Annotation of src/usr.bin/mandoc/html.c, Revision 1.129
1.129 ! schwarze 1: /* $OpenBSD: html.c,v 1.128 2019/09/01 15:12:03 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;
1.127 schwarze 43: #define HTML_INPHRASE (1 << 0) /* Can appear in phrasing context. */
44: #define HTML_TOPHRASE (1 << 1) /* Establishes phrasing context. */
45: #define HTML_NOSTACK (1 << 2) /* Does not have an end tag. */
46: #define HTML_NLBEFORE (1 << 3) /* Output line break before opening. */
47: #define HTML_NLBEGIN (1 << 4) /* Output line break after opening. */
48: #define HTML_NLEND (1 << 5) /* Output line break before closing. */
49: #define HTML_NLAFTER (1 << 6) /* Output line break after closing. */
1.66 schwarze 50: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
51: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
52: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
1.127 schwarze 53: #define HTML_INDENT (1 << 7) /* Indent content by two spaces. */
54: #define HTML_NOINDENT (1 << 8) /* Exception: never indent content. */
1.1 schwarze 55: };
56:
57: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 58: {"html", HTML_NLALL},
59: {"head", HTML_NLALL | HTML_INDENT},
1.127 schwarze 60: {"meta", HTML_NOSTACK | HTML_NLALL},
61: {"link", HTML_NOSTACK | HTML_NLALL},
62: {"style", HTML_NLALL | HTML_INDENT},
63: {"title", HTML_NLAROUND},
1.66 schwarze 64: {"body", HTML_NLALL},
65: {"div", HTML_NLAROUND},
1.123 schwarze 66: {"section", HTML_NLALL},
1.66 schwarze 67: {"table", HTML_NLALL | HTML_INDENT},
68: {"tr", HTML_NLALL | HTML_INDENT},
69: {"td", HTML_NLAROUND},
70: {"li", HTML_NLAROUND | HTML_INDENT},
71: {"ul", HTML_NLALL | HTML_INDENT},
72: {"ol", HTML_NLALL | HTML_INDENT},
73: {"dl", HTML_NLALL | HTML_INDENT},
74: {"dt", HTML_NLAROUND},
75: {"dd", HTML_NLAROUND | HTML_INDENT},
1.127 schwarze 76: {"h1", HTML_TOPHRASE | HTML_NLAROUND},
77: {"h2", HTML_TOPHRASE | HTML_NLAROUND},
78: {"p", HTML_TOPHRASE | HTML_NLAROUND | HTML_INDENT},
79: {"pre", HTML_TOPHRASE | HTML_NLALL | HTML_NOINDENT},
80: {"a", HTML_INPHRASE | HTML_TOPHRASE},
81: {"b", HTML_INPHRASE | HTML_TOPHRASE},
82: {"cite", HTML_INPHRASE | HTML_TOPHRASE},
83: {"code", HTML_INPHRASE | HTML_TOPHRASE},
84: {"i", HTML_INPHRASE | HTML_TOPHRASE},
85: {"small", HTML_INPHRASE | HTML_TOPHRASE},
86: {"span", HTML_INPHRASE | HTML_TOPHRASE},
87: {"var", HTML_INPHRASE | HTML_TOPHRASE},
88: {"br", HTML_INPHRASE | HTML_NOSTACK | HTML_NLALL},
89: {"math", HTML_INPHRASE | HTML_NLALL | HTML_INDENT},
1.66 schwarze 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: {
1.129 ! schwarze 272: struct tag *this, *next;
1.119 schwarze 273:
1.129 ! schwarze 274: this = h->tag;
! 275: for (;;) {
! 276: next = this->next;
! 277: if (htmltags[this->tag].flags &
! 278: (HTML_INPHRASE | HTML_TOPHRASE))
! 279: print_ctag(h, this);
! 280: if ((htmltags[this->tag].flags & HTML_INPHRASE) == 0)
1.119 schwarze 281: break;
1.129 ! schwarze 282: this = next;
1.119 schwarze 283: }
284: }
285:
1.118 schwarze 286: /*
287: * ROFF_nf switches to no-fill mode, ROFF_fi to fill mode.
288: * TOKEN_NONE does not switch. The old mode is returned.
289: */
290: enum roff_tok
291: html_fillmode(struct html *h, enum roff_tok want)
292: {
293: struct tag *t;
294: enum roff_tok had;
295:
296: for (t = h->tag; t != NULL; t = t->next)
297: if (t->tag == TAG_PRE)
298: break;
299:
300: had = t == NULL ? ROFF_fi : ROFF_nf;
301:
302: if (want != had) {
303: switch (want) {
304: case ROFF_fi:
305: print_tagq(h, t);
306: break;
307: case ROFF_nf:
1.119 schwarze 308: html_close_paragraph(h);
1.118 schwarze 309: print_otag(h, TAG_PRE, "");
310: break;
311: case TOKEN_NONE:
312: break;
313: default:
314: abort();
315: }
316: }
317: return had;
1.80 schwarze 318: }
319:
320: char *
1.99 schwarze 321: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 322: {
323: const struct roff_node *nch;
1.99 schwarze 324: char *buf, *bufs, *cp;
325: unsigned int slot;
326: int suffix;
1.80 schwarze 327:
328: for (nch = n->child; nch != NULL; nch = nch->next)
329: if (nch->type != ROFFT_TEXT)
330: return NULL;
331:
332: buf = NULL;
333: deroff(&buf, n);
1.90 schwarze 334: if (buf == NULL)
335: return NULL;
1.80 schwarze 336:
1.100 schwarze 337: /*
338: * In ID attributes, only use ASCII characters that are
339: * permitted in URL-fragment strings according to the
340: * explicit list at:
341: * https://url.spec.whatwg.org/#url-fragment-string
342: */
1.80 schwarze 343:
344: for (cp = buf; *cp != '\0'; cp++)
1.100 schwarze 345: if (isalnum((unsigned char)*cp) == 0 &&
346: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.80 schwarze 347: *cp = '_';
348:
1.99 schwarze 349: if (unique == 0)
350: return buf;
351:
352: /* Avoid duplicate HTML id= attributes. */
353:
354: bufs = NULL;
355: suffix = 1;
356: slot = ohash_qlookup(&id_unique, buf);
357: cp = ohash_find(&id_unique, slot);
358: if (cp != NULL) {
359: while (cp != NULL) {
360: free(bufs);
361: if (++suffix > 127) {
362: free(buf);
363: return NULL;
364: }
365: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
366: slot = ohash_qlookup(&id_unique, bufs);
367: cp = ohash_find(&id_unique, slot);
368: }
369: free(buf);
370: buf = bufs;
371: }
372: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 373: return buf;
1.1 schwarze 374: }
375:
1.5 schwarze 376: static int
1.67 schwarze 377: print_escape(struct html *h, char c)
1.38 schwarze 378: {
379:
380: switch (c) {
381: case '<':
1.67 schwarze 382: print_word(h, "<");
1.38 schwarze 383: break;
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 ASCII_NBRSP:
1.67 schwarze 394: print_word(h, " ");
1.38 schwarze 395: break;
396: case ASCII_HYPH:
1.67 schwarze 397: print_byte(h, '-');
1.59 schwarze 398: break;
1.38 schwarze 399: case ASCII_BREAK:
400: break;
401: default:
1.58 schwarze 402: return 0;
1.38 schwarze 403: }
1.58 schwarze 404: return 1;
1.38 schwarze 405: }
406:
407: static int
1.65 schwarze 408: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 409: {
1.67 schwarze 410: char numbuf[16];
1.84 schwarze 411: const char *seq;
1.4 schwarze 412: size_t sz;
1.84 schwarze 413: int c, len, breakline, nospace;
1.26 schwarze 414: enum mandoc_esc esc;
1.84 schwarze 415: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 416: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 417:
1.65 schwarze 418: if (pend == NULL)
419: pend = strchr(p, '\0');
420:
1.84 schwarze 421: breakline = 0;
1.5 schwarze 422: nospace = 0;
1.1 schwarze 423:
1.65 schwarze 424: while (p < pend) {
1.30 schwarze 425: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
426: h->flags &= ~HTML_SKIPCHAR;
427: p++;
428: continue;
429: }
430:
1.67 schwarze 431: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 432: print_byte(h, *p);
433:
434: if (breakline &&
435: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.115 schwarze 436: print_otag(h, TAG_BR, "");
1.84 schwarze 437: breakline = 0;
438: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
439: p++;
440: continue;
441: }
1.4 schwarze 442:
1.65 schwarze 443: if (p >= pend)
1.26 schwarze 444: break;
445:
1.84 schwarze 446: if (*p == ' ') {
447: print_endword(h);
448: p++;
449: continue;
450: }
451:
1.67 schwarze 452: if (print_escape(h, *p++))
1.33 schwarze 453: continue;
1.4 schwarze 454:
1.26 schwarze 455: esc = mandoc_escape(&p, &seq, &len);
456: switch (esc) {
1.35 schwarze 457: case ESCAPE_FONT:
458: case ESCAPE_FONTPREV:
459: case ESCAPE_FONTBOLD:
460: case ESCAPE_FONTITALIC:
461: case ESCAPE_FONTBI:
1.112 schwarze 462: case ESCAPE_FONTCW:
1.35 schwarze 463: case ESCAPE_FONTROMAN:
1.113 schwarze 464: if (0 == norecurse) {
465: h->flags |= HTML_NOSPACE;
1.125 schwarze 466: if (html_setfont(h, esc))
467: print_metaf(h);
1.113 schwarze 468: h->flags &= ~HTML_NOSPACE;
469: }
1.30 schwarze 470: continue;
1.35 schwarze 471: case ESCAPE_SKIPCHAR:
1.30 schwarze 472: h->flags |= HTML_SKIPCHAR;
473: continue;
1.116 schwarze 474: case ESCAPE_ERROR:
475: continue;
1.30 schwarze 476: default:
477: break;
478: }
479:
480: if (h->flags & HTML_SKIPCHAR) {
481: h->flags &= ~HTML_SKIPCHAR;
482: continue;
483: }
484:
485: switch (esc) {
1.35 schwarze 486: case ESCAPE_UNICODE:
1.38 schwarze 487: /* Skip past "u" header. */
1.26 schwarze 488: c = mchars_num2uc(seq + 1, len - 1);
489: break;
1.35 schwarze 490: case ESCAPE_NUMBERED:
1.26 schwarze 491: c = mchars_num2char(seq, len);
1.51 schwarze 492: if (c < 0)
493: continue;
1.26 schwarze 494: break;
1.35 schwarze 495: case ESCAPE_SPECIAL:
1.61 schwarze 496: c = mchars_spec2cp(seq, len);
1.51 schwarze 497: if (c <= 0)
498: continue;
1.116 schwarze 499: break;
500: case ESCAPE_UNDEF:
501: c = *seq;
1.26 schwarze 502: break;
1.109 schwarze 503: case ESCAPE_DEVICE:
504: print_word(h, "html");
505: continue;
1.84 schwarze 506: case ESCAPE_BREAK:
507: breakline = 1;
508: continue;
1.35 schwarze 509: case ESCAPE_NOSPACE:
1.26 schwarze 510: if ('\0' == *p)
511: nospace = 1;
1.49 schwarze 512: continue;
1.55 schwarze 513: case ESCAPE_OVERSTRIKE:
514: if (len == 0)
515: continue;
516: c = seq[len - 1];
517: break;
1.5 schwarze 518: default:
1.49 schwarze 519: continue;
1.5 schwarze 520: }
1.51 schwarze 521: if ((c < 0x20 && c != 0x09) ||
522: (c > 0x7E && c < 0xA0))
1.49 schwarze 523: c = 0xFFFD;
1.67 schwarze 524: if (c > 0x7E) {
1.86 bentley 525: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 526: print_word(h, numbuf);
527: } else if (print_escape(h, c) == 0)
528: print_byte(h, c);
1.1 schwarze 529: }
1.5 schwarze 530:
1.58 schwarze 531: return nospace;
1.1 schwarze 532: }
533:
1.6 schwarze 534: static void
1.65 schwarze 535: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 536: {
1.110 schwarze 537: struct stat sb;
1.65 schwarze 538: const char *p, *pp;
1.110 schwarze 539: char *filename;
540:
541: if (man) {
542: pp = h->base_man1;
543: if (h->base_man2 != NULL) {
544: mandoc_asprintf(&filename, "%s.%s", name, sec);
545: if (stat(filename, &sb) == -1)
546: pp = h->base_man2;
547: free(filename);
548: }
549: } else
550: pp = h->base_includes;
1.65 schwarze 551:
552: while ((p = strchr(pp, '%')) != NULL) {
553: print_encode(h, pp, p, 1);
554: if (man && p[1] == 'S') {
555: if (sec == NULL)
1.67 schwarze 556: print_byte(h, '1');
1.65 schwarze 557: else
558: print_encode(h, sec, NULL, 1);
559: } else if ((man && p[1] == 'N') ||
560: (man == 0 && p[1] == 'I'))
561: print_encode(h, name, NULL, 1);
562: else
563: print_encode(h, p, p + 2, 1);
564: pp = p + 2;
565: }
566: if (*pp != '\0')
567: print_encode(h, pp, NULL, 1);
1.6 schwarze 568: }
569:
1.1 schwarze 570: struct tag *
1.64 schwarze 571: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 572: {
1.64 schwarze 573: va_list ap;
1.1 schwarze 574: struct tag *t;
1.65 schwarze 575: const char *attr;
1.73 schwarze 576: char *arg1, *arg2;
1.114 schwarze 577: int style_written, tflags;
1.66 schwarze 578:
579: tflags = htmltags[tag].flags;
1.1 schwarze 580:
1.127 schwarze 581: /* Flow content is not allowed in phrasing context. */
582:
583: if ((tflags & HTML_INPHRASE) == 0) {
584: for (t = h->tag; t != NULL; t = t->next) {
585: if (t->closed)
586: continue;
587: assert((htmltags[t->tag].flags & HTML_TOPHRASE) == 0);
588: break;
589: }
590: }
591:
1.74 schwarze 592: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 593:
1.66 schwarze 594: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 595: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 596: t->tag = tag;
1.74 schwarze 597: t->next = h->tag;
1.122 schwarze 598: t->refcnt = 0;
599: t->closed = 0;
1.74 schwarze 600: h->tag = t;
1.1 schwarze 601: } else
602: t = NULL;
603:
1.66 schwarze 604: if (tflags & HTML_NLBEFORE)
1.67 schwarze 605: print_endline(h);
606: if (h->col == 0)
607: print_indent(h);
1.66 schwarze 608: else if ((h->flags & HTML_NOSPACE) == 0) {
609: if (h->flags & HTML_KEEP)
1.86 bentley 610: print_word(h, " ");
1.66 schwarze 611: else {
612: if (h->flags & HTML_PREKEEP)
613: h->flags |= HTML_KEEP;
1.67 schwarze 614: print_endword(h);
1.12 schwarze 615: }
1.66 schwarze 616: }
1.1 schwarze 617:
1.13 schwarze 618: if ( ! (h->flags & HTML_NONOSPACE))
619: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 620: else
621: h->flags |= HTML_NOSPACE;
1.13 schwarze 622:
1.6 schwarze 623: /* Print out the tag name and attributes. */
624:
1.67 schwarze 625: print_byte(h, '<');
626: print_word(h, htmltags[tag].name);
1.64 schwarze 627:
628: va_start(ap, fmt);
629:
1.114 schwarze 630: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 631:
1.108 schwarze 632: /* Parse attributes and arguments. */
1.73 schwarze 633:
634: arg1 = va_arg(ap, char *);
1.108 schwarze 635: arg2 = NULL;
1.64 schwarze 636: switch (*fmt++) {
637: case 'c':
1.65 schwarze 638: attr = "class";
1.64 schwarze 639: break;
640: case 'h':
1.65 schwarze 641: attr = "href";
1.64 schwarze 642: break;
643: case 'i':
1.65 schwarze 644: attr = "id";
1.64 schwarze 645: break;
646: case '?':
1.73 schwarze 647: attr = arg1;
648: arg1 = va_arg(ap, char *);
1.64 schwarze 649: break;
650: default:
651: abort();
652: }
1.73 schwarze 653: if (*fmt == 'M')
654: arg2 = va_arg(ap, char *);
655: if (arg1 == NULL)
656: continue;
657:
1.108 schwarze 658: /* Print the attributes. */
1.73 schwarze 659:
1.67 schwarze 660: print_byte(h, ' ');
661: print_word(h, attr);
662: print_byte(h, '=');
663: print_byte(h, '"');
1.65 schwarze 664: switch (*fmt) {
1.78 schwarze 665: case 'I':
666: print_href(h, arg1, NULL, 0);
667: fmt++;
668: break;
1.65 schwarze 669: case 'M':
1.73 schwarze 670: print_href(h, arg1, arg2, 1);
1.65 schwarze 671: fmt++;
672: break;
1.78 schwarze 673: case 'R':
674: print_byte(h, '#');
675: print_encode(h, arg1, NULL, 1);
1.65 schwarze 676: fmt++;
1.78 schwarze 677: break;
1.65 schwarze 678: default:
1.114 schwarze 679: print_encode(h, arg1, NULL, 1);
1.65 schwarze 680: break;
681: }
1.67 schwarze 682: print_byte(h, '"');
1.64 schwarze 683: }
1.114 schwarze 684:
685: style_written = 0;
686: while (*fmt++ == 's') {
687: arg1 = va_arg(ap, char *);
688: arg2 = va_arg(ap, char *);
689: if (arg2 == NULL)
690: continue;
691: print_byte(h, ' ');
692: if (style_written == 0) {
693: print_word(h, "style=\"");
694: style_written = 1;
695: }
696: print_word(h, arg1);
697: print_byte(h, ':');
698: print_byte(h, ' ');
699: print_word(h, arg2);
700: print_byte(h, ';');
701: }
702: if (style_written)
703: print_byte(h, '"');
704:
1.64 schwarze 705: va_end(ap);
1.6 schwarze 706:
1.42 schwarze 707: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 708:
1.127 schwarze 709: if (htmltags[tag].flags & HTML_NOSTACK)
1.67 schwarze 710: print_byte(h, '/');
1.6 schwarze 711:
1.67 schwarze 712: print_byte(h, '>');
1.1 schwarze 713:
1.66 schwarze 714: if (tflags & HTML_NLBEGIN)
1.67 schwarze 715: print_endline(h);
1.66 schwarze 716: else
717: h->flags |= HTML_NOSPACE;
1.18 schwarze 718:
1.66 schwarze 719: if (tflags & HTML_INDENT)
720: h->indent++;
721: if (tflags & HTML_NOINDENT)
722: h->noindent++;
1.18 schwarze 723:
1.58 schwarze 724: return t;
1.1 schwarze 725: }
726:
727: static void
1.54 schwarze 728: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 729: {
1.66 schwarze 730: int tflags;
1.35 schwarze 731:
1.122 schwarze 732: if (tag->closed == 0) {
733: tag->closed = 1;
734: if (tag == h->metaf)
735: h->metaf = NULL;
736: if (tag == h->tblt)
737: h->tblt = NULL;
738:
739: tflags = htmltags[tag->tag].flags;
740: if (tflags & HTML_INDENT)
741: h->indent--;
742: if (tflags & HTML_NOINDENT)
743: h->noindent--;
744: if (tflags & HTML_NLEND)
745: print_endline(h);
746: print_indent(h);
747: print_byte(h, '<');
748: print_byte(h, '/');
749: print_word(h, htmltags[tag->tag].name);
750: print_byte(h, '>');
751: if (tflags & HTML_NLAFTER)
752: print_endline(h);
753: }
754: if (tag->refcnt == 0) {
755: h->tag = tag->next;
756: free(tag);
757: }
1.1 schwarze 758: }
759:
760: void
1.6 schwarze 761: print_gen_decls(struct html *h)
762: {
1.67 schwarze 763: print_word(h, "<!DOCTYPE html>");
764: print_endline(h);
1.91 schwarze 765: }
766:
767: void
768: print_gen_comment(struct html *h, struct roff_node *n)
769: {
770: int wantblank;
771:
772: print_word(h, "<!-- This is an automatically generated file."
773: " Do not edit.");
774: h->indent = 1;
775: wantblank = 0;
776: while (n != NULL && n->type == ROFFT_COMMENT) {
777: if (strstr(n->string, "-->") == NULL &&
778: (wantblank || *n->string != '\0')) {
779: print_endline(h);
780: print_indent(h);
781: print_word(h, n->string);
782: wantblank = *n->string != '\0';
783: }
784: n = n->next;
785: }
786: if (wantblank)
787: print_endline(h);
788: print_word(h, " -->");
789: print_endline(h);
790: h->indent = 0;
1.1 schwarze 791: }
792:
793: void
1.12 schwarze 794: print_text(struct html *h, const char *word)
1.1 schwarze 795: {
1.67 schwarze 796: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 797: if ( ! (HTML_KEEP & h->flags)) {
798: if (HTML_PREKEEP & h->flags)
799: h->flags |= HTML_KEEP;
1.67 schwarze 800: print_endword(h);
1.12 schwarze 801: } else
1.86 bentley 802: print_word(h, " ");
1.12 schwarze 803: }
1.1 schwarze 804:
1.125 schwarze 805: assert(h->metaf == NULL);
806: print_metaf(h);
807: print_indent(h);
1.65 schwarze 808: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 809: if ( ! (h->flags & HTML_NONOSPACE))
810: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 811: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 812: } else
1.53 schwarze 813: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 814:
1.125 schwarze 815: if (h->metaf != NULL) {
1.20 schwarze 816: print_tagq(h, h->metaf);
817: h->metaf = NULL;
818: }
1.17 schwarze 819:
820: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 821: }
822:
823: void
824: print_tagq(struct html *h, const struct tag *until)
825: {
1.122 schwarze 826: struct tag *this, *next;
1.1 schwarze 827:
1.122 schwarze 828: for (this = h->tag; this != NULL; this = next) {
829: next = this == until ? NULL : this->next;
830: print_ctag(h, this);
1.1 schwarze 831: }
832: }
833:
1.120 schwarze 834: /*
835: * Close out all open elements up to but excluding suntil.
836: * Note that a paragraph just inside stays open together with it
837: * because paragraphs include subsequent phrasing content.
838: */
1.1 schwarze 839: void
840: print_stagq(struct html *h, const struct tag *suntil)
841: {
1.122 schwarze 842: struct tag *this, *next;
1.1 schwarze 843:
1.122 schwarze 844: for (this = h->tag; this != NULL; this = next) {
845: next = this->next;
846: if (this == suntil || (next == suntil &&
847: (this->tag == TAG_P || this->tag == TAG_PRE)))
848: break;
849: print_ctag(h, this);
1.1 schwarze 850: }
1.42 schwarze 851: }
852:
1.67 schwarze 853:
854: /***********************************************************************
855: * Low level output functions.
856: * They implement line breaking using a short static buffer.
857: ***********************************************************************/
858:
859: /*
860: * Buffer one HTML output byte.
861: * If the buffer is full, flush and deactivate it and start a new line.
862: * If the buffer is inactive, print directly.
863: */
864: static void
865: print_byte(struct html *h, char c)
866: {
867: if ((h->flags & HTML_BUFFER) == 0) {
868: putchar(c);
869: h->col++;
870: return;
871: }
872:
873: if (h->col + h->bufcol < sizeof(h->buf)) {
874: h->buf[h->bufcol++] = c;
875: return;
876: }
877:
878: putchar('\n');
879: h->col = 0;
880: print_indent(h);
881: putchar(' ');
882: putchar(' ');
883: fwrite(h->buf, h->bufcol, 1, stdout);
884: putchar(c);
885: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
886: h->bufcol = 0;
887: h->flags &= ~HTML_BUFFER;
888: }
889:
1.66 schwarze 890: /*
891: * If something was printed on the current output line, end it.
1.67 schwarze 892: * Not to be called right after print_indent().
1.66 schwarze 893: */
1.72 schwarze 894: void
1.67 schwarze 895: print_endline(struct html *h)
1.66 schwarze 896: {
1.67 schwarze 897: if (h->col == 0)
1.66 schwarze 898: return;
899:
1.67 schwarze 900: if (h->bufcol) {
901: putchar(' ');
902: fwrite(h->buf, h->bufcol, 1, stdout);
903: h->bufcol = 0;
904: }
1.66 schwarze 905: putchar('\n');
1.67 schwarze 906: h->col = 0;
907: h->flags |= HTML_NOSPACE;
908: h->flags &= ~HTML_BUFFER;
909: }
910:
911: /*
912: * Flush the HTML output buffer.
913: * If it is inactive, activate it.
914: */
915: static void
916: print_endword(struct html *h)
917: {
918: if (h->noindent) {
919: print_byte(h, ' ');
920: return;
921: }
922:
923: if ((h->flags & HTML_BUFFER) == 0) {
924: h->col++;
925: h->flags |= HTML_BUFFER;
926: } else if (h->bufcol) {
927: putchar(' ');
928: fwrite(h->buf, h->bufcol, 1, stdout);
929: h->col += h->bufcol + 1;
930: }
931: h->bufcol = 0;
1.66 schwarze 932: }
933:
934: /*
935: * If at the beginning of a new output line,
936: * perform indentation and mark the line as containing output.
937: * Make sure to really produce some output right afterwards,
938: * but do not use print_otag() for producing it.
939: */
940: static void
1.67 schwarze 941: print_indent(struct html *h)
1.66 schwarze 942: {
1.67 schwarze 943: size_t i;
1.66 schwarze 944:
1.67 schwarze 945: if (h->col)
1.66 schwarze 946: return;
947:
1.67 schwarze 948: if (h->noindent == 0) {
949: h->col = h->indent * 2;
950: for (i = 0; i < h->col; i++)
1.66 schwarze 951: putchar(' ');
1.67 schwarze 952: }
953: h->flags &= ~HTML_NOSPACE;
954: }
955:
956: /*
957: * Print or buffer some characters
958: * depending on the current HTML output buffer state.
959: */
960: static void
961: print_word(struct html *h, const char *cp)
962: {
963: while (*cp != '\0')
964: print_byte(h, *cp++);
1.3 schwarze 965: }