Annotation of mandoc/html.c, Revision 1.259
1.259 ! schwarze 1: /* $Id: html.c,v 1.258 2019/09/01 15:12:19 schwarze Exp $ */
1.1 kristaps 2: /*
1.176 schwarze 3: * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
1.249 schwarze 4: * Copyright (c) 2011-2015, 2017-2019 Ingo Schwarze <schwarze@openbsd.org>
1.1 kristaps 5: *
6: * Permission to use, copy, modify, and distribute this software for any
1.29 kristaps 7: * purpose with or without fee is hereby granted, provided that the above
8: * copyright notice and this permission notice appear in all copies.
1.1 kristaps 9: *
1.186 schwarze 10: * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
1.29 kristaps 11: * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
1.186 schwarze 12: * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
1.29 kristaps 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.
1.1 kristaps 17: */
1.92 kristaps 18: #include "config.h"
19:
1.41 kristaps 20: #include <sys/types.h>
1.240 schwarze 21: #include <sys/stat.h>
1.30 kristaps 22:
1.1 kristaps 23: #include <assert.h>
1.68 kristaps 24: #include <ctype.h>
1.76 kristaps 25: #include <stdarg.h>
1.229 schwarze 26: #include <stddef.h>
1.29 kristaps 27: #include <stdio.h>
1.63 kristaps 28: #include <stdint.h>
1.1 kristaps 29: #include <stdlib.h>
1.33 kristaps 30: #include <string.h>
1.45 kristaps 31: #include <unistd.h>
1.1 kristaps 32:
1.210 schwarze 33: #include "mandoc_aux.h"
1.229 schwarze 34: #include "mandoc_ohash.h"
1.100 kristaps 35: #include "mandoc.h"
1.210 schwarze 36: #include "roff.h"
1.58 kristaps 37: #include "out.h"
1.51 kristaps 38: #include "html.h"
1.186 schwarze 39: #include "manconf.h"
1.64 kristaps 40: #include "main.h"
1.63 kristaps 41:
1.29 kristaps 42: struct htmldata {
1.63 kristaps 43: const char *name;
1.29 kristaps 44: int flags;
1.257 schwarze 45: #define HTML_INPHRASE (1 << 0) /* Can appear in phrasing context. */
46: #define HTML_TOPHRASE (1 << 1) /* Establishes phrasing context. */
47: #define HTML_NOSTACK (1 << 2) /* Does not have an end tag. */
48: #define HTML_NLBEFORE (1 << 3) /* Output line break before opening. */
49: #define HTML_NLBEGIN (1 << 4) /* Output line break after opening. */
50: #define HTML_NLEND (1 << 5) /* Output line break before closing. */
51: #define HTML_NLAFTER (1 << 6) /* Output line break after closing. */
1.196 schwarze 52: #define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER)
53: #define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND)
54: #define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE)
1.257 schwarze 55: #define HTML_INDENT (1 << 7) /* Indent content by two spaces. */
56: #define HTML_NOINDENT (1 << 8) /* Exception: never indent content. */
1.29 kristaps 57: };
1.7 kristaps 58:
1.29 kristaps 59: static const struct htmldata htmltags[TAG_MAX] = {
1.196 schwarze 60: {"html", HTML_NLALL},
61: {"head", HTML_NLALL | HTML_INDENT},
1.257 schwarze 62: {"meta", HTML_NOSTACK | HTML_NLALL},
63: {"link", HTML_NOSTACK | HTML_NLALL},
64: {"style", HTML_NLALL | HTML_INDENT},
65: {"title", HTML_NLAROUND},
1.196 schwarze 66: {"body", HTML_NLALL},
67: {"div", HTML_NLAROUND},
1.253 schwarze 68: {"section", HTML_NLALL},
1.196 schwarze 69: {"table", HTML_NLALL | HTML_INDENT},
70: {"tr", HTML_NLALL | HTML_INDENT},
71: {"td", HTML_NLAROUND},
72: {"li", HTML_NLAROUND | HTML_INDENT},
73: {"ul", HTML_NLALL | HTML_INDENT},
74: {"ol", HTML_NLALL | HTML_INDENT},
75: {"dl", HTML_NLALL | HTML_INDENT},
76: {"dt", HTML_NLAROUND},
77: {"dd", HTML_NLAROUND | HTML_INDENT},
1.257 schwarze 78: {"h1", HTML_TOPHRASE | HTML_NLAROUND},
79: {"h2", HTML_TOPHRASE | HTML_NLAROUND},
80: {"p", HTML_TOPHRASE | HTML_NLAROUND | HTML_INDENT},
81: {"pre", HTML_TOPHRASE | HTML_NLALL | HTML_NOINDENT},
82: {"a", HTML_INPHRASE | HTML_TOPHRASE},
83: {"b", HTML_INPHRASE | HTML_TOPHRASE},
84: {"cite", HTML_INPHRASE | HTML_TOPHRASE},
85: {"code", HTML_INPHRASE | HTML_TOPHRASE},
86: {"i", HTML_INPHRASE | HTML_TOPHRASE},
87: {"small", HTML_INPHRASE | HTML_TOPHRASE},
88: {"span", HTML_INPHRASE | HTML_TOPHRASE},
89: {"var", HTML_INPHRASE | HTML_TOPHRASE},
90: {"br", HTML_INPHRASE | HTML_NOSTACK | HTML_NLALL},
91: {"math", HTML_INPHRASE | HTML_NLALL | HTML_INDENT},
1.196 schwarze 92: {"mrow", 0},
93: {"mi", 0},
1.215 schwarze 94: {"mn", 0},
1.196 schwarze 95: {"mo", 0},
96: {"msup", 0},
97: {"msub", 0},
98: {"msubsup", 0},
99: {"mfrac", 0},
100: {"msqrt", 0},
101: {"mfenced", 0},
102: {"mtable", 0},
103: {"mtr", 0},
104: {"mtd", 0},
105: {"munderover", 0},
106: {"munder", 0},
107: {"mover", 0},
1.90 kristaps 108: };
109:
1.229 schwarze 110: /* Avoid duplicate HTML id= attributes. */
111: static struct ohash id_unique;
112:
1.254 schwarze 113: static void html_reset_internal(struct html *);
1.197 schwarze 114: static void print_byte(struct html *, char);
115: static void print_endword(struct html *);
116: static void print_indent(struct html *);
117: static void print_word(struct html *, const char *);
118:
1.184 schwarze 119: static void print_ctag(struct html *, struct tag *);
1.197 schwarze 120: static int print_escape(struct html *, char);
1.195 schwarze 121: static int print_encode(struct html *, const char *, const char *, int);
122: static void print_href(struct html *, const char *, const char *, int);
1.255 schwarze 123: static void print_metaf(struct html *);
1.82 kristaps 124:
1.156 schwarze 125:
1.180 schwarze 126: void *
1.191 schwarze 127: html_alloc(const struct manoutput *outopts)
1.10 kristaps 128: {
1.30 kristaps 129: struct html *h;
130:
1.128 kristaps 131: h = mandoc_calloc(1, sizeof(struct html));
1.10 kristaps 132:
1.204 schwarze 133: h->tag = NULL;
1.186 schwarze 134: h->style = outopts->style;
1.240 schwarze 135: if ((h->base_man1 = outopts->man) == NULL)
136: h->base_man2 = NULL;
137: else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL)
138: *h->base_man2++ = '\0';
1.186 schwarze 139: h->base_includes = outopts->includes;
140: if (outopts->fragment)
141: h->oflags |= HTML_FRAGMENT;
1.241 schwarze 142: if (outopts->toc)
143: h->oflags |= HTML_TOC;
1.43 kristaps 144:
1.229 schwarze 145: mandoc_ohash_init(&id_unique, 4, 0);
146:
1.188 schwarze 147: return h;
1.29 kristaps 148: }
1.10 kristaps 149:
1.254 schwarze 150: static void
151: html_reset_internal(struct html *h)
1.29 kristaps 152: {
1.30 kristaps 153: struct tag *tag;
1.229 schwarze 154: char *cp;
155: unsigned int slot;
1.30 kristaps 156:
1.204 schwarze 157: while ((tag = h->tag) != NULL) {
158: h->tag = tag->next;
1.30 kristaps 159: free(tag);
160: }
1.229 schwarze 161: cp = ohash_first(&id_unique, &slot);
162: while (cp != NULL) {
163: free(cp);
164: cp = ohash_next(&id_unique, &slot);
165: }
166: ohash_delete(&id_unique);
1.254 schwarze 167: }
168:
169: void
170: html_reset(void *p)
171: {
172: html_reset_internal(p);
173: mandoc_ohash_init(&id_unique, 4, 0);
174: }
175:
176: void
177: html_free(void *p)
178: {
179: html_reset_internal(p);
180: free(p);
1.10 kristaps 181: }
1.2 kristaps 182:
1.51 kristaps 183: void
1.29 kristaps 184: print_gen_head(struct html *h)
185: {
1.165 kristaps 186: struct tag *t;
1.41 kristaps 187:
1.194 schwarze 188: print_otag(h, TAG_META, "?", "charset", "utf-8");
1.222 schwarze 189: if (h->style != NULL) {
190: print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet",
191: h->style, "type", "text/css", "media", "all");
192: return;
193: }
1.165 kristaps 194:
1.168 kristaps 195: /*
1.222 schwarze 196: * Print a minimal embedded style sheet.
1.168 kristaps 197: */
1.196 schwarze 198:
1.194 schwarze 199: t = print_otag(h, TAG_STYLE, "");
1.196 schwarze 200: print_text(h, "table.head, table.foot { width: 100%; }");
1.197 schwarze 201: print_endline(h);
1.196 schwarze 202: print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }");
1.197 schwarze 203: print_endline(h);
1.196 schwarze 204: print_text(h, "td.head-vol { text-align: center; }");
1.197 schwarze 205: print_endline(h);
1.256 schwarze 206: print_text(h, ".Nd, .Bf, .Op { display: inline; }");
1.225 schwarze 207: print_endline(h);
1.256 schwarze 208: print_text(h, ".Pa, .Ad { font-style: italic; }");
1.226 schwarze 209: print_endline(h);
1.256 schwarze 210: print_text(h, ".Ms { font-weight: bold; }");
1.228 schwarze 211: print_endline(h);
1.256 schwarze 212: print_text(h, ".Bl-diag ");
1.224 schwarze 213: print_byte(h, '>');
214: print_text(h, " dt { font-weight: bold; }");
1.223 schwarze 215: print_endline(h);
1.256 schwarze 216: print_text(h, "code.Nm, .Fl, .Cm, .Ic, code.In, .Fd, .Fn, .Cd "
217: "{ font-weight: bold; font-family: inherit; }");
1.165 kristaps 218: print_tagq(h, t);
1.4 kristaps 219: }
220:
1.255 schwarze 221: int
222: html_setfont(struct html *h, enum mandoc_esc font)
1.88 kristaps 223: {
1.255 schwarze 224: switch (font) {
1.156 schwarze 225: case ESCAPE_FONTPREV:
1.90 kristaps 226: font = h->metal;
1.88 kristaps 227: break;
1.156 schwarze 228: case ESCAPE_FONTITALIC:
229: case ESCAPE_FONTBOLD:
230: case ESCAPE_FONTBI:
1.242 schwarze 231: case ESCAPE_FONTCW:
1.255 schwarze 232: case ESCAPE_FONTROMAN:
1.242 schwarze 233: break;
1.156 schwarze 234: case ESCAPE_FONT:
1.255 schwarze 235: font = ESCAPE_FONTROMAN;
1.88 kristaps 236: break;
237: default:
1.255 schwarze 238: return 0;
1.88 kristaps 239: }
1.255 schwarze 240: h->metal = h->metac;
241: h->metac = font;
242: return 1;
243: }
1.88 kristaps 244:
1.255 schwarze 245: static void
246: print_metaf(struct html *h)
247: {
1.122 kristaps 248: if (h->metaf) {
249: print_tagq(h, h->metaf);
250: h->metaf = NULL;
251: }
1.255 schwarze 252: switch (h->metac) {
253: case ESCAPE_FONTITALIC:
1.194 schwarze 254: h->metaf = print_otag(h, TAG_I, "");
1.152 schwarze 255: break;
1.255 schwarze 256: case ESCAPE_FONTBOLD:
1.194 schwarze 257: h->metaf = print_otag(h, TAG_B, "");
1.152 schwarze 258: break;
1.255 schwarze 259: case ESCAPE_FONTBI:
1.194 schwarze 260: h->metaf = print_otag(h, TAG_B, "");
261: print_otag(h, TAG_I, "");
1.152 schwarze 262: break;
1.255 schwarze 263: case ESCAPE_FONTCW:
1.242 schwarze 264: h->metaf = print_otag(h, TAG_SPAN, "c", "Li");
265: break;
1.152 schwarze 266: default:
267: break;
268: }
1.248 schwarze 269: }
270:
1.249 schwarze 271: void
272: html_close_paragraph(struct html *h)
273: {
1.259 ! schwarze 274: struct tag *this, *next;
! 275: int flags;
1.249 schwarze 276:
1.259 ! schwarze 277: this = h->tag;
! 278: for (;;) {
! 279: next = this->next;
! 280: flags = htmltags[this->tag].flags;
! 281: if (flags & (HTML_INPHRASE | HTML_TOPHRASE))
! 282: print_ctag(h, this);
! 283: if ((flags & HTML_INPHRASE) == 0)
1.249 schwarze 284: break;
1.259 ! schwarze 285: this = next;
1.249 schwarze 286: }
287: }
288:
1.248 schwarze 289: /*
290: * ROFF_nf switches to no-fill mode, ROFF_fi to fill mode.
291: * TOKEN_NONE does not switch. The old mode is returned.
292: */
293: enum roff_tok
294: html_fillmode(struct html *h, enum roff_tok want)
295: {
296: struct tag *t;
297: enum roff_tok had;
298:
299: for (t = h->tag; t != NULL; t = t->next)
300: if (t->tag == TAG_PRE)
301: break;
302:
303: had = t == NULL ? ROFF_fi : ROFF_nf;
304:
305: if (want != had) {
306: switch (want) {
307: case ROFF_fi:
308: print_tagq(h, t);
309: break;
310: case ROFF_nf:
1.249 schwarze 311: html_close_paragraph(h);
1.248 schwarze 312: print_otag(h, TAG_PRE, "");
313: break;
314: case TOKEN_NONE:
315: break;
316: default:
317: abort();
318: }
319: }
320: return had;
1.210 schwarze 321: }
322:
323: char *
1.229 schwarze 324: html_make_id(const struct roff_node *n, int unique)
1.210 schwarze 325: {
326: const struct roff_node *nch;
1.229 schwarze 327: char *buf, *bufs, *cp;
328: unsigned int slot;
329: int suffix;
1.210 schwarze 330:
331: for (nch = n->child; nch != NULL; nch = nch->next)
332: if (nch->type != ROFFT_TEXT)
333: return NULL;
334:
335: buf = NULL;
336: deroff(&buf, n);
1.220 schwarze 337: if (buf == NULL)
338: return NULL;
1.210 schwarze 339:
1.230 schwarze 340: /*
341: * In ID attributes, only use ASCII characters that are
342: * permitted in URL-fragment strings according to the
343: * explicit list at:
344: * https://url.spec.whatwg.org/#url-fragment-string
345: */
1.210 schwarze 346:
347: for (cp = buf; *cp != '\0'; cp++)
1.230 schwarze 348: if (isalnum((unsigned char)*cp) == 0 &&
349: strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
1.210 schwarze 350: *cp = '_';
351:
1.229 schwarze 352: if (unique == 0)
353: return buf;
354:
355: /* Avoid duplicate HTML id= attributes. */
356:
357: bufs = NULL;
358: suffix = 1;
359: slot = ohash_qlookup(&id_unique, buf);
360: cp = ohash_find(&id_unique, slot);
361: if (cp != NULL) {
362: while (cp != NULL) {
363: free(bufs);
364: if (++suffix > 127) {
365: free(buf);
366: return NULL;
367: }
368: mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
369: slot = ohash_qlookup(&id_unique, bufs);
370: cp = ohash_find(&id_unique, slot);
371: }
372: free(buf);
373: buf = bufs;
374: }
375: ohash_insert(&id_unique, slot, buf);
1.210 schwarze 376: return buf;
1.88 kristaps 377: }
378:
1.85 kristaps 379: static int
1.197 schwarze 380: print_escape(struct html *h, char c)
1.159 schwarze 381: {
382:
383: switch (c) {
384: case '<':
1.197 schwarze 385: print_word(h, "<");
1.159 schwarze 386: break;
387: case '>':
1.197 schwarze 388: print_word(h, ">");
1.159 schwarze 389: break;
390: case '&':
1.197 schwarze 391: print_word(h, "&");
1.159 schwarze 392: break;
393: case '"':
1.197 schwarze 394: print_word(h, """);
1.159 schwarze 395: break;
396: case ASCII_NBRSP:
1.197 schwarze 397: print_word(h, " ");
1.159 schwarze 398: break;
399: case ASCII_HYPH:
1.197 schwarze 400: print_byte(h, '-');
1.189 schwarze 401: break;
1.159 schwarze 402: case ASCII_BREAK:
403: break;
404: default:
1.188 schwarze 405: return 0;
1.159 schwarze 406: }
1.188 schwarze 407: return 1;
1.159 schwarze 408: }
409:
410: static int
1.195 schwarze 411: print_encode(struct html *h, const char *p, const char *pend, int norecurse)
1.29 kristaps 412: {
1.197 schwarze 413: char numbuf[16];
1.214 schwarze 414: const char *seq;
1.77 kristaps 415: size_t sz;
1.214 schwarze 416: int c, len, breakline, nospace;
1.132 kristaps 417: enum mandoc_esc esc;
1.214 schwarze 418: static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"',
1.154 schwarze 419: ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' };
1.14 kristaps 420:
1.195 schwarze 421: if (pend == NULL)
422: pend = strchr(p, '\0');
423:
1.214 schwarze 424: breakline = 0;
1.85 kristaps 425: nospace = 0;
426:
1.195 schwarze 427: while (p < pend) {
1.151 schwarze 428: if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
429: h->flags &= ~HTML_SKIPCHAR;
430: p++;
431: continue;
432: }
433:
1.197 schwarze 434: for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
1.214 schwarze 435: print_byte(h, *p);
436:
437: if (breakline &&
438: (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
1.245 schwarze 439: print_otag(h, TAG_BR, "");
1.214 schwarze 440: breakline = 0;
441: while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
442: p++;
443: continue;
444: }
1.77 kristaps 445:
1.195 schwarze 446: if (p >= pend)
1.132 kristaps 447: break;
448:
1.214 schwarze 449: if (*p == ' ') {
450: print_endword(h);
451: p++;
452: continue;
453: }
454:
1.197 schwarze 455: if (print_escape(h, *p++))
1.154 schwarze 456: continue;
1.77 kristaps 457:
1.132 kristaps 458: esc = mandoc_escape(&p, &seq, &len);
459: switch (esc) {
1.156 schwarze 460: case ESCAPE_FONT:
461: case ESCAPE_FONTPREV:
462: case ESCAPE_FONTBOLD:
463: case ESCAPE_FONTITALIC:
464: case ESCAPE_FONTBI:
1.242 schwarze 465: case ESCAPE_FONTCW:
1.156 schwarze 466: case ESCAPE_FONTROMAN:
1.243 schwarze 467: if (0 == norecurse) {
468: h->flags |= HTML_NOSPACE;
1.255 schwarze 469: if (html_setfont(h, esc))
470: print_metaf(h);
1.243 schwarze 471: h->flags &= ~HTML_NOSPACE;
472: }
1.151 schwarze 473: continue;
1.156 schwarze 474: case ESCAPE_SKIPCHAR:
1.151 schwarze 475: h->flags |= HTML_SKIPCHAR;
476: continue;
1.246 schwarze 477: case ESCAPE_ERROR:
478: continue;
1.151 schwarze 479: default:
480: break;
481: }
482:
483: if (h->flags & HTML_SKIPCHAR) {
484: h->flags &= ~HTML_SKIPCHAR;
485: continue;
486: }
487:
488: switch (esc) {
1.156 schwarze 489: case ESCAPE_UNICODE:
1.159 schwarze 490: /* Skip past "u" header. */
1.144 kristaps 491: c = mchars_num2uc(seq + 1, len - 1);
492: break;
1.156 schwarze 493: case ESCAPE_NUMBERED:
1.141 kristaps 494: c = mchars_num2char(seq, len);
1.181 schwarze 495: if (c < 0)
496: continue;
1.82 kristaps 497: break;
1.156 schwarze 498: case ESCAPE_SPECIAL:
1.191 schwarze 499: c = mchars_spec2cp(seq, len);
1.181 schwarze 500: if (c <= 0)
501: continue;
1.246 schwarze 502: break;
503: case ESCAPE_UNDEF:
504: c = *seq;
1.132 kristaps 505: break;
1.239 schwarze 506: case ESCAPE_DEVICE:
507: print_word(h, "html");
508: continue;
1.214 schwarze 509: case ESCAPE_BREAK:
510: breakline = 1;
511: continue;
1.156 schwarze 512: case ESCAPE_NOSPACE:
1.132 kristaps 513: if ('\0' == *p)
514: nospace = 1;
1.179 schwarze 515: continue;
1.185 schwarze 516: case ESCAPE_OVERSTRIKE:
517: if (len == 0)
518: continue;
519: c = seq[len - 1];
520: break;
1.82 kristaps 521: default:
1.179 schwarze 522: continue;
1.82 kristaps 523: }
1.181 schwarze 524: if ((c < 0x20 && c != 0x09) ||
525: (c > 0x7E && c < 0xA0))
1.179 schwarze 526: c = 0xFFFD;
1.197 schwarze 527: if (c > 0x7E) {
1.216 schwarze 528: (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c);
1.197 schwarze 529: print_word(h, numbuf);
530: } else if (print_escape(h, c) == 0)
531: print_byte(h, c);
1.32 kristaps 532: }
1.85 kristaps 533:
1.188 schwarze 534: return nospace;
1.14 kristaps 535: }
536:
1.94 kristaps 537: static void
1.195 schwarze 538: print_href(struct html *h, const char *name, const char *sec, int man)
1.94 kristaps 539: {
1.240 schwarze 540: struct stat sb;
1.195 schwarze 541: const char *p, *pp;
1.240 schwarze 542: char *filename;
543:
544: if (man) {
545: pp = h->base_man1;
546: if (h->base_man2 != NULL) {
547: mandoc_asprintf(&filename, "%s.%s", name, sec);
548: if (stat(filename, &sb) == -1)
549: pp = h->base_man2;
550: free(filename);
551: }
552: } else
553: pp = h->base_includes;
1.195 schwarze 554:
555: while ((p = strchr(pp, '%')) != NULL) {
556: print_encode(h, pp, p, 1);
557: if (man && p[1] == 'S') {
558: if (sec == NULL)
1.197 schwarze 559: print_byte(h, '1');
1.195 schwarze 560: else
561: print_encode(h, sec, NULL, 1);
562: } else if ((man && p[1] == 'N') ||
563: (man == 0 && p[1] == 'I'))
564: print_encode(h, name, NULL, 1);
565: else
566: print_encode(h, p, p + 2, 1);
567: pp = p + 2;
568: }
569: if (*pp != '\0')
570: print_encode(h, pp, NULL, 1);
1.94 kristaps 571: }
572:
1.51 kristaps 573: struct tag *
1.194 schwarze 574: print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
1.14 kristaps 575: {
1.194 schwarze 576: va_list ap;
1.30 kristaps 577: struct tag *t;
1.195 schwarze 578: const char *attr;
1.203 schwarze 579: char *arg1, *arg2;
1.244 schwarze 580: int style_written, tflags;
1.196 schwarze 581:
582: tflags = htmltags[tag].flags;
1.30 kristaps 583:
1.257 schwarze 584: /* Flow content is not allowed in phrasing context. */
585:
586: if ((tflags & HTML_INPHRASE) == 0) {
587: for (t = h->tag; t != NULL; t = t->next) {
588: if (t->closed)
589: continue;
590: assert((htmltags[t->tag].flags & HTML_TOPHRASE) == 0);
591: break;
592: }
593: }
594:
1.204 schwarze 595: /* Push this tag onto the stack of open scopes. */
1.94 kristaps 596:
1.196 schwarze 597: if ((tflags & HTML_NOSTACK) == 0) {
1.128 kristaps 598: t = mandoc_malloc(sizeof(struct tag));
1.30 kristaps 599: t->tag = tag;
1.204 schwarze 600: t->next = h->tag;
1.252 schwarze 601: t->refcnt = 0;
602: t->closed = 0;
1.204 schwarze 603: h->tag = t;
1.30 kristaps 604: } else
605: t = NULL;
1.29 kristaps 606:
1.196 schwarze 607: if (tflags & HTML_NLBEFORE)
1.197 schwarze 608: print_endline(h);
609: if (h->col == 0)
610: print_indent(h);
1.196 schwarze 611: else if ((h->flags & HTML_NOSPACE) == 0) {
612: if (h->flags & HTML_KEEP)
1.216 schwarze 613: print_word(h, " ");
1.196 schwarze 614: else {
615: if (h->flags & HTML_PREKEEP)
616: h->flags |= HTML_KEEP;
1.197 schwarze 617: print_endword(h);
1.105 kristaps 618: }
1.196 schwarze 619: }
1.29 kristaps 620:
1.109 kristaps 621: if ( ! (h->flags & HTML_NONOSPACE))
622: h->flags &= ~HTML_NOSPACE;
1.110 kristaps 623: else
624: h->flags |= HTML_NOSPACE;
1.109 kristaps 625:
1.94 kristaps 626: /* Print out the tag name and attributes. */
627:
1.197 schwarze 628: print_byte(h, '<');
629: print_word(h, htmltags[tag].name);
1.194 schwarze 630:
631: va_start(ap, fmt);
632:
1.244 schwarze 633: while (*fmt != '\0' && *fmt != 's') {
1.203 schwarze 634:
1.238 schwarze 635: /* Parse attributes and arguments. */
1.203 schwarze 636:
637: arg1 = va_arg(ap, char *);
1.238 schwarze 638: arg2 = NULL;
1.194 schwarze 639: switch (*fmt++) {
640: case 'c':
1.195 schwarze 641: attr = "class";
1.194 schwarze 642: break;
643: case 'h':
1.195 schwarze 644: attr = "href";
1.194 schwarze 645: break;
646: case 'i':
1.195 schwarze 647: attr = "id";
1.194 schwarze 648: break;
649: case '?':
1.203 schwarze 650: attr = arg1;
651: arg1 = va_arg(ap, char *);
1.194 schwarze 652: break;
653: default:
654: abort();
655: }
1.203 schwarze 656: if (*fmt == 'M')
657: arg2 = va_arg(ap, char *);
658: if (arg1 == NULL)
659: continue;
660:
1.238 schwarze 661: /* Print the attributes. */
1.203 schwarze 662:
1.197 schwarze 663: print_byte(h, ' ');
664: print_word(h, attr);
665: print_byte(h, '=');
666: print_byte(h, '"');
1.195 schwarze 667: switch (*fmt) {
1.208 schwarze 668: case 'I':
669: print_href(h, arg1, NULL, 0);
670: fmt++;
671: break;
1.195 schwarze 672: case 'M':
1.203 schwarze 673: print_href(h, arg1, arg2, 1);
1.195 schwarze 674: fmt++;
675: break;
1.208 schwarze 676: case 'R':
677: print_byte(h, '#');
678: print_encode(h, arg1, NULL, 1);
1.195 schwarze 679: fmt++;
1.208 schwarze 680: break;
1.195 schwarze 681: default:
1.244 schwarze 682: print_encode(h, arg1, NULL, 1);
1.195 schwarze 683: break;
684: }
1.197 schwarze 685: print_byte(h, '"');
1.194 schwarze 686: }
1.244 schwarze 687:
688: style_written = 0;
689: while (*fmt++ == 's') {
690: arg1 = va_arg(ap, char *);
691: arg2 = va_arg(ap, char *);
692: if (arg2 == NULL)
693: continue;
694: print_byte(h, ' ');
695: if (style_written == 0) {
696: print_word(h, "style=\"");
697: style_written = 1;
698: }
699: print_word(h, arg1);
700: print_byte(h, ':');
701: print_byte(h, ' ');
702: print_word(h, arg2);
703: print_byte(h, ';');
704: }
705: if (style_written)
706: print_byte(h, '"');
707:
1.194 schwarze 708: va_end(ap);
1.94 kristaps 709:
1.172 kristaps 710: /* Accommodate for "well-formed" singleton escaping. */
1.94 kristaps 711:
1.257 schwarze 712: if (htmltags[tag].flags & HTML_NOSTACK)
1.197 schwarze 713: print_byte(h, '/');
1.93 kristaps 714:
1.197 schwarze 715: print_byte(h, '>');
1.14 kristaps 716:
1.196 schwarze 717: if (tflags & HTML_NLBEGIN)
1.197 schwarze 718: print_endline(h);
1.196 schwarze 719: else
720: h->flags |= HTML_NOSPACE;
1.117 kristaps 721:
1.196 schwarze 722: if (tflags & HTML_INDENT)
723: h->indent++;
724: if (tflags & HTML_NOINDENT)
725: h->noindent++;
1.117 kristaps 726:
1.188 schwarze 727: return t;
1.14 kristaps 728: }
729:
1.29 kristaps 730: static void
1.184 schwarze 731: print_ctag(struct html *h, struct tag *tag)
1.14 kristaps 732: {
1.196 schwarze 733: int tflags;
1.156 schwarze 734:
1.252 schwarze 735: if (tag->closed == 0) {
736: tag->closed = 1;
737: if (tag == h->metaf)
738: h->metaf = NULL;
739: if (tag == h->tblt)
740: h->tblt = NULL;
741:
742: tflags = htmltags[tag->tag].flags;
743: if (tflags & HTML_INDENT)
744: h->indent--;
745: if (tflags & HTML_NOINDENT)
746: h->noindent--;
747: if (tflags & HTML_NLEND)
748: print_endline(h);
749: print_indent(h);
750: print_byte(h, '<');
751: print_byte(h, '/');
752: print_word(h, htmltags[tag->tag].name);
753: print_byte(h, '>');
754: if (tflags & HTML_NLAFTER)
755: print_endline(h);
756: }
757: if (tag->refcnt == 0) {
758: h->tag = tag->next;
759: free(tag);
760: }
1.14 kristaps 761: }
762:
1.51 kristaps 763: void
1.93 kristaps 764: print_gen_decls(struct html *h)
1.1 kristaps 765: {
1.197 schwarze 766: print_word(h, "<!DOCTYPE html>");
767: print_endline(h);
1.221 schwarze 768: }
769:
770: void
771: print_gen_comment(struct html *h, struct roff_node *n)
772: {
773: int wantblank;
774:
775: print_word(h, "<!-- This is an automatically generated file."
776: " Do not edit.");
777: h->indent = 1;
778: wantblank = 0;
779: while (n != NULL && n->type == ROFFT_COMMENT) {
780: if (strstr(n->string, "-->") == NULL &&
781: (wantblank || *n->string != '\0')) {
782: print_endline(h);
783: print_indent(h);
784: print_word(h, n->string);
785: wantblank = *n->string != '\0';
786: }
787: n = n->next;
788: }
789: if (wantblank)
790: print_endline(h);
791: print_word(h, " -->");
792: print_endline(h);
793: h->indent = 0;
1.1 kristaps 794: }
795:
1.51 kristaps 796: void
1.104 kristaps 797: print_text(struct html *h, const char *word)
1.1 kristaps 798: {
1.197 schwarze 799: if (h->col && (h->flags & HTML_NOSPACE) == 0) {
1.105 kristaps 800: if ( ! (HTML_KEEP & h->flags)) {
801: if (HTML_PREKEEP & h->flags)
802: h->flags |= HTML_KEEP;
1.197 schwarze 803: print_endword(h);
1.105 kristaps 804: } else
1.216 schwarze 805: print_word(h, " ");
1.105 kristaps 806: }
1.30 kristaps 807:
1.255 schwarze 808: assert(h->metaf == NULL);
809: print_metaf(h);
810: print_indent(h);
1.195 schwarze 811: if ( ! print_encode(h, word, NULL, 0)) {
1.109 kristaps 812: if ( ! (h->flags & HTML_NONOSPACE))
813: h->flags &= ~HTML_NOSPACE;
1.183 schwarze 814: h->flags &= ~HTML_NONEWLINE;
1.149 kristaps 815: } else
1.183 schwarze 816: h->flags |= HTML_NOSPACE | HTML_NONEWLINE;
1.122 kristaps 817:
1.255 schwarze 818: if (h->metaf != NULL) {
1.122 kristaps 819: print_tagq(h, h->metaf);
820: h->metaf = NULL;
821: }
1.113 schwarze 822:
823: h->flags &= ~HTML_IGNDELIM;
1.1 kristaps 824: }
1.30 kristaps 825:
1.51 kristaps 826: void
1.30 kristaps 827: print_tagq(struct html *h, const struct tag *until)
828: {
1.252 schwarze 829: struct tag *this, *next;
1.30 kristaps 830:
1.252 schwarze 831: for (this = h->tag; this != NULL; this = next) {
832: next = this == until ? NULL : this->next;
833: print_ctag(h, this);
1.30 kristaps 834: }
835: }
836:
1.250 schwarze 837: /*
838: * Close out all open elements up to but excluding suntil.
839: * Note that a paragraph just inside stays open together with it
840: * because paragraphs include subsequent phrasing content.
841: */
1.51 kristaps 842: void
1.30 kristaps 843: print_stagq(struct html *h, const struct tag *suntil)
844: {
1.252 schwarze 845: struct tag *this, *next;
1.30 kristaps 846:
1.252 schwarze 847: for (this = h->tag; this != NULL; this = next) {
848: next = this->next;
849: if (this == suntil || (next == suntil &&
850: (this->tag == TAG_P || this->tag == TAG_PRE)))
851: break;
852: print_ctag(h, this);
1.30 kristaps 853: }
1.171 kristaps 854: }
855:
1.197 schwarze 856:
857: /***********************************************************************
858: * Low level output functions.
859: * They implement line breaking using a short static buffer.
860: ***********************************************************************/
861:
862: /*
863: * Buffer one HTML output byte.
864: * If the buffer is full, flush and deactivate it and start a new line.
865: * If the buffer is inactive, print directly.
866: */
867: static void
868: print_byte(struct html *h, char c)
869: {
870: if ((h->flags & HTML_BUFFER) == 0) {
871: putchar(c);
872: h->col++;
873: return;
874: }
875:
876: if (h->col + h->bufcol < sizeof(h->buf)) {
877: h->buf[h->bufcol++] = c;
878: return;
879: }
880:
881: putchar('\n');
882: h->col = 0;
883: print_indent(h);
884: putchar(' ');
885: putchar(' ');
886: fwrite(h->buf, h->bufcol, 1, stdout);
887: putchar(c);
888: h->col = (h->indent + 1) * 2 + h->bufcol + 1;
889: h->bufcol = 0;
890: h->flags &= ~HTML_BUFFER;
891: }
892:
1.196 schwarze 893: /*
894: * If something was printed on the current output line, end it.
1.197 schwarze 895: * Not to be called right after print_indent().
1.196 schwarze 896: */
1.202 schwarze 897: void
1.197 schwarze 898: print_endline(struct html *h)
1.196 schwarze 899: {
1.197 schwarze 900: if (h->col == 0)
1.196 schwarze 901: return;
902:
1.197 schwarze 903: if (h->bufcol) {
904: putchar(' ');
905: fwrite(h->buf, h->bufcol, 1, stdout);
906: h->bufcol = 0;
907: }
1.196 schwarze 908: putchar('\n');
1.197 schwarze 909: h->col = 0;
910: h->flags |= HTML_NOSPACE;
911: h->flags &= ~HTML_BUFFER;
912: }
913:
914: /*
915: * Flush the HTML output buffer.
916: * If it is inactive, activate it.
917: */
918: static void
919: print_endword(struct html *h)
920: {
921: if (h->noindent) {
922: print_byte(h, ' ');
923: return;
924: }
925:
926: if ((h->flags & HTML_BUFFER) == 0) {
927: h->col++;
928: h->flags |= HTML_BUFFER;
929: } else if (h->bufcol) {
930: putchar(' ');
931: fwrite(h->buf, h->bufcol, 1, stdout);
932: h->col += h->bufcol + 1;
933: }
934: h->bufcol = 0;
1.196 schwarze 935: }
936:
937: /*
938: * If at the beginning of a new output line,
939: * perform indentation and mark the line as containing output.
940: * Make sure to really produce some output right afterwards,
941: * but do not use print_otag() for producing it.
942: */
943: static void
1.197 schwarze 944: print_indent(struct html *h)
1.196 schwarze 945: {
1.197 schwarze 946: size_t i;
1.196 schwarze 947:
1.197 schwarze 948: if (h->col)
1.196 schwarze 949: return;
950:
1.197 schwarze 951: if (h->noindent == 0) {
952: h->col = h->indent * 2;
953: for (i = 0; i < h->col; i++)
1.196 schwarze 954: putchar(' ');
1.197 schwarze 955: }
956: h->flags &= ~HTML_NOSPACE;
957: }
958:
959: /*
960: * Print or buffer some characters
961: * depending on the current HTML output buffer state.
962: */
963: static void
964: print_word(struct html *h, const char *cp)
965: {
966: while (*cp != '\0')
967: print_byte(h, *cp++);
1.68 kristaps 968: }
CVSweb