strfile.c revision 1.37
1 /* $NetBSD: strfile.c,v 1.37 2013/09/19 00:18:52 uwe Exp $ */ 2 3 /*- 4 * Copyright (c) 1989, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * This code is derived from software contributed to Berkeley by 8 * Ken Arnold. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. Neither the name of the University nor the names of its contributors 19 * may be used to endorse or promote products derived from this software 20 * without specific prior written permission. 21 * 22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 32 * SUCH DAMAGE. 33 */ 34 35 #if HAVE_NBTOOL_CONFIG_H 36 #include "nbtool_config.h" 37 #endif 38 39 #ifdef __NetBSD__ 40 #include <sys/cdefs.h> 41 #ifndef lint 42 __COPYRIGHT("@(#) Copyright (c) 1989, 1993\ 43 The Regents of the University of California. All rights reserved."); 44 #endif /* not lint */ 45 46 #ifndef lint 47 #if 0 48 static char sccsid[] = "@(#)strfile.c 8.1 (Berkeley) 5/31/93"; 49 #else 50 __RCSID("$NetBSD: strfile.c,v 1.37 2013/09/19 00:18:52 uwe Exp $"); 51 #endif 52 #endif /* not lint */ 53 #endif /* __NetBSD__ */ 54 55 /* n.b.: this file is used at build-time - i.e. during build.sh. */ 56 57 #include <sys/types.h> 58 #include <sys/param.h> 59 #include <ctype.h> 60 #include <stdio.h> 61 #include <stdlib.h> 62 #include <string.h> 63 #include <time.h> 64 #include <unistd.h> 65 #include <inttypes.h> 66 #include <err.h> 67 68 #include "strfile.h" 69 70 #ifndef MAXPATHLEN 71 #define MAXPATHLEN 1024 72 #endif /* MAXPATHLEN */ 73 74 #if defined(__NetBSD__) || defined(__dead) 75 #define NORETURN __dead 76 #elif defined __GNUC__ 77 #define NORETURN __attribute__((__noreturn__)) 78 #else 79 #define NORETURN 80 #endif 81 82 /* 83 * This program takes a file composed of strings separated by 84 * lines starting with two consecutive delimiting character (default 85 * character is '%') and creates another file which consists of a table 86 * describing the file (structure from "strfile.h"), a table of seek 87 * pointers to the start of the strings, and the strings, each terminated 88 * by a null byte. Usage: 89 * 90 * % strfile [-iorsx] [ -cC ] sourcefile [ datafile ] 91 * 92 * c - Change delimiting character from '%' to 'C' 93 * s - Silent. Give no summary of data processed at the end of 94 * the run. 95 * o - order the strings in alphabetic order 96 * i - if ordering, ignore case 97 * r - randomize the order of the strings 98 * x - set rotated bit 99 * 100 * Ken Arnold Sept. 7, 1978 -- 101 * 102 * Added ordering options. 103 */ 104 105 # define STORING_PTRS (Oflag || Rflag) 106 # define CHUNKSIZE 512 107 108 # define ALLOC(ptr,sz) do { \ 109 if (ptr == NULL) \ 110 ptr = malloc(CHUNKSIZE * sizeof *ptr); \ 111 else if (((sz) + 1) % CHUNKSIZE == 0) \ 112 ptr = realloc(ptr, ((sz) + CHUNKSIZE) * sizeof *ptr); \ 113 if (ptr == NULL) \ 114 err(1, "out of space"); \ 115 } while (0) 116 117 typedef struct { 118 char first; 119 off_t pos; 120 } STR; 121 122 static char *Infile = NULL; /* input file name */ 123 static char Outfile[MAXPATHLEN] = ""; /* output file name */ 124 static char Delimch = '%'; /* delimiting character */ 125 126 static int Sflag = 0; /* silent run flag */ 127 static int Oflag = 0; /* ordering flag */ 128 static int Iflag = 0; /* ignore case flag */ 129 static int Rflag = 0; /* randomize order flag */ 130 static int Xflag = 0; /* set rotated bit */ 131 static long Num_pts = 0; /* number of pointers/strings */ 132 133 static off_t *Seekpts; 134 135 static FILE *Sort_1, *Sort_2; /* pointers for sorting */ 136 137 static STRFILE Tbl; /* statistics table */ 138 139 static STR *Firstch; /* first chars of each string */ 140 141 142 static uint32_t h2nl(uint32_t h); 143 static void getargs(int argc, char **argv); 144 static void usage(void) NORETURN; 145 static void add_offset(FILE *fp, off_t off); 146 static void do_order(void); 147 static int cmp_str(const void *vp1, const void *vp2); 148 static void randomize(void); 149 static void fwrite_be_offt(off_t off, FILE *f); 150 151 152 /* 153 * main: 154 * Drive the sucker. There are two main modes -- either we store 155 * the seek pointers, if the table is to be sorted or randomized, 156 * or we write the pointer directly to the file, if we are to stay 157 * in file order. If the former, we allocate and re-allocate in 158 * CHUNKSIZE blocks; if the latter, we just write each pointer, 159 * and then seek back to the beginning to write in the table. 160 */ 161 int 162 main(int ac, char **av) 163 { 164 char *sp, dc; 165 FILE *inf, *outf; 166 off_t last_off, length, pos; 167 int first; 168 char *nsp; 169 STR *fp; 170 static char string[257]; 171 long i; 172 173 /* sanity test */ 174 if (sizeof(uint32_t) != 4) 175 errx(1, "sizeof(uint32_t) != 4"); 176 177 getargs(ac, av); /* evalute arguments */ 178 dc = Delimch; 179 if ((inf = fopen(Infile, "r")) == NULL) 180 err(1, "open `%s'", Infile); 181 182 if ((outf = fopen(Outfile, "w")) == NULL) 183 err(1, "open `%s'", Outfile); 184 if (!STORING_PTRS) 185 (void) fseek(outf, sizeof Tbl, SEEK_SET); 186 187 /* 188 * Write the strings onto the file 189 */ 190 191 Tbl.str_longlen = 0; 192 Tbl.str_shortlen = (unsigned int) 0x7fffffff; 193 Tbl.str_delim = dc; 194 Tbl.str_version = VERSION; 195 first = Oflag; 196 add_offset(outf, ftell(inf)); 197 last_off = 0; 198 do { 199 sp = fgets(string, 256, inf); 200 if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) { 201 pos = ftell(inf); 202 length = pos - last_off - (sp ? strlen(sp) : 0); 203 last_off = pos; 204 if (!length) 205 continue; 206 add_offset(outf, pos); 207 if ((off_t)Tbl.str_longlen < length) 208 Tbl.str_longlen = length; 209 if ((off_t)Tbl.str_shortlen > length) 210 Tbl.str_shortlen = length; 211 first = Oflag; 212 } 213 else if (first) { 214 for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++) 215 continue; 216 ALLOC(Firstch, Num_pts); 217 fp = &Firstch[Num_pts - 1]; 218 if (Iflag && isupper((unsigned char)*nsp)) 219 fp->first = tolower((unsigned char)*nsp); 220 else 221 fp->first = *nsp; 222 fp->pos = Seekpts[Num_pts - 1]; 223 first = 0; 224 } 225 } while (sp != NULL); 226 227 /* 228 * write the tables in 229 */ 230 231 (void) fclose(inf); 232 233 if (Oflag) 234 do_order(); 235 else if (Rflag) 236 randomize(); 237 238 if (Xflag) 239 Tbl.str_flags |= STR_ROTATED; 240 241 if (!Sflag) { 242 printf("\"%s\" created\n", Outfile); 243 if (Num_pts == 2) 244 puts("There was 1 string"); 245 else 246 printf("There were %d strings\n", (int)(Num_pts - 1)); 247 printf("Longest string: %lu byte%s\n", (unsigned long)Tbl.str_longlen, 248 Tbl.str_longlen == 1 ? "" : "s"); 249 printf("Shortest string: %lu byte%s\n", (unsigned long)Tbl.str_shortlen, 250 Tbl.str_shortlen == 1 ? "" : "s"); 251 } 252 253 (void) fseek(outf, (off_t) 0, SEEK_SET); 254 Tbl.str_version = h2nl(Tbl.str_version); 255 Tbl.str_numstr = h2nl(Num_pts - 1); 256 Tbl.str_longlen = h2nl(Tbl.str_longlen); 257 Tbl.str_shortlen = h2nl(Tbl.str_shortlen); 258 Tbl.str_flags = h2nl(Tbl.str_flags); 259 (void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf); 260 if (STORING_PTRS) { 261 for (i = 0; i < Num_pts; i++) 262 fwrite_be_offt(Seekpts[i], outf); 263 } 264 fflush(outf); 265 if (ferror(outf)) 266 err(1, "fwrite %s", Outfile); 267 (void) fclose(outf); 268 exit(0); 269 } 270 271 /* 272 * This routine evaluates arguments from the command line 273 */ 274 static void 275 getargs(int argc, char **argv) 276 { 277 int ch; 278 extern int optind; 279 extern char *optarg; 280 281 while ((ch = getopt(argc, argv, "c:iorsx")) != -1) 282 switch(ch) { 283 case 'c': /* new delimiting char */ 284 Delimch = *optarg; 285 if (!isascii(Delimch)) { 286 printf("bad delimiting character: '\\%o\n'", 287 Delimch); 288 } 289 break; 290 case 'i': /* ignore case in ordering */ 291 Iflag++; 292 break; 293 case 'o': /* order strings */ 294 Oflag++; 295 break; 296 case 'r': /* randomize pointers */ 297 Rflag++; 298 break; 299 case 's': /* silent */ 300 Sflag++; 301 break; 302 case 'x': /* set the rotated bit */ 303 Xflag++; 304 break; 305 case '?': 306 default: 307 usage(); 308 } 309 argv += optind; 310 311 if (*argv) { 312 Infile = *argv; 313 if (*++argv) 314 (void) strcpy(Outfile, *argv); 315 } 316 if (!Infile) { 317 puts("No input file name"); 318 usage(); 319 } 320 if (*Outfile == '\0') { 321 (void) strcpy(Outfile, Infile); 322 (void) strcat(Outfile, ".dat"); 323 } 324 } 325 326 static void 327 usage(void) 328 { 329 (void) fprintf(stderr, 330 "Usage: %s [-iorsx] [-c char] sourcefile [datafile]\n", 331 getprogname()); 332 exit(1); 333 } 334 335 /* 336 * add_offset: 337 * Add an offset to the list, or write it out, as appropriate. 338 */ 339 static void 340 add_offset(FILE *fp, off_t off) 341 { 342 343 if (!STORING_PTRS) { 344 fwrite_be_offt(off, fp); 345 } else { 346 ALLOC(Seekpts, Num_pts + 1); 347 Seekpts[Num_pts] = off; 348 } 349 Num_pts++; 350 } 351 352 /* 353 * do_order: 354 * Order the strings alphabetically (possibly ignoring case). 355 */ 356 static void 357 do_order(void) 358 { 359 int i; 360 off_t *lp; 361 STR *fp; 362 363 Sort_1 = fopen(Infile, "r"); 364 Sort_2 = fopen(Infile, "r"); 365 qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str); 366 i = Tbl.str_numstr; 367 lp = Seekpts; 368 fp = Firstch; 369 while (i--) 370 *lp++ = fp++->pos; 371 (void) fclose(Sort_1); 372 (void) fclose(Sort_2); 373 Tbl.str_flags |= STR_ORDERED; 374 } 375 376 static int 377 cmp_str(const void *vp1, const void *vp2) 378 { 379 const STR *p1, *p2; 380 int c1, c2; 381 int n1, n2; 382 383 p1 = (const STR *)vp1; 384 p2 = (const STR *)vp2; 385 386 # define SET_N(nf,ch) (nf = (ch == '\n')) 387 # define IS_END(ch,nf) (ch == Delimch && nf) 388 389 c1 = p1->first; 390 c2 = p2->first; 391 if (c1 != c2) 392 return c1 - c2; 393 394 (void) fseek(Sort_1, p1->pos, SEEK_SET); 395 (void) fseek(Sort_2, p2->pos, SEEK_SET); 396 397 n1 = 0; 398 n2 = 0; 399 while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0') 400 SET_N(n1, c1); 401 while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0') 402 SET_N(n2, c2); 403 404 while (!IS_END(c1, n1) && !IS_END(c2, n2)) { 405 if (Iflag) { 406 if (isupper(c1)) 407 c1 = tolower(c1); 408 if (isupper(c2)) 409 c2 = tolower(c2); 410 } 411 if (c1 != c2) 412 return c1 - c2; 413 SET_N(n1, c1); 414 SET_N(n2, c2); 415 c1 = getc(Sort_1); 416 c2 = getc(Sort_2); 417 } 418 if (IS_END(c1, n1)) 419 c1 = 0; 420 if (IS_END(c2, n2)) 421 c2 = 0; 422 return c1 - c2; 423 } 424 425 /* 426 * randomize: 427 * Randomize the order of the string table. We must be careful 428 * not to randomize across delimiter boundaries. All 429 * randomization is done within each block. 430 */ 431 static void 432 randomize(void) 433 { 434 int cnt, i; 435 off_t tmp; 436 off_t *sp; 437 438 srandom((int)(time(NULL) + getpid())); 439 440 Tbl.str_flags |= STR_RANDOM; 441 cnt = Tbl.str_numstr; 442 443 /* 444 * move things around randomly 445 */ 446 447 for (sp = Seekpts; cnt > 0; cnt--, sp++) { 448 i = random() % cnt; 449 tmp = sp[0]; 450 sp[0] = sp[i]; 451 sp[i] = tmp; 452 } 453 } 454 455 /* 456 * fwrite_be_offt: 457 * Write out the off paramater as a 64 bit big endian number 458 */ 459 460 static void 461 fwrite_be_offt(off_t off, FILE *f) 462 { 463 int i; 464 unsigned char c[8]; 465 466 for (i = 7; i >= 0; i--) { 467 c[i] = off & 0xff; 468 off >>= 8; 469 } 470 fwrite(c, sizeof(c), 1, f); 471 } 472 473 static uint32_t 474 h2nl(uint32_t h) 475 { 476 unsigned char c[4]; 477 uint32_t rv; 478 479 c[0] = (h >> 24) & 0xff; 480 c[1] = (h >> 16) & 0xff; 481 c[2] = (h >> 8) & 0xff; 482 c[3] = (h >> 0) & 0xff; 483 memcpy(&rv, c, sizeof rv); 484 485 return (rv); 486 } 487
Indexes created Thu Oct 16 14:10:15 GMT 2025