Annotation of src/usr.bin/mandoc/html.c, Revision 1.126
1.126 ! schwarze 1: /* $OpenBSD: html.c,v 1.125 2019/04/30 15:52:42 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:
1.124 schwarze 111: static void html_reset_internal(struct html *);
1.67 schwarze 112: static void print_byte(struct html *, char);
113: static void print_endword(struct html *);
114: static void print_indent(struct html *);
115: static void print_word(struct html *, const char *);
116:
1.54 schwarze 117: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 118: static int print_escape(struct html *, char);
1.65 schwarze 119: static int print_encode(struct html *, const char *, const char *, int);
120: static void print_href(struct html *, const char *, const char *, int);
1.125 schwarze 121: static void print_metaf(struct html *);
1.5 schwarze 122:
1.35 schwarze 123:
1.50 schwarze 124: void *
1.61 schwarze 125: html_alloc(const struct manoutput *outopts)
1.1 schwarze 126: {
127: struct html *h;
128:
1.24 schwarze 129: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 130:
1.74 schwarze 131: h->tag = NULL;
1.56 schwarze 132: h->style = outopts->style;
1.110 schwarze 133: if ((h->base_man1 = outopts->man) == NULL)
134: h->base_man2 = NULL;
135: else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL)
136: *h->base_man2++ = '\0';
1.56 schwarze 137: h->base_includes = outopts->includes;
138: if (outopts->fragment)
139: h->oflags |= HTML_FRAGMENT;
1.111 schwarze 140: if (outopts->toc)
141: h->oflags |= HTML_TOC;
1.1 schwarze 142:
1.99 schwarze 143: mandoc_ohash_init(&id_unique, 4, 0);
144:
1.58 schwarze 145: return h;
1.1 schwarze 146: }
147:
1.124 schwarze 148: static void
149: html_reset_internal(struct html *h)
1.1 schwarze 150: {
151: struct tag *tag;
1.99 schwarze 152: char *cp;
153: unsigned int slot;
1.1 schwarze 154:
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: cp = ohash_first(&id_unique, &slot);
160: while (cp != NULL) {
161: free(cp);
162: cp = ohash_next(&id_unique, &slot);
163: }
164: ohash_delete(&id_unique);
1.124 schwarze 165: }
166:
167: void
168: html_reset(void *p)
169: {
170: html_reset_internal(p);
171: mandoc_ohash_init(&id_unique, 4, 0);
172: }
173:
174: void
175: html_free(void *p)
176: {
177: html_reset_internal(p);
178: free(p);
1.1 schwarze 179: }
180:
181: void
182: print_gen_head(struct html *h)
183: {
1.42 schwarze 184: struct tag *t;
185:
1.64 schwarze 186: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 187: if (h->style != NULL) {
188: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
189: h->style, "type", "text/css", "media", "all");
190: return;
191: }
1.1 schwarze 192:
1.42 schwarze 193: /*
1.92 schwarze 194: * Print a minimal embedded style sheet.
1.42 schwarze 195: */
1.66 schwarze 196:
1.64 schwarze 197: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 198: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 199: print_endline(h);
1.66 schwarze 200: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 201: print_endline(h);
1.66 schwarze 202: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 203: print_endline(h);
1.126 ! schwarze 204: print_text(h, ".Nd, .Bf, .Op { display: inline; }");
1.95 schwarze 205: print_endline(h);
1.126 ! schwarze 206: print_text(h, ".Pa, .Ad { font-style: italic; }");
1.96 schwarze 207: print_endline(h);
1.126 ! schwarze 208: print_text(h, ".Ms { font-weight: bold; }");
1.98 schwarze 209: print_endline(h);
1.126 ! schwarze 210: print_text(h, ".Bl-diag ");
1.94 schwarze 211: print_byte(h, '>');
212: print_text(h, " dt { font-weight: bold; }");
1.93 schwarze 213: print_endline(h);
1.126 ! schwarze 214: print_text(h, "code.Nm, .Fl, .Cm, .Ic, code.In, .Fd, .Fn, .Cd "
! 215: "{ font-weight: bold; font-family: inherit; }");
1.42 schwarze 216: print_tagq(h, t);
1.1 schwarze 217: }
218:
1.125 schwarze 219: int
220: html_setfont(struct html *h, enum mandoc_esc font)
1.5 schwarze 221: {
1.125 schwarze 222: switch (font) {
1.35 schwarze 223: case ESCAPE_FONTPREV:
1.5 schwarze 224: font = h->metal;
225: break;
1.35 schwarze 226: case ESCAPE_FONTITALIC:
227: case ESCAPE_FONTBOLD:
228: case ESCAPE_FONTBI:
1.112 schwarze 229: case ESCAPE_FONTCW:
1.125 schwarze 230: case ESCAPE_FONTROMAN:
1.112 schwarze 231: break;
1.35 schwarze 232: case ESCAPE_FONT:
1.125 schwarze 233: font = ESCAPE_FONTROMAN;
1.5 schwarze 234: break;
235: default:
1.125 schwarze 236: return 0;
1.1 schwarze 237: }
1.125 schwarze 238: h->metal = h->metac;
239: h->metac = font;
240: return 1;
241: }
1.1 schwarze 242:
1.125 schwarze 243: static void
244: print_metaf(struct html *h)
245: {
1.20 schwarze 246: if (h->metaf) {
247: print_tagq(h, h->metaf);
248: h->metaf = NULL;
249: }
1.125 schwarze 250: switch (h->metac) {
251: case ESCAPE_FONTITALIC:
1.64 schwarze 252: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 253: break;
1.125 schwarze 254: case ESCAPE_FONTBOLD:
1.64 schwarze 255: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 256: break;
1.125 schwarze 257: case ESCAPE_FONTBI:
1.64 schwarze 258: h->metaf = print_otag(h, TAG_B, "");
259: print_otag(h, TAG_I, "");
1.31 schwarze 260: break;
1.125 schwarze 261: case ESCAPE_FONTCW:
1.112 schwarze 262: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
263: break;
1.31 schwarze 264: default:
265: break;
266: }
1.118 schwarze 267: }
268:
1.119 schwarze 269: void
270: html_close_paragraph(struct html *h)
271: {
272: struct tag *t;
273:
1.122 schwarze 274: for (t = h->tag; t != NULL && t->closed == 0; t = t->next) {
275: switch(t->tag) {
276: case TAG_P:
277: case TAG_PRE:
1.119 schwarze 278: print_tagq(h, t);
279: break;
1.122 schwarze 280: case TAG_A:
281: print_tagq(h, t);
282: continue;
283: default:
284: continue;
1.119 schwarze 285: }
1.122 schwarze 286: break;
1.119 schwarze 287: }
288: }
289:
1.118 schwarze 290: /*
291: * ROFF_nf switches to no-fill mode, ROFF_fi to fill mode.
292: * TOKEN_NONE does not switch. The old mode is returned.
293: */
294: enum roff_tok
295: html_fillmode(struct html *h, enum roff_tok want)
296: {
297: struct tag *t;
298: enum roff_tok had;
299:
300: for (t = h->tag; t != NULL; t = t->next)
301: if (t->tag == TAG_PRE)
302: break;
303:
304: had = t == NULL ? ROFF_fi : ROFF_nf;
305:
306: if (want != had) {
307: switch (want) {
308: case ROFF_fi:
309: print_tagq(h, t);
310: break;
311: case ROFF_nf:
1.119 schwarze 312: html_close_paragraph(h);
1.118 schwarze 313: print_otag(h, TAG_PRE, "");
314: break;
315: case TOKEN_NONE:
316: break;
317: default:
318: abort();
319: }
320: }
321: return had;
1.80 schwarze 322: }
323:
324: char *
1.99 schwarze 325: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 326: {
327: const struct roff_node *nch;
1.99 schwarze 328: char *buf, *bufs, *cp;
329: unsigned int slot;
330: int suffix;
1.80 schwarze 331:
332: for (nch = n->child; nch != NULL; nch = nch->next)
333: if (nch->type != ROFFT_TEXT)
334: return NULL;
335:
336: buf = NULL;
337: deroff(&buf, n);
1.90 schwarze 338: if (buf == NULL)
339: return NULL;
1.80 schwarze 340:
1.100 schwarze 341: /*
342: * In ID attributes, only use ASCII characters that are
343: * permitted in URL-fragment strings according to the
344: * explicit list at:
345: * https://url.spec.whatwg.org/#url-fragment-string
346: */
1.80 schwarze 347:
348: for (cp = buf; *cp != '\0'; cp++)
1.100 schwarze 349: if (isalnum((unsigned char)*cp) == 0 &&
350: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.80 schwarze 351: *cp = '_';
352:
1.99 schwarze 353: if (unique == 0)
354: return buf;
355:
356: /* Avoid duplicate HTML id= attributes. */
357:
358: bufs = NULL;
359: suffix = 1;
360: slot = ohash_qlookup(&id_unique, buf);
361: cp = ohash_find(&id_unique, slot);
362: if (cp != NULL) {
363: while (cp != NULL) {
364: free(bufs);
365: if (++suffix > 127) {
366: free(buf);
367: return NULL;
368: }
369: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
370: slot = ohash_qlookup(&id_unique, bufs);
371: cp = ohash_find(&id_unique, slot);
372: }
373: free(buf);
374: buf = bufs;
375: }
376: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 377: return buf;
1.1 schwarze 378: }
379:
1.5 schwarze 380: static int
1.67 schwarze 381: print_escape(struct html *h, char c)
1.38 schwarze 382: {
383:
384: switch (c) {
385: case '<':
1.67 schwarze 386: print_word(h, "<");
1.38 schwarze 387: break;
388: case '>':
1.67 schwarze 389: print_word(h, ">");
1.38 schwarze 390: break;
391: case '&':
1.67 schwarze 392: print_word(h, "&");
1.38 schwarze 393: break;
394: case '"':
1.67 schwarze 395: print_word(h, """);
1.38 schwarze 396: break;
397: case ASCII_NBRSP:
1.67 schwarze 398: print_word(h, " ");
1.38 schwarze 399: break;
400: case ASCII_HYPH:
1.67 schwarze 401: print_byte(h, '-');
1.59 schwarze 402: break;
1.38 schwarze 403: case ASCII_BREAK:
404: break;
405: default:
1.58 schwarze 406: return 0;
1.38 schwarze 407: }
1.58 schwarze 408: return 1;
1.38 schwarze 409: }
410:
411: static int
1.65 schwarze 412: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 413: {
1.67 schwarze 414: char numbuf[16];
1.84 schwarze 415: const char *seq;
1.4 schwarze 416: size_t sz;
1.84 schwarze 417: int c, len, breakline, nospace;
1.26 schwarze 418: enum mandoc_esc esc;
1.84 schwarze 419: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 420: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 421:
1.65 schwarze 422: if (pend == NULL)
423: pend = strchr(p, '\0');
424:
1.84 schwarze 425: breakline = 0;
1.5 schwarze 426: nospace = 0;
1.1 schwarze 427:
1.65 schwarze 428: while (p < pend) {
1.30 schwarze 429: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
430: h->flags &= ~HTML_SKIPCHAR;
431: p++;
432: continue;
433: }
434:
1.67 schwarze 435: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 436: print_byte(h, *p);
437:
438: if (breakline &&
439: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.115 schwarze 440: print_otag(h, TAG_BR, "");
1.84 schwarze 441: breakline = 0;
442: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
443: p++;
444: continue;
445: }
1.4 schwarze 446:
1.65 schwarze 447: if (p >= pend)
1.26 schwarze 448: break;
449:
1.84 schwarze 450: if (*p == ' ') {
451: print_endword(h);
452: p++;
453: continue;
454: }
455:
1.67 schwarze 456: if (print_escape(h, *p++))
1.33 schwarze 457: continue;
1.4 schwarze 458:
1.26 schwarze 459: esc = mandoc_escape(&p, &seq, &len);
460: switch (esc) {
1.35 schwarze 461: case ESCAPE_FONT:
462: case ESCAPE_FONTPREV:
463: case ESCAPE_FONTBOLD:
464: case ESCAPE_FONTITALIC:
465: case ESCAPE_FONTBI:
1.112 schwarze 466: case ESCAPE_FONTCW:
1.35 schwarze 467: case ESCAPE_FONTROMAN:
1.113 schwarze 468: if (0 == norecurse) {
469: h->flags |= HTML_NOSPACE;
1.125 schwarze 470: if (html_setfont(h, esc))
471: print_metaf(h);
1.113 schwarze 472: h->flags &= ~HTML_NOSPACE;
473: }
1.30 schwarze 474: continue;
1.35 schwarze 475: case ESCAPE_SKIPCHAR:
1.30 schwarze 476: h->flags |= HTML_SKIPCHAR;
477: continue;
1.116 schwarze 478: case ESCAPE_ERROR:
479: continue;
1.30 schwarze 480: default:
481: break;
482: }
483:
484: if (h->flags & HTML_SKIPCHAR) {
485: h->flags &= ~HTML_SKIPCHAR;
486: continue;
487: }
488:
489: switch (esc) {
1.35 schwarze 490: case ESCAPE_UNICODE:
1.38 schwarze 491: /* Skip past "u" header. */
1.26 schwarze 492: c = mchars_num2uc(seq + 1, len - 1);
493: break;
1.35 schwarze 494: case ESCAPE_NUMBERED:
1.26 schwarze 495: c = mchars_num2char(seq, len);
1.51 schwarze 496: if (c < 0)
497: continue;
1.26 schwarze 498: break;
1.35 schwarze 499: case ESCAPE_SPECIAL:
1.61 schwarze 500: c = mchars_spec2cp(seq, len);
1.51 schwarze 501: if (c <= 0)
502: continue;
1.116 schwarze 503: break;
504: case ESCAPE_UNDEF:
505: c = *seq;
1.26 schwarze 506: break;
1.109 schwarze 507: case ESCAPE_DEVICE:
508: print_word(h, "html");
509: continue;
1.84 schwarze 510: case ESCAPE_BREAK:
511: breakline = 1;
512: continue;
1.35 schwarze 513: case ESCAPE_NOSPACE:
1.26 schwarze 514: if ('\0' == *p)
515: nospace = 1;
1.49 schwarze 516: continue;
1.55 schwarze 517: case ESCAPE_OVERSTRIKE:
518: if (len == 0)
519: continue;
520: c = seq[len - 1];
521: break;
1.5 schwarze 522: default:
1.49 schwarze 523: continue;
1.5 schwarze 524: }
1.51 schwarze 525: if ((c < 0x20 && c != 0x09) ||
526: (c > 0x7E && c < 0xA0))
1.49 schwarze 527: c = 0xFFFD;
1.67 schwarze 528: if (c > 0x7E) {
1.86 bentley 529: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 530: print_word(h, numbuf);
531: } else if (print_escape(h, c) == 0)
532: print_byte(h, c);
1.1 schwarze 533: }
1.5 schwarze 534:
1.58 schwarze 535: return nospace;
1.1 schwarze 536: }
537:
1.6 schwarze 538: static void
1.65 schwarze 539: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 540: {
1.110 schwarze 541: struct stat sb;
1.65 schwarze 542: const char *p, *pp;
1.110 schwarze 543: char *filename;
544:
545: if (man) {
546: pp = h->base_man1;
547: if (h->base_man2 != NULL) {
548: mandoc_asprintf(&filename, "%s.%s", name, sec);
549: if (stat(filename, &sb) == -1)
550: pp = h->base_man2;
551: free(filename);
552: }
553: } else
554: pp = h->base_includes;
1.65 schwarze 555:
556: while ((p = strchr(pp, '%')) != NULL) {
557: print_encode(h, pp, p, 1);
558: if (man && p[1] == 'S') {
559: if (sec == NULL)
1.67 schwarze 560: print_byte(h, '1');
1.65 schwarze 561: else
562: print_encode(h, sec, NULL, 1);
563: } else if ((man && p[1] == 'N') ||
564: (man == 0 && p[1] == 'I'))
565: print_encode(h, name, NULL, 1);
566: else
567: print_encode(h, p, p + 2, 1);
568: pp = p + 2;
569: }
570: if (*pp != '\0')
571: print_encode(h, pp, NULL, 1);
1.6 schwarze 572: }
573:
1.1 schwarze 574: struct tag *
1.64 schwarze 575: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 576: {
1.64 schwarze 577: va_list ap;
1.1 schwarze 578: struct tag *t;
1.65 schwarze 579: const char *attr;
1.73 schwarze 580: char *arg1, *arg2;
1.114 schwarze 581: int style_written, tflags;
1.66 schwarze 582:
583: tflags = htmltags[tag].flags;
1.1 schwarze 584:
1.74 schwarze 585: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 586:
1.66 schwarze 587: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 588: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 589: t->tag = tag;
1.74 schwarze 590: t->next = h->tag;
1.122 schwarze 591: t->refcnt = 0;
592: t->closed = 0;
1.74 schwarze 593: h->tag = t;
1.1 schwarze 594: } else
595: t = NULL;
596:
1.66 schwarze 597: if (tflags & HTML_NLBEFORE)
1.67 schwarze 598: print_endline(h);
599: if (h->col == 0)
600: print_indent(h);
1.66 schwarze 601: else if ((h->flags & HTML_NOSPACE) == 0) {
602: if (h->flags & HTML_KEEP)
1.86 bentley 603: print_word(h, " ");
1.66 schwarze 604: else {
605: if (h->flags & HTML_PREKEEP)
606: h->flags |= HTML_KEEP;
1.67 schwarze 607: print_endword(h);
1.12 schwarze 608: }
1.66 schwarze 609: }
1.1 schwarze 610:
1.13 schwarze 611: if ( ! (h->flags & HTML_NONOSPACE))
612: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 613: else
614: h->flags |= HTML_NOSPACE;
1.13 schwarze 615:
1.6 schwarze 616: /* Print out the tag name and attributes. */
617:
1.67 schwarze 618: print_byte(h, '<');
619: print_word(h, htmltags[tag].name);
1.64 schwarze 620:
621: va_start(ap, fmt);
622:
1.114 schwarze 623: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 624:
1.108 schwarze 625: /* Parse attributes and arguments. */
1.73 schwarze 626:
627: arg1 = va_arg(ap, char *);
1.108 schwarze 628: arg2 = NULL;
1.64 schwarze 629: switch (*fmt++) {
630: case 'c':
1.65 schwarze 631: attr = "class";
1.64 schwarze 632: break;
633: case 'h':
1.65 schwarze 634: attr = "href";
1.64 schwarze 635: break;
636: case 'i':
1.65 schwarze 637: attr = "id";
1.64 schwarze 638: break;
639: case '?':
1.73 schwarze 640: attr = arg1;
641: arg1 = va_arg(ap, char *);
1.64 schwarze 642: break;
643: default:
644: abort();
645: }
1.73 schwarze 646: if (*fmt == 'M')
647: arg2 = va_arg(ap, char *);
648: if (arg1 == NULL)
649: continue;
650:
1.108 schwarze 651: /* Print the attributes. */
1.73 schwarze 652:
1.67 schwarze 653: print_byte(h, ' ');
654: print_word(h, attr);
655: print_byte(h, '=');
656: print_byte(h, '"');
1.65 schwarze 657: switch (*fmt) {
1.78 schwarze 658: case 'I':
659: print_href(h, arg1, NULL, 0);
660: fmt++;
661: break;
1.65 schwarze 662: case 'M':
1.73 schwarze 663: print_href(h, arg1, arg2, 1);
1.65 schwarze 664: fmt++;
665: break;
1.78 schwarze 666: case 'R':
667: print_byte(h, '#');
668: print_encode(h, arg1, NULL, 1);
1.65 schwarze 669: fmt++;
1.78 schwarze 670: break;
1.65 schwarze 671: default:
1.114 schwarze 672: print_encode(h, arg1, NULL, 1);
1.65 schwarze 673: break;
674: }
1.67 schwarze 675: print_byte(h, '"');
1.64 schwarze 676: }
1.114 schwarze 677:
678: style_written = 0;
679: while (*fmt++ == 's') {
680: arg1 = va_arg(ap, char *);
681: arg2 = va_arg(ap, char *);
682: if (arg2 == NULL)
683: continue;
684: print_byte(h, ' ');
685: if (style_written == 0) {
686: print_word(h, "style=\"");
687: style_written = 1;
688: }
689: print_word(h, arg1);
690: print_byte(h, ':');
691: print_byte(h, ' ');
692: print_word(h, arg2);
693: print_byte(h, ';');
694: }
695: if (style_written)
696: print_byte(h, '"');
697:
1.64 schwarze 698: va_end(ap);
1.6 schwarze 699:
1.42 schwarze 700: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 701:
702: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 703: print_byte(h, '/');
1.6 schwarze 704:
1.67 schwarze 705: print_byte(h, '>');
1.1 schwarze 706:
1.66 schwarze 707: if (tflags & HTML_NLBEGIN)
1.67 schwarze 708: print_endline(h);
1.66 schwarze 709: else
710: h->flags |= HTML_NOSPACE;
1.18 schwarze 711:
1.66 schwarze 712: if (tflags & HTML_INDENT)
713: h->indent++;
714: if (tflags & HTML_NOINDENT)
715: h->noindent++;
1.18 schwarze 716:
1.58 schwarze 717: return t;
1.1 schwarze 718: }
719:
720: static void
1.54 schwarze 721: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 722: {
1.66 schwarze 723: int tflags;
1.35 schwarze 724:
1.122 schwarze 725: if (tag->closed == 0) {
726: tag->closed = 1;
727: if (tag == h->metaf)
728: h->metaf = NULL;
729: if (tag == h->tblt)
730: h->tblt = NULL;
731:
732: tflags = htmltags[tag->tag].flags;
733: if (tflags & HTML_INDENT)
734: h->indent--;
735: if (tflags & HTML_NOINDENT)
736: h->noindent--;
737: if (tflags & HTML_NLEND)
738: print_endline(h);
739: print_indent(h);
740: print_byte(h, '<');
741: print_byte(h, '/');
742: print_word(h, htmltags[tag->tag].name);
743: print_byte(h, '>');
744: if (tflags & HTML_NLAFTER)
745: print_endline(h);
746: }
747: if (tag->refcnt == 0) {
748: h->tag = tag->next;
749: free(tag);
750: }
1.1 schwarze 751: }
752:
753: void
1.6 schwarze 754: print_gen_decls(struct html *h)
755: {
1.67 schwarze 756: print_word(h, "<!DOCTYPE html>");
757: print_endline(h);
1.91 schwarze 758: }
759:
760: void
761: print_gen_comment(struct html *h, struct roff_node *n)
762: {
763: int wantblank;
764:
765: print_word(h, "<!-- This is an automatically generated file."
766: " Do not edit.");
767: h->indent = 1;
768: wantblank = 0;
769: while (n != NULL && n->type == ROFFT_COMMENT) {
770: if (strstr(n->string, "-->") == NULL &&
771: (wantblank || *n->string != '\0')) {
772: print_endline(h);
773: print_indent(h);
774: print_word(h, n->string);
775: wantblank = *n->string != '\0';
776: }
777: n = n->next;
778: }
779: if (wantblank)
780: print_endline(h);
781: print_word(h, " -->");
782: print_endline(h);
783: h->indent = 0;
1.1 schwarze 784: }
785:
786: void
1.12 schwarze 787: print_text(struct html *h, const char *word)
1.1 schwarze 788: {
1.67 schwarze 789: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 790: if ( ! (HTML_KEEP & h->flags)) {
791: if (HTML_PREKEEP & h->flags)
792: h->flags |= HTML_KEEP;
1.67 schwarze 793: print_endword(h);
1.12 schwarze 794: } else
1.86 bentley 795: print_word(h, " ");
1.12 schwarze 796: }
1.1 schwarze 797:
1.125 schwarze 798: assert(h->metaf == NULL);
799: print_metaf(h);
800: print_indent(h);
1.65 schwarze 801: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 802: if ( ! (h->flags & HTML_NONOSPACE))
803: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 804: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 805: } else
1.53 schwarze 806: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 807:
1.125 schwarze 808: if (h->metaf != NULL) {
1.20 schwarze 809: print_tagq(h, h->metaf);
810: h->metaf = NULL;
811: }
1.17 schwarze 812:
813: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 814: }
815:
816: void
817: print_tagq(struct html *h, const struct tag *until)
818: {
1.122 schwarze 819: struct tag *this, *next;
1.1 schwarze 820:
1.122 schwarze 821: for (this = h->tag; this != NULL; this = next) {
822: next = this == until ? NULL : this->next;
823: print_ctag(h, this);
1.1 schwarze 824: }
825: }
826:
1.120 schwarze 827: /*
828: * Close out all open elements up to but excluding suntil.
829: * Note that a paragraph just inside stays open together with it
830: * because paragraphs include subsequent phrasing content.
831: */
1.1 schwarze 832: void
833: print_stagq(struct html *h, const struct tag *suntil)
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->next;
839: if (this == suntil || (next == suntil &&
840: (this->tag == TAG_P || this->tag == TAG_PRE)))
841: break;
842: print_ctag(h, this);
1.1 schwarze 843: }
1.42 schwarze 844: }
845:
1.67 schwarze 846:
847: /***********************************************************************
848: * Low level output functions.
849: * They implement line breaking using a short static buffer.
850: ***********************************************************************/
851:
852: /*
853: * Buffer one HTML output byte.
854: * If the buffer is full, flush and deactivate it and start a new line.
855: * If the buffer is inactive, print directly.
856: */
857: static void
858: print_byte(struct html *h, char c)
859: {
860: if ((h->flags & HTML_BUFFER) == 0) {
861: putchar(c);
862: h->col++;
863: return;
864: }
865:
866: if (h->col + h->bufcol < sizeof(h->buf)) {
867: h->buf[h->bufcol++] = c;
868: return;
869: }
870:
871: putchar('\n');
872: h->col = 0;
873: print_indent(h);
874: putchar(' ');
875: putchar(' ');
876: fwrite(h->buf, h->bufcol, 1, stdout);
877: putchar(c);
878: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
879: h->bufcol = 0;
880: h->flags &= ~HTML_BUFFER;
881: }
882:
1.66 schwarze 883: /*
884: * If something was printed on the current output line, end it.
1.67 schwarze 885: * Not to be called right after print_indent().
1.66 schwarze 886: */
1.72 schwarze 887: void
1.67 schwarze 888: print_endline(struct html *h)
1.66 schwarze 889: {
1.67 schwarze 890: if (h->col == 0)
1.66 schwarze 891: return;
892:
1.67 schwarze 893: if (h->bufcol) {
894: putchar(' ');
895: fwrite(h->buf, h->bufcol, 1, stdout);
896: h->bufcol = 0;
897: }
1.66 schwarze 898: putchar('\n');
1.67 schwarze 899: h->col = 0;
900: h->flags |= HTML_NOSPACE;
901: h->flags &= ~HTML_BUFFER;
902: }
903:
904: /*
905: * Flush the HTML output buffer.
906: * If it is inactive, activate it.
907: */
908: static void
909: print_endword(struct html *h)
910: {
911: if (h->noindent) {
912: print_byte(h, ' ');
913: return;
914: }
915:
916: if ((h->flags & HTML_BUFFER) == 0) {
917: h->col++;
918: h->flags |= HTML_BUFFER;
919: } else if (h->bufcol) {
920: putchar(' ');
921: fwrite(h->buf, h->bufcol, 1, stdout);
922: h->col += h->bufcol + 1;
923: }
924: h->bufcol = 0;
1.66 schwarze 925: }
926:
927: /*
928: * If at the beginning of a new output line,
929: * perform indentation and mark the line as containing output.
930: * Make sure to really produce some output right afterwards,
931: * but do not use print_otag() for producing it.
932: */
933: static void
1.67 schwarze 934: print_indent(struct html *h)
1.66 schwarze 935: {
1.67 schwarze 936: size_t i;
1.66 schwarze 937:
1.67 schwarze 938: if (h->col)
1.66 schwarze 939: return;
940:
1.67 schwarze 941: if (h->noindent == 0) {
942: h->col = h->indent * 2;
943: for (i = 0; i < h->col; i++)
1.66 schwarze 944: putchar(' ');
1.67 schwarze 945: }
946: h->flags &= ~HTML_NOSPACE;
947: }
948:
949: /*
950: * Print or buffer some characters
951: * depending on the current HTML output buffer state.
952: */
953: static void
954: print_word(struct html *h, const char *cp)
955: {
956: while (*cp != '\0')
957: print_byte(h, *cp++);
1.3 schwarze 958: }