root / src / sylfilter.c @ aebfd4cc
History | View | Annotate | Download (7.5 kB)
| 1 | /* SylFilter - a message filter
|
|---|---|
| 2 | * |
| 3 | * Copyright (C) 2011 Hiroyuki Yamamoto |
| 4 | * Copyright (C) 2011 Sylpheed Development Team |
| 5 | */ |
| 6 | |
| 7 | #include <stdio.h> |
| 8 | #include <string.h> |
| 9 | #include <locale.h> |
| 10 | |
| 11 | #include "filter.h" |
| 12 | #include "filter-manager.h" |
| 13 | #include "filter-utils.h" |
| 14 | #include "filter-kvs.h" |
| 15 | |
| 16 | #define USE_QDBM 1 |
| 17 | #ifdef USE_QDBM
|
| 18 | # include "filter-kvs-qdbm.h" |
| 19 | #endif
|
| 20 | |
| 21 | #include "textcontent-filter.h" |
| 22 | #include "blacklist-filter.h" |
| 23 | #include "whitelist-filter.h" |
| 24 | #include "wordsep-filter.h" |
| 25 | #include "ngram-filter.h" |
| 26 | #include "bayes-filter.h" |
| 27 | |
| 28 | enum {
|
| 29 | MODE_TEST_JUNK, |
| 30 | MODE_LEARN_JUNK, |
| 31 | MODE_LEARN_CLEAN, |
| 32 | MODE_UNLEARN_JUNK, |
| 33 | MODE_UNLEARN_CLEAN, |
| 34 | MODE_SHOW_STATUS |
| 35 | }; |
| 36 | |
| 37 | #define DB_DIR ".sylfilter" |
| 38 | |
| 39 | |
| 40 | static int verbose = 0; |
| 41 | |
| 42 | static int learn_filter(int mode, const char *file); |
| 43 | static int test_filter(int mode, const char *file); |
| 44 | static void print_message_data(XMessageData *msgdata); |
| 45 | static void usage(void); |
| 46 | |
| 47 | |
| 48 | int main(int argc, char *argv[]) |
| 49 | {
|
| 50 | int retval = 2; |
| 51 | int i;
|
| 52 | int mode = MODE_TEST_JUNK;
|
| 53 | int count = 0; |
| 54 | char dbpath[1024]; |
| 55 | |
| 56 | setlocale(LC_ALL, "");
|
| 57 | |
| 58 | xfilter_init(); |
| 59 | |
| 60 | for (i = 1; i < argc; i++) { |
| 61 | if (!strcmp(argv[i], "-j")) |
| 62 | mode = MODE_LEARN_JUNK; |
| 63 | else if (!strcmp(argv[i], "-c")) |
| 64 | mode = MODE_LEARN_CLEAN; |
| 65 | if (!strcmp(argv[i], "-J")) |
| 66 | mode = MODE_UNLEARN_JUNK; |
| 67 | else if (!strcmp(argv[i], "-C")) |
| 68 | mode = MODE_UNLEARN_CLEAN; |
| 69 | else if (!strcmp(argv[i], "-t")) |
| 70 | mode = MODE_TEST_JUNK; |
| 71 | else if (!strcmp(argv[i], "-s")) |
| 72 | mode = MODE_SHOW_STATUS; |
| 73 | else if (!strcmp(argv[i], "-v")) |
| 74 | verbose = 1;
|
| 75 | else if (!strcmp(argv[i], "-d")) |
| 76 | xfilter_set_debug_mode(1);
|
| 77 | else if (!strcmp(argv[i], "-h")) { |
| 78 | usage(); |
| 79 | xfilter_done(); |
| 80 | return 0; |
| 81 | } |
| 82 | } |
| 83 | |
| 84 | #ifdef USE_QDBM
|
| 85 | xfilter_kvs_qdbm_set_engine(); |
| 86 | #endif
|
| 87 | |
| 88 | snprintf(dbpath, sizeof(dbpath), "%s/" DB_DIR, xfilter_utils_get_home_dir()); |
| 89 | xfilter_utils_set_base_dir(dbpath); |
| 90 | |
| 91 | if (xfilter_bayes_db_init(dbpath) < 0) { |
| 92 | fprintf(stderr, "Database initialization error.\n");
|
| 93 | xfilter_done(); |
| 94 | return 127; |
| 95 | } |
| 96 | |
| 97 | if (mode == MODE_SHOW_STATUS) {
|
| 98 | retval = xfilter_bayes_db_show_contents(); |
| 99 | } else if (mode == MODE_LEARN_JUNK || mode == MODE_LEARN_CLEAN || |
| 100 | mode == MODE_UNLEARN_JUNK || mode == MODE_UNLEARN_CLEAN) {
|
| 101 | for (i = 1; i < argc; i++) { |
| 102 | if (argv[i][0] != '-') { |
| 103 | retval = learn_filter(mode, argv[i]); |
| 104 | if (retval != 0) |
| 105 | break;
|
| 106 | count++; |
| 107 | } |
| 108 | } |
| 109 | } else {
|
| 110 | for (i = 1; i < argc; i++) { |
| 111 | if (argv[i][0] != '-') { |
| 112 | retval = test_filter(mode, argv[i]); |
| 113 | if (retval == 127) |
| 114 | break;
|
| 115 | count++; |
| 116 | } |
| 117 | } |
| 118 | } |
| 119 | |
| 120 | xfilter_bayes_db_done(); |
| 121 | xfilter_done(); |
| 122 | |
| 123 | if (count == 0) |
| 124 | fprintf(stderr, "No input file.\n");
|
| 125 | |
| 126 | if (verbose)
|
| 127 | printf("return value: %d\n", retval);
|
| 128 | |
| 129 | return retval;
|
| 130 | } |
| 131 | |
| 132 | static int learn_filter(int mode, const char *file) |
| 133 | {
|
| 134 | XFilterManager *mgr; |
| 135 | XMessageData *msgdata; |
| 136 | XMessageData *resdata; |
| 137 | XFilterResult *res; |
| 138 | XFilterStatus status; |
| 139 | int retval = 0; |
| 140 | |
| 141 | static XFilterConstructorFunc learn_junk_ctors[] = {
|
| 142 | xfilter_textcontent_new, |
| 143 | xfilter_wordsep_new, |
| 144 | xfilter_ngram_new, |
| 145 | xfilter_bayes_learn_junk_new, |
| 146 | NULL
|
| 147 | }; |
| 148 | |
| 149 | static XFilterConstructorFunc learn_nojunk_ctors[] = {
|
| 150 | xfilter_textcontent_new, |
| 151 | xfilter_wordsep_new, |
| 152 | xfilter_ngram_new, |
| 153 | xfilter_bayes_learn_nojunk_new, |
| 154 | NULL
|
| 155 | }; |
| 156 | |
| 157 | static XFilterConstructorFunc unlearn_junk_ctors[] = {
|
| 158 | xfilter_textcontent_new, |
| 159 | xfilter_wordsep_new, |
| 160 | xfilter_ngram_new, |
| 161 | xfilter_bayes_unlearn_junk_new, |
| 162 | NULL
|
| 163 | }; |
| 164 | |
| 165 | static XFilterConstructorFunc unlearn_nojunk_ctors[] = {
|
| 166 | xfilter_textcontent_new, |
| 167 | xfilter_wordsep_new, |
| 168 | xfilter_ngram_new, |
| 169 | xfilter_bayes_unlearn_nojunk_new, |
| 170 | NULL
|
| 171 | }; |
| 172 | |
| 173 | if (verbose)
|
| 174 | printf("learning message file: %s\n", file);
|
| 175 | |
| 176 | mgr = xfilter_manager_new(); |
| 177 | |
| 178 | switch (mode) {
|
| 179 | case MODE_LEARN_JUNK:
|
| 180 | xfilter_manager_add_filters(mgr, learn_junk_ctors); |
| 181 | break;
|
| 182 | case MODE_LEARN_CLEAN:
|
| 183 | xfilter_manager_add_filters(mgr, learn_nojunk_ctors); |
| 184 | break;
|
| 185 | case MODE_UNLEARN_JUNK:
|
| 186 | xfilter_manager_add_filters(mgr, unlearn_junk_ctors); |
| 187 | break;
|
| 188 | case MODE_UNLEARN_CLEAN:
|
| 189 | xfilter_manager_add_filters(mgr, unlearn_nojunk_ctors); |
| 190 | break;
|
| 191 | default:
|
| 192 | fprintf(stderr, "Internal error: invalid learn mode\n");
|
| 193 | xfilter_manager_free(mgr); |
| 194 | return 127; |
| 195 | } |
| 196 | |
| 197 | msgdata = xfilter_message_data_read_file(file, "message/rfc822");
|
| 198 | |
| 199 | res = xfilter_manager_run(mgr, msgdata); |
| 200 | if (verbose)
|
| 201 | xfilter_result_print(res); |
| 202 | status = xfilter_result_get_status(res); |
| 203 | if (status == XF_UNSUPPORTED_TYPE || status == XF_ERROR) {
|
| 204 | fprintf(stderr, "%s: Error on learning mail\n", file);
|
| 205 | retval = 127;
|
| 206 | } |
| 207 | |
| 208 | if (xfilter_get_debug_mode()) {
|
| 209 | resdata = xfilter_result_get_message_data(res); |
| 210 | print_message_data(resdata); |
| 211 | } |
| 212 | |
| 213 | xfilter_result_free(res); |
| 214 | xfilter_message_data_free(msgdata); |
| 215 | xfilter_manager_free(mgr); |
| 216 | |
| 217 | return retval;
|
| 218 | } |
| 219 | |
| 220 | static int test_filter(int mode, const char *file) |
| 221 | {
|
| 222 | XFilterManager *mgr; |
| 223 | XMessageData *msgdata; |
| 224 | XMessageData *resdata; |
| 225 | XFilterResult *res; |
| 226 | XFilterStatus status; |
| 227 | int retval = 0; |
| 228 | |
| 229 | XFilterConstructorFunc ctors[] = {
|
| 230 | xfilter_textcontent_new, |
| 231 | xfilter_blacklist_new, |
| 232 | xfilter_whitelist_new, |
| 233 | xfilter_wordsep_new, |
| 234 | xfilter_ngram_new, |
| 235 | xfilter_bayes_new, |
| 236 | NULL
|
| 237 | }; |
| 238 | |
| 239 | if (verbose)
|
| 240 | printf("testing message file: %s\n", file);
|
| 241 | |
| 242 | mgr = xfilter_manager_new(); |
| 243 | xfilter_manager_add_filters(mgr, ctors); |
| 244 | |
| 245 | msgdata = xfilter_message_data_read_file(file, "message/rfc822");
|
| 246 | |
| 247 | res = xfilter_manager_run(mgr, msgdata); |
| 248 | if (verbose)
|
| 249 | xfilter_result_print(res); |
| 250 | status = xfilter_result_get_status(res); |
| 251 | if (status == XF_JUNK) {
|
| 252 | printf("%s: This is a junk mail (prob: %f)\n", file, xfilter_result_get_probability(res));
|
| 253 | retval = 0;
|
| 254 | } else if (status == XF_UNCERTAIN) { |
| 255 | printf("%s: This mail could not be classified (prob: %f)\n", file, xfilter_result_get_probability(res));
|
| 256 | retval = 2;
|
| 257 | } else if (status == XF_UNSUPPORTED_TYPE || status == XF_ERROR) { |
| 258 | printf("%s: Error on testing mail\n", file);
|
| 259 | retval = 127;
|
| 260 | } else {
|
| 261 | printf("%s: This is a clean mail (prob: %f)\n", file, xfilter_result_get_probability(res));
|
| 262 | retval = 1;
|
| 263 | } |
| 264 | |
| 265 | if (xfilter_get_debug_mode()) {
|
| 266 | resdata = xfilter_result_get_message_data(res); |
| 267 | print_message_data(resdata); |
| 268 | } |
| 269 | |
| 270 | xfilter_result_free(res); |
| 271 | xfilter_message_data_free(msgdata); |
| 272 | |
| 273 | xfilter_manager_free(mgr); |
| 274 | |
| 275 | return retval;
|
| 276 | } |
| 277 | |
| 278 | static void print_message_data(XMessageData *msgdata) |
| 279 | {
|
| 280 | const char *content; |
| 281 | |
| 282 | if (!msgdata)
|
| 283 | return;
|
| 284 | |
| 285 | printf("\n");
|
| 286 | |
| 287 | content = xfilter_message_data_get_attribute(msgdata, XM_FROM); |
| 288 | if (content)
|
| 289 | printf("from: %s\n", content);
|
| 290 | content = xfilter_message_data_get_attribute(msgdata, XM_TO); |
| 291 | if (content)
|
| 292 | printf("to: %s\n", content);
|
| 293 | content = xfilter_message_data_get_attribute(msgdata, XM_CC); |
| 294 | if (content)
|
| 295 | printf("cc: %s\n", content);
|
| 296 | content = xfilter_message_data_get_attribute(msgdata, XM_SUBJECT); |
| 297 | if (content)
|
| 298 | printf("subject: %s\n", content);
|
| 299 | content = xfilter_message_data_get_content(msgdata); |
| 300 | printf("content: %s\n", content);
|
| 301 | } |
| 302 | |
| 303 | static void usage(void) |
| 304 | {
|
| 305 | printf("SylFilter (tentative name) version 0.0.8\n");
|
| 306 | printf("\n");
|
| 307 | printf("Usage: sylfilter [options] message [message ...]\n");
|
| 308 | printf("\n");
|
| 309 | printf("Options:\n");
|
| 310 | printf(" -j learn junk (spam) messages\n");
|
| 311 | printf(" -c learn clean (non-spam) messages\n");
|
| 312 | printf(" -J unlearn junk (spam) messages\n");
|
| 313 | printf(" -C unlearn clean (non-spam) messages\n");
|
| 314 | printf(" -t classify messages\n");
|
| 315 | printf(" -v show verbose messages\n");
|
| 316 | printf(" -d show debug messages\n");
|
| 317 | printf(" -h print this help message\n");
|
| 318 | printf("\n");
|
| 319 | printf("Database will be created at ~/.sylfilter/*.db\n");
|
| 320 | printf("\n");
|
| 321 | printf("Return values:\n");
|
| 322 | printf(" 0 junk (spam)\n");
|
| 323 | printf(" 1 clean (non-spam)\n");
|
| 324 | printf(" 2 uncertain\n");
|
| 325 | printf(" 127 other errors\n");
|
| 326 | } |