1 /* Collect URLs from CSS source.
2 Copyright (C) 1998, 2000, 2001, 2002, 2003, 2009, 2010 Free Software
5 This file is part of GNU Wget.
7 GNU Wget is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 3 of the License, or (at
10 your option) any later version.
12 GNU Wget is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with Wget. If not, see <http://www.gnu.org/licenses/>.
20 Additional permission under GNU GPL version 3 section 7
22 If you modify this program, or any covered work, by linking or
23 combining it with the OpenSSL project's OpenSSL library (or a
24 modified version of that library), containing parts covered by the
25 terms of the OpenSSL or SSLeay licenses, the Free Software Foundation
26 grants you additional permission to convey the resulting work.
27 Corresponding Source for a non-source form of such a combination
28 shall include the source code for the parts of OpenSSL used as well
29 as that of the covered work. */
32 Note that this is not an actual CSS parser, but just a lexical
33 scanner with a tiny bit more smarts bolted on top. A full parser
34 is somewhat overkill for this job. The only things we're interested
35 in are @import rules and url() tokens, so it's easy enough to
36 grab those without truly understanding the input. The only downside
37 to this is that we might be coerced into downloading files that
38 a browser would ignore. That might merit some more investigation.
57 #include "css-tokens.h"
62 typedef struct yy_buffer_state *YY_BUFFER_STATE;
63 extern YY_BUFFER_STATE yy_scan_bytes (const char *bytes,int len );
64 extern int yylex (void);
67 const char *token_names[] = {
102 Given a detected URI token, get only the URI specified within.
103 Also adjust the starting position and length of the string.
105 A URI can be specified with or without quotes, and the quotes
106 can be single or double quotes. In addition there can be
107 whitespace after the opening parenthesis and before the closing
111 get_uri_string (const char *at, int *pos, int *length)
115 strncpy(buf,at + *pos, *length);
117 DEBUGP (("get_uri_string: \"%s\"\n", buf));*/
119 if (0 != strncasecmp (at + *pos, "url(", 4))
123 *length -= 5; /* url() */
124 /* skip leading space */
125 while (isspace (at[*pos]))
128 if (--(*length) == 0)
132 /* skip trailing space */
133 while (isspace (at[*pos + *length - 1]))
137 /* trim off quotes */
138 if (at[*pos] == '\'' || at[*pos] == '"')
144 uri = xmalloc (*length + 1);
147 strncpy (uri, at + *pos, *length);
155 get_urls_css (struct map_context *ctx, int offset, int buf_length)
164 strncpy(tmp,ctx->text + offset, buf_length);
165 tmp[buf_length] = '\0';
166 DEBUGP (("get_urls_css: \"%s\"\n", tmp));
169 /* tell flex to scan from this buffer */
170 yy_scan_bytes (ctx->text + offset, buf_length);
172 while((token = yylex()) != CSSEOF)
174 /*DEBUGP (("%s ", token_names[token]));*/
176 or @import url(foo.css)
178 if(token == IMPORT_SYM)
181 buffer_pos += yyleng;
182 } while((token = yylex()) == S);
184 /*DEBUGP (("%s ", token_names[token]));*/
186 if (token == STRING || token == URI)
188 /*DEBUGP (("Got URI "));*/
189 pos = buffer_pos + offset;
194 uri = get_uri_string (ctx->text, &pos, &length);
198 /* cut out quote characters */
201 uri = xmalloc (length + 1);
202 strncpy (uri, yytext + 1, length);
208 struct urlpos *up = append_url (uri, pos, length, ctx);
209 DEBUGP (("Found @import: [%s] at %d [%s]\n", yytext, buffer_pos, uri));
213 up->link_inline_p = 1;
215 up->link_expect_css = 1;
222 /* background-image: url(foo.png)
223 note that we don't care what
224 property this is actually on.
226 else if(token == URI)
228 pos = buffer_pos + offset;
230 uri = get_uri_string (ctx->text, &pos, &length);
234 struct urlpos *up = append_url (uri, pos, length, ctx);
235 DEBUGP (("Found URI: [%s] at %d [%s]\n", yytext, buffer_pos, uri));
238 up->link_inline_p = 1;
245 buffer_pos += yyleng;
251 get_urls_css_file (const char *file, const char *url)
253 struct file_memory *fm;
254 struct map_context ctx;
257 fm = wget_read_file (file);
260 logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
263 DEBUGP (("Loaded %s (size %s).\n", file, number_to_static_string (fm->length)));
265 ctx.text = fm->content;
268 ctx.parent_base = url ? url : opt.base_href;
269 ctx.document_file = file;
272 get_urls_css (&ctx, 0, fm->length);
273 wget_read_file_free (fm);