Annotation of src/usr.bin/mandoc/html.c, Revision 1.93
1.93 ! schwarze 1: /* $OpenBSD: html.c,v 1.92 2018/05/01 23:36:39 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>
1.1 schwarze 23: #include <stdio.h>
24: #include <stdint.h>
25: #include <stdlib.h>
26: #include <string.h>
27: #include <unistd.h>
28:
1.80 schwarze 29: #include "mandoc_aux.h"
1.9 schwarze 30: #include "mandoc.h"
1.80 schwarze 31: #include "roff.h"
1.1 schwarze 32: #include "out.h"
33: #include "html.h"
1.56 schwarze 34: #include "manconf.h"
1.1 schwarze 35: #include "main.h"
36:
37: struct htmldata {
38: const char *name;
39: int flags;
1.66 schwarze 40: #define HTML_NOSTACK (1 << 0)
41: #define HTML_AUTOCLOSE (1 << 1)
42: #define HTML_NLBEFORE (1 << 2)
43: #define HTML_NLBEGIN (1 << 3)
44: #define HTML_NLEND (1 << 4)
45: #define HTML_NLAFTER (1 << 5)
46: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
47: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
48: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
49: #define HTML_INDENT (1 << 6)
50: #define HTML_NOINDENT (1 << 7)
1.1 schwarze 51: };
52:
53: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 54: {"html", HTML_NLALL},
55: {"head", HTML_NLALL | HTML_INDENT},
56: {"body", HTML_NLALL},
57: {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
58: {"title", HTML_NLAROUND},
59: {"div", HTML_NLAROUND},
60: {"h1", HTML_NLAROUND},
61: {"h2", HTML_NLAROUND},
62: {"span", 0},
63: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
64: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
65: {"a", 0},
66: {"table", HTML_NLALL | HTML_INDENT},
1.75 schwarze 67: {"colgroup", HTML_NLALL | HTML_INDENT},
1.66 schwarze 68: {"col", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
69: {"tr", HTML_NLALL | HTML_INDENT},
70: {"td", HTML_NLAROUND},
71: {"li", HTML_NLAROUND | HTML_INDENT},
72: {"ul", HTML_NLALL | HTML_INDENT},
73: {"ol", HTML_NLALL | HTML_INDENT},
74: {"dl", HTML_NLALL | HTML_INDENT},
75: {"dt", HTML_NLAROUND},
76: {"dd", HTML_NLAROUND | HTML_INDENT},
77: {"pre", HTML_NLALL | HTML_NOINDENT},
1.77 schwarze 78: {"var", 0},
1.76 schwarze 79: {"cite", 0},
1.66 schwarze 80: {"b", 0},
81: {"i", 0},
82: {"code", 0},
83: {"small", 0},
84: {"style", HTML_NLALL | HTML_INDENT},
85: {"math", HTML_NLALL | HTML_INDENT},
86: {"mrow", 0},
87: {"mi", 0},
1.85 schwarze 88: {"mn", 0},
1.66 schwarze 89: {"mo", 0},
90: {"msup", 0},
91: {"msub", 0},
92: {"msubsup", 0},
93: {"mfrac", 0},
94: {"msqrt", 0},
95: {"mfenced", 0},
96: {"mtable", 0},
97: {"mtr", 0},
98: {"mtd", 0},
99: {"munderover", 0},
100: {"munder", 0},
101: {"mover", 0},
1.5 schwarze 102: };
103:
1.26 schwarze 104: static const char *const roffscales[SCALE_MAX] = {
105: "cm", /* SCALE_CM */
106: "in", /* SCALE_IN */
107: "pc", /* SCALE_PC */
108: "pt", /* SCALE_PT */
109: "em", /* SCALE_EM */
110: "em", /* SCALE_MM */
111: "ex", /* SCALE_EN */
112: "ex", /* SCALE_BU */
113: "em", /* SCALE_VS */
114: "ex", /* SCALE_FS */
115: };
1.5 schwarze 116:
1.64 schwarze 117: static void a2width(const char *, struct roffsu *);
1.67 schwarze 118: static void print_byte(struct html *, char);
119: static void print_endword(struct html *);
120: static void print_indent(struct html *);
121: static void print_word(struct html *, const char *);
122:
1.54 schwarze 123: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 124: static int print_escape(struct html *, char);
1.65 schwarze 125: static int print_encode(struct html *, const char *, const char *, int);
126: static void print_href(struct html *, const char *, const char *, int);
1.26 schwarze 127: static void print_metaf(struct html *, enum mandoc_esc);
1.5 schwarze 128:
1.35 schwarze 129:
1.50 schwarze 130: void *
1.61 schwarze 131: html_alloc(const struct manoutput *outopts)
1.1 schwarze 132: {
133: struct html *h;
134:
1.24 schwarze 135: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 136:
1.74 schwarze 137: h->tag = NULL;
1.56 schwarze 138: h->style = outopts->style;
139: h->base_man = outopts->man;
140: h->base_includes = outopts->includes;
141: if (outopts->fragment)
142: h->oflags |= HTML_FRAGMENT;
1.1 schwarze 143:
1.58 schwarze 144: return h;
1.1 schwarze 145: }
146:
147: void
148: html_free(void *p)
149: {
150: struct tag *tag;
151: struct html *h;
152:
153: h = (struct html *)p;
154:
1.74 schwarze 155: while ((tag = h->tag) != NULL) {
156: h->tag = tag->next;
1.1 schwarze 157: free(tag);
158: }
159:
160: free(h);
161: }
162:
163: void
164: print_gen_head(struct html *h)
165: {
1.42 schwarze 166: struct tag *t;
167:
1.64 schwarze 168: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 169: if (h->style != NULL) {
170: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
171: h->style, "type", "text/css", "media", "all");
172: return;
173: }
1.1 schwarze 174:
1.42 schwarze 175: /*
1.92 schwarze 176: * Print a minimal embedded style sheet.
1.42 schwarze 177: */
1.66 schwarze 178:
1.64 schwarze 179: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 180: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 181: print_endline(h);
1.66 schwarze 182: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 183: print_endline(h);
1.66 schwarze 184: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 185: print_endline(h);
1.68 schwarze 186: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.93 ! schwarze 187: print_endline(h);
! 188: print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, "
! 189: "code.In, code.Fd, code.Fn,");
! 190: print_endline(h);
! 191: print_text(h, "code.Cd { font-weight: bold; "
! 192: "font-family: inherit; }");
1.42 schwarze 193: print_tagq(h, t);
1.1 schwarze 194: }
195:
1.5 schwarze 196: static void
1.26 schwarze 197: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 198: {
199: enum htmlfont font;
1.1 schwarze 200:
1.5 schwarze 201: switch (deco) {
1.35 schwarze 202: case ESCAPE_FONTPREV:
1.5 schwarze 203: font = h->metal;
204: break;
1.35 schwarze 205: case ESCAPE_FONTITALIC:
1.5 schwarze 206: font = HTMLFONT_ITALIC;
207: break;
1.35 schwarze 208: case ESCAPE_FONTBOLD:
1.5 schwarze 209: font = HTMLFONT_BOLD;
210: break;
1.35 schwarze 211: case ESCAPE_FONTBI:
1.31 schwarze 212: font = HTMLFONT_BI;
213: break;
1.35 schwarze 214: case ESCAPE_FONT:
215: case ESCAPE_FONTROMAN:
1.5 schwarze 216: font = HTMLFONT_NONE;
217: break;
218: default:
219: abort();
1.1 schwarze 220: }
221:
1.20 schwarze 222: if (h->metaf) {
223: print_tagq(h, h->metaf);
224: h->metaf = NULL;
225: }
226:
227: h->metal = h->metac;
228: h->metac = font;
229:
1.31 schwarze 230: switch (font) {
1.35 schwarze 231: case HTMLFONT_ITALIC:
1.64 schwarze 232: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 233: break;
1.35 schwarze 234: case HTMLFONT_BOLD:
1.64 schwarze 235: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 236: break;
1.35 schwarze 237: case HTMLFONT_BI:
1.64 schwarze 238: h->metaf = print_otag(h, TAG_B, "");
239: print_otag(h, TAG_I, "");
1.31 schwarze 240: break;
241: default:
242: break;
243: }
1.80 schwarze 244: }
245:
246: char *
247: html_make_id(const struct roff_node *n)
248: {
249: const struct roff_node *nch;
250: char *buf, *cp;
251:
252: for (nch = n->child; nch != NULL; nch = nch->next)
253: if (nch->type != ROFFT_TEXT)
254: return NULL;
255:
256: buf = NULL;
257: deroff(&buf, n);
1.90 schwarze 258: if (buf == NULL)
259: return NULL;
1.80 schwarze 260:
261: /* http://www.w3.org/TR/html5/dom.html#the-id-attribute */
262:
263: for (cp = buf; *cp != '\0'; cp++)
264: if (*cp == ' ')
265: *cp = '_';
266:
267: return buf;
1.1 schwarze 268: }
269:
1.26 schwarze 270: int
271: html_strlen(const char *cp)
272: {
1.30 schwarze 273: size_t rsz;
274: int skip, sz;
1.26 schwarze 275:
276: /*
277: * Account for escaped sequences within string length
278: * calculations. This follows the logic in term_strlen() as we
279: * must calculate the width of produced strings.
280: * Assume that characters are always width of "1". This is
281: * hacky, but it gets the job done for approximation of widths.
282: */
283:
284: sz = 0;
1.30 schwarze 285: skip = 0;
286: while (1) {
287: rsz = strcspn(cp, "\\");
288: if (rsz) {
289: cp += rsz;
290: if (skip) {
291: skip = 0;
292: rsz--;
293: }
294: sz += rsz;
295: }
296: if ('\0' == *cp)
297: break;
298: cp++;
299: switch (mandoc_escape(&cp, NULL, NULL)) {
1.35 schwarze 300: case ESCAPE_ERROR:
1.58 schwarze 301: return sz;
1.35 schwarze 302: case ESCAPE_UNICODE:
303: case ESCAPE_NUMBERED:
304: case ESCAPE_SPECIAL:
1.55 schwarze 305: case ESCAPE_OVERSTRIKE:
1.30 schwarze 306: if (skip)
307: skip = 0;
308: else
309: sz++;
310: break;
1.35 schwarze 311: case ESCAPE_SKIPCHAR:
1.30 schwarze 312: skip = 1;
1.26 schwarze 313: break;
314: default:
315: break;
316: }
317: }
1.58 schwarze 318: return sz;
1.26 schwarze 319: }
1.1 schwarze 320:
1.5 schwarze 321: static int
1.67 schwarze 322: print_escape(struct html *h, char c)
1.38 schwarze 323: {
324:
325: switch (c) {
326: case '<':
1.67 schwarze 327: print_word(h, "<");
1.38 schwarze 328: break;
329: case '>':
1.67 schwarze 330: print_word(h, ">");
1.38 schwarze 331: break;
332: case '&':
1.67 schwarze 333: print_word(h, "&");
1.38 schwarze 334: break;
335: case '"':
1.67 schwarze 336: print_word(h, """);
1.38 schwarze 337: break;
338: case ASCII_NBRSP:
1.67 schwarze 339: print_word(h, " ");
1.38 schwarze 340: break;
341: case ASCII_HYPH:
1.67 schwarze 342: print_byte(h, '-');
1.59 schwarze 343: break;
1.38 schwarze 344: case ASCII_BREAK:
345: break;
346: default:
1.58 schwarze 347: return 0;
1.38 schwarze 348: }
1.58 schwarze 349: return 1;
1.38 schwarze 350: }
351:
352: static int
1.65 schwarze 353: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 354: {
1.67 schwarze 355: char numbuf[16];
1.84 schwarze 356: struct tag *t;
357: const char *seq;
1.4 schwarze 358: size_t sz;
1.84 schwarze 359: int c, len, breakline, nospace;
1.26 schwarze 360: enum mandoc_esc esc;
1.84 schwarze 361: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 362: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 363:
1.65 schwarze 364: if (pend == NULL)
365: pend = strchr(p, '\0');
366:
1.84 schwarze 367: breakline = 0;
1.5 schwarze 368: nospace = 0;
1.1 schwarze 369:
1.65 schwarze 370: while (p < pend) {
1.30 schwarze 371: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
372: h->flags &= ~HTML_SKIPCHAR;
373: p++;
374: continue;
375: }
376:
1.67 schwarze 377: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 378: print_byte(h, *p);
379:
380: if (breakline &&
381: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
382: t = print_otag(h, TAG_DIV, "");
383: print_text(h, "\\~");
384: print_tagq(h, t);
385: breakline = 0;
386: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
387: p++;
388: continue;
389: }
1.4 schwarze 390:
1.65 schwarze 391: if (p >= pend)
1.26 schwarze 392: break;
393:
1.84 schwarze 394: if (*p == ' ') {
395: print_endword(h);
396: p++;
397: continue;
398: }
399:
1.67 schwarze 400: if (print_escape(h, *p++))
1.33 schwarze 401: continue;
1.4 schwarze 402:
1.26 schwarze 403: esc = mandoc_escape(&p, &seq, &len);
404: if (ESCAPE_ERROR == esc)
405: break;
1.5 schwarze 406:
1.26 schwarze 407: switch (esc) {
1.35 schwarze 408: case ESCAPE_FONT:
409: case ESCAPE_FONTPREV:
410: case ESCAPE_FONTBOLD:
411: case ESCAPE_FONTITALIC:
412: case ESCAPE_FONTBI:
413: case ESCAPE_FONTROMAN:
1.30 schwarze 414: if (0 == norecurse)
415: print_metaf(h, esc);
416: continue;
1.35 schwarze 417: case ESCAPE_SKIPCHAR:
1.30 schwarze 418: h->flags |= HTML_SKIPCHAR;
419: continue;
420: default:
421: break;
422: }
423:
424: if (h->flags & HTML_SKIPCHAR) {
425: h->flags &= ~HTML_SKIPCHAR;
426: continue;
427: }
428:
429: switch (esc) {
1.35 schwarze 430: case ESCAPE_UNICODE:
1.38 schwarze 431: /* Skip past "u" header. */
1.26 schwarze 432: c = mchars_num2uc(seq + 1, len - 1);
433: break;
1.35 schwarze 434: case ESCAPE_NUMBERED:
1.26 schwarze 435: c = mchars_num2char(seq, len);
1.51 schwarze 436: if (c < 0)
437: continue;
1.26 schwarze 438: break;
1.35 schwarze 439: case ESCAPE_SPECIAL:
1.61 schwarze 440: c = mchars_spec2cp(seq, len);
1.51 schwarze 441: if (c <= 0)
442: continue;
1.26 schwarze 443: break;
1.84 schwarze 444: case ESCAPE_BREAK:
445: breakline = 1;
446: continue;
1.35 schwarze 447: case ESCAPE_NOSPACE:
1.26 schwarze 448: if ('\0' == *p)
449: nospace = 1;
1.49 schwarze 450: continue;
1.55 schwarze 451: case ESCAPE_OVERSTRIKE:
452: if (len == 0)
453: continue;
454: c = seq[len - 1];
455: break;
1.5 schwarze 456: default:
1.49 schwarze 457: continue;
1.5 schwarze 458: }
1.51 schwarze 459: if ((c < 0x20 && c != 0x09) ||
460: (c > 0x7E && c < 0xA0))
1.49 schwarze 461: c = 0xFFFD;
1.67 schwarze 462: if (c > 0x7E) {
1.86 bentley 463: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 464: print_word(h, numbuf);
465: } else if (print_escape(h, c) == 0)
466: print_byte(h, c);
1.1 schwarze 467: }
1.5 schwarze 468:
1.58 schwarze 469: return nospace;
1.1 schwarze 470: }
471:
1.6 schwarze 472: static void
1.65 schwarze 473: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 474: {
1.65 schwarze 475: const char *p, *pp;
476:
477: pp = man ? h->base_man : h->base_includes;
478: while ((p = strchr(pp, '%')) != NULL) {
479: print_encode(h, pp, p, 1);
480: if (man && p[1] == 'S') {
481: if (sec == NULL)
1.67 schwarze 482: print_byte(h, '1');
1.65 schwarze 483: else
484: print_encode(h, sec, NULL, 1);
485: } else if ((man && p[1] == 'N') ||
486: (man == 0 && p[1] == 'I'))
487: print_encode(h, name, NULL, 1);
488: else
489: print_encode(h, p, p + 2, 1);
490: pp = p + 2;
491: }
492: if (*pp != '\0')
493: print_encode(h, pp, NULL, 1);
1.6 schwarze 494: }
495:
1.1 schwarze 496: struct tag *
1.64 schwarze 497: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 498: {
1.64 schwarze 499: va_list ap;
500: struct roffsu mysu, *su;
1.67 schwarze 501: char numbuf[16];
1.1 schwarze 502: struct tag *t;
1.65 schwarze 503: const char *attr;
1.73 schwarze 504: char *arg1, *arg2;
1.65 schwarze 505: double v;
1.66 schwarze 506: int i, have_style, tflags;
507:
508: tflags = htmltags[tag].flags;
1.1 schwarze 509:
1.74 schwarze 510: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 511:
1.66 schwarze 512: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 513: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 514: t->tag = tag;
1.74 schwarze 515: t->next = h->tag;
516: h->tag = t;
1.1 schwarze 517: } else
518: t = NULL;
519:
1.66 schwarze 520: if (tflags & HTML_NLBEFORE)
1.67 schwarze 521: print_endline(h);
522: if (h->col == 0)
523: print_indent(h);
1.66 schwarze 524: else if ((h->flags & HTML_NOSPACE) == 0) {
525: if (h->flags & HTML_KEEP)
1.86 bentley 526: print_word(h, " ");
1.66 schwarze 527: else {
528: if (h->flags & HTML_PREKEEP)
529: h->flags |= HTML_KEEP;
1.67 schwarze 530: print_endword(h);
1.12 schwarze 531: }
1.66 schwarze 532: }
1.1 schwarze 533:
1.13 schwarze 534: if ( ! (h->flags & HTML_NONOSPACE))
535: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 536: else
537: h->flags |= HTML_NOSPACE;
1.13 schwarze 538:
1.6 schwarze 539: /* Print out the tag name and attributes. */
540:
1.67 schwarze 541: print_byte(h, '<');
542: print_word(h, htmltags[tag].name);
1.64 schwarze 543:
544: va_start(ap, fmt);
545:
546: have_style = 0;
547: while (*fmt != '\0') {
548: if (*fmt == 's') {
549: have_style = 1;
550: fmt++;
551: break;
552: }
1.73 schwarze 553:
554: /* Parse a non-style attribute and its arguments. */
555:
556: arg1 = va_arg(ap, char *);
1.64 schwarze 557: switch (*fmt++) {
558: case 'c':
1.65 schwarze 559: attr = "class";
1.64 schwarze 560: break;
561: case 'h':
1.65 schwarze 562: attr = "href";
1.64 schwarze 563: break;
564: case 'i':
1.65 schwarze 565: attr = "id";
1.64 schwarze 566: break;
567: case '?':
1.73 schwarze 568: attr = arg1;
569: arg1 = va_arg(ap, char *);
1.64 schwarze 570: break;
571: default:
572: abort();
573: }
1.73 schwarze 574: arg2 = NULL;
575: if (*fmt == 'M')
576: arg2 = va_arg(ap, char *);
577: if (arg1 == NULL)
578: continue;
579:
580: /* Print the non-style attributes. */
581:
1.67 schwarze 582: print_byte(h, ' ');
583: print_word(h, attr);
584: print_byte(h, '=');
585: print_byte(h, '"');
1.65 schwarze 586: switch (*fmt) {
1.78 schwarze 587: case 'I':
588: print_href(h, arg1, NULL, 0);
589: fmt++;
590: break;
1.65 schwarze 591: case 'M':
1.73 schwarze 592: print_href(h, arg1, arg2, 1);
1.65 schwarze 593: fmt++;
594: break;
1.78 schwarze 595: case 'R':
596: print_byte(h, '#');
597: print_encode(h, arg1, NULL, 1);
1.65 schwarze 598: fmt++;
599: break;
1.78 schwarze 600: case 'T':
601: print_encode(h, arg1, NULL, 1);
602: print_word(h, "\" title=\"");
603: print_encode(h, arg1, NULL, 1);
1.65 schwarze 604: fmt++;
1.78 schwarze 605: break;
1.65 schwarze 606: default:
1.73 schwarze 607: print_encode(h, arg1, NULL, 1);
1.65 schwarze 608: break;
609: }
1.67 schwarze 610: print_byte(h, '"');
1.64 schwarze 611: }
612:
613: /* Print out styles. */
614:
615: while (*fmt != '\0') {
1.73 schwarze 616: arg1 = NULL;
617: su = NULL;
1.64 schwarze 618:
619: /* First letter: input argument type. */
620:
621: switch (*fmt++) {
622: case 'h':
623: i = va_arg(ap, int);
1.73 schwarze 624: su = &mysu;
1.64 schwarze 625: SCALE_HS_INIT(su, i);
626: break;
627: case 's':
1.73 schwarze 628: arg1 = va_arg(ap, char *);
1.64 schwarze 629: break;
630: case 'u':
631: su = va_arg(ap, struct roffsu *);
632: break;
633: case 'w':
1.89 schwarze 634: if ((arg2 = va_arg(ap, char *)) != NULL) {
635: su = &mysu;
636: a2width(arg2, su);
637: }
638: if (*fmt == '*') {
639: if (su != NULL && su->unit == SCALE_EN &&
640: su->scale > 5.9 && su->scale < 6.1)
641: su = NULL;
642: fmt++;
1.88 schwarze 643: }
1.81 schwarze 644: if (*fmt == '+') {
1.89 schwarze 645: if (su != NULL) {
646: /* Make even bold text fit. */
647: su->scale *= 1.2;
648: /* Add padding. */
649: su->scale += 3.0;
650: }
1.81 schwarze 651: fmt++;
652: }
653: if (*fmt == '-') {
1.89 schwarze 654: if (su != NULL)
655: su->scale *= -1.0;
1.81 schwarze 656: fmt++;
657: }
1.64 schwarze 658: break;
659: default:
660: abort();
661: }
662:
663: /* Second letter: style name. */
664:
665: switch (*fmt++) {
666: case 'h':
1.65 schwarze 667: attr = "height";
1.64 schwarze 668: break;
669: case 'i':
1.65 schwarze 670: attr = "text-indent";
1.64 schwarze 671: break;
672: case 'l':
1.65 schwarze 673: attr = "margin-left";
1.64 schwarze 674: break;
675: case 'w':
1.65 schwarze 676: attr = "width";
1.64 schwarze 677: break;
678: case 'W':
1.65 schwarze 679: attr = "min-width";
1.64 schwarze 680: break;
681: case '?':
1.73 schwarze 682: attr = arg1;
683: arg1 = va_arg(ap, char *);
684: break;
1.64 schwarze 685: default:
686: abort();
687: }
1.73 schwarze 688: if (su == NULL && arg1 == NULL)
689: continue;
690:
691: if (have_style == 1)
692: print_word(h, " style=\"");
693: else
694: print_byte(h, ' ');
1.67 schwarze 695: print_word(h, attr);
696: print_byte(h, ':');
697: print_byte(h, ' ');
1.73 schwarze 698: if (su != NULL) {
699: v = su->scale;
700: if (su->unit == SCALE_MM && (v /= 100.0) == 0.0)
701: v = 1.0;
702: else if (su->unit == SCALE_BU)
703: v /= 24.0;
704: (void)snprintf(numbuf, sizeof(numbuf), "%.2f", v);
705: print_word(h, numbuf);
706: print_word(h, roffscales[su->unit]);
707: } else
708: print_word(h, arg1);
1.67 schwarze 709: print_byte(h, ';');
1.73 schwarze 710: have_style = 2;
1.64 schwarze 711: }
1.73 schwarze 712: if (have_style == 2)
1.67 schwarze 713: print_byte(h, '"');
1.64 schwarze 714:
715: va_end(ap);
1.6 schwarze 716:
1.42 schwarze 717: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 718:
719: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 720: print_byte(h, '/');
1.6 schwarze 721:
1.67 schwarze 722: print_byte(h, '>');
1.1 schwarze 723:
1.66 schwarze 724: if (tflags & HTML_NLBEGIN)
1.67 schwarze 725: print_endline(h);
1.66 schwarze 726: else
727: h->flags |= HTML_NOSPACE;
1.18 schwarze 728:
1.66 schwarze 729: if (tflags & HTML_INDENT)
730: h->indent++;
731: if (tflags & HTML_NOINDENT)
732: h->noindent++;
1.18 schwarze 733:
1.58 schwarze 734: return t;
1.1 schwarze 735: }
736:
737: static void
1.54 schwarze 738: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 739: {
1.66 schwarze 740: int tflags;
1.35 schwarze 741:
1.54 schwarze 742: /*
743: * Remember to close out and nullify the current
744: * meta-font and table, if applicable.
745: */
746: if (tag == h->metaf)
747: h->metaf = NULL;
748: if (tag == h->tblt)
749: h->tblt = NULL;
750:
1.66 schwarze 751: tflags = htmltags[tag->tag].flags;
752:
753: if (tflags & HTML_INDENT)
754: h->indent--;
755: if (tflags & HTML_NOINDENT)
756: h->noindent--;
757: if (tflags & HTML_NLEND)
1.67 schwarze 758: print_endline(h);
759: print_indent(h);
760: print_byte(h, '<');
761: print_byte(h, '/');
762: print_word(h, htmltags[tag->tag].name);
763: print_byte(h, '>');
1.66 schwarze 764: if (tflags & HTML_NLAFTER)
1.67 schwarze 765: print_endline(h);
1.54 schwarze 766:
1.74 schwarze 767: h->tag = tag->next;
1.54 schwarze 768: free(tag);
1.1 schwarze 769: }
770:
771: void
1.6 schwarze 772: print_gen_decls(struct html *h)
773: {
1.67 schwarze 774: print_word(h, "<!DOCTYPE html>");
775: print_endline(h);
1.91 schwarze 776: }
777:
778: void
779: print_gen_comment(struct html *h, struct roff_node *n)
780: {
781: int wantblank;
782:
783: print_word(h, "<!-- This is an automatically generated file."
784: " Do not edit.");
785: h->indent = 1;
786: wantblank = 0;
787: while (n != NULL && n->type == ROFFT_COMMENT) {
788: if (strstr(n->string, "-->") == NULL &&
789: (wantblank || *n->string != '\0')) {
790: print_endline(h);
791: print_indent(h);
792: print_word(h, n->string);
793: wantblank = *n->string != '\0';
794: }
795: n = n->next;
796: }
797: if (wantblank)
798: print_endline(h);
799: print_word(h, " -->");
800: print_endline(h);
801: h->indent = 0;
1.1 schwarze 802: }
803:
804: void
1.12 schwarze 805: print_text(struct html *h, const char *word)
1.1 schwarze 806: {
1.67 schwarze 807: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 808: if ( ! (HTML_KEEP & h->flags)) {
809: if (HTML_PREKEEP & h->flags)
810: h->flags |= HTML_KEEP;
1.67 schwarze 811: print_endword(h);
1.12 schwarze 812: } else
1.86 bentley 813: print_word(h, " ");
1.12 schwarze 814: }
1.1 schwarze 815:
1.20 schwarze 816: assert(NULL == h->metaf);
1.31 schwarze 817: switch (h->metac) {
1.35 schwarze 818: case HTMLFONT_ITALIC:
1.64 schwarze 819: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 820: break;
1.35 schwarze 821: case HTMLFONT_BOLD:
1.64 schwarze 822: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 823: break;
1.35 schwarze 824: case HTMLFONT_BI:
1.64 schwarze 825: h->metaf = print_otag(h, TAG_B, "");
826: print_otag(h, TAG_I, "");
1.31 schwarze 827: break;
828: default:
1.67 schwarze 829: print_indent(h);
1.31 schwarze 830: break;
831: }
1.20 schwarze 832:
1.12 schwarze 833: assert(word);
1.65 schwarze 834: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 835: if ( ! (h->flags & HTML_NONOSPACE))
836: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 837: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 838: } else
1.53 schwarze 839: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 840:
841: if (h->metaf) {
842: print_tagq(h, h->metaf);
843: h->metaf = NULL;
844: }
1.17 schwarze 845:
846: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 847: }
848:
849: void
850: print_tagq(struct html *h, const struct tag *until)
851: {
852: struct tag *tag;
853:
1.74 schwarze 854: while ((tag = h->tag) != NULL) {
1.54 schwarze 855: print_ctag(h, tag);
1.1 schwarze 856: if (until && tag == until)
857: return;
858: }
859: }
860:
861: void
862: print_stagq(struct html *h, const struct tag *suntil)
863: {
864: struct tag *tag;
865:
1.74 schwarze 866: while ((tag = h->tag) != NULL) {
1.1 schwarze 867: if (suntil && tag == suntil)
868: return;
1.54 schwarze 869: print_ctag(h, tag);
1.1 schwarze 870: }
871: }
1.42 schwarze 872:
873: void
874: print_paragraph(struct html *h)
875: {
876: struct tag *t;
877:
1.68 schwarze 878: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 879: print_tagq(h, t);
880: }
881:
1.67 schwarze 882:
883: /***********************************************************************
884: * Low level output functions.
885: * They implement line breaking using a short static buffer.
886: ***********************************************************************/
887:
888: /*
889: * Buffer one HTML output byte.
890: * If the buffer is full, flush and deactivate it and start a new line.
891: * If the buffer is inactive, print directly.
892: */
893: static void
894: print_byte(struct html *h, char c)
895: {
896: if ((h->flags & HTML_BUFFER) == 0) {
897: putchar(c);
898: h->col++;
899: return;
900: }
901:
902: if (h->col + h->bufcol < sizeof(h->buf)) {
903: h->buf[h->bufcol++] = c;
904: return;
905: }
906:
907: putchar('\n');
908: h->col = 0;
909: print_indent(h);
910: putchar(' ');
911: putchar(' ');
912: fwrite(h->buf, h->bufcol, 1, stdout);
913: putchar(c);
914: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
915: h->bufcol = 0;
916: h->flags &= ~HTML_BUFFER;
917: }
918:
1.66 schwarze 919: /*
920: * If something was printed on the current output line, end it.
1.67 schwarze 921: * Not to be called right after print_indent().
1.66 schwarze 922: */
1.72 schwarze 923: void
1.67 schwarze 924: print_endline(struct html *h)
1.66 schwarze 925: {
1.67 schwarze 926: if (h->col == 0)
1.66 schwarze 927: return;
928:
1.67 schwarze 929: if (h->bufcol) {
930: putchar(' ');
931: fwrite(h->buf, h->bufcol, 1, stdout);
932: h->bufcol = 0;
933: }
1.66 schwarze 934: putchar('\n');
1.67 schwarze 935: h->col = 0;
936: h->flags |= HTML_NOSPACE;
937: h->flags &= ~HTML_BUFFER;
938: }
939:
940: /*
941: * Flush the HTML output buffer.
942: * If it is inactive, activate it.
943: */
944: static void
945: print_endword(struct html *h)
946: {
947: if (h->noindent) {
948: print_byte(h, ' ');
949: return;
950: }
951:
952: if ((h->flags & HTML_BUFFER) == 0) {
953: h->col++;
954: h->flags |= HTML_BUFFER;
955: } else if (h->bufcol) {
956: putchar(' ');
957: fwrite(h->buf, h->bufcol, 1, stdout);
958: h->col += h->bufcol + 1;
959: }
960: h->bufcol = 0;
1.66 schwarze 961: }
962:
963: /*
964: * If at the beginning of a new output line,
965: * perform indentation and mark the line as containing output.
966: * Make sure to really produce some output right afterwards,
967: * but do not use print_otag() for producing it.
968: */
969: static void
1.67 schwarze 970: print_indent(struct html *h)
1.66 schwarze 971: {
1.67 schwarze 972: size_t i;
1.66 schwarze 973:
1.67 schwarze 974: if (h->col)
1.66 schwarze 975: return;
976:
1.67 schwarze 977: if (h->noindent == 0) {
978: h->col = h->indent * 2;
979: for (i = 0; i < h->col; i++)
1.66 schwarze 980: putchar(' ');
1.67 schwarze 981: }
982: h->flags &= ~HTML_NOSPACE;
983: }
984:
985: /*
986: * Print or buffer some characters
987: * depending on the current HTML output buffer state.
988: */
989: static void
990: print_word(struct html *h, const char *cp)
991: {
992: while (*cp != '\0')
993: print_byte(h, *cp++);
1.66 schwarze 994: }
1.64 schwarze 995:
996: /*
997: * Calculate the scaling unit passed in a `-width' argument. This uses
998: * either a native scaling unit (e.g., 1i, 2m) or the string length of
999: * the value.
1000: */
1001: static void
1002: a2width(const char *p, struct roffsu *su)
1003: {
1.83 schwarze 1004: const char *end;
1005:
1006: end = a2roffsu(p, su, SCALE_MAX);
1007: if (end == NULL || *end != '\0') {
1.64 schwarze 1008: su->unit = SCALE_EN;
1009: su->scale = html_strlen(p);
1010: } else if (su->scale < 0.0)
1011: su->scale = 0.0;
1.3 schwarze 1012: }