src/html.c

   1 /*
   2  * Sylpheed -- a GTK+ based, lightweight, and fast e-mail client
   3  * Copyright (C) 1999-2007 Hiroyuki Yamamoto and the Claws Mail team
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License
  16  * along with this program; if not, write to the Free Software
  17  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  */
  19
  20 #include <glib.h>
  21 #include <stdio.h>
  22 #include <string.h>
  23 #include <ctype.h>
  24
  25 #include "html.h"
  26 #include "codeconv.h"
  27 #include "utils.h"
  28
  29 #define SC_HTMLBUFSIZE  8192
  30 #define HR_STR          "------------------------------------------------"
  31
  32 typedef struct _SC_HTMLSymbol   SC_HTMLSymbol;
  33
  34 struct _SC_HTMLSymbol
  35 {
  36         gchar *const key;
  37         gchar *const val;
  38 };
  39
  40 static SC_HTMLSymbol symbol_list[] = {
  41         {"&lt;"    , "<"},
  42         {"&gt;"    , ">"},
  43         {"&amp;"   , "&"},
  44         {"&quot;"  , "\""},
  45         {"&lsquo;",  "'"},
  46         {"&rsquo;",  "'"},
  47         {"&ldquo;",  "\""},
  48         {"&rdquo;",  "\""},
  49         {"&nbsp;"  , " "},
  50         {"&trade;" , "(TM)"},
  51         {"&hellip;", "..."},
  52         {"&bull;", "*"},
  53         {"&ndash;", "-"},
  54         {"&mdash;", "--"},
  55         {"&euro;", "EUR"},
  56         {"&cent;", "c"},
  57         {"&pound;", "£"},
  58 };
  59
  60 static SC_HTMLSymbol ascii_symbol_list[] = {
  61         {"&iexcl;" , "\302\241"},
  62         {"&brvbar;", "\302\246"},
  63         {"&copy;"  , "\302\251"},
  64         {"&laquo;" , "\302\253"},
  65         {"&reg;"   , "\302\256"},
  66
  67         {"&sup2;"  , "\302\262"},
  68         {"&sup3;"  , "\302\263"},
  69         {"&acute;" , "\302\264"},
  70         {"&cedil;" , "\302\270"},
  71         {"&sup1;"  , "\302\271"},
  72         {"&raquo;" , "\302\273"},
  73         {"&frac14;", "\302\274"},
  74         {"&frac12;", "\302\275"},
  75         {"&frac34;", "\302\276"},
  76         {"&iquest;", "\302\277"},
  77
  78         {"&Agrave;", "\303\200"},
  79         {"&Aacute;", "\303\201"},
  80         {"&Acirc;" , "\303\202"},
  81         {"&Atilde;", "\303\203"},
  82         {"&AElig;" , "\303\206"},
  83         {"&Egrave;", "\303\210"},
  84         {"&Eacute;", "\303\211"},
  85         {"&Ecirc;" , "\303\212"},
  86         {"&Igrave;", "\303\214"},
  87         {"&Iacute;", "\303\215"},
  88         {"&Icirc;" , "\303\216"},
  89
  90         {"&Ntilde;", "\303\221"},
  91         {"&Ograve;", "\303\222"},
  92         {"&Oacute;", "\303\223"},
  93         {"&Ocirc;" , "\303\224"},
  94         {"&Otilde;", "\303\225"},
  95         {"&Ugrave;", "\303\231"},
  96         {"&Uacute;", "\303\232"},
  97         {"&Ucirc;" , "\303\233"},
  98         {"&Yacute;", "\303\235"},
  99
 100         {"&agrave;", "\303\240"},
 101         {"&aacute;", "\303\241"},
 102         {"&acirc;" , "\303\242"},
 103         {"&atilde;", "\303\243"},
 104         {"&aelig;" , "\303\246"},
 105         {"&egrave;", "\303\250"},
 106         {"&eacute;", "\303\251"},
 107         {"&ecirc;" , "\303\252"},
 108         {"&igrave;", "\303\254"},
 109         {"&iacute;", "\303\255"},
 110         {"&icirc;" , "\303\256"},
 111
 112         {"&ntilde;", "\303\261"},
 113         {"&ograve;", "\303\262"},
 114         {"&oacute;", "\303\263"},
 115         {"&ocirc;" , "\303\264"},
 116         {"&otilde;", "\303\265"},
 117         {"&ugrave;", "\303\271"},
 118         {"&uacute;", "\303\272"},
 119         {"&ucirc;" , "\303\273"},
 120         {"&yacute;", "\303\275"},
 121 };
 122
 123 static GHashTable *default_symbol_table;
 124
 125 static SC_HTMLState sc_html_read_line   (SC_HTMLParser  *parser);
 126 static void sc_html_append_char                 (SC_HTMLParser  *parser,
 127                                          gchar           ch);
 128 static void sc_html_append_str                  (SC_HTMLParser  *parser,
 129                                          const gchar    *str,
 130                                          gint            len);
 131 static SC_HTMLState sc_html_parse_tag   (SC_HTMLParser  *parser);
 132 static void sc_html_parse_special               (SC_HTMLParser  *parser);
 133 static void sc_html_get_parenthesis             (SC_HTMLParser  *parser,
 134                                          gchar          *buf,
 135                                          gint            len);
 136
 137
 138 SC_HTMLParser *sc_html_parser_new(FILE *fp, CodeConverter *conv)
 139 {
 140         SC_HTMLParser *parser;
 141
 142         g_return_val_if_fail(fp != NULL, NULL);
 143         g_return_val_if_fail(conv != NULL, NULL);
 144
 145         parser = g_new0(SC_HTMLParser, 1);
 146         parser->fp = fp;
 147         parser->conv = conv;
 148         parser->str = g_string_new(NULL);
 149         parser->buf = g_string_new(NULL);
 150         parser->bufp = parser->buf->str;
 151         parser->state = SC_HTML_NORMAL;
 152         parser->href = NULL;
 153         parser->newline = TRUE;
 154         parser->empty_line = TRUE;
 155         parser->space = FALSE;
 156         parser->pre = FALSE;
 157
 158 #define SYMBOL_TABLE_ADD(table, list) \
 159 { \
 160         gint i; \
 161  \
 162         for (i = 0; i < sizeof(list) / sizeof(list[0]); i++) \
 163                 g_hash_table_insert(table, list[i].key, list[i].val); \
 164 }
 165
 166         if (!default_symbol_table) {
 167                 default_symbol_table =
 168                         g_hash_table_new(g_str_hash, g_str_equal);
 169                 SYMBOL_TABLE_ADD(default_symbol_table, symbol_list);
 170                 SYMBOL_TABLE_ADD(default_symbol_table, ascii_symbol_list);
 171         }
 172
 173 #undef SYMBOL_TABLE_ADD
 174
 175         parser->symbol_table = default_symbol_table;
 176
 177         return parser;
 178 }
 179
 180 void sc_html_parser_destroy(SC_HTMLParser *parser)
 181 {
 182         g_string_free(parser->str, TRUE);
 183         g_string_free(parser->buf, TRUE);
 184         g_free(parser->href);
 185         g_free(parser);
 186 }
 187
 188 gchar *sc_html_parse(SC_HTMLParser *parser)
 189 {
 190         parser->state = SC_HTML_NORMAL;
 191         g_string_truncate(parser->str, 0);
 192
 193         if (*parser->bufp == '\0') {
 194                 g_string_truncate(parser->buf, 0);
 195                 parser->bufp = parser->buf->str;
 196                 if (sc_html_read_line(parser) == SC_HTML_EOF)
 197                         return NULL;
 198         }
 199
 200         while (*parser->bufp != '\0') {
 201                 switch (*parser->bufp) {
 202                 case '<': {
 203                         SC_HTMLState st;
 204                         st = sc_html_parse_tag(parser);
 205                         /* when we see an href, we need to flush the str
 206                          * buffer.  Then collect all the chars until we
 207                          * see the end anchor tag
 208                          */
 209                         if (SC_HTML_HREF_BEG == st || SC_HTML_HREF == st)
 210                                 return parser->str->str;
 211                         }
 212                         break;
 213                 case '&':
 214                         sc_html_parse_special(parser);
 215                         break;
 216                 case ' ':
 217                 case '\t':
 218                 case '\r':
 219                 case '\n':
 220                         if (parser->bufp[0] == '\r' && parser->bufp[1] == '\n')
 221                                 parser->bufp++;
 222
 223                         if (!parser->pre) {
 224                                 if (!parser->newline)
 225                                         parser->space = TRUE;
 226
 227                                 parser->bufp++;
 228                                 break;
 229                         }
 230                         /* fallthrough */
 231                 default:
 232                         sc_html_append_char(parser, *parser->bufp++);
 233                 }
 234         }
 235
 236         return parser->str->str;
 237 }
 238
 239 static SC_HTMLState sc_html_read_line(SC_HTMLParser *parser)
 240 {
 241         gchar buf[SC_HTMLBUFSIZE];
 242         gchar buf2[SC_HTMLBUFSIZE];
 243         gint index;
 244
 245         if (fgets(buf, sizeof(buf), parser->fp) == NULL) {
 246                 parser->state = SC_HTML_EOF;
 247                 return SC_HTML_EOF;
 248         }
 249
 250         if (conv_convert(parser->conv, buf2, sizeof(buf2), buf) < 0) {
 251                 index = parser->bufp - parser->buf->str;
 252
 253                 conv_utf8todisp(buf2, sizeof(buf2), buf);
 254                 g_string_append(parser->buf, buf2);
 255
 256                 parser->bufp = parser->buf->str + index;
 257
 258                 return SC_HTML_CONV_FAILED;
 259         }
 260
 261         index = parser->bufp - parser->buf->str;
 262
 263         g_string_append(parser->buf, buf2);
 264
 265         parser->bufp = parser->buf->str + index;
 266
 267         return SC_HTML_NORMAL;
 268 }
 269
 270 static void sc_html_append_char(SC_HTMLParser *parser, gchar ch)
 271 {
 272         GString *str = parser->str;
 273
 274         if (!parser->pre && parser->space) {
 275                 g_string_append_c(str, ' ');
 276                 parser->space = FALSE;
 277         }
 278
 279         g_string_append_c(str, ch);
 280
 281         parser->empty_line = FALSE;
 282         if (ch == '\n') {
 283                 parser->newline = TRUE;
 284                 if (str->len > 1 && str->str[str->len - 2] == '\n')
 285                         parser->empty_line = TRUE;
 286         } else
 287                 parser->newline = FALSE;
 288 }
 289
 290 static void sc_html_append_str(SC_HTMLParser *parser, const gchar *str, gint len)
 291 {
 292         GString *string = parser->str;
 293
 294         if (!parser->pre && parser->space) {
 295                 g_string_append_c(string, ' ');
 296                 parser->space = FALSE;
 297         }
 298
 299         if (len == 0) return;
 300         if (len < 0)
 301                 g_string_append(string, str);
 302         else {
 303                 gchar *s;
 304                 Xstrndup_a(s, str, len, return);
 305                 g_string_append(string, s);
 306         }
 307
 308         parser->empty_line = FALSE;
 309         if (string->len > 0 && string->str[string->len - 1] == '\n') {
 310                 parser->newline = TRUE;
 311                 if (string->len > 1 && string->str[string->len - 2] == '\n')
 312                         parser->empty_line = TRUE;
 313         } else
 314                 parser->newline = FALSE;
 315 }
 316
 317 static SC_HTMLTag *sc_html_get_tag(const gchar *str)
 318 {
 319         SC_HTMLTag *tag;
 320         gchar *tmp;
 321         guchar *tmpp;
 322
 323         g_return_val_if_fail(str != NULL, NULL);
 324
 325         if (*str == '\0' || *str == '!') return NULL;
 326
 327         Xstrdup_a(tmp, str, return NULL);
 328
 329         tag = g_new0(SC_HTMLTag, 1);
 330
 331         for (tmpp = tmp; *tmpp != '\0' && !g_ascii_isspace(*tmpp); tmpp++)
 332                 ;
 333
 334         if (*tmpp == '\0') {
 335                 g_strdown(tmp);
 336                 tag->name = g_strdup(tmp);
 337                 return tag;
 338         } else {
 339                 *tmpp++ = '\0';
 340                 g_strdown(tmp);
 341                 tag->name = g_strdup(tmp);
 342         }
 343
 344         while (*tmpp != '\0') {
 345                 SC_HTMLAttr *attr;
 346                 gchar *attr_name;
 347                 gchar *attr_value;
 348                 gchar *p;
 349                 gchar quote;
 350
 351                 while (g_ascii_isspace(*tmpp)) tmpp++;
 352                 attr_name = tmpp;
 353
 354                 while (*tmpp != '\0' && !g_ascii_isspace(*tmpp) &&
 355                        *tmpp != '=')
 356                         tmpp++;
 357                 if (*tmpp != '\0' && *tmpp != '=') {
 358                         *tmpp++ = '\0';
 359                         while (g_ascii_isspace(*tmpp)) tmpp++;
 360                 }
 361
 362                 if (*tmpp == '=') {
 363                         *tmpp++ = '\0';
 364                         while (g_ascii_isspace(*tmpp)) tmpp++;
 365
 366                         if (*tmpp == '"' || *tmpp == '\'') {
 367                                 /* name="value" */
 368                                 quote = *tmpp;
 369                                 tmpp++;
 370                                 attr_value = tmpp;
 371                                 if ((p = strchr(attr_value, quote)) == NULL) {
 372                                         g_warning("sc_html_get_tag(): syntax error in tag: '%s'\n", str);
 373                                         return tag;
 374                                 }
 375                                 tmpp = p;
 376                                 *tmpp++ = '\0';
 377                                 while (g_ascii_isspace(*tmpp)) tmpp++;
 378                         } else {
 379                                 /* name=value */
 380                                 attr_value = tmpp;
 381                                 while (*tmpp != '\0' && !g_ascii_isspace(*tmpp)) tmpp++;
 382                                 if (*tmpp != '\0')
 383                                         *tmpp++ = '\0';
 384                         }
 385                 } else
 386                         attr_value = "";
 387
 388                 g_strchomp(attr_name);
 389                 g_strdown(attr_name);
 390                 attr = g_new(SC_HTMLAttr, 1);
 391                 attr->name = g_strdup(attr_name);
 392                 attr->value = g_strdup(attr_value);
 393                 tag->attr = g_list_append(tag->attr, attr);
 394         }
 395
 396         return tag;
 397 }
 398
 399 static void sc_html_free_tag(SC_HTMLTag *tag)
 400 {
 401         if (!tag) return;
 402
 403         g_free(tag->name);
 404         while (tag->attr != NULL) {
 405                 SC_HTMLAttr *attr = (SC_HTMLAttr *)tag->attr->data;
 406                 g_free(attr->name);
 407                 g_free(attr->value);
 408                 g_free(attr);
 409                 tag->attr = g_list_remove(tag->attr, tag->attr->data);
 410         }
 411         g_free(tag);
 412 }
 413
 414 static SC_HTMLState sc_html_parse_tag(SC_HTMLParser *parser)
 415 {
 416         gchar buf[SC_HTMLBUFSIZE];
 417         SC_HTMLTag *tag;
 418
 419         sc_html_get_parenthesis(parser, buf, sizeof(buf));
 420
 421         tag = sc_html_get_tag(buf);
 422
 423         parser->state = SC_HTML_UNKNOWN;
 424         if (!tag) return SC_HTML_UNKNOWN;
 425
 426         if (!strcmp(tag->name, "br")) {
 427                 parser->space = FALSE;
 428                 sc_html_append_char(parser, '\n');
 429                 parser->state = SC_HTML_BR;
 430         } else if (!strcmp(tag->name, "a")) {
 431                 GList *cur;
 432                 for (cur = tag->attr; cur != NULL; cur = cur->next) {
 433                         if (cur->data && !strcmp(((SC_HTMLAttr *)cur->data)->name, "href")) {
 434                                 g_free(parser->href);
 435                                 parser->href = g_strdup(((SC_HTMLAttr *)cur->data)->value);
 436                                 parser->state = SC_HTML_HREF_BEG;
 437                                 break;
 438                         }
 439                 }
 440         } else if (!strcmp(tag->name, "/a")) {
 441                 parser->state = SC_HTML_HREF;
 442         } else if (!strcmp(tag->name, "p")) {
 443                 parser->space = FALSE;
 444                 if (!parser->empty_line) {
 445                         parser->space = FALSE;
 446                         if (!parser->newline) sc_html_append_char(parser, '\n');
 447                         sc_html_append_char(parser, '\n');
 448                 }
 449                 parser->state = SC_HTML_PAR;
 450         } else if (!strcmp(tag->name, "pre")) {
 451                 parser->pre = TRUE;
 452                 parser->state = SC_HTML_PRE;
 453         } else if (!strcmp(tag->name, "/pre")) {
 454                 parser->pre = FALSE;
 455                 parser->state = SC_HTML_NORMAL;
 456         } else if (!strcmp(tag->name, "hr")) {
 457                 if (!parser->newline) {
 458                         parser->space = FALSE;
 459                         sc_html_append_char(parser, '\n');
 460                 }
 461                 sc_html_append_str(parser, HR_STR "\n", -1);
 462                 parser->state = SC_HTML_HR;
 463         } else if (!strcmp(tag->name, "div")    ||
 464                    !strcmp(tag->name, "ul")     ||
 465                    !strcmp(tag->name, "li")     ||
 466                    !strcmp(tag->name, "table")  ||
 467                    !strcmp(tag->name, "tr")     ||
 468                    (tag->name[0] == 'h' && g_ascii_isdigit(tag->name[1]))) {
 469                 if (!parser->newline) {
 470                         parser->space = FALSE;
 471                         sc_html_append_char(parser, '\n');
 472                 }
 473                 parser->state = SC_HTML_NORMAL;
 474         } else if (!strcmp(tag->name, "/table") ||
 475                    (tag->name[0] == '/' &&
 476                     tag->name[1] == 'h' &&
 477                     g_ascii_isdigit(tag->name[1]))) {
 478                 if (!parser->empty_line) {
 479                         parser->space = FALSE;
 480                         if (!parser->newline) sc_html_append_char(parser, '\n');
 481                         sc_html_append_char(parser, '\n');
 482                 }
 483                 parser->state = SC_HTML_NORMAL;
 484         } else if (!strcmp(tag->name, "/div")   ||
 485                    !strcmp(tag->name, "/ul")    ||
 486                    !strcmp(tag->name, "/li")) {
 487                 if (!parser->newline) {
 488                         parser->space = FALSE;
 489                         sc_html_append_char(parser, '\n');
 490                 }
 491                 parser->state = SC_HTML_NORMAL;
 492                         }
 493
 494         sc_html_free_tag(tag);
 495
 496         return parser->state;
 497 }
 498
 499 static void sc_html_parse_special(SC_HTMLParser *parser)
 500 {
 501         gchar symbol_name[9];
 502         gint n;
 503         const gchar *val;
 504
 505         parser->state = SC_HTML_UNKNOWN;
 506         g_return_if_fail(*parser->bufp == '&');
 507
 508         /* &foo; */
 509         for (n = 0; parser->bufp[n] != '\0' && parser->bufp[n] != ';'; n++)
 510                 ;
 511         if (n > 7 || parser->bufp[n] != ';') {
 512                 /* output literal `&' */
 513                 sc_html_append_char(parser, *parser->bufp++);
 514                 parser->state = SC_HTML_NORMAL;
 515                 return;
 516         }
 517         strncpy2(symbol_name, parser->bufp, n + 2);
 518         parser->bufp += n + 1;
 519
 520         if ((val = g_hash_table_lookup(parser->symbol_table, symbol_name))
 521             != NULL) {
 522                 sc_html_append_str(parser, val, -1);
 523                 parser->state = SC_HTML_NORMAL;
 524                 return;
 525         } else if (symbol_name[1] == '#' && g_ascii_isdigit(symbol_name[2])) {
 526                 gint ch;
 527
 528                 ch = atoi(symbol_name + 2);
 529                 if ((ch > 0 && ch <= 127) ||
 530                     (ch >= 128 && ch <= 255 &&
 531                      parser->conv->charset == C_ISO_8859_1)) {
 532                         sc_html_append_char(parser, ch);
 533                         parser->state = SC_HTML_NORMAL;
 534                         return;
 535                 } else {
 536                         char *symb = NULL;
 537                         switch (ch) {
 538                         /* http://www.w3schools.com/html/html_entitiesref.asp */
 539                         case 96:        /* backtick  */
 540                                 symb = "`";
 541                                 break;
 542                         case 153:       /* trademark */
 543                                 symb = "(TM)";
 544                                 break;
 545                         case 162:       /* cent (currency)      &cent; */
 546                                 symb = "c";
 547                                 break;
 548                         case 163:       /* pound (currency)     &pound; */
 549                                 symb = "£";
 550                                 break;
 551                         case 338:       /* capital ligature OE  &OElig;  */
 552                                 symb = "OE";
 553                                 break;
 554                         case 339:       /* small ligature OE    &oelig;  */
 555                                 symb = "oe";
 556                                 break;
 557                         case 352:       /* capital S w/caron    &Scaron; */
 558                         case 353:       /* small S w/caron      &scaron; */
 559                         case 376:       /* cap Y w/ diaeres     &Yuml;   */
 560                                 break;
 561                         case 710:       /* circumflex accent    &circ;   */
 562                                 symb = "^";
 563                                 break;
 564                         case 732:       /* small tilde          &tilde;  */
 565                                 symb = "~";
 566                                 break;
 567                         case 8194:      /* en space             &ensp;   */
 568                         case 8195:      /* em space             &emsp;   */
 569                         case 8201:      /* thin space           &thinsp; */
 570                                 symb = " ";
 571                                 break;
 572                         case 8204:      /* zero width non-joiner &zwnj;  */
 573                         case 8205:      /* zero width joiner    &zwj;   */
 574                         case 8206:      /* l-t-r mark           &lrm;   */
 575                         case 8207:      /* r-t-l mark           &rlm     */
 576                                 break;
 577                         case 8211:      /* en dash              &ndash;  */
 578                                 symb = "-";
 579                                 break;
 580                         case 8212:      /* em dash              &mdash;  */
 581                                 symb = "--";
 582                                 break;
 583                         case 8216:      /* l single quot mark   &lsquo;  */
 584                         case 8217:      /* r single quot mark   &rsquo;  */
 585                                 symb = "'";
 586                                 break;
 587                         case 8218:      /* single low-9 quot    &sbquo;  */
 588                                 symb = ",";
 589                                 break;
 590                         case 8220:      /* l double quot mark   &ldquo;  */
 591                         case 8221:      /* r double quot mark   &rdquo;  */
 592                                 symb = "\"";
 593                                 break;
 594                         case 8222:      /* double low-9 quot    &bdquo;  */
 595                                 symb = ",,";
 596                                 break;
 597                         case 8224:      /* dagger               &dagger; */
 598                         case 8225:      /* double dagger        &Dagger; */
 599                                 break;
 600                         case 8226:      /* bullet       &bull;  */
 601                                 symb = "*";
 602                                 break;
 603                         case 8230:      /* horizontal ellipsis  &hellip; */
 604                                 symb = "...";
 605                                 break;
 606                         case 8240:      /* per mile             &permil; */
 607                                 symb = "\%o";
 608                                 break;
 609                         case 8249:      /* l-pointing angle quot &lsaquo; */
 610                                 symb = "<";
 611                                 break;
 612                         case 8250:      /* r-pointing angle quot &rsaquo; */
 613                                 symb = ">";
 614                                 break;
 615                         case 8364:      /* euro                 &euro;   */
 616                                 symb = "EUR";
 617                                 break;
 618                         case 8482:      /* trademark            &trade;  */
 619                                 symb  = "(TM)";
 620                                 break;
 621                         default:
 622                                 break;
 623                         }
 624                         if (symb) {
 625                                 sc_html_append_str(parser, symb, -1);
 626                                 parser->state = SC_HTML_NORMAL;
 627                                 return;
 628                         }
 629                 }
 630         }
 631
 632         sc_html_append_str(parser, symbol_name, -1);
 633 }
 634
 635 static void sc_html_get_parenthesis(SC_HTMLParser *parser, gchar *buf, gint len)
 636 {
 637         gchar *p;
 638
 639         buf[0] = '\0';
 640         g_return_if_fail(*parser->bufp == '<');
 641
 642         /* ignore comment / CSS / script stuff */
 643         if (!strncmp(parser->bufp, "<!--", 4)) {
 644                 parser->bufp += 4;
 645                 while ((p = strstr(parser->bufp, "-->")) == NULL)
 646                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 647                 parser->bufp = p + 3;
 648                 return;
 649         }
 650         if (!g_ascii_strncasecmp(parser->bufp, "<style", 6)) {
 651                 parser->bufp += 6;
 652                 while ((p = strcasestr(parser->bufp, "</style>")) == NULL)
 653                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 654                 parser->bufp = p + 8;
 655                 return;
 656         }
 657         if (!g_ascii_strncasecmp(parser->bufp, "<script", 7)) {
 658                 parser->bufp += 7;
 659                 while ((p = strcasestr(parser->bufp, "</script>")) == NULL)
 660                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 661                 parser->bufp = p + 9;
 662                 return;
 663         }
 664
 665         parser->bufp++;
 666         while ((p = strchr(parser->bufp, '>')) == NULL)
 667                 if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 668
 669         strncpy2(buf, parser->bufp, MIN(p - parser->bufp + 1, len));
 670         g_strstrip(buf);
 671         parser->bufp = p + 1;
 672 }