Annotation of src/usr.bin/mandoc/html.c, Revision 1.122
1.122 ! schwarze 1: /* $OpenBSD: html.c,v 1.121 2019/01/11 12:44:10 schwarze Exp $ */
1.1 schwarze 2: /*
1.42 schwarze 3: * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.119 schwarze 4: * Copyright (c) 2011-2015, 2017-2019 Ingo Schwarze <schwarze@openbsd.org>
1.1 schwarze 5: *
6: * Permission to use, copy, modify, and distribute this software for any
7: * purpose with or without fee is hereby granted, provided that the above
8: * copyright notice and this permission notice appear in all copies.
9: *
1.56 schwarze 10: * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.1 schwarze 11: * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.56 schwarze 12: * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.1 schwarze 13: * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
14: * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
15: * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
16: * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
17: */
18: #include <sys/types.h>
1.110 schwarze 19: #include <sys/stat.h>
1.1 schwarze 20:
21: #include <assert.h>
1.3 schwarze 22: #include <ctype.h>
1.4 schwarze 23: #include <stdarg.h>
1.99 schwarze 24: #include <stddef.h>
1.1 schwarze 25: #include <stdio.h>
26: #include <stdint.h>
27: #include <stdlib.h>
28: #include <string.h>
29: #include <unistd.h>
30:
1.80 schwarze 31: #include "mandoc_aux.h"
1.99 schwarze 32: #include "mandoc_ohash.h"
1.9 schwarze 33: #include "mandoc.h"
1.80 schwarze 34: #include "roff.h"
1.1 schwarze 35: #include "out.h"
36: #include "html.h"
1.56 schwarze 37: #include "manconf.h"
1.1 schwarze 38: #include "main.h"
39:
40: struct htmldata {
41: const char *name;
42: int flags;
1.66 schwarze 43: #define HTML_NOSTACK (1 << 0)
44: #define HTML_AUTOCLOSE (1 << 1)
45: #define HTML_NLBEFORE (1 << 2)
46: #define HTML_NLBEGIN (1 << 3)
47: #define HTML_NLEND (1 << 4)
48: #define HTML_NLAFTER (1 << 5)
49: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
50: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
51: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
52: #define HTML_INDENT (1 << 6)
53: #define HTML_NOINDENT (1 << 7)
1.1 schwarze 54: };
55:
56: static const struct htmldata htmltags[TAG_MAX] = {
1.66 schwarze 57: {"html", HTML_NLALL},
58: {"head", HTML_NLALL | HTML_INDENT},
59: {"body", HTML_NLALL},
60: {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
61: {"title", HTML_NLAROUND},
62: {"div", HTML_NLAROUND},
1.95 schwarze 63: {"div", 0},
1.66 schwarze 64: {"h1", HTML_NLAROUND},
65: {"h2", HTML_NLAROUND},
66: {"span", 0},
67: {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
68: {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL},
69: {"a", 0},
70: {"table", HTML_NLALL | HTML_INDENT},
71: {"tr", HTML_NLALL | HTML_INDENT},
72: {"td", HTML_NLAROUND},
73: {"li", HTML_NLAROUND | HTML_INDENT},
74: {"ul", HTML_NLALL | HTML_INDENT},
75: {"ol", HTML_NLALL | HTML_INDENT},
76: {"dl", HTML_NLALL | HTML_INDENT},
77: {"dt", HTML_NLAROUND},
78: {"dd", HTML_NLAROUND | HTML_INDENT},
1.119 schwarze 79: {"p", HTML_NLAROUND | HTML_INDENT},
1.66 schwarze 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:
1.99 schwarze 107: /* Avoid duplicate HTML id= attributes. */
108: static struct ohash id_unique;
109:
1.67 schwarze 110: static void print_byte(struct html *, char);
111: static void print_endword(struct html *);
112: static void print_indent(struct html *);
113: static void print_word(struct html *, const char *);
114:
1.54 schwarze 115: static void print_ctag(struct html *, struct tag *);
1.67 schwarze 116: static int print_escape(struct html *, char);
1.65 schwarze 117: static int print_encode(struct html *, const char *, const char *, int);
118: static void print_href(struct html *, const char *, const char *, int);
1.5 schwarze 119:
1.35 schwarze 120:
1.50 schwarze 121: void *
1.61 schwarze 122: html_alloc(const struct manoutput *outopts)
1.1 schwarze 123: {
124: struct html *h;
125:
1.24 schwarze 126: h = mandoc_calloc(1, sizeof(struct html));
1.1 schwarze 127:
1.74 schwarze 128: h->tag = NULL;
1.56 schwarze 129: h->style = outopts->style;
1.110 schwarze 130: if ((h->base_man1 = outopts->man) == NULL)
131: h->base_man2 = NULL;
132: else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL)
133: *h->base_man2++ = '\0';
1.56 schwarze 134: h->base_includes = outopts->includes;
135: if (outopts->fragment)
136: h->oflags |= HTML_FRAGMENT;
1.111 schwarze 137: if (outopts->toc)
138: h->oflags |= HTML_TOC;
1.1 schwarze 139:
1.99 schwarze 140: mandoc_ohash_init(&id_unique, 4, 0);
141:
1.58 schwarze 142: return h;
1.1 schwarze 143: }
144:
145: void
146: html_free(void *p)
147: {
148: struct tag *tag;
149: struct html *h;
1.99 schwarze 150: char *cp;
151: unsigned int slot;
1.1 schwarze 152:
153: h = (struct html *)p;
1.74 schwarze 154: while ((tag = h->tag) != NULL) {
155: h->tag = tag->next;
1.1 schwarze 156: free(tag);
157: }
1.99 schwarze 158: free(h);
1.1 schwarze 159:
1.99 schwarze 160: cp = ohash_first(&id_unique, &slot);
161: while (cp != NULL) {
162: free(cp);
163: cp = ohash_next(&id_unique, &slot);
164: }
165: ohash_delete(&id_unique);
1.1 schwarze 166: }
167:
168: void
169: print_gen_head(struct html *h)
170: {
1.42 schwarze 171: struct tag *t;
172:
1.64 schwarze 173: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.92 schwarze 174: if (h->style != NULL) {
175: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
176: h->style, "type", "text/css", "media", "all");
177: return;
178: }
1.1 schwarze 179:
1.42 schwarze 180: /*
1.92 schwarze 181: * Print a minimal embedded style sheet.
1.42 schwarze 182: */
1.66 schwarze 183:
1.64 schwarze 184: t = print_otag(h, TAG_STYLE, "");
1.66 schwarze 185: print_text(h, "table.head, table.foot { width: 100%; }");
1.67 schwarze 186: print_endline(h);
1.66 schwarze 187: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.67 schwarze 188: print_endline(h);
1.66 schwarze 189: print_text(h, "td.head-vol { text-align: center; }");
1.67 schwarze 190: print_endline(h);
1.68 schwarze 191: print_text(h, "div.Pp { margin: 1ex 0ex; }");
1.95 schwarze 192: print_endline(h);
193: print_text(h, "div.Nd, div.Bf, div.Op { display: inline; }");
1.96 schwarze 194: print_endline(h);
1.97 schwarze 195: print_text(h, "span.Pa, span.Ad { font-style: italic; }");
1.98 schwarze 196: print_endline(h);
197: print_text(h, "span.Ms { font-weight: bold; }");
1.94 schwarze 198: print_endline(h);
199: print_text(h, "dl.Bl-diag ");
200: print_byte(h, '>');
201: print_text(h, " dt { font-weight: bold; }");
1.93 schwarze 202: print_endline(h);
203: print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, "
204: "code.In, code.Fd, code.Fn,");
205: print_endline(h);
206: print_text(h, "code.Cd { font-weight: bold; "
207: "font-family: inherit; }");
1.42 schwarze 208: print_tagq(h, t);
1.1 schwarze 209: }
210:
1.117 schwarze 211: void
1.26 schwarze 212: print_metaf(struct html *h, enum mandoc_esc deco)
1.5 schwarze 213: {
214: enum htmlfont font;
1.1 schwarze 215:
1.5 schwarze 216: switch (deco) {
1.35 schwarze 217: case ESCAPE_FONTPREV:
1.5 schwarze 218: font = h->metal;
219: break;
1.35 schwarze 220: case ESCAPE_FONTITALIC:
1.5 schwarze 221: font = HTMLFONT_ITALIC;
222: break;
1.35 schwarze 223: case ESCAPE_FONTBOLD:
1.5 schwarze 224: font = HTMLFONT_BOLD;
225: break;
1.35 schwarze 226: case ESCAPE_FONTBI:
1.31 schwarze 227: font = HTMLFONT_BI;
228: break;
1.112 schwarze 229: case ESCAPE_FONTCW:
230: font = HTMLFONT_CW;
231: break;
1.35 schwarze 232: case ESCAPE_FONT:
233: case ESCAPE_FONTROMAN:
1.5 schwarze 234: font = HTMLFONT_NONE;
235: break;
236: default:
1.117 schwarze 237: return;
1.1 schwarze 238: }
239:
1.20 schwarze 240: if (h->metaf) {
241: print_tagq(h, h->metaf);
242: h->metaf = NULL;
243: }
244:
245: h->metal = h->metac;
246: h->metac = font;
247:
1.31 schwarze 248: switch (font) {
1.35 schwarze 249: case HTMLFONT_ITALIC:
1.64 schwarze 250: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 251: break;
1.35 schwarze 252: case HTMLFONT_BOLD:
1.64 schwarze 253: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 254: break;
1.35 schwarze 255: case HTMLFONT_BI:
1.64 schwarze 256: h->metaf = print_otag(h, TAG_B, "");
257: print_otag(h, TAG_I, "");
1.31 schwarze 258: break;
1.112 schwarze 259: case HTMLFONT_CW:
260: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
261: break;
1.31 schwarze 262: default:
263: break;
264: }
1.118 schwarze 265: }
266:
1.119 schwarze 267: void
268: html_close_paragraph(struct html *h)
269: {
270: struct tag *t;
271:
1.122 ! schwarze 272: for (t = h->tag; t != NULL && t->closed == 0; t = t->next) {
! 273: switch(t->tag) {
! 274: case TAG_P:
! 275: case TAG_PRE:
1.119 schwarze 276: print_tagq(h, t);
277: break;
1.122 ! schwarze 278: case TAG_A:
! 279: print_tagq(h, t);
! 280: continue;
! 281: default:
! 282: continue;
1.119 schwarze 283: }
1.122 ! schwarze 284: break;
1.119 schwarze 285: }
286: }
287:
1.118 schwarze 288: /*
289: * ROFF_nf switches to no-fill mode, ROFF_fi to fill mode.
290: * TOKEN_NONE does not switch. The old mode is returned.
291: */
292: enum roff_tok
293: html_fillmode(struct html *h, enum roff_tok want)
294: {
295: struct tag *t;
296: enum roff_tok had;
297:
298: for (t = h->tag; t != NULL; t = t->next)
299: if (t->tag == TAG_PRE)
300: break;
301:
302: had = t == NULL ? ROFF_fi : ROFF_nf;
303:
304: if (want != had) {
305: switch (want) {
306: case ROFF_fi:
307: print_tagq(h, t);
308: break;
309: case ROFF_nf:
1.119 schwarze 310: html_close_paragraph(h);
1.118 schwarze 311: print_otag(h, TAG_PRE, "");
312: break;
313: case TOKEN_NONE:
314: break;
315: default:
316: abort();
317: }
318: }
319: return had;
1.80 schwarze 320: }
321:
322: char *
1.99 schwarze 323: html_make_id(const struct roff_node *n, int unique)
1.80 schwarze 324: {
325: const struct roff_node *nch;
1.99 schwarze 326: char *buf, *bufs, *cp;
327: unsigned int slot;
328: int suffix;
1.80 schwarze 329:
330: for (nch = n->child; nch != NULL; nch = nch->next)
331: if (nch->type != ROFFT_TEXT)
332: return NULL;
333:
334: buf = NULL;
335: deroff(&buf, n);
1.90 schwarze 336: if (buf == NULL)
337: return NULL;
1.80 schwarze 338:
1.100 schwarze 339: /*
340: * In ID attributes, only use ASCII characters that are
341: * permitted in URL-fragment strings according to the
342: * explicit list at:
343: * https://url.spec.whatwg.org/#url-fragment-string
344: */
1.80 schwarze 345:
346: for (cp = buf; *cp != '\0'; cp++)
1.100 schwarze 347: if (isalnum((unsigned char)*cp) == 0 &&
348: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.80 schwarze 349: *cp = '_';
350:
1.99 schwarze 351: if (unique == 0)
352: return buf;
353:
354: /* Avoid duplicate HTML id= attributes. */
355:
356: bufs = NULL;
357: suffix = 1;
358: slot = ohash_qlookup(&id_unique, buf);
359: cp = ohash_find(&id_unique, slot);
360: if (cp != NULL) {
361: while (cp != NULL) {
362: free(bufs);
363: if (++suffix > 127) {
364: free(buf);
365: return NULL;
366: }
367: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
368: slot = ohash_qlookup(&id_unique, bufs);
369: cp = ohash_find(&id_unique, slot);
370: }
371: free(buf);
372: buf = bufs;
373: }
374: ohash_insert(&id_unique, slot, buf);
1.80 schwarze 375: return buf;
1.1 schwarze 376: }
377:
1.5 schwarze 378: static int
1.67 schwarze 379: print_escape(struct html *h, char c)
1.38 schwarze 380: {
381:
382: switch (c) {
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 '&':
1.67 schwarze 390: print_word(h, "&");
1.38 schwarze 391: break;
392: case '"':
1.67 schwarze 393: print_word(h, """);
1.38 schwarze 394: break;
395: case ASCII_NBRSP:
1.67 schwarze 396: print_word(h, " ");
1.38 schwarze 397: break;
398: case ASCII_HYPH:
1.67 schwarze 399: print_byte(h, '-');
1.59 schwarze 400: break;
1.38 schwarze 401: case ASCII_BREAK:
402: break;
403: default:
1.58 schwarze 404: return 0;
1.38 schwarze 405: }
1.58 schwarze 406: return 1;
1.38 schwarze 407: }
408:
409: static int
1.65 schwarze 410: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.1 schwarze 411: {
1.67 schwarze 412: char numbuf[16];
1.84 schwarze 413: const char *seq;
1.4 schwarze 414: size_t sz;
1.84 schwarze 415: int c, len, breakline, nospace;
1.26 schwarze 416: enum mandoc_esc esc;
1.84 schwarze 417: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.33 schwarze 418: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.5 schwarze 419:
1.65 schwarze 420: if (pend == NULL)
421: pend = strchr(p, '\0');
422:
1.84 schwarze 423: breakline = 0;
1.5 schwarze 424: nospace = 0;
1.1 schwarze 425:
1.65 schwarze 426: while (p < pend) {
1.30 schwarze 427: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
428: h->flags &= ~HTML_SKIPCHAR;
429: p++;
430: continue;
431: }
432:
1.67 schwarze 433: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.84 schwarze 434: print_byte(h, *p);
435:
436: if (breakline &&
437: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.115 schwarze 438: print_otag(h, TAG_BR, "");
1.84 schwarze 439: breakline = 0;
440: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
441: p++;
442: continue;
443: }
1.4 schwarze 444:
1.65 schwarze 445: if (p >= pend)
1.26 schwarze 446: break;
447:
1.84 schwarze 448: if (*p == ' ') {
449: print_endword(h);
450: p++;
451: continue;
452: }
453:
1.67 schwarze 454: if (print_escape(h, *p++))
1.33 schwarze 455: continue;
1.4 schwarze 456:
1.26 schwarze 457: esc = mandoc_escape(&p, &seq, &len);
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:
1.112 schwarze 464: case ESCAPE_FONTCW:
1.35 schwarze 465: case ESCAPE_FONTROMAN:
1.113 schwarze 466: if (0 == norecurse) {
467: h->flags |= HTML_NOSPACE;
1.30 schwarze 468: print_metaf(h, esc);
1.113 schwarze 469: h->flags &= ~HTML_NOSPACE;
470: }
1.30 schwarze 471: continue;
1.35 schwarze 472: case ESCAPE_SKIPCHAR:
1.30 schwarze 473: h->flags |= HTML_SKIPCHAR;
474: continue;
1.116 schwarze 475: case ESCAPE_ERROR:
476: continue;
1.30 schwarze 477: default:
478: break;
479: }
480:
481: if (h->flags & HTML_SKIPCHAR) {
482: h->flags &= ~HTML_SKIPCHAR;
483: continue;
484: }
485:
486: switch (esc) {
1.35 schwarze 487: case ESCAPE_UNICODE:
1.38 schwarze 488: /* Skip past "u" header. */
1.26 schwarze 489: c = mchars_num2uc(seq + 1, len - 1);
490: break;
1.35 schwarze 491: case ESCAPE_NUMBERED:
1.26 schwarze 492: c = mchars_num2char(seq, len);
1.51 schwarze 493: if (c < 0)
494: continue;
1.26 schwarze 495: break;
1.35 schwarze 496: case ESCAPE_SPECIAL:
1.61 schwarze 497: c = mchars_spec2cp(seq, len);
1.51 schwarze 498: if (c <= 0)
499: continue;
1.116 schwarze 500: break;
501: case ESCAPE_UNDEF:
502: c = *seq;
1.26 schwarze 503: break;
1.109 schwarze 504: case ESCAPE_DEVICE:
505: print_word(h, "html");
506: continue;
1.84 schwarze 507: case ESCAPE_BREAK:
508: breakline = 1;
509: continue;
1.35 schwarze 510: case ESCAPE_NOSPACE:
1.26 schwarze 511: if ('\0' == *p)
512: nospace = 1;
1.49 schwarze 513: continue;
1.55 schwarze 514: case ESCAPE_OVERSTRIKE:
515: if (len == 0)
516: continue;
517: c = seq[len - 1];
518: break;
1.5 schwarze 519: default:
1.49 schwarze 520: continue;
1.5 schwarze 521: }
1.51 schwarze 522: if ((c < 0x20 && c != 0x09) ||
523: (c > 0x7E && c < 0xA0))
1.49 schwarze 524: c = 0xFFFD;
1.67 schwarze 525: if (c > 0x7E) {
1.86 bentley 526: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.67 schwarze 527: print_word(h, numbuf);
528: } else if (print_escape(h, c) == 0)
529: print_byte(h, c);
1.1 schwarze 530: }
1.5 schwarze 531:
1.58 schwarze 532: return nospace;
1.1 schwarze 533: }
534:
1.6 schwarze 535: static void
1.65 schwarze 536: print_href(struct html *h, const char *name, const char *sec, int man)
1.6 schwarze 537: {
1.110 schwarze 538: struct stat sb;
1.65 schwarze 539: const char *p, *pp;
1.110 schwarze 540: char *filename;
541:
542: if (man) {
543: pp = h->base_man1;
544: if (h->base_man2 != NULL) {
545: mandoc_asprintf(&filename, "%s.%s", name, sec);
546: if (stat(filename, &sb) == -1)
547: pp = h->base_man2;
548: free(filename);
549: }
550: } else
551: pp = h->base_includes;
1.65 schwarze 552:
553: while ((p = strchr(pp, '%')) != NULL) {
554: print_encode(h, pp, p, 1);
555: if (man && p[1] == 'S') {
556: if (sec == NULL)
1.67 schwarze 557: print_byte(h, '1');
1.65 schwarze 558: else
559: print_encode(h, sec, NULL, 1);
560: } else if ((man && p[1] == 'N') ||
561: (man == 0 && p[1] == 'I'))
562: print_encode(h, name, NULL, 1);
563: else
564: print_encode(h, p, p + 2, 1);
565: pp = p + 2;
566: }
567: if (*pp != '\0')
568: print_encode(h, pp, NULL, 1);
1.6 schwarze 569: }
570:
1.1 schwarze 571: struct tag *
1.64 schwarze 572: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.1 schwarze 573: {
1.64 schwarze 574: va_list ap;
1.1 schwarze 575: struct tag *t;
1.65 schwarze 576: const char *attr;
1.73 schwarze 577: char *arg1, *arg2;
1.114 schwarze 578: int style_written, tflags;
1.66 schwarze 579:
580: tflags = htmltags[tag].flags;
1.1 schwarze 581:
1.74 schwarze 582: /* Push this tag onto the stack of open scopes. */
1.6 schwarze 583:
1.66 schwarze 584: if ((tflags & HTML_NOSTACK) == 0) {
1.24 schwarze 585: t = mandoc_malloc(sizeof(struct tag));
1.1 schwarze 586: t->tag = tag;
1.74 schwarze 587: t->next = h->tag;
1.122 ! schwarze 588: t->refcnt = 0;
! 589: t->closed = 0;
1.74 schwarze 590: h->tag = t;
1.1 schwarze 591: } else
592: t = NULL;
593:
1.66 schwarze 594: if (tflags & HTML_NLBEFORE)
1.67 schwarze 595: print_endline(h);
596: if (h->col == 0)
597: print_indent(h);
1.66 schwarze 598: else if ((h->flags & HTML_NOSPACE) == 0) {
599: if (h->flags & HTML_KEEP)
1.86 bentley 600: print_word(h, " ");
1.66 schwarze 601: else {
602: if (h->flags & HTML_PREKEEP)
603: h->flags |= HTML_KEEP;
1.67 schwarze 604: print_endword(h);
1.12 schwarze 605: }
1.66 schwarze 606: }
1.1 schwarze 607:
1.13 schwarze 608: if ( ! (h->flags & HTML_NONOSPACE))
609: h->flags &= ~HTML_NOSPACE;
1.14 schwarze 610: else
611: h->flags |= HTML_NOSPACE;
1.13 schwarze 612:
1.6 schwarze 613: /* Print out the tag name and attributes. */
614:
1.67 schwarze 615: print_byte(h, '<');
616: print_word(h, htmltags[tag].name);
1.64 schwarze 617:
618: va_start(ap, fmt);
619:
1.114 schwarze 620: while (*fmt != '\0' && *fmt != 's') {
1.73 schwarze 621:
1.108 schwarze 622: /* Parse attributes and arguments. */
1.73 schwarze 623:
624: arg1 = va_arg(ap, char *);
1.108 schwarze 625: arg2 = NULL;
1.64 schwarze 626: switch (*fmt++) {
627: case 'c':
1.65 schwarze 628: attr = "class";
1.64 schwarze 629: break;
630: case 'h':
1.65 schwarze 631: attr = "href";
1.64 schwarze 632: break;
633: case 'i':
1.65 schwarze 634: attr = "id";
1.64 schwarze 635: break;
636: case '?':
1.73 schwarze 637: attr = arg1;
638: arg1 = va_arg(ap, char *);
1.64 schwarze 639: break;
640: default:
641: abort();
642: }
1.73 schwarze 643: if (*fmt == 'M')
644: arg2 = va_arg(ap, char *);
645: if (arg1 == NULL)
646: continue;
647:
1.108 schwarze 648: /* Print the attributes. */
1.73 schwarze 649:
1.67 schwarze 650: print_byte(h, ' ');
651: print_word(h, attr);
652: print_byte(h, '=');
653: print_byte(h, '"');
1.65 schwarze 654: switch (*fmt) {
1.78 schwarze 655: case 'I':
656: print_href(h, arg1, NULL, 0);
657: fmt++;
658: break;
1.65 schwarze 659: case 'M':
1.73 schwarze 660: print_href(h, arg1, arg2, 1);
1.65 schwarze 661: fmt++;
662: break;
1.78 schwarze 663: case 'R':
664: print_byte(h, '#');
665: print_encode(h, arg1, NULL, 1);
1.65 schwarze 666: fmt++;
1.78 schwarze 667: break;
1.65 schwarze 668: default:
1.114 schwarze 669: print_encode(h, arg1, NULL, 1);
1.65 schwarze 670: break;
671: }
1.67 schwarze 672: print_byte(h, '"');
1.64 schwarze 673: }
1.114 schwarze 674:
675: style_written = 0;
676: while (*fmt++ == 's') {
677: arg1 = va_arg(ap, char *);
678: arg2 = va_arg(ap, char *);
679: if (arg2 == NULL)
680: continue;
681: print_byte(h, ' ');
682: if (style_written == 0) {
683: print_word(h, "style=\"");
684: style_written = 1;
685: }
686: print_word(h, arg1);
687: print_byte(h, ':');
688: print_byte(h, ' ');
689: print_word(h, arg2);
690: print_byte(h, ';');
691: }
692: if (style_written)
693: print_byte(h, '"');
694:
1.64 schwarze 695: va_end(ap);
1.6 schwarze 696:
1.42 schwarze 697: /* Accommodate for "well-formed" singleton escaping. */
1.6 schwarze 698:
699: if (HTML_AUTOCLOSE & htmltags[tag].flags)
1.67 schwarze 700: print_byte(h, '/');
1.6 schwarze 701:
1.67 schwarze 702: print_byte(h, '>');
1.1 schwarze 703:
1.66 schwarze 704: if (tflags & HTML_NLBEGIN)
1.67 schwarze 705: print_endline(h);
1.66 schwarze 706: else
707: h->flags |= HTML_NOSPACE;
1.18 schwarze 708:
1.66 schwarze 709: if (tflags & HTML_INDENT)
710: h->indent++;
711: if (tflags & HTML_NOINDENT)
712: h->noindent++;
1.18 schwarze 713:
1.58 schwarze 714: return t;
1.1 schwarze 715: }
716:
717: static void
1.54 schwarze 718: print_ctag(struct html *h, struct tag *tag)
1.1 schwarze 719: {
1.66 schwarze 720: int tflags;
1.35 schwarze 721:
1.122 ! schwarze 722: if (tag->closed == 0) {
! 723: tag->closed = 1;
! 724: if (tag == h->metaf)
! 725: h->metaf = NULL;
! 726: if (tag == h->tblt)
! 727: h->tblt = NULL;
! 728:
! 729: tflags = htmltags[tag->tag].flags;
! 730: if (tflags & HTML_INDENT)
! 731: h->indent--;
! 732: if (tflags & HTML_NOINDENT)
! 733: h->noindent--;
! 734: if (tflags & HTML_NLEND)
! 735: print_endline(h);
! 736: print_indent(h);
! 737: print_byte(h, '<');
! 738: print_byte(h, '/');
! 739: print_word(h, htmltags[tag->tag].name);
! 740: print_byte(h, '>');
! 741: if (tflags & HTML_NLAFTER)
! 742: print_endline(h);
! 743: }
! 744: if (tag->refcnt == 0) {
! 745: h->tag = tag->next;
! 746: free(tag);
! 747: }
1.1 schwarze 748: }
749:
750: void
1.6 schwarze 751: print_gen_decls(struct html *h)
752: {
1.67 schwarze 753: print_word(h, "<!DOCTYPE html>");
754: print_endline(h);
1.91 schwarze 755: }
756:
757: void
758: print_gen_comment(struct html *h, struct roff_node *n)
759: {
760: int wantblank;
761:
762: print_word(h, "<!-- This is an automatically generated file."
763: " Do not edit.");
764: h->indent = 1;
765: wantblank = 0;
766: while (n != NULL && n->type == ROFFT_COMMENT) {
767: if (strstr(n->string, "-->") == NULL &&
768: (wantblank || *n->string != '\0')) {
769: print_endline(h);
770: print_indent(h);
771: print_word(h, n->string);
772: wantblank = *n->string != '\0';
773: }
774: n = n->next;
775: }
776: if (wantblank)
777: print_endline(h);
778: print_word(h, " -->");
779: print_endline(h);
780: h->indent = 0;
1.1 schwarze 781: }
782:
783: void
1.12 schwarze 784: print_text(struct html *h, const char *word)
1.1 schwarze 785: {
1.67 schwarze 786: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.12 schwarze 787: if ( ! (HTML_KEEP & h->flags)) {
788: if (HTML_PREKEEP & h->flags)
789: h->flags |= HTML_KEEP;
1.67 schwarze 790: print_endword(h);
1.12 schwarze 791: } else
1.86 bentley 792: print_word(h, " ");
1.12 schwarze 793: }
1.1 schwarze 794:
1.20 schwarze 795: assert(NULL == h->metaf);
1.31 schwarze 796: switch (h->metac) {
1.35 schwarze 797: case HTMLFONT_ITALIC:
1.64 schwarze 798: h->metaf = print_otag(h, TAG_I, "");
1.31 schwarze 799: break;
1.35 schwarze 800: case HTMLFONT_BOLD:
1.64 schwarze 801: h->metaf = print_otag(h, TAG_B, "");
1.31 schwarze 802: break;
1.35 schwarze 803: case HTMLFONT_BI:
1.64 schwarze 804: h->metaf = print_otag(h, TAG_B, "");
805: print_otag(h, TAG_I, "");
1.112 schwarze 806: break;
807: case HTMLFONT_CW:
808: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
1.31 schwarze 809: break;
810: default:
1.67 schwarze 811: print_indent(h);
1.31 schwarze 812: break;
813: }
1.20 schwarze 814:
1.12 schwarze 815: assert(word);
1.65 schwarze 816: if ( ! print_encode(h, word, NULL, 0)) {
1.13 schwarze 817: if ( ! (h->flags & HTML_NONOSPACE))
818: h->flags &= ~HTML_NOSPACE;
1.53 schwarze 819: h->flags &= ~HTML_NONEWLINE;
1.28 schwarze 820: } else
1.53 schwarze 821: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.20 schwarze 822:
823: if (h->metaf) {
824: print_tagq(h, h->metaf);
825: h->metaf = NULL;
826: }
1.17 schwarze 827:
828: h->flags &= ~HTML_IGNDELIM;
1.1 schwarze 829: }
830:
831: void
832: print_tagq(struct html *h, const struct tag *until)
833: {
1.122 ! schwarze 834: struct tag *this, *next;
1.1 schwarze 835:
1.122 ! schwarze 836: for (this = h->tag; this != NULL; this = next) {
! 837: next = this == until ? NULL : this->next;
! 838: print_ctag(h, this);
1.1 schwarze 839: }
840: }
841:
1.120 schwarze 842: /*
843: * Close out all open elements up to but excluding suntil.
844: * Note that a paragraph just inside stays open together with it
845: * because paragraphs include subsequent phrasing content.
846: */
1.1 schwarze 847: void
848: print_stagq(struct html *h, const struct tag *suntil)
849: {
1.122 ! schwarze 850: struct tag *this, *next;
1.1 schwarze 851:
1.122 ! schwarze 852: for (this = h->tag; this != NULL; this = next) {
! 853: next = this->next;
! 854: if (this == suntil || (next == suntil &&
! 855: (this->tag == TAG_P || this->tag == TAG_PRE)))
! 856: break;
! 857: print_ctag(h, this);
1.1 schwarze 858: }
1.42 schwarze 859: }
860:
1.67 schwarze 861:
862: /***********************************************************************
863: * Low level output functions.
864: * They implement line breaking using a short static buffer.
865: ***********************************************************************/
866:
867: /*
868: * Buffer one HTML output byte.
869: * If the buffer is full, flush and deactivate it and start a new line.
870: * If the buffer is inactive, print directly.
871: */
872: static void
873: print_byte(struct html *h, char c)
874: {
875: if ((h->flags & HTML_BUFFER) == 0) {
876: putchar(c);
877: h->col++;
878: return;
879: }
880:
881: if (h->col + h->bufcol < sizeof(h->buf)) {
882: h->buf[h->bufcol++] = c;
883: return;
884: }
885:
886: putchar('\n');
887: h->col = 0;
888: print_indent(h);
889: putchar(' ');
890: putchar(' ');
891: fwrite(h->buf, h->bufcol, 1, stdout);
892: putchar(c);
893: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
894: h->bufcol = 0;
895: h->flags &= ~HTML_BUFFER;
896: }
897:
1.66 schwarze 898: /*
899: * If something was printed on the current output line, end it.
1.67 schwarze 900: * Not to be called right after print_indent().
1.66 schwarze 901: */
1.72 schwarze 902: void
1.67 schwarze 903: print_endline(struct html *h)
1.66 schwarze 904: {
1.67 schwarze 905: if (h->col == 0)
1.66 schwarze 906: return;
907:
1.67 schwarze 908: if (h->bufcol) {
909: putchar(' ');
910: fwrite(h->buf, h->bufcol, 1, stdout);
911: h->bufcol = 0;
912: }
1.66 schwarze 913: putchar('\n');
1.67 schwarze 914: h->col = 0;
915: h->flags |= HTML_NOSPACE;
916: h->flags &= ~HTML_BUFFER;
917: }
918:
919: /*
920: * Flush the HTML output buffer.
921: * If it is inactive, activate it.
922: */
923: static void
924: print_endword(struct html *h)
925: {
926: if (h->noindent) {
927: print_byte(h, ' ');
928: return;
929: }
930:
931: if ((h->flags & HTML_BUFFER) == 0) {
932: h->col++;
933: h->flags |= HTML_BUFFER;
934: } else if (h->bufcol) {
935: putchar(' ');
936: fwrite(h->buf, h->bufcol, 1, stdout);
937: h->col += h->bufcol + 1;
938: }
939: h->bufcol = 0;
1.66 schwarze 940: }
941:
942: /*
943: * If at the beginning of a new output line,
944: * perform indentation and mark the line as containing output.
945: * Make sure to really produce some output right afterwards,
946: * but do not use print_otag() for producing it.
947: */
948: static void
1.67 schwarze 949: print_indent(struct html *h)
1.66 schwarze 950: {
1.67 schwarze 951: size_t i;
1.66 schwarze 952:
1.67 schwarze 953: if (h->col)
1.66 schwarze 954: return;
955:
1.67 schwarze 956: if (h->noindent == 0) {
957: h->col = h->indent * 2;
958: for (i = 0; i < h->col; i++)
1.66 schwarze 959: putchar(' ');
1.67 schwarze 960: }
961: h->flags &= ~HTML_NOSPACE;
962: }
963:
964: /*
965: * Print or buffer some characters
966: * depending on the current HTML output buffer state.
967: */
968: static void
969: print_word(struct html *h, const char *cp)
970: {
971: while (*cp != '\0')
972: print_byte(h, *cp++);
1.3 schwarze 973: }