Annotation of src/usr.bin/mandoc/html.c, Revision 1.99
1.99 ! schwarze 1: /* $OpenBSD: html.c,v 1.98 2018/05/21 01:10:06 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>
19:
20: #include <assert.h>
1.3 schwarze 21: #include <ctype.h>
1.4 schwarze 22: #include <stdarg.h>
1.99 ! schwarze 23: #include <stddef.h>
1.1 schwarze 24: #include <stdio.h>
25: #include <stdint.h>
26: #include <stdlib.h>
27: #include <string.h>
28: #include <unistd.h>
29:
1.80 schwarze 30: #include "mandoc_aux.h"
1.99 ! schwarze 31: #include "mandoc_ohash.h"
1.9 schwarze 32: #include "mandoc.h"
1.80 schwarze 33: #include "roff.h"
1.1 schwarze 34: #include "out.h"
35: #include "html.h"
1.56 schwarze 36: #include "manconf.h"
1.1 schwarze 37: #include "main.h"
38:
39: struct htmldata {
40: const char *name;
41: int flags;
1.66 schwarze 42: #define HTML_NOSTACK (1 << 0)
43: #define HTML_AUTOCLOSE (1 << 1)
44: #define HTML_NLBEFORE (1 << 2)
45: #define HTML_NLBEGIN (1 << 3)
46: #define HTML_NLEND (1 << 4)
47: #define HTML_NLAFTER (1 << 5)
48: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
49: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
50: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
51: #define HTML_INDENT (1 << 6)
52: #define HTML_NOINDENT (1 << 7)
1.1 schwarze 53: };
54:
55: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 56: {"html", HTML_NLALL},
57: {"head", HTML_NLALL | HTML_INDENT},
58: {"body", HTML_NLALL},
59: {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
60: {"title", HTML_NLAROUND},
61: {"div", HTML_NLAROUND},
1.95 schwarze 62: {"div", 0},
1.66 schwarze 63: {"h1", HTML_NLAROUND},
64: {"h2", HTML_NLAROUND},
65: {"span", 0},
66: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
67: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
68: {"a", 0},
69: {"table", HTML_NLALL | HTML_INDENT},
1.75 schwarze 70: {"colgroup", HTML_NLALL | HTML_INDENT},
1.66 schwarze 71: {"col", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
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},
80: {"pre", HTML_NLALL | HTML_NOINDENT},
1.77 schwarze 81: {"var", 0},
1.76 schwarze 82: {"cite", 0},
1.66 schwarze 83: {"b", 0},
84: {"i", 0},
85: {"code", 0},
86: {"small", 0},
87: {"style", HTML_NLALL | HTML_INDENT},
88: {"math", HTML_NLALL | HTML_INDENT},
89: {"mrow", 0},
90: {"mi", 0},
1.85 schwarze 91: {"mn", 0},
1.66 schwarze 92: {"mo", 0},
93: {"msup", 0},
94: {"msub", 0},
95: {"msubsup", 0},
96: {"mfrac", 0},
97: {"msqrt", 0},
98: {"mfenced", 0},
99: {"mtable", 0},
100: {"mtr", 0},
101: {"mtd", 0},
102: {"munderover", 0},
103: {"munder", 0},
104: {"mover", 0},
1.5 schwarze 105: };
106:
1.26 schwarze 107: static const char *const roffscales[SCALE_MAX] = {
108: "cm", /* SCALE_CM */
109: "in", /* SCALE_IN */
110: "pc", /* SCALE_PC */
111: "pt", /* SCALE_PT */
112: "em", /* SCALE_EM */
113: "em", /* SCALE_MM */
114: "ex", /* SCALE_EN */
115: "ex", /* SCALE_BU */
116: "em", /* SCALE_VS */
117: "ex", /* SCALE_FS */
118: };
1.5 schwarze 119:
1.99 ! schwarze 120: /* Avoid duplicate HTML id= attributes. */
! 121: static struct ohash id_unique;
! 122:
1.64 schwarze 123: static void a2width(const char *, struct roffsu *);
1.67 schwarze 124: static void print_byte(struct html *, char);
125: static void print_endword(struct html *);
126: static void print_indent(struct html *);
127: static void print_word(struct html *, const char *);
128:
1.54 schwarze 129: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 130: static int print_escape(struct html *, char);
1.65 schwarze 131: static int print_encode(struct html *, const char *, const char *, int);
132: static void print_href(struct html *, const char *, const char *, int);
1.26 schwarze 133: static void print_metaf(struct html *, enum mandoc_esc);
1.5 schwarze 134:
1.35 schwarze 135:
1.50 schwarze 136: void *
1.61 schwarze 137: html_alloc(const struct manoutput *outopts)
1.1 schwarze 138: {
139: struct html *h;
140:
1.24 schwarze 141: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 142:
1.74 schwarze 143: h->tag = NULL;
1.56 schwarze 144: h->style = outopts->style;
145: h->base_man = outopts->man;
146: h->base_includes = outopts->includes;
147: if (outopts->fragment)
148: h->oflags |= HTML_FRAGMENT;
1.1 schwarze 149:
1.99 ! schwarze 150: mandoc_ohash_init(&id_unique, 4, 0);
! 151:
1.58 schwarze 152: return h;
1.1 schwarze 153: }
154:
155: void
156: html_free(void *p)
157: {
158: struct tag *tag;
159: struct html *h;
1.99 ! schwarze 160: char *cp;
! 161: unsigned int slot;
1.1 schwarze 162:
163: h = (struct html *)p;
1.74 schwarze 164: while ((tag = h->tag) != NULL) {
165: h->tag = tag->next;
1.1 schwarze 166: free(tag);
167: }
1.99 ! schwarze 168: free(h);
1.1 schwarze 169:
1.99 ! schwarze 170: cp = ohash_first(&id_unique, &slot);
! 171: while (cp != NULL) {
! 172: free(cp);
! 173: cp = ohash_next(&id_unique, &slot);
! 174: }
! 175: ohash_delete(&id_unique);
1.1 schwarze 176: }
177:
178: void
179: print_gen_head(struct html *h)
180: {
1.42 schwarze 181: struct tag *t;
182:
1.64 schwarze 183: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 184: if (h->style != NULL) {
185: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
186: h->style, "type", "text/css", "media", "all");
187: return;
188: }
1.1 schwarze 189:
1.42 schwarze 190: /*
1.92 schwarze 191: * Print a minimal embedded style sheet.
1.42 schwarze 192: */
1.66 schwarze 193:
1.64 schwarze 194: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 195: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 196: print_endline(h);
1.66 schwarze 197: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 198: print_endline(h);
1.66 schwarze 199: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 200: print_endline(h);
1.68 schwarze 201: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.95 schwarze 202: print_endline(h);
203: print_text(h, "div.Nd, div.Bf, div.Op { display: inline; }");
1.96 schwarze 204: print_endline(h);
1.97 schwarze 205: print_text(h, "span.Pa, span.Ad { font-style: italic; }");
1.98 schwarze 206: print_endline(h);
207: print_text(h, "span.Ms { font-weight: bold; }");
1.94 schwarze 208: print_endline(h);
209: print_text(h, "dl.Bl-diag ");
210: print_byte(h, '>');
211: print_text(h, " dt { font-weight: bold; }");
1.93 schwarze 212: print_endline(h);
213: print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, "
214: "code.In, code.Fd, code.Fn,");
215: print_endline(h);
216: print_text(h, "code.Cd { font-weight: bold; "
217: "font-family: inherit; }");
1.42 schwarze 218: print_tagq(h, t);
1.1 schwarze 219: }
220:
1.5 schwarze 221: static void
1.26 schwarze 222: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 223: {
224: enum htmlfont font;
1.1 schwarze 225:
1.5 schwarze 226: switch (deco) {
1.35 schwarze 227: case ESCAPE_FONTPREV:
1.5 schwarze 228: font = h->metal;
229: break;
1.35 schwarze 230: case ESCAPE_FONTITALIC:
1.5 schwarze 231: font = HTMLFONT_ITALIC;
232: break;
1.35 schwarze 233: case ESCAPE_FONTBOLD:
1.5 schwarze 234: font = HTMLFONT_BOLD;
235: break;
1.35 schwarze 236: case ESCAPE_FONTBI:
1.31 schwarze 237: font = HTMLFONT_BI;
238: break;
1.35 schwarze 239: case ESCAPE_FONT:
240: case ESCAPE_FONTROMAN:
1.5 schwarze 241: font = HTMLFONT_NONE;
242: break;
243: default:
244: abort();
1.1 schwarze 245: }
246:
1.20 schwarze 247: if (h->metaf) {
248: print_tagq(h, h->metaf);
249: h->metaf = NULL;
250: }
251:
252: h->metal = h->metac;
253: h->metac = font;
254:
1.31 schwarze 255: switch (font) {
1.35 schwarze 256: case HTMLFONT_ITALIC:
1.64 schwarze 257: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 258: break;
1.35 schwarze 259: case HTMLFONT_BOLD:
1.64 schwarze 260: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 261: break;
1.35 schwarze 262: case HTMLFONT_BI:
1.64 schwarze 263: h->metaf = print_otag(h, TAG_B, "");
264: print_otag(h, TAG_I, "");
1.31 schwarze 265: break;
266: default:
267: break;
268: }
1.80 schwarze 269: }
270:
271: char *
1.99 ! schwarze 272: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 273: {
274: const struct roff_node *nch;
1.99 ! schwarze 275: char *buf, *bufs, *cp;
! 276: unsigned int slot;
! 277: int suffix;
1.80 schwarze 278:
279: for (nch = n->child; nch != NULL; nch = nch->next)
280: if (nch->type != ROFFT_TEXT)
281: return NULL;
282:
283: buf = NULL;
284: deroff(&buf, n);
1.90 schwarze 285: if (buf == NULL)
286: return NULL;
1.80 schwarze 287:
288: /* http://www.w3.org/TR/html5/dom.html#the-id-attribute */
289:
290: for (cp = buf; *cp != '\0'; cp++)
291: if (*cp == ' ')
292: *cp = '_';
293:
1.99 ! schwarze 294: if (unique == 0)
! 295: return buf;
! 296:
! 297: /* Avoid duplicate HTML id= attributes. */
! 298:
! 299: bufs = NULL;
! 300: suffix = 1;
! 301: slot = ohash_qlookup(&id_unique, buf);
! 302: cp = ohash_find(&id_unique, slot);
! 303: if (cp != NULL) {
! 304: while (cp != NULL) {
! 305: free(bufs);
! 306: if (++suffix > 127) {
! 307: free(buf);
! 308: return NULL;
! 309: }
! 310: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
! 311: slot = ohash_qlookup(&id_unique, bufs);
! 312: cp = ohash_find(&id_unique, slot);
! 313: }
! 314: free(buf);
! 315: buf = bufs;
! 316: }
! 317: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 318: return buf;
1.1 schwarze 319: }
320:
1.26 schwarze 321: int
322: html_strlen(const char *cp)
323: {
1.30 schwarze 324: size_t rsz;
325: int skip, sz;
1.26 schwarze 326:
327: /*
328: * Account for escaped sequences within string length
329: * calculations. This follows the logic in term_strlen() as we
330: * must calculate the width of produced strings.
331: * Assume that characters are always width of "1". This is
332: * hacky, but it gets the job done for approximation of widths.
333: */
334:
335: sz = 0;
1.30 schwarze 336: skip = 0;
337: while (1) {
338: rsz = strcspn(cp, "\\");
339: if (rsz) {
340: cp += rsz;
341: if (skip) {
342: skip = 0;
343: rsz--;
344: }
345: sz += rsz;
346: }
347: if ('\0' == *cp)
348: break;
349: cp++;
350: switch (mandoc_escape(&cp, NULL, NULL)) {
1.35 schwarze 351: case ESCAPE_ERROR:
1.58 schwarze 352: return sz;
1.35 schwarze 353: case ESCAPE_UNICODE:
354: case ESCAPE_NUMBERED:
355: case ESCAPE_SPECIAL:
1.55 schwarze 356: case ESCAPE_OVERSTRIKE:
1.30 schwarze 357: if (skip)
358: skip = 0;
359: else
360: sz++;
361: break;
1.35 schwarze 362: case ESCAPE_SKIPCHAR:
1.30 schwarze 363: skip = 1;
1.26 schwarze 364: break;
365: default:
366: break;
367: }
368: }
1.58 schwarze 369: return sz;
1.26 schwarze 370: }
1.1 schwarze 371:
1.5 schwarze 372: static int
1.67 schwarze 373: print_escape(struct html *h, char c)
1.38 schwarze 374: {
375:
376: switch (c) {
377: case '<':
1.67 schwarze 378: print_word(h, "<");
1.38 schwarze 379: break;
380: case '>':
1.67 schwarze 381: print_word(h, ">");
1.38 schwarze 382: break;
383: case '&':
1.67 schwarze 384: print_word(h, "&");
1.38 schwarze 385: break;
386: case '"':
1.67 schwarze 387: print_word(h, """);
1.38 schwarze 388: break;
389: case ASCII_NBRSP:
1.67 schwarze 390: print_word(h, " ");
1.38 schwarze 391: break;
392: case ASCII_HYPH:
1.67 schwarze 393: print_byte(h, '-');
1.59 schwarze 394: break;
1.38 schwarze 395: case ASCII_BREAK:
396: break;
397: default:
1.58 schwarze 398: return 0;
1.38 schwarze 399: }
1.58 schwarze 400: return 1;
1.38 schwarze 401: }
402:
403: static int
1.65 schwarze 404: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 405: {
1.67 schwarze 406: char numbuf[16];
1.84 schwarze 407: struct tag *t;
408: const char *seq;
1.4 schwarze 409: size_t sz;
1.84 schwarze 410: int c, len, breakline, nospace;
1.26 schwarze 411: enum mandoc_esc esc;
1.84 schwarze 412: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 413: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 414:
1.65 schwarze 415: if (pend == NULL)
416: pend = strchr(p, '\0');
417:
1.84 schwarze 418: breakline = 0;
1.5 schwarze 419: nospace = 0;
1.1 schwarze 420:
1.65 schwarze 421: while (p < pend) {
1.30 schwarze 422: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
423: h->flags &= ~HTML_SKIPCHAR;
424: p++;
425: continue;
426: }
427:
1.67 schwarze 428: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 429: print_byte(h, *p);
430:
431: if (breakline &&
432: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
433: t = print_otag(h, TAG_DIV, "");
434: print_text(h, "\\~");
435: print_tagq(h, t);
436: breakline = 0;
437: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
438: p++;
439: continue;
440: }
1.4 schwarze 441:
1.65 schwarze 442: if (p >= pend)
1.26 schwarze 443: break;
444:
1.84 schwarze 445: if (*p == ' ') {
446: print_endword(h);
447: p++;
448: continue;
449: }
450:
1.67 schwarze 451: if (print_escape(h, *p++))
1.33 schwarze 452: continue;
1.4 schwarze 453:
1.26 schwarze 454: esc = mandoc_escape(&p, &seq, &len);
455: if (ESCAPE_ERROR == esc)
456: break;
1.5 schwarze 457:
1.26 schwarze 458: switch (esc) {
1.35 schwarze 459: case ESCAPE_FONT:
460: case ESCAPE_FONTPREV:
461: case ESCAPE_FONTBOLD:
462: case ESCAPE_FONTITALIC:
463: case ESCAPE_FONTBI:
464: case ESCAPE_FONTROMAN:
1.30 schwarze 465: if (0 == norecurse)
466: print_metaf(h, esc);
467: continue;
1.35 schwarze 468: case ESCAPE_SKIPCHAR:
1.30 schwarze 469: h->flags |= HTML_SKIPCHAR;
470: continue;
471: default:
472: break;
473: }
474:
475: if (h->flags & HTML_SKIPCHAR) {
476: h->flags &= ~HTML_SKIPCHAR;
477: continue;
478: }
479:
480: switch (esc) {
1.35 schwarze 481: case ESCAPE_UNICODE:
1.38 schwarze 482: /* Skip past "u" header. */
1.26 schwarze 483: c = mchars_num2uc(seq + 1, len - 1);
484: break;
1.35 schwarze 485: case ESCAPE_NUMBERED:
1.26 schwarze 486: c = mchars_num2char(seq, len);
1.51 schwarze 487: if (c < 0)
488: continue;
1.26 schwarze 489: break;
1.35 schwarze 490: case ESCAPE_SPECIAL:
1.61 schwarze 491: c = mchars_spec2cp(seq, len);
1.51 schwarze 492: if (c <= 0)
493: continue;
1.26 schwarze 494: break;
1.84 schwarze 495: case ESCAPE_BREAK:
496: breakline = 1;
497: continue;
1.35 schwarze 498: case ESCAPE_NOSPACE:
1.26 schwarze 499: if ('\0' == *p)
500: nospace = 1;
1.49 schwarze 501: continue;
1.55 schwarze 502: case ESCAPE_OVERSTRIKE:
503: if (len == 0)
504: continue;
505: c = seq[len - 1];
506: break;
1.5 schwarze 507: default:
1.49 schwarze 508: continue;
1.5 schwarze 509: }
1.51 schwarze 510: if ((c < 0x20 && c != 0x09) ||
511: (c > 0x7E && c < 0xA0))
1.49 schwarze 512: c = 0xFFFD;
1.67 schwarze 513: if (c > 0x7E) {
1.86 bentley 514: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 515: print_word(h, numbuf);
516: } else if (print_escape(h, c) == 0)
517: print_byte(h, c);
1.1 schwarze 518: }
1.5 schwarze 519:
1.58 schwarze 520: return nospace;
1.1 schwarze 521: }
522:
1.6 schwarze 523: static void
1.65 schwarze 524: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 525: {
1.65 schwarze 526: const char *p, *pp;
527:
528: pp = man ? h->base_man : h->base_includes;
529: while ((p = strchr(pp, '%')) != NULL) {
530: print_encode(h, pp, p, 1);
531: if (man && p[1] == 'S') {
532: if (sec == NULL)
1.67 schwarze 533: print_byte(h, '1');
1.65 schwarze 534: else
535: print_encode(h, sec, NULL, 1);
536: } else if ((man && p[1] == 'N') ||
537: (man == 0 && p[1] == 'I'))
538: print_encode(h, name, NULL, 1);
539: else
540: print_encode(h, p, p + 2, 1);
541: pp = p + 2;
542: }
543: if (*pp != '\0')
544: print_encode(h, pp, NULL, 1);
1.6 schwarze 545: }
546:
1.1 schwarze 547: struct tag *
1.64 schwarze 548: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 549: {
1.64 schwarze 550: va_list ap;
551: struct roffsu mysu, *su;
1.67 schwarze 552: char numbuf[16];
1.1 schwarze 553: struct tag *t;
1.65 schwarze 554: const char *attr;
1.73 schwarze 555: char *arg1, *arg2;
1.65 schwarze 556: double v;
1.66 schwarze 557: int i, have_style, tflags;
558:
559: tflags = htmltags[tag].flags;
1.1 schwarze 560:
1.74 schwarze 561: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 562:
1.66 schwarze 563: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 564: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 565: t->tag = tag;
1.74 schwarze 566: t->next = h->tag;
567: h->tag = t;
1.1 schwarze 568: } else
569: t = NULL;
570:
1.66 schwarze 571: if (tflags & HTML_NLBEFORE)
1.67 schwarze 572: print_endline(h);
573: if (h->col == 0)
574: print_indent(h);
1.66 schwarze 575: else if ((h->flags & HTML_NOSPACE) == 0) {
576: if (h->flags & HTML_KEEP)
1.86 bentley 577: print_word(h, " ");
1.66 schwarze 578: else {
579: if (h->flags & HTML_PREKEEP)
580: h->flags |= HTML_KEEP;
1.67 schwarze 581: print_endword(h);
1.12 schwarze 582: }
1.66 schwarze 583: }
1.1 schwarze 584:
1.13 schwarze 585: if ( ! (h->flags & HTML_NONOSPACE))
586: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 587: else
588: h->flags |= HTML_NOSPACE;
1.13 schwarze 589:
1.6 schwarze 590: /* Print out the tag name and attributes. */
591:
1.67 schwarze 592: print_byte(h, '<');
593: print_word(h, htmltags[tag].name);
1.64 schwarze 594:
595: va_start(ap, fmt);
596:
597: have_style = 0;
598: while (*fmt != '\0') {
599: if (*fmt == 's') {
600: have_style = 1;
601: fmt++;
602: break;
603: }
1.73 schwarze 604:
605: /* Parse a non-style attribute and its arguments. */
606:
607: arg1 = va_arg(ap, char *);
1.64 schwarze 608: switch (*fmt++) {
609: case 'c':
1.65 schwarze 610: attr = "class";
1.64 schwarze 611: break;
612: case 'h':
1.65 schwarze 613: attr = "href";
1.64 schwarze 614: break;
615: case 'i':
1.65 schwarze 616: attr = "id";
1.64 schwarze 617: break;
618: case '?':
1.73 schwarze 619: attr = arg1;
620: arg1 = va_arg(ap, char *);
1.64 schwarze 621: break;
622: default:
623: abort();
624: }
1.73 schwarze 625: arg2 = NULL;
626: if (*fmt == 'M')
627: arg2 = va_arg(ap, char *);
628: if (arg1 == NULL)
629: continue;
630:
631: /* Print the non-style attributes. */
632:
1.67 schwarze 633: print_byte(h, ' ');
634: print_word(h, attr);
635: print_byte(h, '=');
636: print_byte(h, '"');
1.65 schwarze 637: switch (*fmt) {
1.78 schwarze 638: case 'I':
639: print_href(h, arg1, NULL, 0);
640: fmt++;
641: break;
1.65 schwarze 642: case 'M':
1.73 schwarze 643: print_href(h, arg1, arg2, 1);
1.65 schwarze 644: fmt++;
645: break;
1.78 schwarze 646: case 'R':
647: print_byte(h, '#');
648: print_encode(h, arg1, NULL, 1);
1.65 schwarze 649: fmt++;
650: break;
1.78 schwarze 651: case 'T':
652: print_encode(h, arg1, NULL, 1);
653: print_word(h, "\" title=\"");
654: print_encode(h, arg1, NULL, 1);
1.65 schwarze 655: fmt++;
1.78 schwarze 656: break;
1.65 schwarze 657: default:
1.73 schwarze 658: print_encode(h, arg1, NULL, 1);
1.65 schwarze 659: break;
660: }
1.67 schwarze 661: print_byte(h, '"');
1.64 schwarze 662: }
663:
664: /* Print out styles. */
665:
666: while (*fmt != '\0') {
1.73 schwarze 667: arg1 = NULL;
668: su = NULL;
1.64 schwarze 669:
670: /* First letter: input argument type. */
671:
672: switch (*fmt++) {
673: case 'h':
674: i = va_arg(ap, int);
1.73 schwarze 675: su = &mysu;
1.64 schwarze 676: SCALE_HS_INIT(su, i);
677: break;
678: case 's':
1.73 schwarze 679: arg1 = va_arg(ap, char *);
1.64 schwarze 680: break;
681: case 'u':
682: su = va_arg(ap, struct roffsu *);
683: break;
684: case 'w':
1.89 schwarze 685: if ((arg2 = va_arg(ap, char *)) != NULL) {
686: su = &mysu;
687: a2width(arg2, su);
688: }
689: if (*fmt == '*') {
690: if (su != NULL && su->unit == SCALE_EN &&
691: su->scale > 5.9 && su->scale < 6.1)
692: su = NULL;
693: fmt++;
1.88 schwarze 694: }
1.81 schwarze 695: if (*fmt == '+') {
1.89 schwarze 696: if (su != NULL) {
697: /* Make even bold text fit. */
698: su->scale *= 1.2;
699: /* Add padding. */
700: su->scale += 3.0;
701: }
1.81 schwarze 702: fmt++;
703: }
704: if (*fmt == '-') {
1.89 schwarze 705: if (su != NULL)
706: su->scale *= -1.0;
1.81 schwarze 707: fmt++;
708: }
1.64 schwarze 709: break;
710: default:
711: abort();
712: }
713:
714: /* Second letter: style name. */
715:
716: switch (*fmt++) {
717: case 'h':
1.65 schwarze 718: attr = "height";
1.64 schwarze 719: break;
720: case 'i':
1.65 schwarze 721: attr = "text-indent";
1.64 schwarze 722: break;
723: case 'l':
1.65 schwarze 724: attr = "margin-left";
1.64 schwarze 725: break;
726: case 'w':
1.65 schwarze 727: attr = "width";
1.64 schwarze 728: break;
729: case 'W':
1.65 schwarze 730: attr = "min-width";
1.64 schwarze 731: break;
732: case '?':
1.73 schwarze 733: attr = arg1;
734: arg1 = va_arg(ap, char *);
735: break;
1.64 schwarze 736: default:
737: abort();
738: }
1.73 schwarze 739: if (su == NULL && arg1 == NULL)
740: continue;
741:
742: if (have_style == 1)
743: print_word(h, " style=\"");
744: else
745: print_byte(h, ' ');
1.67 schwarze 746: print_word(h, attr);
747: print_byte(h, ':');
748: print_byte(h, ' ');
1.73 schwarze 749: if (su != NULL) {
750: v = su->scale;
751: if (su->unit == SCALE_MM && (v /= 100.0) == 0.0)
752: v = 1.0;
753: else if (su->unit == SCALE_BU)
754: v /= 24.0;
755: (void)snprintf(numbuf, sizeof(numbuf), "%.2f", v);
756: print_word(h, numbuf);
757: print_word(h, roffscales[su->unit]);
758: } else
759: print_word(h, arg1);
1.67 schwarze 760: print_byte(h, ';');
1.73 schwarze 761: have_style = 2;
1.64 schwarze 762: }
1.73 schwarze 763: if (have_style == 2)
1.67 schwarze 764: print_byte(h, '"');
1.64 schwarze 765:
766: va_end(ap);
1.6 schwarze 767:
1.42 schwarze 768: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 769:
770: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 771: print_byte(h, '/');
1.6 schwarze 772:
1.67 schwarze 773: print_byte(h, '>');
1.1 schwarze 774:
1.66 schwarze 775: if (tflags & HTML_NLBEGIN)
1.67 schwarze 776: print_endline(h);
1.66 schwarze 777: else
778: h->flags |= HTML_NOSPACE;
1.18 schwarze 779:
1.66 schwarze 780: if (tflags & HTML_INDENT)
781: h->indent++;
782: if (tflags & HTML_NOINDENT)
783: h->noindent++;
1.18 schwarze 784:
1.58 schwarze 785: return t;
1.1 schwarze 786: }
787:
788: static void
1.54 schwarze 789: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 790: {
1.66 schwarze 791: int tflags;
1.35 schwarze 792:
1.54 schwarze 793: /*
794: * Remember to close out and nullify the current
795: * meta-font and table, if applicable.
796: */
797: if (tag == h->metaf)
798: h->metaf = NULL;
799: if (tag == h->tblt)
800: h->tblt = NULL;
801:
1.66 schwarze 802: tflags = htmltags[tag->tag].flags;
803:
804: if (tflags & HTML_INDENT)
805: h->indent--;
806: if (tflags & HTML_NOINDENT)
807: h->noindent--;
808: if (tflags & HTML_NLEND)
1.67 schwarze 809: print_endline(h);
810: print_indent(h);
811: print_byte(h, '<');
812: print_byte(h, '/');
813: print_word(h, htmltags[tag->tag].name);
814: print_byte(h, '>');
1.66 schwarze 815: if (tflags & HTML_NLAFTER)
1.67 schwarze 816: print_endline(h);
1.54 schwarze 817:
1.74 schwarze 818: h->tag = tag->next;
1.54 schwarze 819: free(tag);
1.1 schwarze 820: }
821:
822: void
1.6 schwarze 823: print_gen_decls(struct html *h)
824: {
1.67 schwarze 825: print_word(h, "<!DOCTYPE html>");
826: print_endline(h);
1.91 schwarze 827: }
828:
829: void
830: print_gen_comment(struct html *h, struct roff_node *n)
831: {
832: int wantblank;
833:
834: print_word(h, "<!-- This is an automatically generated file."
835: " Do not edit.");
836: h->indent = 1;
837: wantblank = 0;
838: while (n != NULL && n->type == ROFFT_COMMENT) {
839: if (strstr(n->string, "-->") == NULL &&
840: (wantblank || *n->string != '\0')) {
841: print_endline(h);
842: print_indent(h);
843: print_word(h, n->string);
844: wantblank = *n->string != '\0';
845: }
846: n = n->next;
847: }
848: if (wantblank)
849: print_endline(h);
850: print_word(h, " -->");
851: print_endline(h);
852: h->indent = 0;
1.1 schwarze 853: }
854:
855: void
1.12 schwarze 856: print_text(struct html *h, const char *word)
1.1 schwarze 857: {
1.67 schwarze 858: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 859: if ( ! (HTML_KEEP & h->flags)) {
860: if (HTML_PREKEEP & h->flags)
861: h->flags |= HTML_KEEP;
1.67 schwarze 862: print_endword(h);
1.12 schwarze 863: } else
1.86 bentley 864: print_word(h, " ");
1.12 schwarze 865: }
1.1 schwarze 866:
1.20 schwarze 867: assert(NULL == h->metaf);
1.31 schwarze 868: switch (h->metac) {
1.35 schwarze 869: case HTMLFONT_ITALIC:
1.64 schwarze 870: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 871: break;
1.35 schwarze 872: case HTMLFONT_BOLD:
1.64 schwarze 873: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 874: break;
1.35 schwarze 875: case HTMLFONT_BI:
1.64 schwarze 876: h->metaf = print_otag(h, TAG_B, "");
877: print_otag(h, TAG_I, "");
1.31 schwarze 878: break;
879: default:
1.67 schwarze 880: print_indent(h);
1.31 schwarze 881: break;
882: }
1.20 schwarze 883:
1.12 schwarze 884: assert(word);
1.65 schwarze 885: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 886: if ( ! (h->flags & HTML_NONOSPACE))
887: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 888: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 889: } else
1.53 schwarze 890: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 891:
892: if (h->metaf) {
893: print_tagq(h, h->metaf);
894: h->metaf = NULL;
895: }
1.17 schwarze 896:
897: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 898: }
899:
900: void
901: print_tagq(struct html *h, const struct tag *until)
902: {
903: struct tag *tag;
904:
1.74 schwarze 905: while ((tag = h->tag) != NULL) {
1.54 schwarze 906: print_ctag(h, tag);
1.1 schwarze 907: if (until && tag == until)
908: return;
909: }
910: }
911:
912: void
913: print_stagq(struct html *h, const struct tag *suntil)
914: {
915: struct tag *tag;
916:
1.74 schwarze 917: while ((tag = h->tag) != NULL) {
1.1 schwarze 918: if (suntil && tag == suntil)
919: return;
1.54 schwarze 920: print_ctag(h, tag);
1.1 schwarze 921: }
922: }
1.42 schwarze 923:
924: void
925: print_paragraph(struct html *h)
926: {
927: struct tag *t;
928:
1.68 schwarze 929: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 930: print_tagq(h, t);
931: }
932:
1.67 schwarze 933:
934: /***********************************************************************
935: * Low level output functions.
936: * They implement line breaking using a short static buffer.
937: ***********************************************************************/
938:
939: /*
940: * Buffer one HTML output byte.
941: * If the buffer is full, flush and deactivate it and start a new line.
942: * If the buffer is inactive, print directly.
943: */
944: static void
945: print_byte(struct html *h, char c)
946: {
947: if ((h->flags & HTML_BUFFER) == 0) {
948: putchar(c);
949: h->col++;
950: return;
951: }
952:
953: if (h->col + h->bufcol < sizeof(h->buf)) {
954: h->buf[h->bufcol++] = c;
955: return;
956: }
957:
958: putchar('\n');
959: h->col = 0;
960: print_indent(h);
961: putchar(' ');
962: putchar(' ');
963: fwrite(h->buf, h->bufcol, 1, stdout);
964: putchar(c);
965: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
966: h->bufcol = 0;
967: h->flags &= ~HTML_BUFFER;
968: }
969:
1.66 schwarze 970: /*
971: * If something was printed on the current output line, end it.
1.67 schwarze 972: * Not to be called right after print_indent().
1.66 schwarze 973: */
1.72 schwarze 974: void
1.67 schwarze 975: print_endline(struct html *h)
1.66 schwarze 976: {
1.67 schwarze 977: if (h->col == 0)
1.66 schwarze 978: return;
979:
1.67 schwarze 980: if (h->bufcol) {
981: putchar(' ');
982: fwrite(h->buf, h->bufcol, 1, stdout);
983: h->bufcol = 0;
984: }
1.66 schwarze 985: putchar('\n');
1.67 schwarze 986: h->col = 0;
987: h->flags |= HTML_NOSPACE;
988: h->flags &= ~HTML_BUFFER;
989: }
990:
991: /*
992: * Flush the HTML output buffer.
993: * If it is inactive, activate it.
994: */
995: static void
996: print_endword(struct html *h)
997: {
998: if (h->noindent) {
999: print_byte(h, ' ');
1000: return;
1001: }
1002:
1003: if ((h->flags & HTML_BUFFER) == 0) {
1004: h->col++;
1005: h->flags |= HTML_BUFFER;
1006: } else if (h->bufcol) {
1007: putchar(' ');
1008: fwrite(h->buf, h->bufcol, 1, stdout);
1009: h->col += h->bufcol + 1;
1010: }
1011: h->bufcol = 0;
1.66 schwarze 1012: }
1013:
1014: /*
1015: * If at the beginning of a new output line,
1016: * perform indentation and mark the line as containing output.
1017: * Make sure to really produce some output right afterwards,
1018: * but do not use print_otag() for producing it.
1019: */
1020: static void
1.67 schwarze 1021: print_indent(struct html *h)
1.66 schwarze 1022: {
1.67 schwarze 1023: size_t i;
1.66 schwarze 1024:
1.67 schwarze 1025: if (h->col)
1.66 schwarze 1026: return;
1027:
1.67 schwarze 1028: if (h->noindent == 0) {
1029: h->col = h->indent * 2;
1030: for (i = 0; i < h->col; i++)
1.66 schwarze 1031: putchar(' ');
1.67 schwarze 1032: }
1033: h->flags &= ~HTML_NOSPACE;
1034: }
1035:
1036: /*
1037: * Print or buffer some characters
1038: * depending on the current HTML output buffer state.
1039: */
1040: static void
1041: print_word(struct html *h, const char *cp)
1042: {
1043: while (*cp != '\0')
1044: print_byte(h, *cp++);
1.66 schwarze 1045: }
1.64 schwarze 1046:
1047: /*
1048: * Calculate the scaling unit passed in a `-width' argument. This uses
1049: * either a native scaling unit (e.g., 1i, 2m) or the string length of
1050: * the value.
1051: */
1052: static void
1053: a2width(const char *p, struct roffsu *su)
1054: {
1.83 schwarze 1055: const char *end;
1056:
1057: end = a2roffsu(p, su, SCALE_MAX);
1058: if (end == NULL || *end != '\0') {
1.64 schwarze 1059: su->unit = SCALE_EN;
1060: su->scale = html_strlen(p);
1061: } else if (su->scale < 0.0)
1062: su->scale = 0.0;
1.3 schwarze 1063: }