strfile.c revision 1.36
1 /* $NetBSD: strfile.c,v 1.36 2013/09/18 23:18:14 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 1.36 2013/09/18 23:18:14 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 TRUE 1 106 # define FALSE 0 107 108 # define STORING_PTRS (Oflag || Rflag) 109 # define CHUNKSIZE 512 110 111 # define ALLOC(ptr,sz) do { \ 112 if (ptr == NULL) \ 113 ptr = malloc(CHUNKSIZE * sizeof *ptr); \ 114 else if (((sz) + 1) % CHUNKSIZE == 0) \ 115 ptr = realloc(ptr, ((sz) + CHUNKSIZE) * sizeof *ptr); \ 116 if (ptr == NULL) \ 117 err(1, "out of space"); \ 118 } while (0) 119 120 typedef struct { 121 char first; 122 off_t pos; 123 } STR; 124 125 static char *Infile = NULL; /* input file name */ 126 static char Outfile[MAXPATHLEN] = ""; /* output file name */ 127 static char Delimch = '%'; /* delimiting character */ 128 129 static int Sflag = FALSE; /* silent run flag */ 130 static int Oflag = FALSE; /* ordering flag */ 131 static int Iflag = FALSE; /* ignore case flag */ 132 static int Rflag = FALSE; /* randomize order flag */ 133 static int Xflag = FALSE; /* set rotated bit */ 134 static long Num_pts = 0; /* number of pointers/strings */ 135 136 static off_t *Seekpts; 137 138 static FILE *Sort_1, *Sort_2; /* pointers for sorting */ 139 140 static STRFILE Tbl; /* statistics table */ 141 142 static STR *Firstch; /* first chars of each string */ 143 144 145 static uint32_t h2nl(uint32_t h); 146 static void getargs(int argc, char **argv); 147 static void usage(void) NORETURN; 148 static void add_offset(FILE *fp, off_t off); 149 static void do_order(void); 150 static int cmp_str(const void *vp1, const void *vp2); 151 static void randomize(void); 152 static void fwrite_be_offt(off_t off, FILE *f); 153 154 155 /* 156 * main: 157 * Drive the sucker. There are two main modes -- either we store 158 * the seek pointers, if the table is to be sorted or randomized, 159 * or we write the pointer directly to the file, if we are to stay 160 * in file order. If the former, we allocate and re-allocate in 161 * CHUNKSIZE blocks; if the latter, we just write each pointer, 162 * and then seek back to the beginning to write in the table. 163 */ 164 int 165 main(int ac, char **av) 166 { 167 char *sp, dc; 168 FILE *inf, *outf; 169 off_t last_off, length, pos; 170 int first; 171 char *nsp; 172 STR *fp; 173 static char string[257]; 174 long i; 175 176 /* sanity test */ 177 if (sizeof(uint32_t) != 4) 178 errx(1, "sizeof(uint32_t) != 4"); 179 180 getargs(ac, av); /* evalute arguments */ 181 dc = Delimch; 182 if ((inf = fopen(Infile, "r")) == NULL) 183 err(1, "open `%s'", Infile); 184 185 if ((outf = fopen(Outfile, "w")) == NULL) 186 err(1, "open `%s'", Outfile); 187 if (!STORING_PTRS) 188 (void) fseek(outf, sizeof Tbl, SEEK_SET); 189 190 /* 191 * Write the strings onto the file 192 */ 193 194 Tbl.str_longlen = 0; 195 Tbl.str_shortlen = (unsigned int) 0x7fffffff; 196 Tbl.str_delim = dc; 197 Tbl.str_version = VERSION; 198 first = Oflag; 199 add_offset(outf, ftell(inf)); 200 last_off = 0; 201 do { 202 sp = fgets(string, 256, inf); 203 if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) { 204 pos = ftell(inf); 205 length = pos - last_off - (sp ? strlen(sp) : 0); 206 last_off = pos; 207 if (!length) 208 continue; 209 add_offset(outf, pos); 210 if ((off_t)Tbl.str_longlen < length) 211 Tbl.str_longlen = length; 212 if ((off_t)Tbl.str_shortlen > length) 213 Tbl.str_shortlen = length; 214 first = Oflag; 215 } 216 else if (first) { 217 for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++) 218 continue; 219 ALLOC(Firstch, Num_pts); 220 fp = &Firstch[Num_pts - 1]; 221 if (Iflag && isupper((unsigned char)*nsp)) 222 fp->first = tolower((unsigned char)*nsp); 223 else 224 fp->first = *nsp; 225 fp->pos = Seekpts[Num_pts - 1]; 226 first = FALSE; 227 } 228 } while (sp != NULL); 229 230 /* 231 * write the tables in 232 */ 233 234 (void) fclose(inf); 235 236 if (Oflag) 237 do_order(); 238 else if (Rflag) 239 randomize(); 240 241 if (Xflag) 242 Tbl.str_flags |= STR_ROTATED; 243 244 if (!Sflag) { 245 printf("\"%s\" created\n", Outfile); 246 if (Num_pts == 2) 247 puts("There was 1 string"); 248 else 249 printf("There were %d strings\n", (int)(Num_pts - 1)); 250 printf("Longest string: %lu byte%s\n", (unsigned long)Tbl.str_longlen, 251 Tbl.str_longlen == 1 ? "" : "s"); 252 printf("Shortest string: %lu byte%s\n", (unsigned long)Tbl.str_shortlen, 253 Tbl.str_shortlen == 1 ? "" : "s"); 254 } 255 256 (void) fseek(outf, (off_t) 0, SEEK_SET); 257 Tbl.str_version = h2nl(Tbl.str_version); 258 Tbl.str_numstr = h2nl(Num_pts - 1); 259 Tbl.str_longlen = h2nl(Tbl.str_longlen); 260 Tbl.str_shortlen = h2nl(Tbl.str_shortlen); 261 Tbl.str_flags = h2nl(Tbl.str_flags); 262 (void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf); 263 if (STORING_PTRS) { 264 for (i = 0; i < Num_pts; i++) 265 fwrite_be_offt(Seekpts[i], outf); 266 } 267 fflush(outf); 268 if (ferror(outf)) 269 err(1, "fwrite %s", Outfile); 270 (void) fclose(outf); 271 exit(0); 272 } 273 274 /* 275 * This routine evaluates arguments from the command line 276 */ 277 static void 278 getargs(int argc, char **argv) 279 { 280 int ch; 281 extern int optind; 282 extern char *optarg; 283 284 while ((ch = getopt(argc, argv, "c:iorsx")) != -1) 285 switch(ch) { 286 case 'c': /* new delimiting char */ 287 Delimch = *optarg; 288 if (!isascii(Delimch)) { 289 printf("bad delimiting character: '\\%o\n'", 290 Delimch); 291 } 292 break; 293 case 'i': /* ignore case in ordering */ 294 Iflag++; 295 break; 296 case 'o': /* order strings */ 297 Oflag++; 298 break; 299 case 'r': /* randomize pointers */ 300 Rflag++; 301 break; 302 case 's': /* silent */ 303 Sflag++; 304 break; 305 case 'x': /* set the rotated bit */ 306 Xflag++; 307 break; 308 case '?': 309 default: 310 usage(); 311 } 312 argv += optind; 313 314 if (*argv) { 315 Infile = *argv; 316 if (*++argv) 317 (void) strcpy(Outfile, *argv); 318 } 319 if (!Infile) { 320 puts("No input file name"); 321 usage(); 322 } 323 if (*Outfile == '\0') { 324 (void) strcpy(Outfile, Infile); 325 (void) strcat(Outfile, ".dat"); 326 } 327 } 328 329 static void 330 usage(void) 331 { 332 (void) fprintf(stderr, 333 "Usage: %s [-iorsx] [-c char] sourcefile [datafile]\n", 334 getprogname()); 335 exit(1); 336 } 337 338 /* 339 * add_offset: 340 * Add an offset to the list, or write it out, as appropriate. 341 */ 342 static void 343 add_offset(FILE *fp, off_t off) 344 { 345 346 if (!STORING_PTRS) { 347 fwrite_be_offt(off, fp); 348 } else { 349 ALLOC(Seekpts, Num_pts + 1); 350 Seekpts[Num_pts] = off; 351 } 352 Num_pts++; 353 } 354 355 /* 356 * do_order: 357 * Order the strings alphabetically (possibly ignoring case). 358 */ 359 static void 360 do_order(void) 361 { 362 int i; 363 off_t *lp; 364 STR *fp; 365 366 Sort_1 = fopen(Infile, "r"); 367 Sort_2 = fopen(Infile, "r"); 368 qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str); 369 i = Tbl.str_numstr; 370 lp = Seekpts; 371 fp = Firstch; 372 while (i--) 373 *lp++ = fp++->pos; 374 (void) fclose(Sort_1); 375 (void) fclose(Sort_2); 376 Tbl.str_flags |= STR_ORDERED; 377 } 378 379 static int 380 cmp_str(const void *vp1, const void *vp2) 381 { 382 const STR *p1, *p2; 383 int c1, c2; 384 int n1, n2; 385 386 p1 = (const STR *)vp1; 387 p2 = (const STR *)vp2; 388 389 # define SET_N(nf,ch) (nf = (ch == '\n')) 390 # define IS_END(ch,nf) (ch == Delimch && nf) 391 392 c1 = p1->first; 393 c2 = p2->first; 394 if (c1 != c2) 395 return c1 - c2; 396 397 (void) fseek(Sort_1, p1->pos, SEEK_SET); 398 (void) fseek(Sort_2, p2->pos, SEEK_SET); 399 400 n1 = FALSE; 401 n2 = FALSE; 402 while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0') 403 SET_N(n1, c1); 404 while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0') 405 SET_N(n2, c2); 406 407 while (!IS_END(c1, n1) && !IS_END(c2, n2)) { 408 if (Iflag) { 409 if (isupper(c1)) 410 c1 = tolower(c1); 411 if (isupper(c2)) 412 c2 = tolower(c2); 413 } 414 if (c1 != c2) 415 return c1 - c2; 416 SET_N(n1, c1); 417 SET_N(n2, c2); 418 c1 = getc(Sort_1); 419 c2 = getc(Sort_2); 420 } 421 if (IS_END(c1, n1)) 422 c1 = 0; 423 if (IS_END(c2, n2)) 424 c2 = 0; 425 return c1 - c2; 426 } 427 428 /* 429 * randomize: 430 * Randomize the order of the string table. We must be careful 431 * not to randomize across delimiter boundaries. All 432 * randomization is done within each block. 433 */ 434 static void 435 randomize(void) 436 { 437 int cnt, i; 438 off_t tmp; 439 off_t *sp; 440 441 srandom((int)(time(NULL) + getpid())); 442 443 Tbl.str_flags |= STR_RANDOM; 444 cnt = Tbl.str_numstr; 445 446 /* 447 * move things around randomly 448 */ 449 450 for (sp = Seekpts; cnt > 0; cnt--, sp++) { 451 i = random() % cnt; 452 tmp = sp[0]; 453 sp[0] = sp[i]; 454 sp[i] = tmp; 455 } 456 } 457 458 /* 459 * fwrite_be_offt: 460 * Write out the off paramater as a 64 bit big endian number 461 */ 462 463 static void 464 fwrite_be_offt(off_t off, FILE *f) 465 { 466 int i; 467 unsigned char c[8]; 468 469 for (i = 7; i >= 0; i--) { 470 c[i] = off & 0xff; 471 off >>= 8; 472 } 473 fwrite(c, sizeof(c), 1, f); 474 } 475 476 static uint32_t 477 h2nl(uint32_t h) 478 { 479 unsigned char c[4]; 480 uint32_t rv; 481 482 c[0] = (h >> 24) & 0xff; 483 c[1] = (h >> 16) & 0xff; 484 c[2] = (h >> 8) & 0xff; 485 c[3] = (h >> 0) & 0xff; 486 memcpy(&rv, c, sizeof rv); 487 488 return (rv); 489 } 490
Indexes created Mon Oct 13 21:09:55 GMT 2025