| 1 | /* $Id: read_config.c,v 1.197 2011/06/28 00:13:48 sbajic Exp $ */ |
|---|
| 2 | |
|---|
| 3 | /* |
|---|
| 4 | DSPAM |
|---|
| 5 | COPYRIGHT (C) 2002-2012 DSPAM PROJECT |
|---|
| 6 | |
|---|
| 7 | This program is free software: you can redistribute it and/or modify |
|---|
| 8 | it under the terms of the GNU Affero General Public License as |
|---|
| 9 | published by the Free Software Foundation, either version 3 of the |
|---|
| 10 | License, or (at your option) any later version. |
|---|
| 11 | |
|---|
| 12 | This program 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 Affero General Public License for more details. |
|---|
| 16 | |
|---|
| 17 | You should have received a copy of the GNU Affero General Public License |
|---|
| 18 | along with this program. If not, see <http://www.gnu.org/licenses/>. |
|---|
| 19 | |
|---|
| 20 | */ |
|---|
| 21 | |
|---|
| 22 | #ifdef HAVE_CONFIG_H |
|---|
| 23 | #include <auto-config.h> |
|---|
| 24 | #endif |
|---|
| 25 | |
|---|
| 26 | #include <stdio.h> |
|---|
| 27 | #include <errno.h> |
|---|
| 28 | #include <stdlib.h> |
|---|
| 29 | #ifdef SPLIT_CONFIG |
|---|
| 30 | #include <dirent.h> |
|---|
| 31 | #endif |
|---|
| 32 | #ifdef HAVE_STRINGS_H |
|---|
| 33 | #include <strings.h> |
|---|
| 34 | #endif |
|---|
| 35 | #include <string.h> |
|---|
| 36 | |
|---|
| 37 | #include "config_shared.h" |
|---|
| 38 | #include "read_config.h" |
|---|
| 39 | #include "config.h" |
|---|
| 40 | #include "error.h" |
|---|
| 41 | #include "language.h" |
|---|
| 42 | #include "libdspam.h" |
|---|
| 43 | #include "pref.h" |
|---|
| 44 | #include "util.h" |
|---|
| 45 | |
|---|
| 46 | #ifdef SPLIT_CONFIG |
|---|
| 47 | long dirread(const char *path, config_t *attrib, long num_root); |
|---|
| 48 | long fileread(const char *path, config_t *attrib, long num_root); |
|---|
| 49 | #endif |
|---|
| 50 | |
|---|
| 51 | static char *next_normal_token(char **p) |
|---|
| 52 | { |
|---|
| 53 | char *start = *p; |
|---|
| 54 | while (**p) { |
|---|
| 55 | if (**p == ' ' || **p == '\t') { |
|---|
| 56 | **p = 0; |
|---|
| 57 | *p += 1; |
|---|
| 58 | break; |
|---|
| 59 | } |
|---|
| 60 | *p += 1; |
|---|
| 61 | } |
|---|
| 62 | return start; |
|---|
| 63 | } |
|---|
| 64 | |
|---|
| 65 | static char *next_quoted_token(char **p) |
|---|
| 66 | { |
|---|
| 67 | char *start = *p; |
|---|
| 68 | while (**p) { |
|---|
| 69 | if (**p == '\"') { |
|---|
| 70 | **p = 0; |
|---|
| 71 | *p += 1; |
|---|
| 72 | break; |
|---|
| 73 | } |
|---|
| 74 | *p += 1; |
|---|
| 75 | } |
|---|
| 76 | return start; |
|---|
| 77 | } |
|---|
| 78 | |
|---|
| 79 | static char *tokenize(char *text, char **next) |
|---|
| 80 | { |
|---|
| 81 | /* Initialize */ |
|---|
| 82 | if (text) |
|---|
| 83 | *next = text; |
|---|
| 84 | |
|---|
| 85 | while (**next) { |
|---|
| 86 | /* Skip leading whitespace */ |
|---|
| 87 | if (**next == ' ' || **next == '\t') { |
|---|
| 88 | *next += 1; |
|---|
| 89 | continue; |
|---|
| 90 | } |
|---|
| 91 | |
|---|
| 92 | /* Strip off one token */ |
|---|
| 93 | if (**next == '\"') { |
|---|
| 94 | *next += 1; |
|---|
| 95 | return next_quoted_token(next); |
|---|
| 96 | } else |
|---|
| 97 | return next_normal_token(next); |
|---|
| 98 | } |
|---|
| 99 | |
|---|
| 100 | return NULL; |
|---|
| 101 | } |
|---|
| 102 | |
|---|
| 103 | #ifdef SPLIT_CONFIG |
|---|
| 104 | // Read the files in the directory and pass it to fileread |
|---|
| 105 | // or if it is a file, pass it to fileread. |
|---|
| 106 | long dirread(const char *path, config_t *attrib, long num_root) { |
|---|
| 107 | DIR *dir_p; |
|---|
| 108 | char *fulldir = NULL; |
|---|
| 109 | struct dirent *dir_entry_p; |
|---|
| 110 | |
|---|
| 111 | if (path == NULL) return 0; |
|---|
| 112 | |
|---|
| 113 | // Strip "\n" |
|---|
| 114 | char *ptr = strrchr(path, '\n'); |
|---|
| 115 | if (ptr) |
|---|
| 116 | *ptr = '\0'; |
|---|
| 117 | |
|---|
| 118 | if ((dir_p = opendir(path)) == NULL) { |
|---|
| 119 | LOG(LOG_ERR, ERR_IO_FILE_OPEN, path, strerror(errno)); |
|---|
| 120 | return 0; |
|---|
| 121 | } |
|---|
| 122 | |
|---|
| 123 | if ((dir_p = opendir(path))) { |
|---|
| 124 | while((dir_entry_p = readdir(dir_p))) |
|---|
| 125 | { |
|---|
| 126 | // We don't need the . and .. |
|---|
| 127 | if (strcmp(dir_entry_p->d_name, ".") == 0 || |
|---|
| 128 | strcmp(dir_entry_p->d_name, "..") == 0) |
|---|
| 129 | continue; |
|---|
| 130 | |
|---|
| 131 | int n = strlen(dir_entry_p->d_name); |
|---|
| 132 | |
|---|
| 133 | // only use files which end in .conf: |
|---|
| 134 | if (n < 5) continue; |
|---|
| 135 | if (strncmp(dir_entry_p->d_name + n - 5, ".conf", 5) != 0) continue; |
|---|
| 136 | |
|---|
| 137 | int m = strlen(path); |
|---|
| 138 | |
|---|
| 139 | fulldir = (char *)malloc(n+m+2); |
|---|
| 140 | if (fulldir == NULL) { |
|---|
| 141 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 142 | return 0; |
|---|
| 143 | } |
|---|
| 144 | |
|---|
| 145 | strcpy(fulldir, (char *)path); |
|---|
| 146 | strcat(fulldir, "/"); |
|---|
| 147 | strcat(fulldir, dir_entry_p->d_name); |
|---|
| 148 | num_root = fileread((const char *)fulldir, attrib, num_root); |
|---|
| 149 | free(fulldir); |
|---|
| 150 | } |
|---|
| 151 | closedir(dir_p); |
|---|
| 152 | } else { |
|---|
| 153 | // Could be a file. |
|---|
| 154 | return fileread((const char *)path, attrib, num_root); |
|---|
| 155 | } |
|---|
| 156 | |
|---|
| 157 | return num_root; |
|---|
| 158 | } |
|---|
| 159 | |
|---|
| 160 | // Read the file and check if there is an Include directive, if so then pass |
|---|
| 161 | // it to dirread. |
|---|
| 162 | long fileread(const char *path, config_t *attrib, long num_root) { |
|---|
| 163 | config_t ptr; |
|---|
| 164 | #else |
|---|
| 165 | config_t read_config(const char *path) { |
|---|
| 166 | config_t attrib, ptr; |
|---|
| 167 | #endif |
|---|
| 168 | FILE *file; |
|---|
| 169 | #ifdef SPLIT_CONFIG |
|---|
| 170 | long attrib_size = 128; |
|---|
| 171 | #else |
|---|
| 172 | long attrib_size = 128, num_root = 0; |
|---|
| 173 | #endif |
|---|
| 174 | char buffer[1024]; |
|---|
| 175 | char *a, *c, *v, *bufptr = buffer; |
|---|
| 176 | |
|---|
| 177 | #ifndef SPLIT_CONFIG |
|---|
| 178 | attrib = calloc(1, attrib_size*sizeof(attribute_t)); |
|---|
| 179 | if (attrib == NULL) { |
|---|
| 180 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 181 | return NULL; |
|---|
| 182 | } |
|---|
| 183 | #endif |
|---|
| 184 | |
|---|
| 185 | if (path == NULL) |
|---|
| 186 | file = fopen(CONFIG_DEFAULT, "r"); |
|---|
| 187 | else |
|---|
| 188 | file = fopen(path, "r"); |
|---|
| 189 | |
|---|
| 190 | if (file == NULL) { |
|---|
| 191 | #ifdef SPLIT_CONFIG |
|---|
| 192 | if (path == NULL) { |
|---|
| 193 | LOG(LOG_ERR, ERR_IO_FILE_OPEN, CONFIG_DEFAULT, strerror(errno)); |
|---|
| 194 | } else { |
|---|
| 195 | LOG(LOG_ERR, ERR_IO_FILE_OPEN, path, strerror(errno)); |
|---|
| 196 | } |
|---|
| 197 | return 0; |
|---|
| 198 | #else |
|---|
| 199 | LOG(LOG_ERR, ERR_IO_FILE_OPEN, CONFIG_DEFAULT, strerror(errno)); |
|---|
| 200 | free(attrib); |
|---|
| 201 | return NULL; |
|---|
| 202 | #endif |
|---|
| 203 | } |
|---|
| 204 | |
|---|
| 205 | while(fgets(buffer, sizeof(buffer), file)!=NULL) { |
|---|
| 206 | |
|---|
| 207 | chomp(buffer); |
|---|
| 208 | |
|---|
| 209 | /* Remove comments */ |
|---|
| 210 | if ((c = strchr(buffer, '#')) || (c = strchr(buffer, ';'))) |
|---|
| 211 | *c = 0; |
|---|
| 212 | |
|---|
| 213 | /* Parse attribute name */ |
|---|
| 214 | if (!(a = tokenize(buffer, &bufptr))) |
|---|
| 215 | continue; /* Ignore whitespace-only lines */ |
|---|
| 216 | |
|---|
| 217 | while ((v = tokenize(NULL, &bufptr)) != NULL) { |
|---|
| 218 | #ifdef SPLIT_CONFIG |
|---|
| 219 | // Check for include directive |
|---|
| 220 | if (strcmp(a, "Include") == 0) { |
|---|
| 221 | // Give v (value) to dirraed |
|---|
| 222 | num_root = dirread(v, attrib, num_root); |
|---|
| 223 | } else { |
|---|
| 224 | if (_ds_find_attribute((*attrib), a)!=NULL) { |
|---|
| 225 | _ds_add_attribute((*attrib), a, v); |
|---|
| 226 | } |
|---|
| 227 | else { |
|---|
| 228 | num_root++; |
|---|
| 229 | if (num_root >= attrib_size) { |
|---|
| 230 | attrib_size *=2; |
|---|
| 231 | ptr = realloc((*attrib), attrib_size*sizeof(attribute_t)); |
|---|
| 232 | if (ptr) |
|---|
| 233 | *attrib = ptr; |
|---|
| 234 | else { |
|---|
| 235 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 236 | fclose(file); |
|---|
| 237 | return 0; |
|---|
| 238 | } |
|---|
| 239 | } |
|---|
| 240 | _ds_add_attribute((*attrib), a, v); |
|---|
| 241 | } |
|---|
| 242 | #else |
|---|
| 243 | if (_ds_find_attribute(attrib, a)!=NULL) { |
|---|
| 244 | _ds_add_attribute(attrib, a, v); |
|---|
| 245 | } |
|---|
| 246 | else { |
|---|
| 247 | num_root++; |
|---|
| 248 | if (num_root >= attrib_size) { |
|---|
| 249 | attrib_size *=2; |
|---|
| 250 | ptr = realloc(attrib, attrib_size*sizeof(attribute_t)); |
|---|
| 251 | if (ptr) |
|---|
| 252 | attrib = ptr; |
|---|
| 253 | else { |
|---|
| 254 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 255 | fclose(file); |
|---|
| 256 | return NULL; |
|---|
| 257 | } |
|---|
| 258 | } |
|---|
| 259 | _ds_add_attribute(attrib, a, v); |
|---|
| 260 | #endif |
|---|
| 261 | } |
|---|
| 262 | } |
|---|
| 263 | } |
|---|
| 264 | |
|---|
| 265 | fclose(file); |
|---|
| 266 | |
|---|
| 267 | #ifdef SPLIT_CONFIG |
|---|
| 268 | return num_root; |
|---|
| 269 | } |
|---|
| 270 | |
|---|
| 271 | config_t read_config(const char *path) { |
|---|
| 272 | config_t attrib; |
|---|
| 273 | long attrib_size = 128, num_root = 0; |
|---|
| 274 | |
|---|
| 275 | attrib = calloc(1, attrib_size*sizeof(attribute_t)); |
|---|
| 276 | if (attrib == NULL) { |
|---|
| 277 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 278 | return NULL; |
|---|
| 279 | } |
|---|
| 280 | |
|---|
| 281 | if (fileread(path, &attrib, num_root) == 0) |
|---|
| 282 | return NULL; |
|---|
| 283 | #else |
|---|
| 284 | ptr = realloc(attrib, ((num_root+1)*sizeof(attribute_t))+1); |
|---|
| 285 | if (ptr) |
|---|
| 286 | return ptr; |
|---|
| 287 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 288 | #endif |
|---|
| 289 | |
|---|
| 290 | return attrib; |
|---|
| 291 | } |
|---|
| 292 | |
|---|
| 293 | int configure_algorithms(DSPAM_CTX *CTX) { |
|---|
| 294 | if (_ds_read_attribute(agent_config, "Algorithm")) |
|---|
| 295 | CTX->algorithms = 0; |
|---|
| 296 | |
|---|
| 297 | if (_ds_match_attribute(agent_config, "Algorithm", "graham")) |
|---|
| 298 | CTX->algorithms |= DSA_GRAHAM; |
|---|
| 299 | |
|---|
| 300 | if (_ds_match_attribute(agent_config, "Algorithm", "burton")) |
|---|
| 301 | CTX->algorithms |= DSA_BURTON; |
|---|
| 302 | |
|---|
| 303 | if (_ds_match_attribute(agent_config, "Algorithm", "robinson")) |
|---|
| 304 | CTX->algorithms |= DSA_ROBINSON; |
|---|
| 305 | |
|---|
| 306 | if (_ds_match_attribute(agent_config, "Algorithm", "naive")) |
|---|
| 307 | CTX->algorithms |= DSA_NAIVE; |
|---|
| 308 | |
|---|
| 309 | if (_ds_match_attribute(agent_config, "PValue", "robinson")) |
|---|
| 310 | CTX->algorithms |= DSP_ROBINSON; |
|---|
| 311 | else if (_ds_match_attribute(agent_config, "PValue", "markov")) |
|---|
| 312 | CTX->algorithms |= DSP_MARKOV; |
|---|
| 313 | else |
|---|
| 314 | CTX->algorithms |= DSP_GRAHAM; |
|---|
| 315 | |
|---|
| 316 | if (_ds_match_attribute(agent_config, "Tokenizer", "word")) |
|---|
| 317 | CTX->tokenizer = DSZ_WORD; |
|---|
| 318 | else if (_ds_match_attribute(agent_config, "Tokenizer", "chain") || |
|---|
| 319 | _ds_match_attribute(agent_config, "Tokenizer", "chained")) |
|---|
| 320 | CTX->tokenizer = DSZ_CHAIN; |
|---|
| 321 | else if (_ds_match_attribute(agent_config, "Tokenizer", "sbph")) |
|---|
| 322 | CTX->tokenizer = DSZ_SBPH; |
|---|
| 323 | else if (_ds_match_attribute(agent_config, "Tokenizer", "osb")) |
|---|
| 324 | CTX->tokenizer = DSZ_OSB; |
|---|
| 325 | |
|---|
| 326 | if (_ds_match_attribute(agent_config, "Algorithm", "chi-square")) |
|---|
| 327 | { |
|---|
| 328 | if (CTX->algorithms != 0 && CTX->algorithms != DSP_ROBINSON) { |
|---|
| 329 | LOG(LOG_WARNING, "Warning: Chi-Square algorithm enabled with other algorithms. False positives may ensue."); |
|---|
| 330 | } |
|---|
| 331 | CTX->algorithms |= DSA_CHI_SQUARE; |
|---|
| 332 | } |
|---|
| 333 | |
|---|
| 334 | return 0; |
|---|
| 335 | } |
|---|
| 336 | |
|---|
| 337 | agent_pref_t pref_config(void) |
|---|
| 338 | { |
|---|
| 339 | agent_pref_t PTX = malloc(sizeof(agent_attrib_t)*PREF_MAX); |
|---|
| 340 | agent_pref_t ptr; |
|---|
| 341 | attribute_t attrib; |
|---|
| 342 | char *p, *q; |
|---|
| 343 | char *ptrptr = NULL; |
|---|
| 344 | int i = 0; |
|---|
| 345 | |
|---|
| 346 | if (PTX == NULL) { |
|---|
| 347 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 348 | return NULL; |
|---|
| 349 | } |
|---|
| 350 | PTX[0] = NULL; |
|---|
| 351 | |
|---|
| 352 | /* Apply default preferences from dspam.conf */ |
|---|
| 353 | |
|---|
| 354 | attrib = _ds_find_attribute(agent_config, "Preference"); |
|---|
| 355 | |
|---|
| 356 | LOGDEBUG("Loading preferences from dspam.conf"); |
|---|
| 357 | |
|---|
| 358 | while(attrib != NULL) { |
|---|
| 359 | char *pcopy = strdup(attrib->value); |
|---|
| 360 | |
|---|
| 361 | p = strtok_r(pcopy, "=", &ptrptr); |
|---|
| 362 | if (p == NULL) { |
|---|
| 363 | free(pcopy); |
|---|
| 364 | continue; |
|---|
| 365 | } |
|---|
| 366 | q = p + strlen(p)+1; |
|---|
| 367 | |
|---|
| 368 | PTX[i] = _ds_pref_new(p, q); |
|---|
| 369 | PTX[i+1] = NULL; |
|---|
| 370 | |
|---|
| 371 | i++; |
|---|
| 372 | attrib = attrib->next; |
|---|
| 373 | free(pcopy); |
|---|
| 374 | } |
|---|
| 375 | |
|---|
| 376 | ptr = realloc(PTX, sizeof(agent_attrib_t)*(i+1)); |
|---|
| 377 | if (ptr) |
|---|
| 378 | return ptr; |
|---|
| 379 | |
|---|
| 380 | LOG(LOG_CRIT, ERR_MEM_ALLOC); |
|---|
| 381 | return PTX; |
|---|
| 382 | } |
|---|