| 1 | /* -*-c-*- |
| 2 | * |
| 3 | * $Id: mkphrase.c,v 1.2 2000/12/06 20:33:27 mdw Exp $ |
| 4 | * |
| 5 | * Generate passphrases from word lists |
| 6 | * |
| 7 | * (c) 2000 Straylight/Edgeware |
| 8 | */ |
| 9 | |
| 10 | /*----- Licensing notice --------------------------------------------------* |
| 11 | * |
| 12 | * This file is part of Catacomb. |
| 13 | * |
| 14 | * Catacomb is free software; you can redistribute it and/or modify |
| 15 | * it under the terms of the GNU Library General Public License as |
| 16 | * published by the Free Software Foundation; either version 2 of the |
| 17 | * License, or (at your option) any later version. |
| 18 | * |
| 19 | * Catacomb is distributed in the hope that it will be useful, |
| 20 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 21 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 22 | * GNU Library General Public License for more details. |
| 23 | * |
| 24 | * You should have received a copy of the GNU Library General Public |
| 25 | * License along with Catacomb; if not, write to the Free |
| 26 | * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, |
| 27 | * MA 02111-1307, USA. |
| 28 | */ |
| 29 | |
| 30 | /*----- Revision history --------------------------------------------------* |
| 31 | * |
| 32 | * $Log: mkphrase.c,v $ |
| 33 | * Revision 1.2 2000/12/06 20:33:27 mdw |
| 34 | * Make flags be macros rather than enumerations, to ensure that they're |
| 35 | * unsigned. |
| 36 | * |
| 37 | * Revision 1.1 2000/08/06 10:50:55 mdw |
| 38 | * (mkphrase): New program for generating random passphrases with measured |
| 39 | * strength. |
| 40 | * |
| 41 | */ |
| 42 | |
| 43 | /*----- Header files ------------------------------------------------------*/ |
| 44 | |
| 45 | #include "config.h" |
| 46 | |
| 47 | #include <ctype.h> |
| 48 | #include <errno.h> |
| 49 | #include <math.h> |
| 50 | #include <stdio.h> |
| 51 | #include <stdlib.h> |
| 52 | #include <string.h> |
| 53 | |
| 54 | #include <mLib/alloc.h> |
| 55 | #include <mLib/bits.h> |
| 56 | #include <mLib/darray.h> |
| 57 | #include <mLib/dstr.h> |
| 58 | #include <mLib/mdwopt.h> |
| 59 | #include <mLib/quis.h> |
| 60 | #include <mLib/report.h> |
| 61 | #include <mLib/sym.h> |
| 62 | |
| 63 | #include "grand.h" |
| 64 | #include "noise.h" |
| 65 | #include "rand.h" |
| 66 | |
| 67 | /*----- Global state ------------------------------------------------------*/ |
| 68 | |
| 69 | static unsigned min = 0, max = 256; /* Word length bounds */ |
| 70 | static unsigned bits = 128; /* Minimum acceptable entropy */ |
| 71 | static unsigned count = 1; /* How many passphrases to make */ |
| 72 | |
| 73 | static const char wchars[] = "abcdefghijklmnopqrstuvwxyz'"; |
| 74 | |
| 75 | typedef struct ppgen_ops { |
| 76 | const char *name; /* Name of the generator */ |
| 77 | void *(*init)(void); /* Initialize generator */ |
| 78 | void (*scan)(FILE */*fp*/, void */*p*/); /* Scan an input word list */ |
| 79 | void (*endscan)(void */*p*/); /* Scanning phase completed */ |
| 80 | double (*gen)(dstr */*d*/, grand */*r*/, void */*p*/); |
| 81 | /* Emit word and return entropy */ |
| 82 | void (*done)(void */*p*/); /* Close down generator */ |
| 83 | } ppgen_ops; |
| 84 | |
| 85 | /*----- Word list ---------------------------------------------------------*/ |
| 86 | |
| 87 | #ifndef STRING_V |
| 88 | # define STRING_V |
| 89 | DA_DECL(string_v, char *); |
| 90 | #endif |
| 91 | |
| 92 | typedef struct wlist { |
| 93 | string_v sv; |
| 94 | sym_table tab; |
| 95 | char *buf; |
| 96 | double logp; |
| 97 | } wlist; |
| 98 | |
| 99 | static void *wordlist_init(void) |
| 100 | { |
| 101 | wlist *w = xmalloc(sizeof(wlist)); |
| 102 | sym_create(&w->tab); |
| 103 | w->logp = 0; |
| 104 | return (w); |
| 105 | } |
| 106 | |
| 107 | static void wordlist_scan(FILE *fp, void *p) |
| 108 | { |
| 109 | wlist *w = p; |
| 110 | dstr d = DSTR_INIT; |
| 111 | unsigned f = 0; |
| 112 | |
| 113 | for (;;) { |
| 114 | int ch = getc(fp); |
| 115 | if (ch == EOF || isspace(ch)) { |
| 116 | DPUTZ(&d); |
| 117 | if (f && d.len >= min && d.len <= max) |
| 118 | sym_find(&w->tab, d.buf, d.len + 1, sizeof(sym_base), 0); |
| 119 | f = 0; |
| 120 | DRESET(&d); |
| 121 | if (ch == EOF) |
| 122 | break; |
| 123 | continue; |
| 124 | } |
| 125 | ch = tolower(ch); |
| 126 | if (strchr(wchars, ch)) { |
| 127 | DPUTC(&d, ch); |
| 128 | f = 1; |
| 129 | } |
| 130 | } |
| 131 | |
| 132 | dstr_destroy(&d); |
| 133 | } |
| 134 | |
| 135 | static void wordlist_endscan(void *p) |
| 136 | { |
| 137 | wlist *w = p; |
| 138 | size_t buflen = 0; |
| 139 | sym_iter i; |
| 140 | sym_base *b; |
| 141 | char *q; |
| 142 | |
| 143 | for (sym_mkiter(&i, &w->tab); (b = sym_next(&i)) != 0; ) |
| 144 | buflen += b->len; |
| 145 | w->buf = xmalloc(buflen); |
| 146 | q = w->buf; |
| 147 | DA_CREATE(&w->sv); |
| 148 | for (sym_mkiter(&i, &w->tab); (b = sym_next(&i)) != 0; ) { |
| 149 | memcpy(q, SYM_NAME(b), b->len); |
| 150 | DA_PUSH(&w->sv, q); |
| 151 | q += b->len; |
| 152 | } |
| 153 | sym_destroy(&w->tab); |
| 154 | w->logp = log(DA_LEN(&w->sv))/log(2); |
| 155 | } |
| 156 | |
| 157 | static double wordlist_gen(dstr *d, grand *r, void *p) |
| 158 | { |
| 159 | wlist *w = p; |
| 160 | uint32 i = r->ops->range(r, DA_LEN(&w->sv)); |
| 161 | DPUTS(d, DA(&w->sv)[i]); |
| 162 | return (w->logp); |
| 163 | } |
| 164 | |
| 165 | static void wordlist_done(void *p) |
| 166 | { |
| 167 | wlist *w = p; |
| 168 | xfree(w->buf); |
| 169 | DA_DESTROY(&w->sv); |
| 170 | xfree(w); |
| 171 | } |
| 172 | |
| 173 | static ppgen_ops wordlist_ops = { |
| 174 | "wordlist", |
| 175 | wordlist_init, wordlist_scan, wordlist_endscan, wordlist_gen, wordlist_done |
| 176 | }; |
| 177 | |
| 178 | /*----- Markov word model -------------------------------------------------*/ |
| 179 | |
| 180 | enum { |
| 181 | C_START = 27, |
| 182 | C_END, |
| 183 | VECSZ |
| 184 | }; |
| 185 | |
| 186 | typedef struct node { |
| 187 | uint32 count; |
| 188 | uint32 p[VECSZ]; |
| 189 | } node; |
| 190 | |
| 191 | static void *markov_init(void) |
| 192 | { |
| 193 | node (*model)[VECSZ][VECSZ][VECSZ] = xmalloc(sizeof(*model)); |
| 194 | unsigned i, j, k, l; |
| 195 | |
| 196 | for (i = 0; i < VECSZ; i++) { |
| 197 | for (j = 0; j < VECSZ; j++) { |
| 198 | for (k = 0; k < VECSZ; k++) { |
| 199 | node *n = &(*model)[i][j][k]; |
| 200 | n->count = 0; |
| 201 | for (l = 0; l < VECSZ; l++) |
| 202 | n->p[l] = 0; |
| 203 | } |
| 204 | } |
| 205 | } |
| 206 | |
| 207 | return (model); |
| 208 | } |
| 209 | |
| 210 | static void markov_scan(FILE *fp, void *p) |
| 211 | { |
| 212 | node (*model)[VECSZ][VECSZ][VECSZ] = p; |
| 213 | unsigned i = C_START, j = C_START, k = C_START, l = C_END; |
| 214 | |
| 215 | for (;;) { |
| 216 | int ch = getc(fp); |
| 217 | const char *q; |
| 218 | node *n = &(*model)[i][j][k]; |
| 219 | |
| 220 | if (ch == EOF || isspace(ch)) { |
| 221 | if (l != C_END) { |
| 222 | l = C_END; |
| 223 | n->count++; |
| 224 | n->p[l]++; |
| 225 | i = j = k = C_START; |
| 226 | } |
| 227 | if (ch == EOF) |
| 228 | break; |
| 229 | continue; |
| 230 | } |
| 231 | |
| 232 | if ((q = strchr(wchars, tolower(ch))) == 0) |
| 233 | continue; |
| 234 | l = q - wchars; |
| 235 | n->count++; |
| 236 | n->p[l]++; |
| 237 | i = j; j = k; k = l; |
| 238 | } |
| 239 | } |
| 240 | |
| 241 | static double markov_gen(dstr *d, grand *r, void *p) |
| 242 | { |
| 243 | node (*model)[VECSZ][VECSZ][VECSZ] = p; |
| 244 | unsigned i = C_START, j = C_START, k = C_START, l; |
| 245 | double logp = 0; |
| 246 | double log2 = log(2); |
| 247 | |
| 248 | for (;;) { |
| 249 | node *n = &(*model)[i][j][k]; |
| 250 | uint32 z = r->ops->range(r, n->count); |
| 251 | for (l = 0; z >= n->p[l]; z -= n->p[l++]) |
| 252 | ; |
| 253 | logp -= log((double)n->p[l]/(double)n->count)/log2; |
| 254 | if (l == C_END) |
| 255 | break; |
| 256 | DPUTC(d, wchars[l]); |
| 257 | i = j; j = k; k = l; |
| 258 | } |
| 259 | |
| 260 | return (logp); |
| 261 | } |
| 262 | |
| 263 | static void markov_done(void *p) |
| 264 | { |
| 265 | node (*model)[VECSZ][VECSZ][VECSZ] = p; |
| 266 | xfree(model); |
| 267 | } |
| 268 | |
| 269 | static ppgen_ops markov_ops = { |
| 270 | "markov", |
| 271 | markov_init, markov_scan, 0, markov_gen, markov_done |
| 272 | }; |
| 273 | |
| 274 | /*----- Main code ---------------------------------------------------------*/ |
| 275 | |
| 276 | static ppgen_ops *ppgentab[] = { |
| 277 | &markov_ops, |
| 278 | &wordlist_ops, |
| 279 | 0 |
| 280 | }; |
| 281 | |
| 282 | static void version(FILE *fp) |
| 283 | { |
| 284 | pquis(fp, "$, Catacomb version " VERSION "\n"); |
| 285 | } |
| 286 | |
| 287 | static void usage(FILE *fp) |
| 288 | { |
| 289 | pquis(fp, "\ |
| 290 | Usage: $ [-p] [-b bits] [-g gen] [-n count] [-r [min-]max] wordlist...\n\ |
| 291 | "); |
| 292 | } |
| 293 | |
| 294 | static void help(FILE *fp) |
| 295 | { |
| 296 | ppgen_ops **ops; |
| 297 | version(fp); |
| 298 | fputc('\n', fp); |
| 299 | usage(fp); |
| 300 | pquis(fp, "\n\ |
| 301 | Generates random passphrases with the requested level of entropy. Options\n\ |
| 302 | supported are:\n\ |
| 303 | \n\ |
| 304 | -h, --help Show this help text.\n\ |
| 305 | -v, --version Show the program's version number.\n\ |
| 306 | -u, --usage Show a terse usage summary.\n\ |
| 307 | -b, --bits=BITS Produce at least BITS bits of entropy.\n\ |
| 308 | -g, --generator=GEN Use passphrase generator GEN.\n\ |
| 309 | -n, --count=COUNT Generate COUNT passphrases.\n\ |
| 310 | -p, --probability Show -log_2 of probability for each phrase.\n\ |
| 311 | -r, --range=[MIN-]MAX Supply minimum and maximum word lengths.\n\ |
| 312 | \n\ |
| 313 | Generators currently available:"); |
| 314 | for (ops = ppgentab; *ops; ops++) |
| 315 | fprintf(fp, " %s", (*ops)->name); |
| 316 | fputc('\n', fp); |
| 317 | } |
| 318 | |
| 319 | int main(int argc, char *argv[]) |
| 320 | { |
| 321 | ppgen_ops *ops = ppgentab[0]; |
| 322 | unsigned f = 0; |
| 323 | void *ctx; |
| 324 | dstr d = DSTR_INIT; |
| 325 | dstr dd = DSTR_INIT; |
| 326 | unsigned i; |
| 327 | |
| 328 | #define f_bogus 1u |
| 329 | #define f_showp 2u |
| 330 | |
| 331 | ego(argv[0]); |
| 332 | for (;;) { |
| 333 | static struct option opts[] = { |
| 334 | { "help", 0, 0, 'h' }, |
| 335 | { "version", 0, 0, 'v' }, |
| 336 | { "usage", 0, 0, 'u' }, |
| 337 | { "bits", OPTF_ARGREQ, 0, 'b' }, |
| 338 | { "generator", OPTF_ARGREQ, 0, 'g' }, |
| 339 | { "count", OPTF_ARGREQ, 0, 'n' }, |
| 340 | { "probability", 0, 0, 'p' }, |
| 341 | { "range", OPTF_ARGREQ, 0, 'r' }, |
| 342 | { 0, 0, 0, 0 } |
| 343 | }; |
| 344 | int i = mdwopt(argc, argv, "hvu b:g:n:pr:", opts, 0, 0, 0); |
| 345 | |
| 346 | if (i < 0) |
| 347 | break; |
| 348 | switch (i) { |
| 349 | case 'h': |
| 350 | help(stdout); |
| 351 | exit(0); |
| 352 | case 'v': |
| 353 | version(stdout); |
| 354 | exit(0); |
| 355 | case 'u': |
| 356 | usage(stdout); |
| 357 | exit(0); |
| 358 | case 'b': { |
| 359 | char *p; |
| 360 | unsigned long n = strtoul(optarg, &p, 0); |
| 361 | if (*p) |
| 362 | die(EXIT_FAILURE, "bad integer `%s'", optarg); |
| 363 | bits = n; |
| 364 | } break; |
| 365 | case 'g': { |
| 366 | ppgen_ops **p; |
| 367 | size_t n = strlen(optarg); |
| 368 | ops = 0; |
| 369 | for (p = ppgentab; *p; p++) { |
| 370 | if (strncmp(optarg, (*p)->name, n) == 0) { |
| 371 | if (!(*p)->name[n]) { |
| 372 | ops = *p; |
| 373 | break; |
| 374 | } else if (ops) |
| 375 | die(EXIT_FAILURE, "ambiguous generator name `%s'", optarg); |
| 376 | ops = *p; |
| 377 | } |
| 378 | } |
| 379 | if (!ops) |
| 380 | die(EXIT_FAILURE, "unknown generator name `%s'", optarg); |
| 381 | } break; |
| 382 | case 'n': { |
| 383 | char *p; |
| 384 | unsigned long n = strtoul(optarg, &p, 0); |
| 385 | if (*p) |
| 386 | die(EXIT_FAILURE, "bad integer `%s'", optarg); |
| 387 | count = n; |
| 388 | } break; |
| 389 | case 'p': |
| 390 | f |= f_showp; |
| 391 | break; |
| 392 | case 'r': { |
| 393 | char *p; |
| 394 | unsigned long n = min, nn = max; |
| 395 | nn = strtoul(optarg, &p, 0); |
| 396 | if (*p == '-') { |
| 397 | n = nn; |
| 398 | nn = strtoul(p + 1, &p, 0); |
| 399 | } |
| 400 | if (*p) |
| 401 | die(EXIT_FAILURE, "bad range string `%s'", optarg); |
| 402 | min = n; max = nn; |
| 403 | } break; |
| 404 | default: |
| 405 | f |= f_bogus; |
| 406 | break; |
| 407 | } |
| 408 | } |
| 409 | |
| 410 | argc -= optind; |
| 411 | argv += optind; |
| 412 | if ((f & f_bogus) || !argc) { |
| 413 | usage(stderr); |
| 414 | exit(EXIT_FAILURE); |
| 415 | } |
| 416 | |
| 417 | rand_noisesrc(RAND_GLOBAL, &noise_source); |
| 418 | rand_seed(RAND_GLOBAL, 160); |
| 419 | |
| 420 | ctx = ops->init(); |
| 421 | while (*argv) { |
| 422 | if (strcmp(*argv, "-") == 0) |
| 423 | ops->scan(ctx, stdin); |
| 424 | else { |
| 425 | FILE *fp = fopen(*argv, "r"); |
| 426 | if (!fp) { |
| 427 | die(EXIT_FAILURE, "error opening file `%s': %s", |
| 428 | *argv, strerror(errno)); |
| 429 | } |
| 430 | ops->scan(fp, ctx); |
| 431 | fclose(fp); |
| 432 | } |
| 433 | argv++; |
| 434 | } |
| 435 | if (ops->endscan) |
| 436 | ops->endscan(ctx); |
| 437 | |
| 438 | for (i = 0; !count || i < count; i++) { |
| 439 | double logp = 0; |
| 440 | DRESET(&d); |
| 441 | while (logp < bits) { |
| 442 | double pp; |
| 443 | DRESET(&dd); |
| 444 | pp = ops->gen(&dd, &rand_global, ctx); |
| 445 | if (!pp || dd.len < min || dd.len > max) |
| 446 | continue; |
| 447 | if (logp) |
| 448 | DPUTC(&d, ' '); |
| 449 | DPUTD(&d, &dd); |
| 450 | logp += pp; |
| 451 | } |
| 452 | dstr_write(&d, stdout); |
| 453 | if (f & f_showp) |
| 454 | printf(" [%g]", logp); |
| 455 | fputc('\n', stdout); |
| 456 | } |
| 457 | |
| 458 | ops->done(ctx); |
| 459 | dstr_destroy(&d); |
| 460 | dstr_destroy(&dd); |
| 461 | return (0); |
| 462 | } |
| 463 | |
| 464 | /*----- That's all, folks -------------------------------------------------*/ |