Annotation of src/usr.bin/mandoc/html.c, Revision 1.69
1.69 ! schwarze 1: /* $OpenBSD: html.c,v 1.68 2017/01/19 15:27:26 schwarze Exp $ */
1.1 schwarze 2: /*
1.42 schwarze 3: * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.64 schwarze 4: * Copyright (c) 2011-2015, 2017 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:
1.9 schwarze 29: #include "mandoc.h"
1.34 schwarze 30: #include "mandoc_aux.h"
1.1 schwarze 31: #include "out.h"
32: #include "html.h"
1.56 schwarze 33: #include "manconf.h"
1.1 schwarze 34: #include "main.h"
35:
36: struct htmldata {
37: const char *name;
38: int flags;
1.66 schwarze 39: #define HTML_NOSTACK (1 << 0)
40: #define HTML_AUTOCLOSE (1 << 1)
41: #define HTML_NLBEFORE (1 << 2)
42: #define HTML_NLBEGIN (1 << 3)
43: #define HTML_NLEND (1 << 4)
44: #define HTML_NLAFTER (1 << 5)
45: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
46: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
47: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
48: #define HTML_INDENT (1 << 6)
49: #define HTML_NOINDENT (1 << 7)
1.1 schwarze 50: };
51:
52: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 53: {"html", HTML_NLALL},
54: {"head", HTML_NLALL | HTML_INDENT},
55: {"body", HTML_NLALL},
56: {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
57: {"title", HTML_NLAROUND},
58: {"div", HTML_NLAROUND},
59: {"h1", HTML_NLAROUND},
60: {"h2", HTML_NLAROUND},
61: {"span", 0},
62: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
63: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
64: {"a", 0},
65: {"table", HTML_NLALL | HTML_INDENT},
66: {"tbody", HTML_NLALL | HTML_INDENT},
67: {"col", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
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},
76: {"pre", HTML_NLALL | HTML_NOINDENT},
77: {"b", 0},
78: {"i", 0},
79: {"code", 0},
80: {"small", 0},
81: {"style", HTML_NLALL | HTML_INDENT},
82: {"math", HTML_NLALL | HTML_INDENT},
83: {"mrow", 0},
84: {"mi", 0},
85: {"mo", 0},
86: {"msup", 0},
87: {"msub", 0},
88: {"msubsup", 0},
89: {"mfrac", 0},
90: {"msqrt", 0},
91: {"mfenced", 0},
92: {"mtable", 0},
93: {"mtr", 0},
94: {"mtd", 0},
95: {"munderover", 0},
96: {"munder", 0},
97: {"mover", 0},
1.5 schwarze 98: };
99:
1.26 schwarze 100: static const char *const roffscales[SCALE_MAX] = {
101: "cm", /* SCALE_CM */
102: "in", /* SCALE_IN */
103: "pc", /* SCALE_PC */
104: "pt", /* SCALE_PT */
105: "em", /* SCALE_EM */
106: "em", /* SCALE_MM */
107: "ex", /* SCALE_EN */
108: "ex", /* SCALE_BU */
109: "em", /* SCALE_VS */
110: "ex", /* SCALE_FS */
111: };
1.5 schwarze 112:
1.64 schwarze 113: static void a2width(const char *, struct roffsu *);
1.67 schwarze 114: static void print_byte(struct html *, char);
115: static void print_endline(struct html *);
116: static void print_endword(struct html *);
117: static void print_indent(struct html *);
118: static void print_word(struct html *, const char *);
119:
1.54 schwarze 120: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 121: static int print_escape(struct html *, char);
1.65 schwarze 122: static int print_encode(struct html *, const char *, const char *, int);
123: static void print_href(struct html *, const char *, const char *, int);
1.26 schwarze 124: static void print_metaf(struct html *, enum mandoc_esc);
1.5 schwarze 125:
1.35 schwarze 126:
1.50 schwarze 127: void *
1.61 schwarze 128: html_alloc(const struct manoutput *outopts)
1.1 schwarze 129: {
130: struct html *h;
131:
1.24 schwarze 132: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 133:
1.2 schwarze 134: h->tags.head = NULL;
1.56 schwarze 135: h->style = outopts->style;
136: h->base_man = outopts->man;
137: h->base_includes = outopts->includes;
138: if (outopts->fragment)
139: h->oflags |= HTML_FRAGMENT;
1.1 schwarze 140:
1.58 schwarze 141: return h;
1.1 schwarze 142: }
143:
144: void
145: html_free(void *p)
146: {
147: struct tag *tag;
148: struct html *h;
149:
150: h = (struct html *)p;
151:
1.2 schwarze 152: while ((tag = h->tags.head) != NULL) {
1.35 schwarze 153: h->tags.head = tag->next;
1.1 schwarze 154: free(tag);
155: }
156:
157: free(h);
158: }
159:
160: void
161: print_gen_head(struct html *h)
162: {
1.42 schwarze 163: struct tag *t;
164:
1.64 schwarze 165: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.1 schwarze 166:
1.42 schwarze 167: /*
168: * Print a default style-sheet.
169: */
1.66 schwarze 170:
1.64 schwarze 171: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 172: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 173: print_endline(h);
1.66 schwarze 174: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 175: print_endline(h);
1.66 schwarze 176: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 177: print_endline(h);
1.66 schwarze 178: print_text(h, "table.foot td { width: 50%; }");
1.67 schwarze 179: print_endline(h);
1.66 schwarze 180: print_text(h, "table.head td { width: 33%; }");
1.67 schwarze 181: print_endline(h);
1.68 schwarze 182: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.42 schwarze 183: print_tagq(h, t);
1.1 schwarze 184:
1.64 schwarze 185: if (h->style)
186: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
187: h->style, "type", "text/css", "media", "all");
1.1 schwarze 188: }
189:
1.5 schwarze 190: static void
1.26 schwarze 191: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 192: {
193: enum htmlfont font;
1.1 schwarze 194:
1.5 schwarze 195: switch (deco) {
1.35 schwarze 196: case ESCAPE_FONTPREV:
1.5 schwarze 197: font = h->metal;
198: break;
1.35 schwarze 199: case ESCAPE_FONTITALIC:
1.5 schwarze 200: font = HTMLFONT_ITALIC;
201: break;
1.35 schwarze 202: case ESCAPE_FONTBOLD:
1.5 schwarze 203: font = HTMLFONT_BOLD;
204: break;
1.35 schwarze 205: case ESCAPE_FONTBI:
1.31 schwarze 206: font = HTMLFONT_BI;
207: break;
1.35 schwarze 208: case ESCAPE_FONT:
209: case ESCAPE_FONTROMAN:
1.5 schwarze 210: font = HTMLFONT_NONE;
211: break;
212: default:
213: abort();
1.1 schwarze 214: }
215:
1.20 schwarze 216: if (h->metaf) {
217: print_tagq(h, h->metaf);
218: h->metaf = NULL;
219: }
220:
221: h->metal = h->metac;
222: h->metac = font;
223:
1.31 schwarze 224: switch (font) {
1.35 schwarze 225: case HTMLFONT_ITALIC:
1.64 schwarze 226: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 227: break;
1.35 schwarze 228: case HTMLFONT_BOLD:
1.64 schwarze 229: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 230: break;
1.35 schwarze 231: case HTMLFONT_BI:
1.64 schwarze 232: h->metaf = print_otag(h, TAG_B, "");
233: print_otag(h, TAG_I, "");
1.31 schwarze 234: break;
235: default:
236: break;
237: }
1.1 schwarze 238: }
239:
1.26 schwarze 240: int
241: html_strlen(const char *cp)
242: {
1.30 schwarze 243: size_t rsz;
244: int skip, sz;
1.26 schwarze 245:
246: /*
247: * Account for escaped sequences within string length
248: * calculations. This follows the logic in term_strlen() as we
249: * must calculate the width of produced strings.
250: * Assume that characters are always width of "1". This is
251: * hacky, but it gets the job done for approximation of widths.
252: */
253:
254: sz = 0;
1.30 schwarze 255: skip = 0;
256: while (1) {
257: rsz = strcspn(cp, "\\");
258: if (rsz) {
259: cp += rsz;
260: if (skip) {
261: skip = 0;
262: rsz--;
263: }
264: sz += rsz;
265: }
266: if ('\0' == *cp)
267: break;
268: cp++;
269: switch (mandoc_escape(&cp, NULL, NULL)) {
1.35 schwarze 270: case ESCAPE_ERROR:
1.58 schwarze 271: return sz;
1.35 schwarze 272: case ESCAPE_UNICODE:
273: case ESCAPE_NUMBERED:
274: case ESCAPE_SPECIAL:
1.55 schwarze 275: case ESCAPE_OVERSTRIKE:
1.30 schwarze 276: if (skip)
277: skip = 0;
278: else
279: sz++;
280: break;
1.35 schwarze 281: case ESCAPE_SKIPCHAR:
1.30 schwarze 282: skip = 1;
1.26 schwarze 283: break;
284: default:
285: break;
286: }
287: }
1.58 schwarze 288: return sz;
1.26 schwarze 289: }
1.1 schwarze 290:
1.5 schwarze 291: static int
1.67 schwarze 292: print_escape(struct html *h, char c)
1.38 schwarze 293: {
294:
295: switch (c) {
296: case '<':
1.67 schwarze 297: print_word(h, "<");
1.38 schwarze 298: break;
299: case '>':
1.67 schwarze 300: print_word(h, ">");
1.38 schwarze 301: break;
302: case '&':
1.67 schwarze 303: print_word(h, "&");
1.38 schwarze 304: break;
305: case '"':
1.67 schwarze 306: print_word(h, """);
1.38 schwarze 307: break;
308: case ASCII_NBRSP:
1.67 schwarze 309: print_word(h, " ");
1.38 schwarze 310: break;
311: case ASCII_HYPH:
1.67 schwarze 312: print_byte(h, '-');
1.59 schwarze 313: break;
1.38 schwarze 314: case ASCII_BREAK:
315: break;
316: default:
1.58 schwarze 317: return 0;
1.38 schwarze 318: }
1.58 schwarze 319: return 1;
1.38 schwarze 320: }
321:
322: static int
1.65 schwarze 323: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 324: {
1.67 schwarze 325: char numbuf[16];
1.4 schwarze 326: size_t sz;
1.26 schwarze 327: int c, len, nospace;
1.5 schwarze 328: const char *seq;
1.26 schwarze 329: enum mandoc_esc esc;
1.37 schwarze 330: static const char rejs[9] = { '\\', '<', '>', '&', '"',
1.33 schwarze 331: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 332:
1.65 schwarze 333: if (pend == NULL)
334: pend = strchr(p, '\0');
335:
1.5 schwarze 336: nospace = 0;
1.1 schwarze 337:
1.65 schwarze 338: while (p < pend) {
1.30 schwarze 339: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
340: h->flags &= ~HTML_SKIPCHAR;
341: p++;
342: continue;
343: }
344:
1.67 schwarze 345: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
346: if (*p == ' ')
347: print_endword(h);
348: else
349: print_byte(h, *p);
1.4 schwarze 350:
1.65 schwarze 351: if (p >= pend)
1.26 schwarze 352: break;
353:
1.67 schwarze 354: if (print_escape(h, *p++))
1.33 schwarze 355: continue;
1.4 schwarze 356:
1.26 schwarze 357: esc = mandoc_escape(&p, &seq, &len);
358: if (ESCAPE_ERROR == esc)
359: break;
1.5 schwarze 360:
1.26 schwarze 361: switch (esc) {
1.35 schwarze 362: case ESCAPE_FONT:
363: case ESCAPE_FONTPREV:
364: case ESCAPE_FONTBOLD:
365: case ESCAPE_FONTITALIC:
366: case ESCAPE_FONTBI:
367: case ESCAPE_FONTROMAN:
1.30 schwarze 368: if (0 == norecurse)
369: print_metaf(h, esc);
370: continue;
1.35 schwarze 371: case ESCAPE_SKIPCHAR:
1.30 schwarze 372: h->flags |= HTML_SKIPCHAR;
373: continue;
374: default:
375: break;
376: }
377:
378: if (h->flags & HTML_SKIPCHAR) {
379: h->flags &= ~HTML_SKIPCHAR;
380: continue;
381: }
382:
383: switch (esc) {
1.35 schwarze 384: case ESCAPE_UNICODE:
1.38 schwarze 385: /* Skip past "u" header. */
1.26 schwarze 386: c = mchars_num2uc(seq + 1, len - 1);
387: break;
1.35 schwarze 388: case ESCAPE_NUMBERED:
1.26 schwarze 389: c = mchars_num2char(seq, len);
1.51 schwarze 390: if (c < 0)
391: continue;
1.26 schwarze 392: break;
1.35 schwarze 393: case ESCAPE_SPECIAL:
1.61 schwarze 394: c = mchars_spec2cp(seq, len);
1.51 schwarze 395: if (c <= 0)
396: continue;
1.26 schwarze 397: break;
1.35 schwarze 398: case ESCAPE_NOSPACE:
1.26 schwarze 399: if ('\0' == *p)
400: nospace = 1;
1.49 schwarze 401: continue;
1.55 schwarze 402: case ESCAPE_OVERSTRIKE:
403: if (len == 0)
404: continue;
405: c = seq[len - 1];
406: break;
1.5 schwarze 407: default:
1.49 schwarze 408: continue;
1.5 schwarze 409: }
1.51 schwarze 410: if ((c < 0x20 && c != 0x09) ||
411: (c > 0x7E && c < 0xA0))
1.49 schwarze 412: c = 0xFFFD;
1.67 schwarze 413: if (c > 0x7E) {
414: (void)snprintf(numbuf, sizeof(numbuf), "&#%d;", c);
415: print_word(h, numbuf);
416: } else if (print_escape(h, c) == 0)
417: print_byte(h, c);
1.1 schwarze 418: }
1.5 schwarze 419:
1.58 schwarze 420: return nospace;
1.1 schwarze 421: }
422:
1.6 schwarze 423: static void
1.65 schwarze 424: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 425: {
1.65 schwarze 426: const char *p, *pp;
427:
428: pp = man ? h->base_man : h->base_includes;
429: while ((p = strchr(pp, '%')) != NULL) {
430: print_encode(h, pp, p, 1);
431: if (man && p[1] == 'S') {
432: if (sec == NULL)
1.67 schwarze 433: print_byte(h, '1');
1.65 schwarze 434: else
435: print_encode(h, sec, NULL, 1);
436: } else if ((man && p[1] == 'N') ||
437: (man == 0 && p[1] == 'I'))
438: print_encode(h, name, NULL, 1);
439: else
440: print_encode(h, p, p + 2, 1);
441: pp = p + 2;
442: }
443: if (*pp != '\0')
444: print_encode(h, pp, NULL, 1);
1.6 schwarze 445: }
446:
1.1 schwarze 447: struct tag *
1.64 schwarze 448: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 449: {
1.64 schwarze 450: va_list ap;
451: struct roffsu mysu, *su;
1.67 schwarze 452: char numbuf[16];
1.1 schwarze 453: struct tag *t;
1.65 schwarze 454: const char *attr;
1.64 schwarze 455: char *s;
1.65 schwarze 456: double v;
1.66 schwarze 457: int i, have_style, tflags;
458:
459: tflags = htmltags[tag].flags;
1.1 schwarze 460:
1.6 schwarze 461: /* Push this tags onto the stack of open scopes. */
462:
1.66 schwarze 463: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 464: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 465: t->tag = tag;
1.2 schwarze 466: t->next = h->tags.head;
467: h->tags.head = t;
1.1 schwarze 468: } else
469: t = NULL;
470:
1.66 schwarze 471: if (tflags & HTML_NLBEFORE)
1.67 schwarze 472: print_endline(h);
473: if (h->col == 0)
474: print_indent(h);
1.66 schwarze 475: else if ((h->flags & HTML_NOSPACE) == 0) {
476: if (h->flags & HTML_KEEP)
1.67 schwarze 477: print_word(h, " ");
1.66 schwarze 478: else {
479: if (h->flags & HTML_PREKEEP)
480: h->flags |= HTML_KEEP;
1.67 schwarze 481: print_endword(h);
1.12 schwarze 482: }
1.66 schwarze 483: }
1.1 schwarze 484:
1.13 schwarze 485: if ( ! (h->flags & HTML_NONOSPACE))
486: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 487: else
488: h->flags |= HTML_NOSPACE;
1.13 schwarze 489:
1.6 schwarze 490: /* Print out the tag name and attributes. */
491:
1.67 schwarze 492: print_byte(h, '<');
493: print_word(h, htmltags[tag].name);
1.64 schwarze 494:
495: va_start(ap, fmt);
496:
497: have_style = 0;
498: while (*fmt != '\0') {
499: if (*fmt == 's') {
1.67 schwarze 500: print_word(h, " style=\"");
1.64 schwarze 501: have_style = 1;
502: fmt++;
503: break;
504: }
505: s = va_arg(ap, char *);
506: switch (*fmt++) {
507: case 'c':
1.65 schwarze 508: attr = "class";
1.64 schwarze 509: break;
510: case 'h':
1.65 schwarze 511: attr = "href";
1.64 schwarze 512: break;
513: case 'i':
1.65 schwarze 514: attr = "id";
1.64 schwarze 515: break;
516: case '?':
1.65 schwarze 517: attr = s;
518: s = va_arg(ap, char *);
1.64 schwarze 519: break;
520: default:
521: abort();
522: }
1.67 schwarze 523: print_byte(h, ' ');
524: print_word(h, attr);
525: print_byte(h, '=');
526: print_byte(h, '"');
1.65 schwarze 527: switch (*fmt) {
528: case 'M':
529: print_href(h, s, va_arg(ap, char *), 1);
530: fmt++;
531: break;
532: case 'I':
533: print_href(h, s, NULL, 0);
534: fmt++;
535: break;
536: case 'R':
1.67 schwarze 537: print_byte(h, '#');
1.65 schwarze 538: fmt++;
539: /* FALLTHROUGH */
540: default:
541: print_encode(h, s, NULL, 1);
542: break;
543: }
1.67 schwarze 544: print_byte(h, '"');
1.64 schwarze 545: }
546:
547: /* Print out styles. */
548:
549: s = NULL;
550: su = &mysu;
551: while (*fmt != '\0') {
552:
553: /* First letter: input argument type. */
554:
555: switch (*fmt++) {
556: case 'h':
557: i = va_arg(ap, int);
558: SCALE_HS_INIT(su, i);
559: break;
560: case 's':
561: s = va_arg(ap, char *);
562: break;
563: case 'u':
564: su = va_arg(ap, struct roffsu *);
565: break;
566: case 'v':
567: i = va_arg(ap, int);
568: SCALE_VS_INIT(su, i);
569: break;
570: case 'w':
571: s = va_arg(ap, char *);
572: a2width(s, su);
573: break;
574: default:
575: abort();
576: }
577:
578: /* Second letter: style name. */
579:
580: switch (*fmt++) {
581: case 'b':
1.65 schwarze 582: attr = "margin-bottom";
1.64 schwarze 583: break;
584: case 'h':
1.65 schwarze 585: attr = "height";
1.64 schwarze 586: break;
587: case 'i':
1.65 schwarze 588: attr = "text-indent";
1.64 schwarze 589: break;
590: case 'l':
1.65 schwarze 591: attr = "margin-left";
1.64 schwarze 592: break;
593: case 't':
1.65 schwarze 594: attr = "margin-top";
1.64 schwarze 595: break;
596: case 'w':
1.65 schwarze 597: attr = "width";
1.64 schwarze 598: break;
599: case 'W':
1.65 schwarze 600: attr = "min-width";
1.64 schwarze 601: break;
602: case '?':
1.67 schwarze 603: print_word(h, s);
604: print_byte(h, ':');
605: print_byte(h, ' ');
606: print_word(h, va_arg(ap, char *));
607: print_byte(h, ';');
608: if (*fmt != '\0')
609: print_byte(h, ' ');
1.65 schwarze 610: continue;
1.64 schwarze 611: default:
612: abort();
613: }
1.65 schwarze 614: v = su->scale;
615: if (su->unit == SCALE_MM && (v /= 100.0) == 0.0)
616: v = 1.0;
617: else if (su->unit == SCALE_BU)
618: v /= 24.0;
1.67 schwarze 619: print_word(h, attr);
620: print_byte(h, ':');
621: print_byte(h, ' ');
622: (void)snprintf(numbuf, sizeof(numbuf), "%.2f", v);
623: print_word(h, numbuf);
624: print_word(h, roffscales[su->unit]);
625: print_byte(h, ';');
626: if (*fmt != '\0')
627: print_byte(h, ' ');
1.64 schwarze 628: }
629: if (have_style)
1.67 schwarze 630: print_byte(h, '"');
1.64 schwarze 631:
632: va_end(ap);
1.6 schwarze 633:
1.42 schwarze 634: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 635:
636: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 637: print_byte(h, '/');
1.6 schwarze 638:
1.67 schwarze 639: print_byte(h, '>');
1.1 schwarze 640:
1.66 schwarze 641: if (tflags & HTML_NLBEGIN)
1.67 schwarze 642: print_endline(h);
1.66 schwarze 643: else
644: h->flags |= HTML_NOSPACE;
1.18 schwarze 645:
1.66 schwarze 646: if (tflags & HTML_INDENT)
647: h->indent++;
648: if (tflags & HTML_NOINDENT)
649: h->noindent++;
1.18 schwarze 650:
1.58 schwarze 651: return t;
1.1 schwarze 652: }
653:
654: static void
1.54 schwarze 655: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 656: {
1.66 schwarze 657: int tflags;
1.35 schwarze 658:
1.54 schwarze 659: /*
660: * Remember to close out and nullify the current
661: * meta-font and table, if applicable.
662: */
663: if (tag == h->metaf)
664: h->metaf = NULL;
665: if (tag == h->tblt)
666: h->tblt = NULL;
667:
1.66 schwarze 668: tflags = htmltags[tag->tag].flags;
669:
670: if (tflags & HTML_INDENT)
671: h->indent--;
672: if (tflags & HTML_NOINDENT)
673: h->noindent--;
674: if (tflags & HTML_NLEND)
1.67 schwarze 675: print_endline(h);
676: print_indent(h);
677: print_byte(h, '<');
678: print_byte(h, '/');
679: print_word(h, htmltags[tag->tag].name);
680: print_byte(h, '>');
1.66 schwarze 681: if (tflags & HTML_NLAFTER)
1.67 schwarze 682: print_endline(h);
1.54 schwarze 683:
684: h->tags.head = tag->next;
685: free(tag);
1.1 schwarze 686: }
687:
688: void
1.6 schwarze 689: print_gen_decls(struct html *h)
690: {
1.67 schwarze 691: print_word(h, "<!DOCTYPE html>");
692: print_endline(h);
1.1 schwarze 693: }
694:
695: void
1.12 schwarze 696: print_text(struct html *h, const char *word)
1.1 schwarze 697: {
1.67 schwarze 698: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 699: if ( ! (HTML_KEEP & h->flags)) {
700: if (HTML_PREKEEP & h->flags)
701: h->flags |= HTML_KEEP;
1.67 schwarze 702: print_endword(h);
1.12 schwarze 703: } else
1.67 schwarze 704: print_word(h, " ");
1.12 schwarze 705: }
1.1 schwarze 706:
1.20 schwarze 707: assert(NULL == h->metaf);
1.31 schwarze 708: switch (h->metac) {
1.35 schwarze 709: case HTMLFONT_ITALIC:
1.64 schwarze 710: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 711: break;
1.35 schwarze 712: case HTMLFONT_BOLD:
1.64 schwarze 713: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 714: break;
1.35 schwarze 715: case HTMLFONT_BI:
1.64 schwarze 716: h->metaf = print_otag(h, TAG_B, "");
717: print_otag(h, TAG_I, "");
1.31 schwarze 718: break;
719: default:
1.67 schwarze 720: print_indent(h);
1.31 schwarze 721: break;
722: }
1.20 schwarze 723:
1.12 schwarze 724: assert(word);
1.65 schwarze 725: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 726: if ( ! (h->flags & HTML_NONOSPACE))
727: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 728: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 729: } else
1.53 schwarze 730: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 731:
732: if (h->metaf) {
733: print_tagq(h, h->metaf);
734: h->metaf = NULL;
735: }
1.17 schwarze 736:
737: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 738: }
739:
740: void
741: print_tagq(struct html *h, const struct tag *until)
742: {
743: struct tag *tag;
744:
1.2 schwarze 745: while ((tag = h->tags.head) != NULL) {
1.54 schwarze 746: print_ctag(h, tag);
1.1 schwarze 747: if (until && tag == until)
748: return;
749: }
750: }
751:
752: void
753: print_stagq(struct html *h, const struct tag *suntil)
754: {
755: struct tag *tag;
756:
1.2 schwarze 757: while ((tag = h->tags.head) != NULL) {
1.1 schwarze 758: if (suntil && tag == suntil)
759: return;
1.54 schwarze 760: print_ctag(h, tag);
1.1 schwarze 761: }
762: }
1.42 schwarze 763:
764: void
765: print_paragraph(struct html *h)
766: {
767: struct tag *t;
768:
1.68 schwarze 769: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 770: print_tagq(h, t);
771: }
772:
1.67 schwarze 773:
774: /***********************************************************************
775: * Low level output functions.
776: * They implement line breaking using a short static buffer.
777: ***********************************************************************/
778:
779: /*
780: * Buffer one HTML output byte.
781: * If the buffer is full, flush and deactivate it and start a new line.
782: * If the buffer is inactive, print directly.
783: */
784: static void
785: print_byte(struct html *h, char c)
786: {
787: if ((h->flags & HTML_BUFFER) == 0) {
788: putchar(c);
789: h->col++;
790: return;
791: }
792:
793: if (h->col + h->bufcol < sizeof(h->buf)) {
794: h->buf[h->bufcol++] = c;
795: return;
796: }
797:
798: putchar('\n');
799: h->col = 0;
800: print_indent(h);
801: putchar(' ');
802: putchar(' ');
803: fwrite(h->buf, h->bufcol, 1, stdout);
804: putchar(c);
805: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
806: h->bufcol = 0;
807: h->flags &= ~HTML_BUFFER;
808: }
809:
1.66 schwarze 810: /*
811: * If something was printed on the current output line, end it.
1.67 schwarze 812: * Not to be called right after print_indent().
1.66 schwarze 813: */
814: static void
1.67 schwarze 815: print_endline(struct html *h)
1.66 schwarze 816: {
1.67 schwarze 817: if (h->col == 0)
1.66 schwarze 818: return;
819:
1.67 schwarze 820: if (h->bufcol) {
821: putchar(' ');
822: fwrite(h->buf, h->bufcol, 1, stdout);
823: h->bufcol = 0;
824: }
1.66 schwarze 825: putchar('\n');
1.67 schwarze 826: h->col = 0;
827: h->flags |= HTML_NOSPACE;
828: h->flags &= ~HTML_BUFFER;
829: }
830:
831: /*
832: * Flush the HTML output buffer.
833: * If it is inactive, activate it.
834: */
835: static void
836: print_endword(struct html *h)
837: {
838: if (h->noindent) {
839: print_byte(h, ' ');
840: return;
841: }
842:
843: if ((h->flags & HTML_BUFFER) == 0) {
844: h->col++;
845: h->flags |= HTML_BUFFER;
846: } else if (h->bufcol) {
847: putchar(' ');
848: fwrite(h->buf, h->bufcol, 1, stdout);
849: h->col += h->bufcol + 1;
850: }
851: h->bufcol = 0;
1.66 schwarze 852: }
853:
854: /*
855: * If at the beginning of a new output line,
856: * perform indentation and mark the line as containing output.
857: * Make sure to really produce some output right afterwards,
858: * but do not use print_otag() for producing it.
859: */
860: static void
1.67 schwarze 861: print_indent(struct html *h)
1.66 schwarze 862: {
1.67 schwarze 863: size_t i;
1.66 schwarze 864:
1.67 schwarze 865: if (h->col)
1.66 schwarze 866: return;
867:
1.67 schwarze 868: if (h->noindent == 0) {
869: h->col = h->indent * 2;
870: for (i = 0; i < h->col; i++)
1.66 schwarze 871: putchar(' ');
1.67 schwarze 872: }
873: h->flags &= ~HTML_NOSPACE;
874: }
875:
876: /*
877: * Print or buffer some characters
878: * depending on the current HTML output buffer state.
879: */
880: static void
881: print_word(struct html *h, const char *cp)
882: {
883: while (*cp != '\0')
884: print_byte(h, *cp++);
1.66 schwarze 885: }
1.64 schwarze 886:
887: /*
888: * Calculate the scaling unit passed in a `-width' argument. This uses
889: * either a native scaling unit (e.g., 1i, 2m) or the string length of
890: * the value.
891: */
892: static void
893: a2width(const char *p, struct roffsu *su)
894: {
895: if (a2roffsu(p, su, SCALE_MAX) < 2) {
896: su->unit = SCALE_EN;
897: su->scale = html_strlen(p);
898: } else if (su->scale < 0.0)
899: su->scale = 0.0;
1.3 schwarze 900: }