src/html.c

   1 /*
   2  * Claws Mail -- a GTK+ based, lightweight, and fast e-mail client
   3  * Copyright (C) 1999-2016 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 3 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, see <http://www.gnu.org/licenses/>.
  17  */
  18
  19 #ifdef HAVE_CONFIG_H
  20 #  include "config.h"
  21 #include "claws-features.h"
  22 #endif
  23
  24 #include <glib.h>
  25 #include <stdio.h>
  26 #include <string.h>
  27 #include <ctype.h>
  28
  29 #include "html.h"
  30 #include "codeconv.h"
  31 #include "utils.h"
  32 #include "entity.h"
  33 #include "file-utils.h"
  34
  35 #define SC_HTMLBUFSIZE  8192
  36 #define HR_STR          "────────────────────────────────────────────────"
  37 #define LI_STR          "• "
  38
  39 static SC_HTMLState sc_html_read_line   (SC_HTMLParser  *parser);
  40 static void sc_html_append_char                 (SC_HTMLParser  *parser,
  41                                          gchar           ch);
  42 static void sc_html_append_str                  (SC_HTMLParser  *parser,
  43                                          const gchar    *str,
  44                                          gint            len);
  45 static SC_HTMLState sc_html_parse_tag   (SC_HTMLParser  *parser);
  46 static void sc_html_parse_special               (SC_HTMLParser  *parser);
  47 static void sc_html_get_parenthesis             (SC_HTMLParser  *parser,
  48                                          gchar          *buf,
  49                                          gint            len);
  50
  51
  52 SC_HTMLParser *sc_html_parser_new(FILE *fp, CodeConverter *conv)
  53 {
  54         SC_HTMLParser *parser;
  55
  56         cm_return_val_if_fail(fp != NULL, NULL);
  57         cm_return_val_if_fail(conv != NULL, NULL);
  58
  59         parser = g_new0(SC_HTMLParser, 1);
  60         parser->fp = fp;
  61         parser->conv = conv;
  62         parser->str = g_string_new(NULL);
  63         parser->buf = g_string_new(NULL);
  64         parser->bufp = parser->buf->str;
  65         parser->state = SC_HTML_NORMAL;
  66         parser->href = NULL;
  67         parser->newline = TRUE;
  68         parser->empty_line = TRUE;
  69         parser->space = FALSE;
  70         parser->pre = FALSE;
  71         parser->indent = 0;
  72
  73         return parser;
  74 }
  75
  76 void sc_html_parser_destroy(SC_HTMLParser *parser)
  77 {
  78         g_string_free(parser->str, TRUE);
  79         g_string_free(parser->buf, TRUE);
  80         g_free(parser->href);
  81         g_free(parser);
  82 }
  83
  84 gchar *sc_html_parse(SC_HTMLParser *parser)
  85 {
  86         parser->state = SC_HTML_NORMAL;
  87         g_string_truncate(parser->str, 0);
  88
  89         if (*parser->bufp == '\0') {
  90                 g_string_truncate(parser->buf, 0);
  91                 parser->bufp = parser->buf->str;
  92                 if (sc_html_read_line(parser) == SC_HTML_EOF)
  93                         return NULL;
  94         }
  95
  96         while (*parser->bufp != '\0') {
  97                 switch (*parser->bufp) {
  98                 case '<': {
  99                         SC_HTMLState st;
 100                         st = sc_html_parse_tag(parser);
 101                         /* when we see an href, we need to flush the str
 102                          * buffer.  Then collect all the chars until we
 103                          * see the end anchor tag
 104                          */
 105                         if (SC_HTML_HREF_BEG == st || SC_HTML_HREF == st)
 106                                 return parser->str->str;
 107                         }
 108                         break;
 109                 case '&':
 110                         sc_html_parse_special(parser);
 111                         break;
 112                 case ' ':
 113                 case '\t':
 114                 case '\r':
 115                 case '\n':
 116                         if (parser->bufp[0] == '\r' && parser->bufp[1] == '\n')
 117                                 parser->bufp++;
 118
 119                         if (!parser->pre) {
 120                                 if (!parser->newline)
 121                                         parser->space = TRUE;
 122
 123                                 parser->bufp++;
 124                                 break;
 125                         }
 126                         /* fallthrough */
 127                 default:
 128                         sc_html_append_char(parser, *parser->bufp++);
 129                 }
 130         }
 131
 132         return parser->str->str;
 133 }
 134
 135 static SC_HTMLState sc_html_read_line(SC_HTMLParser *parser)
 136 {
 137         gchar buf[SC_HTMLBUFSIZE];
 138         gchar buf2[SC_HTMLBUFSIZE];
 139         gint index;
 140         gint n;
 141
 142         if (parser->fp == NULL)
 143                 return SC_HTML_EOF;
 144
 145         n = claws_fread(buf, 1, sizeof(buf) - 1, parser->fp);
 146         if (n == 0) {
 147                 parser->state = SC_HTML_EOF;
 148                 return SC_HTML_EOF;
 149         } else
 150                 buf[n] = '\0';
 151
 152         if (conv_convert(parser->conv, buf2, sizeof(buf2), buf) < 0) {
 153                 index = parser->bufp - parser->buf->str;
 154
 155                 conv_utf8todisp(buf2, sizeof(buf2), buf);
 156                 g_string_append(parser->buf, buf2);
 157
 158                 parser->bufp = parser->buf->str + index;
 159
 160                 return SC_HTML_CONV_FAILED;
 161         }
 162
 163         index = parser->bufp - parser->buf->str;
 164
 165         g_string_append(parser->buf, buf2);
 166
 167         parser->bufp = parser->buf->str + index;
 168
 169         return SC_HTML_NORMAL;
 170 }
 171
 172 static void sc_html_append_char(SC_HTMLParser *parser, gchar ch)
 173 {
 174         GString *str = parser->str;
 175
 176         if (!parser->pre && parser->space) {
 177                 g_string_append_c(str, ' ');
 178                 parser->space = FALSE;
 179         }
 180
 181         g_string_append_c(str, ch);
 182
 183         parser->empty_line = FALSE;
 184         if (ch == '\n') {
 185                 parser->newline = TRUE;
 186                 if (str->len > 1 && str->str[str->len - 2] == '\n')
 187                         parser->empty_line = TRUE;
 188                 if (parser->indent > 0) {
 189                         gint i, n = parser->indent;
 190                         for (i = 0; i < n; i++)
 191                                 g_string_append_c(str, '>');
 192                         g_string_append_c(str, ' ');
 193                 }
 194         } else
 195                 parser->newline = FALSE;
 196 }
 197
 198 static void sc_html_append_str(SC_HTMLParser *parser, const gchar *str, gint len)
 199 {
 200         GString *string = parser->str;
 201
 202         if (!parser->pre && parser->space) {
 203                 g_string_append_c(string, ' ');
 204                 parser->space = FALSE;
 205         }
 206
 207         if (len == 0) return;
 208         if (len < 0)
 209                 g_string_append(string, str);
 210         else {
 211                 gchar *s;
 212                 Xstrndup_a(s, str, len, return);
 213                 g_string_append(string, s);
 214         }
 215
 216         parser->empty_line = FALSE;
 217         if (string->len > 0 && string->str[string->len - 1] == '\n') {
 218                 parser->newline = TRUE;
 219                 if (string->len > 1 && string->str[string->len - 2] == '\n')
 220                         parser->empty_line = TRUE;
 221         } else
 222                 parser->newline = FALSE;
 223 }
 224
 225 static SC_HTMLTag *sc_html_get_tag(const gchar *str)
 226 {
 227         SC_HTMLTag *tag;
 228         gchar *tmp;
 229         guchar *tmpp;
 230
 231         cm_return_val_if_fail(str != NULL, NULL);
 232
 233         if (*str == '\0' || *str == '!') return NULL;
 234
 235         Xstrdup_a(tmp, str, return NULL);
 236
 237         tag = g_new0(SC_HTMLTag, 1);
 238
 239         for (tmpp = tmp; *tmpp != '\0' && !g_ascii_isspace(*tmpp); tmpp++)
 240                 ;
 241
 242         if (*tmpp == '\0') {
 243                 tag->name = g_utf8_strdown(tmp, -1);
 244                 return tag;
 245         } else {
 246                 *tmpp++ = '\0';
 247                 tag->name = g_utf8_strdown(tmp, -1);
 248         }
 249
 250         while (*tmpp != '\0') {
 251                 SC_HTMLAttr *attr;
 252                 gchar *attr_name;
 253                 gchar *attr_value;
 254                 gchar *p;
 255                 gchar quote;
 256
 257                 while (g_ascii_isspace(*tmpp)) tmpp++;
 258                 attr_name = tmpp;
 259
 260                 while (*tmpp != '\0' && !g_ascii_isspace(*tmpp) &&
 261                        *tmpp != '=')
 262                         tmpp++;
 263                 if (*tmpp != '\0' && *tmpp != '=') {
 264                         *tmpp++ = '\0';
 265                         while (g_ascii_isspace(*tmpp)) tmpp++;
 266                 }
 267
 268                 if (*tmpp == '=') {
 269                         *tmpp++ = '\0';
 270                         while (g_ascii_isspace(*tmpp)) tmpp++;
 271
 272                         if (*tmpp == '"' || *tmpp == '\'') {
 273                                 /* name="value" */
 274                                 quote = *tmpp;
 275                                 tmpp++;
 276                                 attr_value = tmpp;
 277                                 if ((p = strchr(attr_value, quote)) == NULL) {
 278                                         if (debug_get_mode()) {
 279                                                 g_warning("sc_html_get_tag(): syntax error in tag: '%s'",
 280                                                                   str);
 281                                         } else {
 282                                                 gchar *cut = g_strndup(str, 100);
 283                                                 g_warning("sc_html_get_tag(): syntax error in tag: '%s%s'",
 284                                                                   cut, strlen(str)>100?"...":".");
 285                                                 g_free(cut);
 286                                         }
 287                                         return tag;
 288                                 }
 289                                 tmpp = p;
 290                                 *tmpp++ = '\0';
 291                                 while (g_ascii_isspace(*tmpp)) tmpp++;
 292                         } else {
 293                                 /* name=value */
 294                                 attr_value = tmpp;
 295                                 while (*tmpp != '\0' && !g_ascii_isspace(*tmpp)) tmpp++;
 296                                 if (*tmpp != '\0')
 297                                         *tmpp++ = '\0';
 298                         }
 299                 } else
 300                         attr_value = "";
 301
 302                 g_strchomp(attr_name);
 303                 attr = g_new(SC_HTMLAttr, 1);
 304                 attr->name = g_utf8_strdown(attr_name, -1);
 305                 attr->value = g_strdup(attr_value);
 306                 tag->attr = g_list_append(tag->attr, attr);
 307         }
 308
 309         return tag;
 310 }
 311
 312 static void sc_html_free_tag(SC_HTMLTag *tag)
 313 {
 314         if (!tag) return;
 315
 316         g_free(tag->name);
 317         while (tag->attr != NULL) {
 318                 SC_HTMLAttr *attr = (SC_HTMLAttr *)tag->attr->data;
 319                 g_free(attr->name);
 320                 g_free(attr->value);
 321                 g_free(attr);
 322                 tag->attr = g_list_remove(tag->attr, tag->attr->data);
 323         }
 324         g_free(tag);
 325 }
 326
 327 static void decode_href(SC_HTMLParser *parser)
 328 {
 329         gchar *tmp;
 330         SC_HTMLParser *tparser = g_new0(SC_HTMLParser, 1);
 331
 332         tparser->str = g_string_new(NULL);
 333         tparser->buf = g_string_new(parser->href);
 334         tparser->bufp = tparser->buf->str;
 335
 336         tmp = sc_html_parse(tparser);
 337
 338         g_free(parser->href);
 339         parser->href = g_strdup(tmp);
 340
 341         sc_html_parser_destroy(tparser);
 342 }
 343
 344 static SC_HTMLState sc_html_parse_tag(SC_HTMLParser *parser)
 345 {
 346         gchar buf[SC_HTMLBUFSIZE];
 347         SC_HTMLTag *tag;
 348
 349         sc_html_get_parenthesis(parser, buf, sizeof(buf));
 350
 351         tag = sc_html_get_tag(buf);
 352
 353         parser->state = SC_HTML_UNKNOWN;
 354         if (!tag) return SC_HTML_UNKNOWN;
 355
 356         if (!strcmp(tag->name, "br") || !strcmp(tag->name, "br/")) {
 357                 parser->space = FALSE;
 358                 sc_html_append_char(parser, '\n');
 359                 parser->state = SC_HTML_BR;
 360         } else if (!strcmp(tag->name, "a")) {
 361                 GList *cur;
 362                 if (parser->href != NULL) {
 363                         g_free(parser->href);
 364                         parser->href = NULL;
 365                 }
 366                 for (cur = tag->attr; cur != NULL; cur = cur->next) {
 367                         if (cur->data && !strcmp(((SC_HTMLAttr *)cur->data)->name, "href")) {
 368                                 g_free(parser->href);
 369                                 parser->href = g_strdup(((SC_HTMLAttr *)cur->data)->value);
 370                                 decode_href(parser);
 371                                 parser->state = SC_HTML_HREF_BEG;
 372                                 break;
 373                         }
 374                 }
 375                 if (parser->href == NULL)
 376                         parser->href = g_strdup("");
 377                 parser->state = SC_HTML_HREF_BEG;
 378         } else if (!strcmp(tag->name, "/a")) {
 379                 parser->state = SC_HTML_HREF;
 380         } else if (!strcmp(tag->name, "p")) {
 381                 parser->space = FALSE;
 382                 if (!parser->empty_line) {
 383                         parser->space = FALSE;
 384                         if (!parser->newline) sc_html_append_char(parser, '\n');
 385                         sc_html_append_char(parser, '\n');
 386                 }
 387                 parser->state = SC_HTML_PAR;
 388         } else if (!strcmp(tag->name, "pre")) {
 389                 parser->pre = TRUE;
 390                 parser->state = SC_HTML_PRE;
 391         } else if (!strcmp(tag->name, "/pre")) {
 392                 parser->pre = FALSE;
 393                 parser->state = SC_HTML_NORMAL;
 394         } else if (!strcmp(tag->name, "hr")) {
 395                 if (!parser->newline) {
 396                         parser->space = FALSE;
 397                         sc_html_append_char(parser, '\n');
 398                 }
 399                 sc_html_append_str(parser, HR_STR, -1);
 400                 sc_html_append_char(parser, '\n');
 401                 parser->state = SC_HTML_HR;
 402         } else if (!strcmp(tag->name, "div")    ||
 403                    !strcmp(tag->name, "ul")     ||
 404                    !strcmp(tag->name, "li")     ||
 405                    !strcmp(tag->name, "table")  ||
 406                    !strcmp(tag->name, "dd")     ||
 407                    !strcmp(tag->name, "tr")) {
 408                 if (!parser->newline) {
 409                         parser->space = FALSE;
 410                         sc_html_append_char(parser, '\n');
 411                 }
 412                 if (!strcmp(tag->name, "li")) {
 413                         sc_html_append_str(parser, LI_STR, -1);
 414                 }
 415                 parser->state = SC_HTML_NORMAL;
 416         } else if (tag->name[0] == 'h' && g_ascii_isdigit(tag->name[1])) {
 417                 if (!parser->newline) {
 418                         parser->space = FALSE;
 419                         sc_html_append_char(parser, '\n');
 420                 }
 421                 sc_html_append_char(parser, '\n');
 422         } else if (!strcmp(tag->name, "blockquote")) {
 423                 parser->state = SC_HTML_NORMAL;
 424                 parser->indent++;
 425         } else if (!strcmp(tag->name, "/blockquote")) {
 426                 parser->state = SC_HTML_NORMAL;
 427                 parser->indent--;
 428         } else if (!strcmp(tag->name, "/table") ||
 429                    (tag->name[0] == '/' &&
 430                     tag->name[1] == 'h' &&
 431                     g_ascii_isdigit(tag->name[2]))) {
 432                 if (!parser->empty_line) {
 433                         parser->space = FALSE;
 434                         if (!parser->newline) sc_html_append_char(parser, '\n');
 435                         sc_html_append_char(parser, '\n');
 436                 }
 437                 parser->state = SC_HTML_NORMAL;
 438         } else if (!strcmp(tag->name, "/div")   ||
 439                    !strcmp(tag->name, "/ul")    ||
 440                    !strcmp(tag->name, "/li")) {
 441                 if (!parser->newline) {
 442                         parser->space = FALSE;
 443                         sc_html_append_char(parser, '\n');
 444                 }
 445                 parser->state = SC_HTML_NORMAL;
 446                         }
 447
 448         sc_html_free_tag(tag);
 449
 450         return parser->state;
 451 }
 452
 453 static void sc_html_parse_special(SC_HTMLParser *parser)
 454 {
 455         gchar *entity;
 456
 457         parser->state = SC_HTML_UNKNOWN;
 458         cm_return_if_fail(*parser->bufp == '&');
 459
 460         entity = entity_decode(parser->bufp);
 461         if (entity != NULL) {
 462                 sc_html_append_str(parser, entity, -1);
 463                 g_free(entity);
 464                 while (*parser->bufp++ != ';');
 465         } else {
 466                 /* output literal `&' */
 467                 sc_html_append_char(parser, *parser->bufp++);
 468         }
 469         parser->state = SC_HTML_NORMAL;
 470 }
 471
 472 static gchar *sc_html_find_tag(SC_HTMLParser *parser, const gchar *tag)
 473 {
 474         gchar *cur = parser->bufp;
 475         gint len = strlen(tag);
 476
 477         if (cur == NULL)
 478                 return NULL;
 479
 480         while ((cur = strstr(cur, "<")) != NULL) {
 481                 if (!g_ascii_strncasecmp(cur, tag, len))
 482                         return cur;
 483                 cur += 2;
 484         }
 485         return NULL;
 486 }
 487
 488 static void sc_html_get_parenthesis(SC_HTMLParser *parser, gchar *buf, gint len)
 489 {
 490         gchar *p;
 491
 492         buf[0] = '\0';
 493         cm_return_if_fail(*parser->bufp == '<');
 494
 495         /* ignore comment / CSS / script stuff */
 496         if (!strncmp(parser->bufp, "<!--", 4)) {
 497                 parser->bufp += 4;
 498                 while ((p = strstr(parser->bufp, "-->")) == NULL)
 499                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 500                 parser->bufp = p + 3;
 501                 return;
 502         }
 503         if (!g_ascii_strncasecmp(parser->bufp, "<style", 6)) {
 504                 parser->bufp += 6;
 505                 while ((p = sc_html_find_tag(parser, "</style>")) == NULL)
 506                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 507                 parser->bufp = p + 8;
 508                 return;
 509         }
 510         if (!g_ascii_strncasecmp(parser->bufp, "<script", 7)) {
 511                 parser->bufp += 7;
 512                 while ((p = sc_html_find_tag(parser, "</script>")) == NULL)
 513                         if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 514                 parser->bufp = p + 9;
 515                 return;
 516         }
 517
 518         parser->bufp++;
 519         while ((p = strchr(parser->bufp, '>')) == NULL)
 520                 if (sc_html_read_line(parser) == SC_HTML_EOF) return;
 521
 522         strncpy2(buf, parser->bufp, MIN(p - parser->bufp + 1, len));
 523         g_strstrip(buf);
 524         parser->bufp = p + 1;
 525 }