1 /*
2 * Copyright(c) 2014-2018 Tim Ruehsen
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice shall be included in
12 * all copies or substantial portions of the Software.
13 *
14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
15 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
17 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
18 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
19 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
20 * DEALINGS IN THE SOFTWARE.
21 *
22 * This file is part of libpsl.
23 *
24 * Using the libpsl functions via command line
25 *
26 * Changelog
27 * 11.04.2014 Tim Ruehsen created
28 *
29 */
30
31 #if HAVE_CONFIG_H
32 # include <config.h>
33 #endif
34
35 #ifdef HAVE_UNISTD_H
36 # include <unistd.h>
37 #endif
38
39 #include <stdlib.h>
40 #include <string.h>
41 #include <ctype.h>
42 #include <locale.h>
43
44 #include <libpsl.h>
45
usage(int err,FILE * f)46 static void usage(int err, FILE* f)
47 {
48 fprintf(f, "Usage: psl [options] <domains...>\n");
49 fprintf(f, "\n");
50 fprintf(f, "Options:\n");
51 fprintf(f, " --version show library version information\n");
52 fprintf(f, " --use-latest-data use the latest PSL data available [default]\n");
53 fprintf(f, " --use-builtin-data use the builtin PSL data\n");
54 fprintf(f, " --no-star-rule do not apply the prevailing star rule\n");
55 fprintf(f, " (only applies to --is-public-suffix)\n");
56 fprintf(f, " --load-psl-file <filename> load PSL data from file\n");
57 fprintf(f, " --is-public-suffix check if domains are public suffixes [default]\n");
58 fprintf(f, " --is-cookie-domain-acceptable <cookie-domain>\n");
59 fprintf(f, " check if cookie-domain is acceptable for domains\n");
60 fprintf(f, " --print-unreg-domain print the longest public suffix part\n");
61 fprintf(f, " --print-reg-domain print the shortest private suffix part\n");
62 fprintf(f, " --print-info print info about library builtin data\n");
63 fprintf(f, " -b, --batch don't print leading domain\n");
64 fprintf(f, "\n");
65
66 exit(err);
67 }
68
69 /* RFC 2822-compliant date format */
time2str(time_t t)70 static const char *time2str(time_t t)
71 {
72 static char buf[64];
73 struct tm *tp = localtime(&t);
74
75 strftime(buf, sizeof(buf), "%a, %d %b %Y %H:%M:%S %Z", tp);
76 return buf;
77 }
78
main(int argc,const char * const * argv)79 int main(int argc, const char *const *argv)
80 {
81 int mode = 1, no_star_rule = 0, batch_mode = 0;
82 const char *const *arg, *psl_file = NULL, *cookie_domain = NULL;
83 psl_ctx_t *psl = (psl_ctx_t *) psl_latest(NULL);
84
85 /* set current locale according to the environment variables */
86 setlocale(LC_ALL, "");
87
88 for (arg = argv + 1; arg < argv + argc; arg++) {
89 if (**arg == '-') {
90 if (!strcmp(*arg, "--is-public-suffix"))
91 mode = 1;
92 else if (!strcmp(*arg, "--print-unreg-domain"))
93 mode = 2;
94 else if (!strcmp(*arg, "--print-reg-domain"))
95 mode = 3;
96 else if (!strcmp(*arg, "--print-info"))
97 mode = 99;
98 else if (!strcmp(*arg, "--is-cookie-domain-acceptable") && arg < argv + argc - 1) {
99 mode = 4;
100 cookie_domain = *(++arg);
101 }
102 else if (!strcmp(*arg, "--use-latest-data")) {
103 psl_free(psl);
104 if (psl_file) {
105 fprintf(stderr, "Dropped data from %s\n", psl_file);
106 psl_file = NULL;
107 }
108 if (!(psl = (psl_ctx_t *) psl_latest(NULL)))
109 printf("No PSL data available\n");
110 }
111 else if (!strcmp(*arg, "--use-builtin-data")) {
112 psl_free(psl);
113 if (psl_file) {
114 fprintf(stderr, "Dropped data from %s\n", psl_file);
115 psl_file = NULL;
116 }
117 if (!(psl = (psl_ctx_t *) psl_builtin()))
118 printf("No builtin PSL data available\n");
119 }
120 else if (!strcmp(*arg, "--no-star-rule")) {
121 no_star_rule = 1;
122 }
123 else if (!strcmp(*arg, "--load-psl-file") && arg < argv + argc - 1) {
124 psl_free(psl);
125 if (psl_file) {
126 fprintf(stderr, "Dropped data from %s\n", psl_file);
127 psl_file = NULL;
128 }
129 if (!(psl = psl_load_file(psl_file = *(++arg)))) {
130 fprintf(stderr, "Failed to load PSL data from %s\n\n", psl_file);
131 psl_file = NULL;
132 }
133 }
134 else if (!strcmp(*arg, "--batch") || !strcmp(*arg, "-b")) {
135 batch_mode = 1;
136 }
137 else if (!strcmp(*arg, "--help")) {
138 fprintf(stdout, "`psl' explores the Public Suffix List\n\n");
139 usage(0, stdout);
140 }
141 else if (!strcmp(*arg, "--version")) {
142 printf("psl %s (0x%06x)\n", PACKAGE_VERSION, psl_check_version_number(0));
143 printf("libpsl %s\n", psl_get_version());
144 printf("\n");
145 printf("Copyright (C) 2014-2018 Tim Ruehsen\n");
146 printf("License: MIT\n");
147 exit(0);
148 }
149 else if (!strcmp(*arg, "--")) {
150 arg++;
151 break;
152 }
153 else {
154 fprintf(stderr, "Unknown option '%s'\n", *arg);
155 usage(1, stderr);
156 }
157 } else
158 break;
159 }
160
161 if (mode != 99) {
162 if (mode != 1 && no_star_rule) {
163 fprintf(stderr, "--no-star-rule only combines with --is-public-suffix\n");
164 usage(1, stderr);
165 }
166 if (!psl) {
167 fprintf(stderr, "No PSL data available - aborting\n");
168 exit(2);
169 }
170 if (arg >= argv + argc) {
171 char buf[256], *domain, *lower;
172 size_t len;
173 psl_error_t rc;
174
175 /* read URLs from STDIN */
176 while (fgets(buf, sizeof(buf), stdin)) {
177 for (domain = buf; isspace(*domain); domain++); /* skip leading spaces */
178 if (*domain == '#' || !*domain) continue; /* skip empty lines and comments */
179 for (len = strlen(domain); len && isspace(domain[len - 1]); len--); /* skip trailing spaces */
180 domain[len] = 0;
181
182 if ((rc = psl_str_to_utf8lower(domain, NULL, NULL, &lower)) != PSL_SUCCESS) {
183 fprintf(stderr, "%s: Failed to convert to lowercase UTF-8 (%d)\n", domain, rc);
184 continue;
185 }
186
187 if (!batch_mode && mode != 4)
188 printf("%s: ", domain);
189
190 if (mode == 1) {
191 if (no_star_rule)
192 printf("%d", psl_is_public_suffix2(psl, lower, PSL_TYPE_ANY|PSL_TYPE_NO_STAR_RULE));
193 else
194 printf("%d", psl_is_public_suffix(psl, lower));
195
196 if (!batch_mode)
197 printf(" (%s)\n", lower);
198 else
199 putchar('\n');
200 }
201 else if (mode == 2) {
202 const char *dom = psl_unregistrable_domain(psl, lower);
203 printf("%s\n", dom ? dom : "(null)");
204 }
205 else if (mode == 3) {
206 const char *dom = psl_registrable_domain(psl, lower);
207 printf("%s\n", dom ? dom : "(null)");
208 }
209 else if (mode == 4) {
210 char *cookie_domain_lower;
211
212 if ((rc = psl_str_to_utf8lower(domain, NULL, NULL, &cookie_domain_lower)) == PSL_SUCCESS) {
213 if (!batch_mode)
214 printf("%s: ", domain);
215 printf("%d\n", psl_is_cookie_domain_acceptable(psl, lower, cookie_domain));
216 free(cookie_domain_lower);
217 } else
218 fprintf(stderr, "%s: Failed to convert cookie domain '%s' to lowercase UTF-8 (%d)\n", domain, cookie_domain, rc);
219 }
220
221 psl_free_string(lower);
222 }
223
224 psl_free(psl);
225 exit(0);
226 }
227 }
228
229 if (mode == 1) {
230 for (; arg < argv + argc; arg++) {
231 if (!batch_mode)
232 printf("%s: ", *arg);
233 if (no_star_rule)
234 printf("%d\n", psl_is_public_suffix2(psl, *arg, PSL_TYPE_ANY|PSL_TYPE_NO_STAR_RULE));
235 else
236 printf("%d\n", psl_is_public_suffix(psl, *arg));
237 }
238 }
239 else if (mode == 2) {
240 for (; arg < argv + argc; arg++) {
241 const char *dom = psl_unregistrable_domain(psl, *arg);
242 if (!batch_mode)
243 printf("%s: ", *arg);
244 printf("%s\n", dom ? dom : "(null)");
245 }
246 }
247 else if (mode == 3) {
248 for (; arg < argv + argc; arg++) {
249 const char *dom = psl_registrable_domain(psl, *arg);
250 if (!batch_mode)
251 printf("%s: ", *arg);
252 printf("%s\n", dom ? dom : "(null)");
253 }
254 }
255 else if (mode == 4) {
256 for (; arg < argv + argc; arg++) {
257 if (!batch_mode)
258 printf("%s: ", *arg);
259 printf("%d\n", psl_is_cookie_domain_acceptable(psl, *arg, cookie_domain));
260 }
261 }
262 else if (mode == 99) {
263 printf("dist filename: %s\n", psl_dist_filename());
264
265 if (psl && psl != psl_builtin()) {
266 static char not_avail[] = "- information not available -";
267 int n;
268
269 if ((n = psl_suffix_count(psl)) >= 0)
270 printf("suffixes: %d\n", n);
271 else
272 printf("suffixes: %s\n", not_avail);
273
274 if ((n = psl_suffix_exception_count(psl)) >= 0)
275 printf("exceptions: %d\n", n);
276 else
277 printf("exceptions: %s\n", not_avail);
278
279 if ((n = psl_suffix_wildcard_count(psl)) >= 0)
280 printf("wildcards: %d\n", n);
281 else
282 printf("wildcards: %s\n", not_avail);
283 }
284
285 psl_free(psl);
286 psl = (psl_ctx_t *) psl_builtin();
287
288 if (psl) {
289 printf("builtin suffixes: %d\n", psl_suffix_count(psl));
290 printf("builtin exceptions: %d\n", psl_suffix_exception_count(psl));
291 printf("builtin wildcards: %d\n", psl_suffix_wildcard_count(psl));
292 printf("builtin filename: %s\n", psl_builtin_filename());
293 printf("builtin file time: %ld (%s)\n", (long) psl_builtin_file_time(), time2str(psl_builtin_file_time()));
294 printf("builtin SHA1 file hash: %s\n", psl_builtin_sha1sum());
295 printf("builtin outdated: %d\n", psl_builtin_outdated());
296 } else
297 printf("No builtin PSL data available\n");
298 }
299
300 psl_free(psl);
301
302 return 0;
303 }
304