Annotation of src/usr.bin/mandoc/html.c, Revision 1.116
1.116 ! schwarze 1: /* $OpenBSD: html.c,v 1.115 2018/12/04 18:29:18 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: const char *seq;
1.4 schwarze 359: size_t sz;
1.84 schwarze 360: int c, len, breakline, nospace;
1.26 schwarze 361: enum mandoc_esc esc;
1.84 schwarze 362: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 363: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 364:
1.65 schwarze 365: if (pend == NULL)
366: pend = strchr(p, '\0');
367:
1.84 schwarze 368: breakline = 0;
1.5 schwarze 369: nospace = 0;
1.1 schwarze 370:
1.65 schwarze 371: while (p < pend) {
1.30 schwarze 372: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
373: h->flags &= ~HTML_SKIPCHAR;
374: p++;
375: continue;
376: }
377:
1.67 schwarze 378: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 379: print_byte(h, *p);
380:
381: if (breakline &&
382: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.115 schwarze 383: print_otag(h, TAG_BR, "");
1.84 schwarze 384: breakline = 0;
385: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
386: p++;
387: continue;
388: }
1.4 schwarze 389:
1.65 schwarze 390: if (p >= pend)
1.26 schwarze 391: break;
392:
1.84 schwarze 393: if (*p == ' ') {
394: print_endword(h);
395: p++;
396: continue;
397: }
398:
1.67 schwarze 399: if (print_escape(h, *p++))
1.33 schwarze 400: continue;
1.4 schwarze 401:
1.26 schwarze 402: esc = mandoc_escape(&p, &seq, &len);
403: switch (esc) {
1.35 schwarze 404: case ESCAPE_FONT:
405: case ESCAPE_FONTPREV:
406: case ESCAPE_FONTBOLD:
407: case ESCAPE_FONTITALIC:
408: case ESCAPE_FONTBI:
1.112 schwarze 409: case ESCAPE_FONTCW:
1.35 schwarze 410: case ESCAPE_FONTROMAN:
1.113 schwarze 411: if (0 == norecurse) {
412: h->flags |= HTML_NOSPACE;
1.30 schwarze 413: print_metaf(h, esc);
1.113 schwarze 414: h->flags &= ~HTML_NOSPACE;
415: }
1.30 schwarze 416: continue;
1.35 schwarze 417: case ESCAPE_SKIPCHAR:
1.30 schwarze 418: h->flags |= HTML_SKIPCHAR;
419: continue;
1.116 ! schwarze 420: case ESCAPE_ERROR:
! 421: continue;
1.30 schwarze 422: default:
423: break;
424: }
425:
426: if (h->flags & HTML_SKIPCHAR) {
427: h->flags &= ~HTML_SKIPCHAR;
428: continue;
429: }
430:
431: switch (esc) {
1.35 schwarze 432: case ESCAPE_UNICODE:
1.38 schwarze 433: /* Skip past "u" header. */
1.26 schwarze 434: c = mchars_num2uc(seq + 1, len - 1);
435: break;
1.35 schwarze 436: case ESCAPE_NUMBERED:
1.26 schwarze 437: c = mchars_num2char(seq, len);
1.51 schwarze 438: if (c < 0)
439: continue;
1.26 schwarze 440: break;
1.35 schwarze 441: case ESCAPE_SPECIAL:
1.61 schwarze 442: c = mchars_spec2cp(seq, len);
1.51 schwarze 443: if (c <= 0)
444: continue;
1.116 ! schwarze 445: break;
! 446: case ESCAPE_UNDEF:
! 447: c = *seq;
1.26 schwarze 448: break;
1.109 schwarze 449: case ESCAPE_DEVICE:
450: print_word(h, "html");
451: continue;
1.84 schwarze 452: case ESCAPE_BREAK:
453: breakline = 1;
454: continue;
1.35 schwarze 455: case ESCAPE_NOSPACE:
1.26 schwarze 456: if ('\0' == *p)
457: nospace = 1;
1.49 schwarze 458: continue;
1.55 schwarze 459: case ESCAPE_OVERSTRIKE:
460: if (len == 0)
461: continue;
462: c = seq[len - 1];
463: break;
1.5 schwarze 464: default:
1.49 schwarze 465: continue;
1.5 schwarze 466: }
1.51 schwarze 467: if ((c < 0x20 && c != 0x09) ||
468: (c > 0x7E && c < 0xA0))
1.49 schwarze 469: c = 0xFFFD;
1.67 schwarze 470: if (c > 0x7E) {
1.86 bentley 471: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 472: print_word(h, numbuf);
473: } else if (print_escape(h, c) == 0)
474: print_byte(h, c);
1.1 schwarze 475: }
1.5 schwarze 476:
1.58 schwarze 477: return nospace;
1.1 schwarze 478: }
479:
1.6 schwarze 480: static void
1.65 schwarze 481: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 482: {
1.110 schwarze 483: struct stat sb;
1.65 schwarze 484: const char *p, *pp;
1.110 schwarze 485: char *filename;
486:
487: if (man) {
488: pp = h->base_man1;
489: if (h->base_man2 != NULL) {
490: mandoc_asprintf(&filename, "%s.%s", name, sec);
491: if (stat(filename, &sb) == -1)
492: pp = h->base_man2;
493: free(filename);
494: }
495: } else
496: pp = h->base_includes;
1.65 schwarze 497:
498: while ((p = strchr(pp, '%')) != NULL) {
499: print_encode(h, pp, p, 1);
500: if (man && p[1] == 'S') {
501: if (sec == NULL)
1.67 schwarze 502: print_byte(h, '1');
1.65 schwarze 503: else
504: print_encode(h, sec, NULL, 1);
505: } else if ((man && p[1] == 'N') ||
506: (man == 0 && p[1] == 'I'))
507: print_encode(h, name, NULL, 1);
508: else
509: print_encode(h, p, p + 2, 1);
510: pp = p + 2;
511: }
512: if (*pp != '\0')
513: print_encode(h, pp, NULL, 1);
1.6 schwarze 514: }
515:
1.1 schwarze 516: struct tag *
1.64 schwarze 517: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 518: {
1.64 schwarze 519: va_list ap;
1.1 schwarze 520: struct tag *t;
1.65 schwarze 521: const char *attr;
1.73 schwarze 522: char *arg1, *arg2;
1.114 schwarze 523: int style_written, tflags;
1.66 schwarze 524:
525: tflags = htmltags[tag].flags;
1.1 schwarze 526:
1.74 schwarze 527: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 528:
1.66 schwarze 529: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 530: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 531: t->tag = tag;
1.74 schwarze 532: t->next = h->tag;
533: h->tag = t;
1.1 schwarze 534: } else
535: t = NULL;
536:
1.66 schwarze 537: if (tflags & HTML_NLBEFORE)
1.67 schwarze 538: print_endline(h);
539: if (h->col == 0)
540: print_indent(h);
1.66 schwarze 541: else if ((h->flags & HTML_NOSPACE) == 0) {
542: if (h->flags & HTML_KEEP)
1.86 bentley 543: print_word(h, " ");
1.66 schwarze 544: else {
545: if (h->flags & HTML_PREKEEP)
546: h->flags |= HTML_KEEP;
1.67 schwarze 547: print_endword(h);
1.12 schwarze 548: }
1.66 schwarze 549: }
1.1 schwarze 550:
1.13 schwarze 551: if ( ! (h->flags & HTML_NONOSPACE))
552: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 553: else
554: h->flags |= HTML_NOSPACE;
1.13 schwarze 555:
1.6 schwarze 556: /* Print out the tag name and attributes. */
557:
1.67 schwarze 558: print_byte(h, '<');
559: print_word(h, htmltags[tag].name);
1.64 schwarze 560:
561: va_start(ap, fmt);
562:
1.114 schwarze 563: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 564:
1.108 schwarze 565: /* Parse attributes and arguments. */
1.73 schwarze 566:
567: arg1 = va_arg(ap, char *);
1.108 schwarze 568: arg2 = NULL;
1.64 schwarze 569: switch (*fmt++) {
570: case 'c':
1.65 schwarze 571: attr = "class";
1.64 schwarze 572: break;
573: case 'h':
1.65 schwarze 574: attr = "href";
1.64 schwarze 575: break;
576: case 'i':
1.65 schwarze 577: attr = "id";
1.64 schwarze 578: break;
579: case '?':
1.73 schwarze 580: attr = arg1;
581: arg1 = va_arg(ap, char *);
1.64 schwarze 582: break;
583: default:
584: abort();
585: }
1.73 schwarze 586: if (*fmt == 'M')
587: arg2 = va_arg(ap, char *);
588: if (arg1 == NULL)
589: continue;
590:
1.108 schwarze 591: /* Print the attributes. */
1.73 schwarze 592:
1.67 schwarze 593: print_byte(h, ' ');
594: print_word(h, attr);
595: print_byte(h, '=');
596: print_byte(h, '"');
1.65 schwarze 597: switch (*fmt) {
1.78 schwarze 598: case 'I':
599: print_href(h, arg1, NULL, 0);
600: fmt++;
601: break;
1.65 schwarze 602: case 'M':
1.73 schwarze 603: print_href(h, arg1, arg2, 1);
1.65 schwarze 604: fmt++;
605: break;
1.78 schwarze 606: case 'R':
607: print_byte(h, '#');
608: print_encode(h, arg1, NULL, 1);
1.65 schwarze 609: fmt++;
610: break;
1.78 schwarze 611: case 'T':
612: print_encode(h, arg1, NULL, 1);
613: print_word(h, "\" title=\"");
614: print_encode(h, arg1, NULL, 1);
1.65 schwarze 615: fmt++;
1.78 schwarze 616: break;
1.65 schwarze 617: default:
1.114 schwarze 618: print_encode(h, arg1, NULL, 1);
1.65 schwarze 619: break;
620: }
1.67 schwarze 621: print_byte(h, '"');
1.64 schwarze 622: }
1.114 schwarze 623:
624: style_written = 0;
625: while (*fmt++ == 's') {
626: arg1 = va_arg(ap, char *);
627: arg2 = va_arg(ap, char *);
628: if (arg2 == NULL)
629: continue;
630: print_byte(h, ' ');
631: if (style_written == 0) {
632: print_word(h, "style=\"");
633: style_written = 1;
634: }
635: print_word(h, arg1);
636: print_byte(h, ':');
637: print_byte(h, ' ');
638: print_word(h, arg2);
639: print_byte(h, ';');
640: }
641: if (style_written)
642: print_byte(h, '"');
643:
1.64 schwarze 644: va_end(ap);
1.6 schwarze 645:
1.42 schwarze 646: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 647:
648: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 649: print_byte(h, '/');
1.6 schwarze 650:
1.67 schwarze 651: print_byte(h, '>');
1.1 schwarze 652:
1.66 schwarze 653: if (tflags & HTML_NLBEGIN)
1.67 schwarze 654: print_endline(h);
1.66 schwarze 655: else
656: h->flags |= HTML_NOSPACE;
1.18 schwarze 657:
1.66 schwarze 658: if (tflags & HTML_INDENT)
659: h->indent++;
660: if (tflags & HTML_NOINDENT)
661: h->noindent++;
1.18 schwarze 662:
1.58 schwarze 663: return t;
1.1 schwarze 664: }
665:
666: static void
1.54 schwarze 667: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 668: {
1.66 schwarze 669: int tflags;
1.35 schwarze 670:
1.54 schwarze 671: /*
672: * Remember to close out and nullify the current
673: * meta-font and table, if applicable.
674: */
675: if (tag == h->metaf)
676: h->metaf = NULL;
677: if (tag == h->tblt)
678: h->tblt = NULL;
679:
1.66 schwarze 680: tflags = htmltags[tag->tag].flags;
681:
682: if (tflags & HTML_INDENT)
683: h->indent--;
684: if (tflags & HTML_NOINDENT)
685: h->noindent--;
686: if (tflags & HTML_NLEND)
1.67 schwarze 687: print_endline(h);
688: print_indent(h);
689: print_byte(h, '<');
690: print_byte(h, '/');
691: print_word(h, htmltags[tag->tag].name);
692: print_byte(h, '>');
1.66 schwarze 693: if (tflags & HTML_NLAFTER)
1.67 schwarze 694: print_endline(h);
1.54 schwarze 695:
1.74 schwarze 696: h->tag = tag->next;
1.54 schwarze 697: free(tag);
1.1 schwarze 698: }
699:
700: void
1.6 schwarze 701: print_gen_decls(struct html *h)
702: {
1.67 schwarze 703: print_word(h, "<!DOCTYPE html>");
704: print_endline(h);
1.91 schwarze 705: }
706:
707: void
708: print_gen_comment(struct html *h, struct roff_node *n)
709: {
710: int wantblank;
711:
712: print_word(h, "<!-- This is an automatically generated file."
713: " Do not edit.");
714: h->indent = 1;
715: wantblank = 0;
716: while (n != NULL && n->type == ROFFT_COMMENT) {
717: if (strstr(n->string, "-->") == NULL &&
718: (wantblank || *n->string != '\0')) {
719: print_endline(h);
720: print_indent(h);
721: print_word(h, n->string);
722: wantblank = *n->string != '\0';
723: }
724: n = n->next;
725: }
726: if (wantblank)
727: print_endline(h);
728: print_word(h, " -->");
729: print_endline(h);
730: h->indent = 0;
1.1 schwarze 731: }
732:
733: void
1.12 schwarze 734: print_text(struct html *h, const char *word)
1.1 schwarze 735: {
1.67 schwarze 736: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 737: if ( ! (HTML_KEEP & h->flags)) {
738: if (HTML_PREKEEP & h->flags)
739: h->flags |= HTML_KEEP;
1.67 schwarze 740: print_endword(h);
1.12 schwarze 741: } else
1.86 bentley 742: print_word(h, " ");
1.12 schwarze 743: }
1.1 schwarze 744:
1.20 schwarze 745: assert(NULL == h->metaf);
1.31 schwarze 746: switch (h->metac) {
1.35 schwarze 747: case HTMLFONT_ITALIC:
1.64 schwarze 748: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 749: break;
1.35 schwarze 750: case HTMLFONT_BOLD:
1.64 schwarze 751: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 752: break;
1.35 schwarze 753: case HTMLFONT_BI:
1.64 schwarze 754: h->metaf = print_otag(h, TAG_B, "");
755: print_otag(h, TAG_I, "");
1.112 schwarze 756: break;
757: case HTMLFONT_CW:
758: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
1.31 schwarze 759: break;
760: default:
1.67 schwarze 761: print_indent(h);
1.31 schwarze 762: break;
763: }
1.20 schwarze 764:
1.12 schwarze 765: assert(word);
1.65 schwarze 766: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 767: if ( ! (h->flags & HTML_NONOSPACE))
768: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 769: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 770: } else
1.53 schwarze 771: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 772:
773: if (h->metaf) {
774: print_tagq(h, h->metaf);
775: h->metaf = NULL;
776: }
1.17 schwarze 777:
778: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 779: }
780:
781: void
782: print_tagq(struct html *h, const struct tag *until)
783: {
784: struct tag *tag;
785:
1.74 schwarze 786: while ((tag = h->tag) != NULL) {
1.54 schwarze 787: print_ctag(h, tag);
1.1 schwarze 788: if (until && tag == until)
789: return;
790: }
791: }
792:
793: void
794: print_stagq(struct html *h, const struct tag *suntil)
795: {
796: struct tag *tag;
797:
1.74 schwarze 798: while ((tag = h->tag) != NULL) {
1.1 schwarze 799: if (suntil && tag == suntil)
800: return;
1.54 schwarze 801: print_ctag(h, tag);
1.1 schwarze 802: }
803: }
1.42 schwarze 804:
805: void
806: print_paragraph(struct html *h)
807: {
808: struct tag *t;
809:
1.68 schwarze 810: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 811: print_tagq(h, t);
812: }
813:
1.67 schwarze 814:
815: /***********************************************************************
816: * Low level output functions.
817: * They implement line breaking using a short static buffer.
818: ***********************************************************************/
819:
820: /*
821: * Buffer one HTML output byte.
822: * If the buffer is full, flush and deactivate it and start a new line.
823: * If the buffer is inactive, print directly.
824: */
825: static void
826: print_byte(struct html *h, char c)
827: {
828: if ((h->flags & HTML_BUFFER) == 0) {
829: putchar(c);
830: h->col++;
831: return;
832: }
833:
834: if (h->col + h->bufcol < sizeof(h->buf)) {
835: h->buf[h->bufcol++] = c;
836: return;
837: }
838:
839: putchar('\n');
840: h->col = 0;
841: print_indent(h);
842: putchar(' ');
843: putchar(' ');
844: fwrite(h->buf, h->bufcol, 1, stdout);
845: putchar(c);
846: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
847: h->bufcol = 0;
848: h->flags &= ~HTML_BUFFER;
849: }
850:
1.66 schwarze 851: /*
852: * If something was printed on the current output line, end it.
1.67 schwarze 853: * Not to be called right after print_indent().
1.66 schwarze 854: */
1.72 schwarze 855: void
1.67 schwarze 856: print_endline(struct html *h)
1.66 schwarze 857: {
1.67 schwarze 858: if (h->col == 0)
1.66 schwarze 859: return;
860:
1.67 schwarze 861: if (h->bufcol) {
862: putchar(' ');
863: fwrite(h->buf, h->bufcol, 1, stdout);
864: h->bufcol = 0;
865: }
1.66 schwarze 866: putchar('\n');
1.67 schwarze 867: h->col = 0;
868: h->flags |= HTML_NOSPACE;
869: h->flags &= ~HTML_BUFFER;
870: }
871:
872: /*
873: * Flush the HTML output buffer.
874: * If it is inactive, activate it.
875: */
876: static void
877: print_endword(struct html *h)
878: {
879: if (h->noindent) {
880: print_byte(h, ' ');
881: return;
882: }
883:
884: if ((h->flags & HTML_BUFFER) == 0) {
885: h->col++;
886: h->flags |= HTML_BUFFER;
887: } else if (h->bufcol) {
888: putchar(' ');
889: fwrite(h->buf, h->bufcol, 1, stdout);
890: h->col += h->bufcol + 1;
891: }
892: h->bufcol = 0;
1.66 schwarze 893: }
894:
895: /*
896: * If at the beginning of a new output line,
897: * perform indentation and mark the line as containing output.
898: * Make sure to really produce some output right afterwards,
899: * but do not use print_otag() for producing it.
900: */
901: static void
1.67 schwarze 902: print_indent(struct html *h)
1.66 schwarze 903: {
1.67 schwarze 904: size_t i;
1.66 schwarze 905:
1.67 schwarze 906: if (h->col)
1.66 schwarze 907: return;
908:
1.67 schwarze 909: if (h->noindent == 0) {
910: h->col = h->indent * 2;
911: for (i = 0; i < h->col; i++)
1.66 schwarze 912: putchar(' ');
1.67 schwarze 913: }
914: h->flags &= ~HTML_NOSPACE;
915: }
916:
917: /*
918: * Print or buffer some characters
919: * depending on the current HTML output buffer state.
920: */
921: static void
922: print_word(struct html *h, const char *cp)
923: {
924: while (*cp != '\0')
925: print_byte(h, *cp++);
1.3 schwarze 926: }