Annotation of src/usr.bin/mandoc/html.c, Revision 1.88
1.88 ! schwarze 1: /* $OpenBSD: html.c,v 1.87 2017/07/14 16:05:52 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:
1.80 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");
1.1 schwarze 169:
1.42 schwarze 170: /*
171: * Print a default style-sheet.
172: */
1.66 schwarze 173:
1.64 schwarze 174: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 175: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 176: print_endline(h);
1.66 schwarze 177: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 178: print_endline(h);
1.66 schwarze 179: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 180: print_endline(h);
1.68 schwarze 181: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.42 schwarze 182: print_tagq(h, t);
1.1 schwarze 183:
1.64 schwarze 184: if (h->style)
185: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
186: h->style, "type", "text/css", "media", "all");
1.1 schwarze 187: }
188:
1.5 schwarze 189: static void
1.26 schwarze 190: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 191: {
192: enum htmlfont font;
1.1 schwarze 193:
1.5 schwarze 194: switch (deco) {
1.35 schwarze 195: case ESCAPE_FONTPREV:
1.5 schwarze 196: font = h->metal;
197: break;
1.35 schwarze 198: case ESCAPE_FONTITALIC:
1.5 schwarze 199: font = HTMLFONT_ITALIC;
200: break;
1.35 schwarze 201: case ESCAPE_FONTBOLD:
1.5 schwarze 202: font = HTMLFONT_BOLD;
203: break;
1.35 schwarze 204: case ESCAPE_FONTBI:
1.31 schwarze 205: font = HTMLFONT_BI;
206: break;
1.35 schwarze 207: case ESCAPE_FONT:
208: case ESCAPE_FONTROMAN:
1.5 schwarze 209: font = HTMLFONT_NONE;
210: break;
211: default:
212: abort();
1.1 schwarze 213: }
214:
1.20 schwarze 215: if (h->metaf) {
216: print_tagq(h, h->metaf);
217: h->metaf = NULL;
218: }
219:
220: h->metal = h->metac;
221: h->metac = font;
222:
1.31 schwarze 223: switch (font) {
1.35 schwarze 224: case HTMLFONT_ITALIC:
1.64 schwarze 225: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 226: break;
1.35 schwarze 227: case HTMLFONT_BOLD:
1.64 schwarze 228: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 229: break;
1.35 schwarze 230: case HTMLFONT_BI:
1.64 schwarze 231: h->metaf = print_otag(h, TAG_B, "");
232: print_otag(h, TAG_I, "");
1.31 schwarze 233: break;
234: default:
235: break;
236: }
1.80 schwarze 237: }
238:
239: char *
240: html_make_id(const struct roff_node *n)
241: {
242: const struct roff_node *nch;
243: char *buf, *cp;
244:
245: for (nch = n->child; nch != NULL; nch = nch->next)
246: if (nch->type != ROFFT_TEXT)
247: return NULL;
248:
249: buf = NULL;
250: deroff(&buf, n);
251:
252: /* http://www.w3.org/TR/html5/dom.html#the-id-attribute */
253:
254: for (cp = buf; *cp != '\0'; cp++)
255: if (*cp == ' ')
256: *cp = '_';
257:
258: return buf;
1.1 schwarze 259: }
260:
1.26 schwarze 261: int
262: html_strlen(const char *cp)
263: {
1.30 schwarze 264: size_t rsz;
265: int skip, sz;
1.26 schwarze 266:
267: /*
268: * Account for escaped sequences within string length
269: * calculations. This follows the logic in term_strlen() as we
270: * must calculate the width of produced strings.
271: * Assume that characters are always width of "1". This is
272: * hacky, but it gets the job done for approximation of widths.
273: */
274:
275: sz = 0;
1.30 schwarze 276: skip = 0;
277: while (1) {
278: rsz = strcspn(cp, "\\");
279: if (rsz) {
280: cp += rsz;
281: if (skip) {
282: skip = 0;
283: rsz--;
284: }
285: sz += rsz;
286: }
287: if ('\0' == *cp)
288: break;
289: cp++;
290: switch (mandoc_escape(&cp, NULL, NULL)) {
1.35 schwarze 291: case ESCAPE_ERROR:
1.58 schwarze 292: return sz;
1.35 schwarze 293: case ESCAPE_UNICODE:
294: case ESCAPE_NUMBERED:
295: case ESCAPE_SPECIAL:
1.55 schwarze 296: case ESCAPE_OVERSTRIKE:
1.30 schwarze 297: if (skip)
298: skip = 0;
299: else
300: sz++;
301: break;
1.35 schwarze 302: case ESCAPE_SKIPCHAR:
1.30 schwarze 303: skip = 1;
1.26 schwarze 304: break;
305: default:
306: break;
307: }
308: }
1.58 schwarze 309: return sz;
1.26 schwarze 310: }
1.1 schwarze 311:
1.5 schwarze 312: static int
1.67 schwarze 313: print_escape(struct html *h, char c)
1.38 schwarze 314: {
315:
316: switch (c) {
317: case '<':
1.67 schwarze 318: print_word(h, "<");
1.38 schwarze 319: break;
320: case '>':
1.67 schwarze 321: print_word(h, ">");
1.38 schwarze 322: break;
323: case '&':
1.67 schwarze 324: print_word(h, "&");
1.38 schwarze 325: break;
326: case '"':
1.67 schwarze 327: print_word(h, """);
1.38 schwarze 328: break;
329: case ASCII_NBRSP:
1.67 schwarze 330: print_word(h, " ");
1.38 schwarze 331: break;
332: case ASCII_HYPH:
1.67 schwarze 333: print_byte(h, '-');
1.59 schwarze 334: break;
1.38 schwarze 335: case ASCII_BREAK:
336: break;
337: default:
1.58 schwarze 338: return 0;
1.38 schwarze 339: }
1.58 schwarze 340: return 1;
1.38 schwarze 341: }
342:
343: static int
1.65 schwarze 344: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 345: {
1.67 schwarze 346: char numbuf[16];
1.84 schwarze 347: struct tag *t;
348: const char *seq;
1.4 schwarze 349: size_t sz;
1.84 schwarze 350: int c, len, breakline, nospace;
1.26 schwarze 351: enum mandoc_esc esc;
1.84 schwarze 352: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 353: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 354:
1.65 schwarze 355: if (pend == NULL)
356: pend = strchr(p, '\0');
357:
1.84 schwarze 358: breakline = 0;
1.5 schwarze 359: nospace = 0;
1.1 schwarze 360:
1.65 schwarze 361: while (p < pend) {
1.30 schwarze 362: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
363: h->flags &= ~HTML_SKIPCHAR;
364: p++;
365: continue;
366: }
367:
1.67 schwarze 368: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 369: print_byte(h, *p);
370:
371: if (breakline &&
372: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
373: t = print_otag(h, TAG_DIV, "");
374: print_text(h, "\\~");
375: print_tagq(h, t);
376: breakline = 0;
377: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
378: p++;
379: continue;
380: }
1.4 schwarze 381:
1.65 schwarze 382: if (p >= pend)
1.26 schwarze 383: break;
384:
1.84 schwarze 385: if (*p == ' ') {
386: print_endword(h);
387: p++;
388: continue;
389: }
390:
1.67 schwarze 391: if (print_escape(h, *p++))
1.33 schwarze 392: continue;
1.4 schwarze 393:
1.26 schwarze 394: esc = mandoc_escape(&p, &seq, &len);
395: if (ESCAPE_ERROR == esc)
396: break;
1.5 schwarze 397:
1.26 schwarze 398: switch (esc) {
1.35 schwarze 399: case ESCAPE_FONT:
400: case ESCAPE_FONTPREV:
401: case ESCAPE_FONTBOLD:
402: case ESCAPE_FONTITALIC:
403: case ESCAPE_FONTBI:
404: case ESCAPE_FONTROMAN:
1.30 schwarze 405: if (0 == norecurse)
406: print_metaf(h, esc);
407: continue;
1.35 schwarze 408: case ESCAPE_SKIPCHAR:
1.30 schwarze 409: h->flags |= HTML_SKIPCHAR;
410: continue;
411: default:
412: break;
413: }
414:
415: if (h->flags & HTML_SKIPCHAR) {
416: h->flags &= ~HTML_SKIPCHAR;
417: continue;
418: }
419:
420: switch (esc) {
1.35 schwarze 421: case ESCAPE_UNICODE:
1.38 schwarze 422: /* Skip past "u" header. */
1.26 schwarze 423: c = mchars_num2uc(seq + 1, len - 1);
424: break;
1.35 schwarze 425: case ESCAPE_NUMBERED:
1.26 schwarze 426: c = mchars_num2char(seq, len);
1.51 schwarze 427: if (c < 0)
428: continue;
1.26 schwarze 429: break;
1.35 schwarze 430: case ESCAPE_SPECIAL:
1.61 schwarze 431: c = mchars_spec2cp(seq, len);
1.51 schwarze 432: if (c <= 0)
433: continue;
1.26 schwarze 434: break;
1.84 schwarze 435: case ESCAPE_BREAK:
436: breakline = 1;
437: continue;
1.35 schwarze 438: case ESCAPE_NOSPACE:
1.26 schwarze 439: if ('\0' == *p)
440: nospace = 1;
1.49 schwarze 441: continue;
1.55 schwarze 442: case ESCAPE_OVERSTRIKE:
443: if (len == 0)
444: continue;
445: c = seq[len - 1];
446: break;
1.5 schwarze 447: default:
1.49 schwarze 448: continue;
1.5 schwarze 449: }
1.51 schwarze 450: if ((c < 0x20 && c != 0x09) ||
451: (c > 0x7E && c < 0xA0))
1.49 schwarze 452: c = 0xFFFD;
1.67 schwarze 453: if (c > 0x7E) {
1.86 bentley 454: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 455: print_word(h, numbuf);
456: } else if (print_escape(h, c) == 0)
457: print_byte(h, c);
1.1 schwarze 458: }
1.5 schwarze 459:
1.58 schwarze 460: return nospace;
1.1 schwarze 461: }
462:
1.6 schwarze 463: static void
1.65 schwarze 464: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 465: {
1.65 schwarze 466: const char *p, *pp;
467:
468: pp = man ? h->base_man : h->base_includes;
469: while ((p = strchr(pp, '%')) != NULL) {
470: print_encode(h, pp, p, 1);
471: if (man && p[1] == 'S') {
472: if (sec == NULL)
1.67 schwarze 473: print_byte(h, '1');
1.65 schwarze 474: else
475: print_encode(h, sec, NULL, 1);
476: } else if ((man && p[1] == 'N') ||
477: (man == 0 && p[1] == 'I'))
478: print_encode(h, name, NULL, 1);
479: else
480: print_encode(h, p, p + 2, 1);
481: pp = p + 2;
482: }
483: if (*pp != '\0')
484: print_encode(h, pp, NULL, 1);
1.6 schwarze 485: }
486:
1.1 schwarze 487: struct tag *
1.64 schwarze 488: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 489: {
1.64 schwarze 490: va_list ap;
491: struct roffsu mysu, *su;
1.67 schwarze 492: char numbuf[16];
1.1 schwarze 493: struct tag *t;
1.65 schwarze 494: const char *attr;
1.73 schwarze 495: char *arg1, *arg2;
1.65 schwarze 496: double v;
1.66 schwarze 497: int i, have_style, tflags;
498:
499: tflags = htmltags[tag].flags;
1.1 schwarze 500:
1.74 schwarze 501: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 502:
1.66 schwarze 503: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 504: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 505: t->tag = tag;
1.74 schwarze 506: t->next = h->tag;
507: h->tag = t;
1.1 schwarze 508: } else
509: t = NULL;
510:
1.66 schwarze 511: if (tflags & HTML_NLBEFORE)
1.67 schwarze 512: print_endline(h);
513: if (h->col == 0)
514: print_indent(h);
1.66 schwarze 515: else if ((h->flags & HTML_NOSPACE) == 0) {
516: if (h->flags & HTML_KEEP)
1.86 bentley 517: print_word(h, " ");
1.66 schwarze 518: else {
519: if (h->flags & HTML_PREKEEP)
520: h->flags |= HTML_KEEP;
1.67 schwarze 521: print_endword(h);
1.12 schwarze 522: }
1.66 schwarze 523: }
1.1 schwarze 524:
1.13 schwarze 525: if ( ! (h->flags & HTML_NONOSPACE))
526: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 527: else
528: h->flags |= HTML_NOSPACE;
1.13 schwarze 529:
1.6 schwarze 530: /* Print out the tag name and attributes. */
531:
1.67 schwarze 532: print_byte(h, '<');
533: print_word(h, htmltags[tag].name);
1.64 schwarze 534:
535: va_start(ap, fmt);
536:
537: have_style = 0;
538: while (*fmt != '\0') {
539: if (*fmt == 's') {
540: have_style = 1;
541: fmt++;
542: break;
543: }
1.73 schwarze 544:
545: /* Parse a non-style attribute and its arguments. */
546:
547: arg1 = va_arg(ap, char *);
1.64 schwarze 548: switch (*fmt++) {
549: case 'c':
1.65 schwarze 550: attr = "class";
1.64 schwarze 551: break;
552: case 'h':
1.65 schwarze 553: attr = "href";
1.64 schwarze 554: break;
555: case 'i':
1.65 schwarze 556: attr = "id";
1.64 schwarze 557: break;
558: case '?':
1.73 schwarze 559: attr = arg1;
560: arg1 = va_arg(ap, char *);
1.64 schwarze 561: break;
562: default:
563: abort();
564: }
1.73 schwarze 565: arg2 = NULL;
566: if (*fmt == 'M')
567: arg2 = va_arg(ap, char *);
568: if (arg1 == NULL)
569: continue;
570:
571: /* Print the non-style attributes. */
572:
1.67 schwarze 573: print_byte(h, ' ');
574: print_word(h, attr);
575: print_byte(h, '=');
576: print_byte(h, '"');
1.65 schwarze 577: switch (*fmt) {
1.78 schwarze 578: case 'I':
579: print_href(h, arg1, NULL, 0);
580: fmt++;
581: break;
1.65 schwarze 582: case 'M':
1.73 schwarze 583: print_href(h, arg1, arg2, 1);
1.65 schwarze 584: fmt++;
585: break;
1.78 schwarze 586: case 'R':
587: print_byte(h, '#');
588: print_encode(h, arg1, NULL, 1);
1.65 schwarze 589: fmt++;
590: break;
1.78 schwarze 591: case 'T':
592: print_encode(h, arg1, NULL, 1);
593: print_word(h, "\" title=\"");
594: print_encode(h, arg1, NULL, 1);
1.65 schwarze 595: fmt++;
1.78 schwarze 596: break;
1.65 schwarze 597: default:
1.73 schwarze 598: print_encode(h, arg1, NULL, 1);
1.65 schwarze 599: break;
600: }
1.67 schwarze 601: print_byte(h, '"');
1.64 schwarze 602: }
603:
604: /* Print out styles. */
605:
606: while (*fmt != '\0') {
1.73 schwarze 607: arg1 = NULL;
608: su = NULL;
1.64 schwarze 609:
610: /* First letter: input argument type. */
611:
612: switch (*fmt++) {
613: case 'h':
614: i = va_arg(ap, int);
1.73 schwarze 615: su = &mysu;
1.64 schwarze 616: SCALE_HS_INIT(su, i);
617: break;
618: case 's':
1.73 schwarze 619: arg1 = va_arg(ap, char *);
1.64 schwarze 620: break;
621: case 'u':
622: su = va_arg(ap, struct roffsu *);
623: break;
624: case 'w':
1.88 ! schwarze 625: if ((arg2 = va_arg(ap, char *)) == NULL) {
! 626: if (*fmt == '+')
! 627: fmt++;
! 628: if (*fmt == '-')
! 629: fmt++;
1.73 schwarze 630: break;
1.88 ! schwarze 631: }
1.73 schwarze 632: su = &mysu;
633: a2width(arg2, su);
1.81 schwarze 634: if (*fmt == '+') {
635: /* Increase to make even bold text fit. */
1.82 schwarze 636: su->scale *= 1.2;
1.81 schwarze 637: /* Add padding. */
638: su->scale += 3.0;
639: fmt++;
640: }
641: if (*fmt == '-') {
1.71 schwarze 642: su->scale *= -1.0;
1.81 schwarze 643: fmt++;
644: }
1.64 schwarze 645: break;
646: default:
647: abort();
648: }
649:
650: /* Second letter: style name. */
651:
652: switch (*fmt++) {
653: case 'h':
1.65 schwarze 654: attr = "height";
1.64 schwarze 655: break;
656: case 'i':
1.65 schwarze 657: attr = "text-indent";
1.64 schwarze 658: break;
659: case 'l':
1.65 schwarze 660: attr = "margin-left";
1.64 schwarze 661: break;
662: case 'w':
1.65 schwarze 663: attr = "width";
1.64 schwarze 664: break;
665: case 'W':
1.65 schwarze 666: attr = "min-width";
1.64 schwarze 667: break;
668: case '?':
1.73 schwarze 669: attr = arg1;
670: arg1 = va_arg(ap, char *);
671: break;
1.64 schwarze 672: default:
673: abort();
674: }
1.73 schwarze 675: if (su == NULL && arg1 == NULL)
676: continue;
677:
678: if (have_style == 1)
679: print_word(h, " style=\"");
680: else
681: print_byte(h, ' ');
1.67 schwarze 682: print_word(h, attr);
683: print_byte(h, ':');
684: print_byte(h, ' ');
1.73 schwarze 685: if (su != NULL) {
686: v = su->scale;
687: if (su->unit == SCALE_MM && (v /= 100.0) == 0.0)
688: v = 1.0;
689: else if (su->unit == SCALE_BU)
690: v /= 24.0;
691: (void)snprintf(numbuf, sizeof(numbuf), "%.2f", v);
692: print_word(h, numbuf);
693: print_word(h, roffscales[su->unit]);
694: } else
695: print_word(h, arg1);
1.67 schwarze 696: print_byte(h, ';');
1.73 schwarze 697: have_style = 2;
1.64 schwarze 698: }
1.73 schwarze 699: if (have_style == 2)
1.67 schwarze 700: print_byte(h, '"');
1.64 schwarze 701:
702: va_end(ap);
1.6 schwarze 703:
1.42 schwarze 704: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 705:
706: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 707: print_byte(h, '/');
1.6 schwarze 708:
1.67 schwarze 709: print_byte(h, '>');
1.1 schwarze 710:
1.66 schwarze 711: if (tflags & HTML_NLBEGIN)
1.67 schwarze 712: print_endline(h);
1.66 schwarze 713: else
714: h->flags |= HTML_NOSPACE;
1.18 schwarze 715:
1.66 schwarze 716: if (tflags & HTML_INDENT)
717: h->indent++;
718: if (tflags & HTML_NOINDENT)
719: h->noindent++;
1.18 schwarze 720:
1.58 schwarze 721: return t;
1.1 schwarze 722: }
723:
724: static void
1.54 schwarze 725: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 726: {
1.66 schwarze 727: int tflags;
1.35 schwarze 728:
1.54 schwarze 729: /*
730: * Remember to close out and nullify the current
731: * meta-font and table, if applicable.
732: */
733: if (tag == h->metaf)
734: h->metaf = NULL;
735: if (tag == h->tblt)
736: h->tblt = NULL;
737:
1.66 schwarze 738: tflags = htmltags[tag->tag].flags;
739:
740: if (tflags & HTML_INDENT)
741: h->indent--;
742: if (tflags & HTML_NOINDENT)
743: h->noindent--;
744: if (tflags & HTML_NLEND)
1.67 schwarze 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, '>');
1.66 schwarze 751: if (tflags & HTML_NLAFTER)
1.67 schwarze 752: print_endline(h);
1.54 schwarze 753:
1.74 schwarze 754: h->tag = tag->next;
1.54 schwarze 755: free(tag);
1.1 schwarze 756: }
757:
758: void
1.6 schwarze 759: print_gen_decls(struct html *h)
760: {
1.67 schwarze 761: print_word(h, "<!DOCTYPE html>");
762: print_endline(h);
1.1 schwarze 763: }
764:
765: void
1.12 schwarze 766: print_text(struct html *h, const char *word)
1.1 schwarze 767: {
1.67 schwarze 768: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 769: if ( ! (HTML_KEEP & h->flags)) {
770: if (HTML_PREKEEP & h->flags)
771: h->flags |= HTML_KEEP;
1.67 schwarze 772: print_endword(h);
1.12 schwarze 773: } else
1.86 bentley 774: print_word(h, " ");
1.12 schwarze 775: }
1.1 schwarze 776:
1.20 schwarze 777: assert(NULL == h->metaf);
1.31 schwarze 778: switch (h->metac) {
1.35 schwarze 779: case HTMLFONT_ITALIC:
1.64 schwarze 780: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 781: break;
1.35 schwarze 782: case HTMLFONT_BOLD:
1.64 schwarze 783: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 784: break;
1.35 schwarze 785: case HTMLFONT_BI:
1.64 schwarze 786: h->metaf = print_otag(h, TAG_B, "");
787: print_otag(h, TAG_I, "");
1.31 schwarze 788: break;
789: default:
1.67 schwarze 790: print_indent(h);
1.31 schwarze 791: break;
792: }
1.20 schwarze 793:
1.12 schwarze 794: assert(word);
1.65 schwarze 795: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 796: if ( ! (h->flags & HTML_NONOSPACE))
797: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 798: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 799: } else
1.53 schwarze 800: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 801:
802: if (h->metaf) {
803: print_tagq(h, h->metaf);
804: h->metaf = NULL;
805: }
1.17 schwarze 806:
807: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 808: }
809:
810: void
811: print_tagq(struct html *h, const struct tag *until)
812: {
813: struct tag *tag;
814:
1.74 schwarze 815: while ((tag = h->tag) != NULL) {
1.54 schwarze 816: print_ctag(h, tag);
1.1 schwarze 817: if (until && tag == until)
818: return;
819: }
820: }
821:
822: void
823: print_stagq(struct html *h, const struct tag *suntil)
824: {
825: struct tag *tag;
826:
1.74 schwarze 827: while ((tag = h->tag) != NULL) {
1.1 schwarze 828: if (suntil && tag == suntil)
829: return;
1.54 schwarze 830: print_ctag(h, tag);
1.1 schwarze 831: }
832: }
1.42 schwarze 833:
834: void
835: print_paragraph(struct html *h)
836: {
837: struct tag *t;
838:
1.68 schwarze 839: t = print_otag(h, TAG_DIV, "c", "Pp");
1.42 schwarze 840: print_tagq(h, t);
841: }
842:
1.67 schwarze 843:
844: /***********************************************************************
845: * Low level output functions.
846: * They implement line breaking using a short static buffer.
847: ***********************************************************************/
848:
849: /*
850: * Buffer one HTML output byte.
851: * If the buffer is full, flush and deactivate it and start a new line.
852: * If the buffer is inactive, print directly.
853: */
854: static void
855: print_byte(struct html *h, char c)
856: {
857: if ((h->flags & HTML_BUFFER) == 0) {
858: putchar(c);
859: h->col++;
860: return;
861: }
862:
863: if (h->col + h->bufcol < sizeof(h->buf)) {
864: h->buf[h->bufcol++] = c;
865: return;
866: }
867:
868: putchar('\n');
869: h->col = 0;
870: print_indent(h);
871: putchar(' ');
872: putchar(' ');
873: fwrite(h->buf, h->bufcol, 1, stdout);
874: putchar(c);
875: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
876: h->bufcol = 0;
877: h->flags &= ~HTML_BUFFER;
878: }
879:
1.66 schwarze 880: /*
881: * If something was printed on the current output line, end it.
1.67 schwarze 882: * Not to be called right after print_indent().
1.66 schwarze 883: */
1.72 schwarze 884: void
1.67 schwarze 885: print_endline(struct html *h)
1.66 schwarze 886: {
1.67 schwarze 887: if (h->col == 0)
1.66 schwarze 888: return;
889:
1.67 schwarze 890: if (h->bufcol) {
891: putchar(' ');
892: fwrite(h->buf, h->bufcol, 1, stdout);
893: h->bufcol = 0;
894: }
1.66 schwarze 895: putchar('\n');
1.67 schwarze 896: h->col = 0;
897: h->flags |= HTML_NOSPACE;
898: h->flags &= ~HTML_BUFFER;
899: }
900:
901: /*
902: * Flush the HTML output buffer.
903: * If it is inactive, activate it.
904: */
905: static void
906: print_endword(struct html *h)
907: {
908: if (h->noindent) {
909: print_byte(h, ' ');
910: return;
911: }
912:
913: if ((h->flags & HTML_BUFFER) == 0) {
914: h->col++;
915: h->flags |= HTML_BUFFER;
916: } else if (h->bufcol) {
917: putchar(' ');
918: fwrite(h->buf, h->bufcol, 1, stdout);
919: h->col += h->bufcol + 1;
920: }
921: h->bufcol = 0;
1.66 schwarze 922: }
923:
924: /*
925: * If at the beginning of a new output line,
926: * perform indentation and mark the line as containing output.
927: * Make sure to really produce some output right afterwards,
928: * but do not use print_otag() for producing it.
929: */
930: static void
1.67 schwarze 931: print_indent(struct html *h)
1.66 schwarze 932: {
1.67 schwarze 933: size_t i;
1.66 schwarze 934:
1.67 schwarze 935: if (h->col)
1.66 schwarze 936: return;
937:
1.67 schwarze 938: if (h->noindent == 0) {
939: h->col = h->indent * 2;
940: for (i = 0; i < h->col; i++)
1.66 schwarze 941: putchar(' ');
1.67 schwarze 942: }
943: h->flags &= ~HTML_NOSPACE;
944: }
945:
946: /*
947: * Print or buffer some characters
948: * depending on the current HTML output buffer state.
949: */
950: static void
951: print_word(struct html *h, const char *cp)
952: {
953: while (*cp != '\0')
954: print_byte(h, *cp++);
1.66 schwarze 955: }
1.64 schwarze 956:
957: /*
958: * Calculate the scaling unit passed in a `-width' argument. This uses
959: * either a native scaling unit (e.g., 1i, 2m) or the string length of
960: * the value.
961: */
962: static void
963: a2width(const char *p, struct roffsu *su)
964: {
1.83 schwarze 965: const char *end;
966:
967: end = a2roffsu(p, su, SCALE_MAX);
968: if (end == NULL || *end != '\0') {
1.64 schwarze 969: su->unit = SCALE_EN;
970: su->scale = html_strlen(p);
971: } else if (su->scale < 0.0)
972: su->scale = 0.0;
1.3 schwarze 973: }