1 /* Collect URLs from CSS source.
2 Copyright (C) 1998, 2000, 2001, 2002, 2003, 2009, 2010, 2011 Free
3 Software Foundation, Inc.
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"
63 typedef struct yy_buffer_state *YY_BUFFER_STATE;
64 extern YY_BUFFER_STATE yy_scan_bytes (const char *bytes,int len );
65 extern int yylex (void);
68 Given a detected URI token, get only the URI specified within.
69 Also adjust the starting position and length of the string.
71 A URI can be specified with or without quotes, and the quotes
72 can be single or double quotes. In addition there can be
73 whitespace after the opening parenthesis and before the closing
77 get_uri_string (const char *at, int *pos, int *length)
81 strncpy(buf,at + *pos, *length);
83 DEBUGP (("get_uri_string: \"%s\"\n", buf));*/
85 if (0 != strncasecmp (at + *pos, "url(", 4))
89 *length -= 5; /* url() */
90 /* skip leading space */
91 while (isspace (at[*pos]))
98 /* skip trailing space */
99 while (isspace (at[*pos + *length - 1]))
103 /* trim off quotes */
104 if (at[*pos] == '\'' || at[*pos] == '"')
110 uri = xmalloc (*length + 1);
113 strncpy (uri, at + *pos, *length);
121 get_urls_css (struct map_context *ctx, int offset, int buf_length)
130 strncpy(tmp,ctx->text + offset, buf_length);
131 tmp[buf_length] = '\0';
132 DEBUGP (("get_urls_css: \"%s\"\n", tmp));
135 /* tell flex to scan from this buffer */
136 yy_scan_bytes (ctx->text + offset, buf_length);
138 while((token = yylex()) != CSSEOF)
140 /*DEBUGP (("%s ", token_names[token]));*/
142 or @import url(foo.css)
144 if(token == IMPORT_SYM)
147 buffer_pos += yyleng;
148 } while((token = yylex()) == S);
150 /*DEBUGP (("%s ", token_names[token]));*/
152 if (token == STRING || token == URI)
154 /*DEBUGP (("Got URI "));*/
155 pos = buffer_pos + offset;
160 uri = get_uri_string (ctx->text, &pos, &length);
164 /* cut out quote characters */
167 uri = xmalloc (length + 1);
168 strncpy (uri, yytext + 1, length);
174 struct urlpos *up = append_url (uri, pos, length, ctx);
175 DEBUGP (("Found @import: [%s] at %d [%s]\n", yytext, buffer_pos, uri));
179 up->link_inline_p = 1;
181 up->link_expect_css = 1;
188 /* background-image: url(foo.png)
189 note that we don't care what
190 property this is actually on.
192 else if(token == URI)
194 pos = buffer_pos + offset;
196 uri = get_uri_string (ctx->text, &pos, &length);
200 struct urlpos *up = append_url (uri, pos, length, ctx);
201 DEBUGP (("Found URI: [%s] at %d [%s]\n", yytext, buffer_pos, uri));
204 up->link_inline_p = 1;
211 buffer_pos += yyleng;
217 get_urls_css_file (const char *file, const char *url)
219 struct file_memory *fm;
220 struct map_context ctx;
223 fm = wget_read_file (file);
226 logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
229 DEBUGP (("Loaded %s (size %s).\n", file, number_to_static_string (fm->length)));
231 ctx.text = fm->content;
234 ctx.parent_base = url ? url : opt.base_href;
235 ctx.document_file = file;
238 get_urls_css (&ctx, 0, fm->length);
239 wget_read_file_free (fm);