PREKEEP & h->flags)
740: h->flags |= HTML_KEEP;
1.67 schwarze 741: print_endword(h);
1.12 schwarze 742: } else
1.86 bentley 743: print_word(h, " ");
1.12 schwarze 744: }
1.1 schwarze 745:
1.20 schwarze 746: assert(NULL == h->metaf);
1.31 schwarze 747: switch (h->metac) {
1.35 schwarze 748: case HTMLFONT_ITALIC:
1.64 schwarze 749: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 750: break;
1.35 schwarze 751: case HTMLFONT_BOLD:
1.64 schwarze 752: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 753: break;
1.35 schwarze 754: case HTMLFONT_BI:
1.64 schwarze 755: h->metaf = print_otag(h, TAG_B, "");
756: print_otag(h, TAG_I, "");
1.112 schwarze 757: break;
758: case HTMLFONT_CW:
759: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
1.31 schwarze 760: break;
761: default:
1.67 schwarze 762: print_indent(h);
1.31 schwarze 763: break;
764: }
1.20 schwarze 765:
1.12 schwarze 766: assert(word);
1.65 schwarze 767: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 768: if ( ! (h->flags & HTML_NONOSPACE))
769: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 770: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 771: } else
1.53 schwarze 772: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 773:
774: if (h->metaf) {
775: print_tagq(h, h->metaf);
776: h->metaf = NULL;
777: }
1.17 schwarze 778:
779: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 780: }
781:
782: void
783: print_tagq(struct html *h, const struct tag *until)
784: {
785: struct tag *tag;
786:
1.74 schwarze 787: while ((tag = h->tag) != NULL) {
1.54 schwarze 788: print_ctag(h, tag);
1.1 schwarze 789: if (until && tag == until)
790: return;
791: }
792: }
793:
794: void
795: print_stagq(struct html *h, const struct tag *suntil)
796: {
797: struct tag *tag;
798:
1.74 schwarze 799: while ((tag = h->tag) != NULL) {
1.1 schwarze 800: if (suntil && tag == suntil)
801: return;
1.54 schwarze 802: print_ctag(h, tag);
1.1 schwarze 803: }
804: }
1.42 schwarze 805:
806: void
807: print_paragraph(struct html *h)
808: {
809: struct tag *t;
810:
1.68 schwarze 811: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 812: print_tagq(h, t);
813: }
814:
1.67 schwarze 815:
816: /***********************************************************************
817: * Low level output functions.
818: * They implement line breaking using a short static buffer.
819: ***********************************************************************/
820:
821: /*
822: * Buffer one HTML output byte.
823: * If the buffer is full, flush and deactivate it and start a new line.
824: * If the buffer is inactive, print directly.
825: */
826: static void
827: print_byte(struct html *h, char c)
828: {
829: if ((h->flags & HTML_BUFFER) == 0) {
830: putchar(c);
831: h->col++;
832: return;
833: }
834:
835: if (h->col + h->bufcol < sizeof(h->buf)) {
836: h->buf[h->bufcol++] = c;
837: return;
838: }
839:
840: putchar('\n');
841: h->col = 0;
842: print_indent(h);
843: putchar(' ');
844: putchar(' ');
845: fwrite(h->buf, h->bufcol, 1, stdout);
846: putchar(c);
847: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
848: h->bufcol = 0;
849: h->flags &= ~HTML_BUFFER;
850: }
851:
1.66 schwarze 852: /*
853: * If something was printed on the current output line, end it.
1.67 schwarze 854: * Not to be called right after print_indent().
1.66 schwarze 855: */
1.72 schwarze 856: void
1.67 schwarze 857: print_endline(struct html *h)
1.66 schwarze 858: {
1.67 schwarze 859: if (h->col == 0)
1.66 schwarze 860: return;
861:
1.67 schwarze 862: if (h->bufcol) {
863: putchar(' ');
864: fwrite(h->buf, h->bufcol, 1, stdout);
865: h->bufcol = 0;
866: }
1.66 schwarze 867: putchar('\n');
1.67 schwarze 868: h->col = 0;
869: h->flags |= HTML_NOSPACE;
870: h->flags &= ~HTML_BUFFER;
871: }
872:
873: /*
874: * Flush the HTML output buffer.
875: * If it is inactive, activate it.
876: */
877: static void
878: print_endword(struct html *h)
879: {
880: if (h->noindent) {
881: print_byte(h, ' ');
882: return;
883: }
884:
885: if ((h->flags & HTML_BUFFER) == 0) {
886: h->col++;
887: h->flags |= HTML_BUFFER;
888: } else if (h->bufcol) {
889: putchar(' ');
890: fwrite(h->buf, h->bufcol, 1, stdout);
891: h->col += h->bufcol + 1;
892: }
893: h->bufcol = 0;
1.66 schwarze 894: }
895:
896: /*
897: * If at the beginning of a new output line,
898: * perform indentation and mark the line as containing output.
899: * Make sure to really produce some output right afterwards,
900: * but do not use print_otag() for producing it.
901: */
902: static void
1.67 schwarze 903: print_indent(struct html *h)
1.66 schwarze 904: {
1.67 schwarze 905: size_t i;
1.66 schwarze 906:
1.67 schwarze 907: if (h->col)
1.66 schwarze 908: return;
909:
1.67 schwarze 910: if (h->noindent == 0) {
911: h->col = h->indent * 2;
912: for (i = 0; i < h->col; i++)
1.66 schwarze 913: putchar(' ');
1.67 schwarze 914: }
915: h->flags &= ~HTML_NOSPACE;
916: }
917:
918: /*
919: * Print or buffer some characters
920: * depending on the current HTML output buffer state.
921: */
922: static void
923: print_word(struct html *h, const char *cp)
924: {
925: while (*cp != '\0')
926: print_byte(h, *cp++);
1.3 schwarze 927: }