Annotation of src/usr.bin/mandoc/html.c, Revision 1.123
1.123 ! schwarze 1: /* $OpenBSD: html.c,v 1.122 2019/01/18 14:36:16 schwarze Exp $ */
1.1 schwarze 2: /*
1.42 schwarze 3: * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.119 schwarze 4: * Copyright (c) 2011-2015, 2017-2019 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},
1.123 ! schwarze 64: {"section", HTML_NLALL},
1.66 schwarze 65: {"h1", HTML_NLAROUND},
66: {"h2", HTML_NLAROUND},
67: {"span", 0},
68: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
69: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
70: {"a", 0},
71: {"table", HTML_NLALL | HTML_INDENT},
72: {"tr", HTML_NLALL | HTML_INDENT},
73: {"td", HTML_NLAROUND},
74: {"li", HTML_NLAROUND | HTML_INDENT},
75: {"ul", HTML_NLALL | HTML_INDENT},
76: {"ol", HTML_NLALL | HTML_INDENT},
77: {"dl", HTML_NLALL | HTML_INDENT},
78: {"dt", HTML_NLAROUND},
79: {"dd", HTML_NLAROUND | HTML_INDENT},
1.119 schwarze 80: {"p", HTML_NLAROUND | HTML_INDENT},
1.66 schwarze 81: {"pre", HTML_NLALL | HTML_NOINDENT},
1.77 schwarze 82: {"var", 0},
1.76 schwarze 83: {"cite", 0},
1.66 schwarze 84: {"b", 0},
85: {"i", 0},
86: {"code", 0},
87: {"small", 0},
88: {"style", HTML_NLALL | HTML_INDENT},
89: {"math", HTML_NLALL | HTML_INDENT},
90: {"mrow", 0},
91: {"mi", 0},
1.85 schwarze 92: {"mn", 0},
1.66 schwarze 93: {"mo", 0},
94: {"msup", 0},
95: {"msub", 0},
96: {"msubsup", 0},
97: {"mfrac", 0},
98: {"msqrt", 0},
99: {"mfenced", 0},
100: {"mtable", 0},
101: {"mtr", 0},
102: {"mtd", 0},
103: {"munderover", 0},
104: {"munder", 0},
105: {"mover", 0},
1.5 schwarze 106: };
107:
1.99 schwarze 108: /* Avoid duplicate HTML id= attributes. */
109: static struct ohash id_unique;
110:
1.67 schwarze 111: static void print_byte(struct html *, char);
112: static void print_endword(struct html *);
113: static void print_indent(struct html *);
114: static void print_word(struct html *, const char *);
115:
1.54 schwarze 116: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 117: static int print_escape(struct html *, char);
1.65 schwarze 118: static int print_encode(struct html *, const char *, const char *, int);
119: static void print_href(struct html *, const char *, const char *, int);
1.5 schwarze 120:
1.35 schwarze 121:
1.50 schwarze 122: void *
1.61 schwarze 123: html_alloc(const struct manoutput *outopts)
1.1 schwarze 124: {
125: struct html *h;
126:
1.24 schwarze 127: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 128:
1.74 schwarze 129: h->tag = NULL;
1.56 schwarze 130: h->style = outopts->style;
1.110 schwarze 131: if ((h->base_man1 = outopts->man) == NULL)
132: h->base_man2 = NULL;
133: else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL)
134: *h->base_man2++ = '\0';
1.56 schwarze 135: h->base_includes = outopts->includes;
136: if (outopts->fragment)
137: h->oflags |= HTML_FRAGMENT;
1.111 schwarze 138: if (outopts->toc)
139: h->oflags |= HTML_TOC;
1.1 schwarze 140:
1.99 schwarze 141: mandoc_ohash_init(&id_unique, 4, 0);
142:
1.58 schwarze 143: return h;
1.1 schwarze 144: }
145:
146: void
147: html_free(void *p)
148: {
149: struct tag *tag;
150: struct html *h;
1.99 schwarze 151: char *cp;
152: unsigned int slot;
1.1 schwarze 153:
154: h = (struct html *)p;
1.74 schwarze 155: while ((tag = h->tag) != NULL) {
156: h->tag = tag->next;
1.1 schwarze 157: free(tag);
158: }
1.99 schwarze 159: free(h);
1.1 schwarze 160:
1.99 schwarze 161: cp = ohash_first(&id_unique, &slot);
162: while (cp != NULL) {
163: free(cp);
164: cp = ohash_next(&id_unique, &slot);
165: }
166: ohash_delete(&id_unique);
1.1 schwarze 167: }
168:
169: void
170: print_gen_head(struct html *h)
171: {
1.42 schwarze 172: struct tag *t;
173:
1.64 schwarze 174: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 175: if (h->style != NULL) {
176: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
177: h->style, "type", "text/css", "media", "all");
178: return;
179: }
1.1 schwarze 180:
1.42 schwarze 181: /*
1.92 schwarze 182: * Print a minimal embedded style sheet.
1.42 schwarze 183: */
1.66 schwarze 184:
1.64 schwarze 185: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 186: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 187: print_endline(h);
1.66 schwarze 188: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 189: print_endline(h);
1.66 schwarze 190: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 191: print_endline(h);
1.68 schwarze 192: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.95 schwarze 193: print_endline(h);
194: print_text(h, "div.Nd, div.Bf, div.Op { display: inline; }");
1.96 schwarze 195: print_endline(h);
1.97 schwarze 196: print_text(h, "span.Pa, span.Ad { font-style: italic; }");
1.98 schwarze 197: print_endline(h);
198: print_text(h, "span.Ms { font-weight: bold; }");
1.94 schwarze 199: print_endline(h);
200: print_text(h, "dl.Bl-diag ");
201: print_byte(h, '>');
202: print_text(h, " dt { font-weight: bold; }");
1.93 schwarze 203: print_endline(h);
204: print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, "
205: "code.In, code.Fd, code.Fn,");
206: print_endline(h);
207: print_text(h, "code.Cd { font-weight: bold; "
208: "font-family: inherit; }");
1.42 schwarze 209: print_tagq(h, t);
1.1 schwarze 210: }
211:
1.117 schwarze 212: void
1.26 schwarze 213: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 214: {
215: enum htmlfont font;
1.1 schwarze 216:
1.5 schwarze 217: switch (deco) {
1.35 schwarze 218: case ESCAPE_FONTPREV:
1.5 schwarze 219: font = h->metal;
220: break;
1.35 schwarze 221: case ESCAPE_FONTITALIC:
1.5 schwarze 222: font = HTMLFONT_ITALIC;
223: break;
1.35 schwarze 224: case ESCAPE_FONTBOLD:
1.5 schwarze 225: font = HTMLFONT_BOLD;
226: break;
1.35 schwarze 227: case ESCAPE_FONTBI:
1.31 schwarze 228: font = HTMLFONT_BI;
229: break;
1.112 schwarze 230: case ESCAPE_FONTCW:
231: font = HTMLFONT_CW;
232: break;
1.35 schwarze 233: case ESCAPE_FONT:
234: case ESCAPE_FONTROMAN:
1.5 schwarze 235: font = HTMLFONT_NONE;
236: break;
237: default:
1.117 schwarze 238: return;
1.1 schwarze 239: }
240:
1.20 schwarze 241: if (h->metaf) {
242: print_tagq(h, h->metaf);
243: h->metaf = NULL;
244: }
245:
246: h->metal = h->metac;
247: h->metac = font;
248:
1.31 schwarze 249: switch (font) {
1.35 schwarze 250: case HTMLFONT_ITALIC:
1.64 schwarze 251: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 252: break;
1.35 schwarze 253: case HTMLFONT_BOLD:
1.64 schwarze 254: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 255: break;
1.35 schwarze 256: case HTMLFONT_BI:
1.64 schwarze 257: h->metaf = print_otag(h, TAG_B, "");
258: print_otag(h, TAG_I, "");
1.31 schwarze 259: break;
1.112 schwarze 260: case HTMLFONT_CW:
261: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
262: break;
1.31 schwarze 263: default:
264: break;
265: }
1.118 schwarze 266: }
267:
1.119 schwarze 268: void
269: html_close_paragraph(struct html *h)
270: {
271: struct tag *t;
272:
1.122 schwarze 273: for (t = h->tag; t != NULL && t->closed == 0; t = t->next) {
274: switch(t->tag) {
275: case TAG_P:
276: case TAG_PRE:
1.119 schwarze 277: print_tagq(h, t);
278: break;
1.122 schwarze 279: case TAG_A:
280: print_tagq(h, t);
281: continue;
282: default:
283: continue;
1.119 schwarze 284: }
1.122 schwarze 285: break;
1.119 schwarze 286: }
287: }
288:
1.118 schwarze 289: /*
290: * ROFF_nf switches to no-fill mode, ROFF_fi to fill mode.
291: * TOKEN_NONE does not switch. The old mode is returned.
292: */
293: enum roff_tok
294: html_fillmode(struct html *h, enum roff_tok want)
295: {
296: struct tag *t;
297: enum roff_tok had;
298:
299: for (t = h->tag; t != NULL; t = t->next)
300: if (t->tag == TAG_PRE)
301: break;
302:
303: had = t == NULL ? ROFF_fi : ROFF_nf;
304:
305: if (want != had) {
306: switch (want) {
307: case ROFF_fi:
308: print_tagq(h, t);
309: break;
310: case ROFF_nf:
1.119 schwarze 311: html_close_paragraph(h);
1.118 schwarze 312: print_otag(h, TAG_PRE, "");
313: break;
314: case TOKEN_NONE:
315: break;
316: default:
317: abort();
318: }
319: }
320: return had;
1.80 schwarze 321: }
322:
323: char *
1.99 schwarze 324: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 325: {
326: const struct roff_node *nch;
1.99 schwarze 327: char *buf, *bufs, *cp;
328: unsigned int slot;
329: int suffix;
1.80 schwarze 330:
331: for (nch = n->child; nch != NULL; nch = nch->next)
332: if (nch->type != ROFFT_TEXT)
333: return NULL;
334:
335: buf = NULL;
336: deroff(&buf, n);
1.90 schwarze 337: if (buf == NULL)
338: return NULL;
1.80 schwarze 339:
1.100 schwarze 340: /*
341: * In ID attributes, only use ASCII characters that are
342: * permitted in URL-fragment strings according to the
343: * explicit list at:
344: * https://url.spec.whatwg.org/#url-fragment-string
345: */
1.80 schwarze 346:
347: for (cp = buf; *cp != '\0'; cp++)
1.100 schwarze 348: if (isalnum((unsigned char)*cp) == 0 &&
349: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.80 schwarze 350: *cp = '_';
351:
1.99 schwarze 352: if (unique == 0)
353: return buf;
354:
355: /* Avoid duplicate HTML id= attributes. */
356:
357: bufs = NULL;
358: suffix = 1;
359: slot = ohash_qlookup(&id_unique, buf);
360: cp = ohash_find(&id_unique, slot);
361: if (cp != NULL) {
362: while (cp != NULL) {
363: free(bufs);
364: if (++suffix > 127) {
365: free(buf);
366: return NULL;
367: }
368: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
369: slot = ohash_qlookup(&id_unique, bufs);
370: cp = ohash_find(&id_unique, slot);
371: }
372: free(buf);
373: buf = bufs;
374: }
375: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 376: return buf;
1.1 schwarze 377: }
378:
1.5 schwarze 379: static int
1.67 schwarze 380: print_escape(struct html *h, char c)
1.38 schwarze 381: {
382:
383: switch (c) {
384: case '<':
1.67 schwarze 385: print_word(h, "<");
1.38 schwarze 386: break;
387: case '>':
1.67 schwarze 388: print_word(h, ">");
1.38 schwarze 389: break;
390: case '&':
1.67 schwarze 391: print_word(h, "&");
1.38 schwarze 392: break;
393: case '"':
1.67 schwarze 394: print_word(h, """);
1.38 schwarze 395: break;
396: case ASCII_NBRSP:
1.67 schwarze 397: print_word(h, " ");
1.38 schwarze 398: break;
399: case ASCII_HYPH:
1.67 schwarze 400: print_byte(h, '-');
1.59 schwarze 401: break;
1.38 schwarze 402: case ASCII_BREAK:
403: break;
404: default:
1.58 schwarze 405: return 0;
1.38 schwarze 406: }
1.58 schwarze 407: return 1;
1.38 schwarze 408: }
409:
410: static int
1.65 schwarze 411: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 412: {
1.67 schwarze 413: char numbuf[16];
1.84 schwarze 414: const char *seq;
1.4 schwarze 415: size_t sz;
1.84 schwarze 416: int c, len, breakline, nospace;
1.26 schwarze 417: enum mandoc_esc esc;
1.84 schwarze 418: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 419: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 420:
1.65 schwarze 421: if (pend == NULL)
422: pend = strchr(p, '\0');
423:
1.84 schwarze 424: breakline = 0;
1.5 schwarze 425: nospace = 0;
1.1 schwarze 426:
1.65 schwarze 427: while (p < pend) {
1.30 schwarze 428: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
429: h->flags &= ~HTML_SKIPCHAR;
430: p++;
431: continue;
432: }
433:
1.67 schwarze 434: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 435: print_byte(h, *p);
436:
437: if (breakline &&
438: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.115 schwarze 439: print_otag(h, TAG_BR, "");
1.84 schwarze 440: breakline = 0;
441: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
442: p++;
443: continue;
444: }
1.4 schwarze 445:
1.65 schwarze 446: if (p >= pend)
1.26 schwarze 447: break;
448:
1.84 schwarze 449: if (*p == ' ') {
450: print_endword(h);
451: p++;
452: continue;
453: }
454:
1.67 schwarze 455: if (print_escape(h, *p++))
1.33 schwarze 456: continue;
1.4 schwarze 457:
1.26 schwarze 458: esc = mandoc_escape(&p, &seq, &len);
459: switch (esc) {
1.35 schwarze 460: case ESCAPE_FONT:
461: case ESCAPE_FONTPREV:
462: case ESCAPE_FONTBOLD:
463: case ESCAPE_FONTITALIC:
464: case ESCAPE_FONTBI:
1.112 schwarze 465: case ESCAPE_FONTCW:
1.35 schwarze 466: case ESCAPE_FONTROMAN:
1.113 schwarze 467: if (0 == norecurse) {
468: h->flags |= HTML_NOSPACE;
1.30 schwarze 469: print_metaf(h, esc);
1.113 schwarze 470: h->flags &= ~HTML_NOSPACE;
471: }
1.30 schwarze 472: continue;
1.35 schwarze 473: case ESCAPE_SKIPCHAR:
1.30 schwarze 474: h->flags |= HTML_SKIPCHAR;
475: continue;
1.116 schwarze 476: case ESCAPE_ERROR:
477: continue;
1.30 schwarze 478: default:
479: break;
480: }
481:
482: if (h->flags & HTML_SKIPCHAR) {
483: h->flags &= ~HTML_SKIPCHAR;
484: continue;
485: }
486:
487: switch (esc) {
1.35 schwarze 488: case ESCAPE_UNICODE:
1.38 schwarze 489: /* Skip past "u" header. */
1.26 schwarze 490: c = mchars_num2uc(seq + 1, len - 1);
491: break;
1.35 schwarze 492: case ESCAPE_NUMBERED:
1.26 schwarze 493: c = mchars_num2char(seq, len);
1.51 schwarze 494: if (c < 0)
495: continue;
1.26 schwarze 496: break;
1.35 schwarze 497: case ESCAPE_SPECIAL:
1.61 schwarze 498: c = mchars_spec2cp(seq, len);
1.51 schwarze 499: if (c <= 0)
500: continue;
1.116 schwarze 501: break;
502: case ESCAPE_UNDEF:
503: c = *seq;
1.26 schwarze 504: break;
1.109 schwarze 505: case ESCAPE_DEVICE:
506: print_word(h, "html");
507: continue;
1.84 schwarze 508: case ESCAPE_BREAK:
509: breakline = 1;
510: continue;
1.35 schwarze 511: case ESCAPE_NOSPACE:
1.26 schwarze 512: if ('\0' == *p)
513: nospace = 1;
1.49 schwarze 514: continue;
1.55 schwarze 515: case ESCAPE_OVERSTRIKE:
516: if (len == 0)
517: continue;
518: c = seq[len - 1];
519: break;
1.5 schwarze 520: default:
1.49 schwarze 521: continue;
1.5 schwarze 522: }
1.51 schwarze 523: if ((c < 0x20 && c != 0x09) ||
524: (c > 0x7E && c < 0xA0))
1.49 schwarze 525: c = 0xFFFD;
1.67 schwarze 526: if (c > 0x7E) {
1.86 bentley 527: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 528: print_word(h, numbuf);
529: } else if (print_escape(h, c) == 0)
530: print_byte(h, c);
1.1 schwarze 531: }
1.5 schwarze 532:
1.58 schwarze 533: return nospace;
1.1 schwarze 534: }
535:
1.6 schwarze 536: static void
1.65 schwarze 537: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 538: {
1.110 schwarze 539: struct stat sb;
1.65 schwarze 540: const char *p, *pp;
1.110 schwarze 541: char *filename;
542:
543: if (man) {
544: pp = h->base_man1;
545: if (h->base_man2 != NULL) {
546: mandoc_asprintf(&filename, "%s.%s", name, sec);
547: if (stat(filename, &sb) == -1)
548: pp = h->base_man2;
549: free(filename);
550: }
551: } else
552: pp = h->base_includes;
1.65 schwarze 553:
554: while ((p = strchr(pp, '%')) != NULL) {
555: print_encode(h, pp, p, 1);
556: if (man && p[1] == 'S') {
557: if (sec == NULL)
1.67 schwarze 558: print_byte(h, '1');
1.65 schwarze 559: else
560: print_encode(h, sec, NULL, 1);
561: } else if ((man && p[1] == 'N') ||
562: (man == 0 && p[1] == 'I'))
563: print_encode(h, name, NULL, 1);
564: else
565: print_encode(h, p, p + 2, 1);
566: pp = p + 2;
567: }
568: if (*pp != '\0')
569: print_encode(h, pp, NULL, 1);
1.6 schwarze 570: }
571:
1.1 schwarze 572: struct tag *
1.64 schwarze 573: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 574: {
1.64 schwarze 575: va_list ap;
1.1 schwarze 576: struct tag *t;
1.65 schwarze 577: const char *attr;
1.73 schwarze 578: char *arg1, *arg2;
1.114 schwarze 579: int style_written, tflags;
1.66 schwarze 580:
581: tflags = htmltags[tag].flags;
1.1 schwarze 582:
1.74 schwarze 583: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 584:
1.66 schwarze 585: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 586: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 587: t->tag = tag;
1.74 schwarze 588: t->next = h->tag;
1.122 schwarze 589: t->refcnt = 0;
590: t->closed = 0;
1.74 schwarze 591: h->tag = t;
1.1 schwarze 592: } else
593: t = NULL;
594:
1.66 schwarze 595: if (tflags & HTML_NLBEFORE)
1.67 schwarze 596: print_endline(h);
597: if (h->col == 0)
598: print_indent(h);
1.66 schwarze 599: else if ((h->flags & HTML_NOSPACE) == 0) {
600: if (h->flags & HTML_KEEP)
1.86 bentley 601: print_word(h, " ");
1.66 schwarze 602: else {
603: if (h->flags & HTML_PREKEEP)
604: h->flags |= HTML_KEEP;
1.67 schwarze 605: print_endword(h);
1.12 schwarze 606: }
1.66 schwarze 607: }
1.1 schwarze 608:
1.13 schwarze 609: if ( ! (h->flags & HTML_NONOSPACE))
610: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 611: else
612: h->flags |= HTML_NOSPACE;
1.13 schwarze 613:
1.6 schwarze 614: /* Print out the tag name and attributes. */
615:
1.67 schwarze 616: print_byte(h, '<');
617: print_word(h, htmltags[tag].name);
1.64 schwarze 618:
619: va_start(ap, fmt);
620:
1.114 schwarze 621: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 622:
1.108 schwarze 623: /* Parse attributes and arguments. */
1.73 schwarze 624:
625: arg1 = va_arg(ap, char *);
1.108 schwarze 626: arg2 = NULL;
1.64 schwarze 627: switch (*fmt++) {
628: case 'c':
1.65 schwarze 629: attr = "class";
1.64 schwarze 630: break;
631: case 'h':
1.65 schwarze 632: attr = "href";
1.64 schwarze 633: break;
634: case 'i':
1.65 schwarze 635: attr = "id";
1.64 schwarze 636: break;
637: case '?':
1.73 schwarze 638: attr = arg1;
639: arg1 = va_arg(ap, char *);
1.64 schwarze 640: break;
641: default:
642: abort();
643: }
1.73 schwarze 644: if (*fmt == 'M')
645: arg2 = va_arg(ap, char *);
646: if (arg1 == NULL)
647: continue;
648:
1.108 schwarze 649: /* Print the attributes. */
1.73 schwarze 650:
1.67 schwarze 651: print_byte(h, ' ');
652: print_word(h, attr);
653: print_byte(h, '=');
654: print_byte(h, '"');
1.65 schwarze 655: switch (*fmt) {
1.78 schwarze 656: case 'I':
657: print_href(h, arg1, NULL, 0);
658: fmt++;
659: break;
1.65 schwarze 660: case 'M':
1.73 schwarze 661: print_href(h, arg1, arg2, 1);
1.65 schwarze 662: fmt++;
663: break;
1.78 schwarze 664: case 'R':
665: print_byte(h, '#');
666: print_encode(h, arg1, NULL, 1);
1.65 schwarze 667: fmt++;
1.78 schwarze 668: break;
1.65 schwarze 669: default:
1.114 schwarze 670: print_encode(h, arg1, NULL, 1);
1.65 schwarze 671: break;
672: }
1.67 schwarze 673: print_byte(h, '"');
1.64 schwarze 674: }
1.114 schwarze 675:
676: style_written = 0;
677: while (*fmt++ == 's') {
678: arg1 = va_arg(ap, char *);
679: arg2 = va_arg(ap, char *);
680: if (arg2 == NULL)
681: continue;
682: print_byte(h, ' ');
683: if (style_written == 0) {
684: print_word(h, "style=\"");
685: style_written = 1;
686: }
687: print_word(h, arg1);
688: print_byte(h, ':');
689: print_byte(h, ' ');
690: print_word(h, arg2);
691: print_byte(h, ';');
692: }
693: if (style_written)
694: print_byte(h, '"');
695:
1.64 schwarze 696: va_end(ap);
1.6 schwarze 697:
1.42 schwarze 698: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 699:
700: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 701: print_byte(h, '/');
1.6 schwarze 702:
1.67 schwarze 703: print_byte(h, '>');
1.1 schwarze 704:
1.66 schwarze 705: if (tflags & HTML_NLBEGIN)
1.67 schwarze 706: print_endline(h);
1.66 schwarze 707: else
708: h->flags |= HTML_NOSPACE;
1.18 schwarze 709:
1.66 schwarze 710: if (tflags & HTML_INDENT)
711: h->indent++;
712: if (tflags & HTML_NOINDENT)
713: h->noindent++;
1.18 schwarze 714:
1.58 schwarze 715: return t;
1.1 schwarze 716: }
717:
718: static void
1.54 schwarze 719: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 720: {
1.66 schwarze 721: int tflags;
1.35 schwarze 722:
1.122 schwarze 723: if (tag->closed == 0) {
724: tag->closed = 1;
725: if (tag == h->metaf)
726: h->metaf = NULL;
727: if (tag == h->tblt)
728: h->tblt = NULL;
729:
730: tflags = htmltags[tag->tag].flags;
731: if (tflags & HTML_INDENT)
732: h->indent--;
733: if (tflags & HTML_NOINDENT)
734: h->noindent--;
735: if (tflags & HTML_NLEND)
736: print_endline(h);
737: print_indent(h);
738: print_byte(h, '<');
739: print_byte(h, '/');
740: print_word(h, htmltags[tag->tag].name);
741: print_byte(h, '>');
742: if (tflags & HTML_NLAFTER)
743: print_endline(h);
744: }
745: if (tag->refcnt == 0) {
746: h->tag = tag->next;
747: free(tag);
748: }
1.1 schwarze 749: }
750:
751: void
1.6 schwarze 752: print_gen_decls(struct html *h)
753: {
1.67 schwarze 754: print_word(h, "<!DOCTYPE html>");
755: print_endline(h);
1.91 schwarze 756: }
757:
758: void
759: print_gen_comment(struct html *h, struct roff_node *n)
760: {
761: int wantblank;
762:
763: print_word(h, "<!-- This is an automatically generated file."
764: " Do not edit.");
765: h->indent = 1;
766: wantblank = 0;
767: while (n != NULL && n->type == ROFFT_COMMENT) {
768: if (strstr(n->string, "-->") == NULL &&
769: (wantblank || *n->string != '\0')) {
770: print_endline(h);
771: print_indent(h);
772: print_word(h, n->string);
773: wantblank = *n->string != '\0';
774: }
775: n = n->next;
776: }
777: if (wantblank)
778: print_endline(h);
779: print_word(h, " -->");
780: print_endline(h);
781: h->indent = 0;
1.1 schwarze 782: }
783:
784: void
1.12 schwarze 785: print_text(struct html *h, const char *word)
1.1 schwarze 786: {
1.67 schwarze 787: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 788: if ( ! (HTML_KEEP & h->flags)) {
789: if (HTML_PREKEEP & h->flags)
790: h->flags |= HTML_KEEP;
1.67 schwarze 791: print_endword(h);
1.12 schwarze 792: } else
1.86 bentley 793: print_word(h, " ");
1.12 schwarze 794: }
1.1 schwarze 795:
1.20 schwarze 796: assert(NULL == h->metaf);
1.31 schwarze 797: switch (h->metac) {
1.35 schwarze 798: case HTMLFONT_ITALIC:
1.64 schwarze 799: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 800: break;
1.35 schwarze 801: case HTMLFONT_BOLD:
1.64 schwarze 802: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 803: break;
1.35 schwarze 804: case HTMLFONT_BI:
1.64 schwarze 805: h->metaf = print_otag(h, TAG_B, "");
806: print_otag(h, TAG_I, "");
1.112 schwarze 807: break;
808: case HTMLFONT_CW:
809: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
1.31 schwarze 810: break;
811: default:
1.67 schwarze 812: print_indent(h);
1.31 schwarze 813: break;
814: }
1.20 schwarze 815:
1.12 schwarze 816: assert(word);
1.65 schwarze 817: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 818: if ( ! (h->flags & HTML_NONOSPACE))
819: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 820: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 821: } else
1.53 schwarze 822: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 823:
824: if (h->metaf) {
825: print_tagq(h, h->metaf);
826: h->metaf = NULL;
827: }
1.17 schwarze 828:
829: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 830: }
831:
832: void
833: print_tagq(struct html *h, const struct tag *until)
834: {
1.122 schwarze 835: struct tag *this, *next;
1.1 schwarze 836:
1.122 schwarze 837: for (this = h->tag; this != NULL; this = next) {
838: next = this == until ? NULL : this->next;
839: print_ctag(h, this);
1.1 schwarze 840: }
841: }
842:
1.120 schwarze 843: /*
844: * Close out all open elements up to but excluding suntil.
845: * Note that a paragraph just inside stays open together with it
846: * because paragraphs include subsequent phrasing content.
847: */
1.1 schwarze 848: void
849: print_stagq(struct html *h, const struct tag *suntil)
850: {
1.122 schwarze 851: struct tag *this, *next;
1.1 schwarze 852:
1.122 schwarze 853: for (this = h->tag; this != NULL; this = next) {
854: next = this->next;
855: if (this == suntil || (next == suntil &&
856: (this->tag == TAG_P || this->tag == TAG_PRE)))
857: break;
858: print_ctag(h, this);
1.1 schwarze 859: }
1.42 schwarze 860: }
861:
1.67 schwarze 862:
863: /***********************************************************************
864: * Low level output functions.
865: * They implement line breaking using a short static buffer.
866: ***********************************************************************/
867:
868: /*
869: * Buffer one HTML output byte.
870: * If the buffer is full, flush and deactivate it and start a new line.
871: * If the buffer is inactive, print directly.
872: */
873: static void
874: print_byte(struct html *h, char c)
875: {
876: if ((h->flags & HTML_BUFFER) == 0) {
877: putchar(c);
878: h->col++;
879: return;
880: }
881:
882: if (h->col + h->bufcol < sizeof(h->buf)) {
883: h->buf[h->bufcol++] = c;
884: return;
885: }
886:
887: putchar('\n');
888: h->col = 0;
889: print_indent(h);
890: putchar(' ');
891: putchar(' ');
892: fwrite(h->buf, h->bufcol, 1, stdout);
893: putchar(c);
894: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
895: h->bufcol = 0;
896: h->flags &= ~HTML_BUFFER;
897: }
898:
1.66 schwarze 899: /*
900: * If something was printed on the current output line, end it.
1.67 schwarze 901: * Not to be called right after print_indent().
1.66 schwarze 902: */
1.72 schwarze 903: void
1.67 schwarze 904: print_endline(struct html *h)
1.66 schwarze 905: {
1.67 schwarze 906: if (h->col == 0)
1.66 schwarze 907: return;
908:
1.67 schwarze 909: if (h->bufcol) {
910: putchar(' ');
911: fwrite(h->buf, h->bufcol, 1, stdout);
912: h->bufcol = 0;
913: }
1.66 schwarze 914: putchar('\n');
1.67 schwarze 915: h->col = 0;
916: h->flags |= HTML_NOSPACE;
917: h->flags &= ~HTML_BUFFER;
918: }
919:
920: /*
921: * Flush the HTML output buffer.
922: * If it is inactive, activate it.
923: */
924: static void
925: print_endword(struct html *h)
926: {
927: if (h->noindent) {
928: print_byte(h, ' ');
929: return;
930: }
931:
932: if ((h->flags & HTML_BUFFER) == 0) {
933: h->col++;
934: h->flags |= HTML_BUFFER;
935: } else if (h->bufcol) {
936: putchar(' ');
937: fwrite(h->buf, h->bufcol, 1, stdout);
938: h->col += h->bufcol + 1;
939: }
940: h->bufcol = 0;
1.66 schwarze 941: }
942:
943: /*
944: * If at the beginning of a new output line,
945: * perform indentation and mark the line as containing output.
946: * Make sure to really produce some output right afterwards,
947: * but do not use print_otag() for producing it.
948: */
949: static void
1.67 schwarze 950: print_indent(struct html *h)
1.66 schwarze 951: {
1.67 schwarze 952: size_t i;
1.66 schwarze 953:
1.67 schwarze 954: if (h->col)
1.66 schwarze 955: return;
956:
1.67 schwarze 957: if (h->noindent == 0) {
958: h->col = h->indent * 2;
959: for (i = 0; i < h->col; i++)
1.66 schwarze 960: putchar(' ');
1.67 schwarze 961: }
962: h->flags &= ~HTML_NOSPACE;
963: }
964:
965: /*
966: * Print or buffer some characters
967: * depending on the current HTML output buffer state.
968: */
969: static void
970: print_word(struct html *h, const char *cp)
971: {
972: while (*cp != '\0')
973: print_byte(h, *cp++);
1.3 schwarze 974: }