usr.bin/xstr/xstr.c

   1 /*      $NetBSD: xstr.c,v 1.25 2011/09/16 15:39:31 joerg Exp $  */
   2
   3 /*
   4  * Copyright (c) 1980, 1993
   5  *      The Regents of the University of California.  All rights reserved.
   6  *
   7  * Redistribution and use in source and binary forms, with or without
   8  * modification, are permitted provided that the following conditions
   9  * are met:
  10  * 1. Redistributions of source code must retain the above copyright
  11  *    notice, this list of conditions and the following disclaimer.
  12  * 2. Redistributions in binary form must reproduce the above copyright
  13  *    notice, this list of conditions and the following disclaimer in the
  14  *    documentation and/or other materials provided with the distribution.
  15  * 3. Neither the name of the University nor the names of its contributors
  16  *    may be used to endorse or promote products derived from this software
  17  *    without specific prior written permission.
  18  *
  19  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  29  * SUCH DAMAGE.
  30  */
  31
  32 #include <sys/cdefs.h>
  33 #ifndef lint
  34 __COPYRIGHT("@(#) Copyright (c) 1980, 1993\
  35  The Regents of the University of California.  All rights reserved.");
  36 #endif /* not lint */
  37
  38 #ifndef lint
  39 #if 0
  40 static char sccsid[] = "@(#)xstr.c      8.1 (Berkeley) 6/9/93";
  41 #else
  42 __RCSID("$NetBSD: xstr.c,v 1.25 2011/09/16 15:39:31 joerg Exp $");
  43 #endif
  44 #endif /* not lint */
  45
  46 #include <sys/param.h>
  47 #include <signal.h>
  48 #include <errno.h>
  49 #include <unistd.h>
  50 #include <stdio.h>
  51 #include <ctype.h>
  52 #include <string.h>
  53 #include <stdlib.h>
  54 #include <err.h>
  55 #include "pathnames.h"
  56
  57 /*
  58  * xstr - extract and hash strings in a C program
  59  *
  60  * Bill Joy UCB
  61  * November, 1978
  62  */
  63
  64 static off_t    hashit(const char *, int);
  65 __dead static void      onintr(int);
  66 static off_t    yankstr(char **);
  67 static int      octdigit(char);
  68 static void     inithash(void);
  69 static int      fgetNUL(char *, int, FILE *);
  70 static int      xgetc(FILE *);
  71 static void     flushsh(void);
  72 static void     found(int, off_t, const char *);
  73 static void     prstr(const char *);
  74 static void     xsdotc(void);
  75 static char     lastchr(const char *);
  76 static int      istail(const char *, const char *);
  77 static void     process(const char *);
  78 __dead static void      usage(void);
  79
  80 static off_t    tellpt;
  81 static off_t    mesgpt;
  82 static char     stringtmpfile[MAXPATHLEN];
  83 static const char *strings =    "strings";
  84 static const char *array =      0;
  85 static int      cflg;
  86 static int      vflg;
  87 static int      readstd;
  88 static char     linebuf[8192];
  89
  90 #define BUCKETS 128
  91
  92 static struct   hash {
  93         off_t   hpt;
  94         char    *hstr;
  95         struct  hash *hnext;
  96         short   hnew;
  97 } bucket[BUCKETS];
  98
  99 int
 100 main(int argc, char *argv[])
 101 {
 102         int c;
 103
 104         while ((c = getopt(argc, argv, "-cvl:")) != -1)
 105                 switch (c) {
 106                 case '-':
 107                         readstd++;
 108                         break;
 109                 case 'c':
 110                         cflg++;
 111                         break;
 112                 case 'v':
 113                         vflg++;
 114                         break;
 115                 case 'l':
 116                         array = optarg;
 117                         break;
 118                 default:
 119                         usage();
 120                 }
 121         argc -= optind;
 122         argv += optind;
 123
 124         if (array == 0)
 125                 array = "xstr";
 126
 127         if (signal(SIGINT, SIG_IGN) == SIG_DFL)
 128                 (void)signal(SIGINT, onintr);
 129         if (cflg || (argc == 0 && !readstd))
 130                 inithash();
 131         else {
 132                 int     fd;
 133
 134                 snprintf(stringtmpfile, sizeof(stringtmpfile),
 135                     "%s%s.XXXXXX", _PATH_TMP, "xstr");
 136                 strings = stringtmpfile;
 137                 fd = mkstemp(stringtmpfile);
 138                 if (fd == -1)
 139                         err(1, "mkstemp failed");
 140                 close(fd);
 141         }
 142         while (readstd || argc > 0) {
 143                 if (freopen("x.c", "w", stdout) == NULL)
 144                         err(1, "Cannot open `%s'", "x.c");
 145                 if (!readstd && freopen(argv[0], "r", stdin) == NULL)
 146                         err(1, "Cannot open `%s'", argv[0]);
 147                 process("x.c");
 148                 if (readstd == 0)
 149                         argc--, argv++;
 150                 else
 151                         readstd = 0;
 152         };
 153         flushsh();
 154         if (cflg == 0)
 155                 xsdotc();
 156         if (strings[0] == '/')
 157                 (void)unlink(strings);
 158         exit(0);
 159 }
 160
 161 static void
 162 process(const char *name)
 163 {
 164         char *cp;
 165         int c;
 166         int incomm = 0;
 167         int inasm = 0;
 168         int asmparnest = 0;
 169         int ret;
 170
 171         printf("extern char\t%s[];\n", array);
 172         for (;;) {
 173                 if (fgets(linebuf, sizeof linebuf, stdin) == NULL) {
 174                         if (ferror(stdin))
 175                                 err(1, "Error reading `%s'", name);
 176                         break;
 177                 }
 178                 if (linebuf[0] == '#') {
 179                         printf("%s", linebuf);
 180                         continue;
 181                 }
 182                 for (cp = linebuf; (c = *cp++);)
 183                         switch (c) {
 184
 185                         case '"':
 186                                 if (incomm || inasm)
 187                                         goto def;
 188                                 if ((ret = (int) yankstr(&cp)) == -1)
 189                                         goto out;
 190                                 printf("(&%s[%d])", array, ret);
 191                                 break;
 192
 193                         case '\'':
 194                                 if (incomm || inasm)
 195                                         goto def;
 196                                 putchar(c);
 197                                 if (*cp)
 198                                         putchar(*cp++);
 199                                 break;
 200
 201                         case '/':
 202                                 if (incomm || *cp != '*')
 203                                         goto def;
 204                                 incomm = 1;
 205                                 cp++;
 206                                 printf("/*");
 207                                 continue;
 208
 209                         case '*':
 210                                 if (incomm && *cp == '/') {
 211                                         incomm = 0;
 212                                         cp++;
 213                                         printf("*/");
 214                                         continue;
 215                                 }
 216                                 goto def;
 217
 218                         case '(':
 219                                 if (!incomm && inasm)
 220                                         asmparnest++;
 221                                 goto def;
 222
 223                         case ')':
 224                                 if (!incomm && inasm && !--asmparnest)
 225                                         inasm = 0;
 226                                 goto def;
 227
 228                         case '_':
 229                                 if (incomm || inasm)
 230                                         goto def;
 231                                 if (!strncmp(cp, "_asm", 4)) {
 232                                         cp += 4;
 233                                         printf("__asm");
 234                                         if (!strncmp(cp, "__", 2)) {
 235                                                 cp += 2;
 236                                                 printf("__");
 237                                         }
 238                                         if (isalnum((unsigned char)*cp) ||
 239                                             *cp == '_')
 240                                                 goto def;
 241                                         asmparnest = 0;
 242                                         inasm = 1;
 243                                 } else
 244                                         goto def;
 245                                 break;
 246 def:
 247                         default:
 248                                 putchar(c);
 249                                 break;
 250                         }
 251         }
 252 out:
 253         if (ferror(stdout)) {
 254                 warn("Error reading `%s'", "x.c");
 255                 onintr(1);
 256         }
 257 }
 258
 259 static off_t
 260 yankstr(char **cpp)
 261 {
 262         char *cp = *cpp;
 263         int c, ch;
 264         char *dbuf, *dp, *edp;
 265         const char *tp;
 266         off_t hash;
 267         size_t bsiz = BUFSIZ;
 268
 269         if ((dp = dbuf = malloc(bsiz)) == NULL)
 270                 err(1, "malloc");
 271         edp = dbuf + bsiz;
 272
 273         while ((c = *cp++) != '\0') {
 274                 switch (c) {
 275
 276                 case '"':
 277                         /* Look for a concatenated string */
 278                         for (;;) {
 279                                 while (isspace((unsigned char)*cp))
 280                                         cp++;
 281                                 if (*cp == '\0') {
 282                                         if (fgets(linebuf,
 283                                             sizeof linebuf, stdin) == NULL) {
 284                                                 if (ferror(stdin))
 285                                                         err(1,
 286                                                         "Error reading `x.c'");
 287                                                 goto out;
 288                                         }
 289                                         cp = linebuf;
 290                                 } else {
 291                                         if (*cp == '"') {
 292                                                 cp++;
 293                                                 if (*cp == '"') {
 294                                                         cp++;
 295                                                         continue;
 296                                                 } else {
 297                                                         c = *cp++;
 298                                                         goto gotc;
 299                                                 }
 300                                         } else {
 301                                                 cp++;
 302                                                 goto out;
 303                                         }
 304                                 }
 305                         }
 306                         /*NOTREACHED*/
 307                 case '\\':
 308                         c = *cp++;
 309                         if (c == 0)
 310                                 break;
 311                         if (c == '\n') {
 312                                 if (fgets(linebuf, sizeof linebuf, stdin)
 313                                     == NULL) {
 314                                         if (ferror(stdin))
 315                                                 err(1, "Error reading `x.c'");
 316                                         return(-1);
 317                                 }
 318                                 cp = linebuf;
 319                                 continue;
 320                         }
 321                         for (tp = "b\bt\tr\rn\nf\f\\\\\"\""; (ch = *tp++); tp++)
 322                                 if (c == ch) {
 323                                         c = *tp;
 324                                         goto gotc;
 325                                 }
 326                         if (!octdigit(c)) {
 327                                 *dp++ = '\\';
 328                                 break;
 329                         }
 330                         c -= '0';
 331                         if (!octdigit(*cp))
 332                                 break;
 333                         c <<= 3, c += *cp++ - '0';
 334                         if (!octdigit(*cp))
 335                                 break;
 336                         c <<= 3, c += *cp++ - '0';
 337                         break;
 338                 }
 339 gotc:
 340                 if (dp >= edp - 1) {
 341                         char *nbuf;
 342                         bsiz += BUFSIZ;
 343                         if ((nbuf = realloc(dbuf, bsiz)) == NULL) {
 344                                 free(dbuf);
 345                                 err(1, "realloc");
 346                         }
 347                         dp = nbuf + (dp - dbuf);
 348                         edp = nbuf + bsiz;
 349                         dbuf = nbuf;
 350                 }
 351                 *dp++ = c;
 352         }
 353 out:
 354         *cpp = --cp;
 355         *dp = '\0';
 356         hash = hashit(dbuf, 1);
 357         free(dbuf);
 358         return hash;
 359 }
 360
 361 static int
 362 octdigit(char c)
 363 {
 364
 365         return (isdigit((unsigned char)c) && c != '8' && c != '9');
 366 }
 367
 368 static void
 369 inithash(void)
 370 {
 371         char buf[BUFSIZ];
 372         FILE *mesgread = fopen(strings, "r");
 373
 374         if (mesgread == NULL)
 375                 return;
 376         for (;;) {
 377                 mesgpt = tellpt;
 378                 if (fgetNUL(buf, sizeof buf, mesgread) == 0)
 379                         break;
 380                 (void)hashit(buf, 0);
 381         }
 382         (void)fclose(mesgread);
 383 }
 384
 385 static int
 386 fgetNUL(char *obuf, int rmdr, FILE *file)
 387 {
 388         int c;
 389         char *buf = obuf;
 390
 391         c = 0;  /* XXXGCC -Wuninitialized */
 392
 393         while (--rmdr > 0 && (c = xgetc(file) != 0 && c != EOF))
 394                 *buf++ = c;
 395         *buf++ = 0;
 396         return (feof(file) || ferror(file)) ? 0 : 1;
 397 }
 398
 399 static int
 400 xgetc(FILE *file)
 401 {
 402
 403         tellpt++;
 404         return getc(file);
 405 }
 406
 407
 408 static off_t
 409 hashit(const char *str, int new)
 410 {
 411         int i;
 412         struct hash *hp, *hp0;
 413
 414         hp = hp0 = &bucket[lastchr(str) & 0177];
 415         while (hp->hnext) {
 416                 hp = hp->hnext;
 417                 i = istail(str, hp->hstr);
 418                 if (i >= 0)
 419                         return (hp->hpt + i);
 420         }
 421         if ((hp = calloc(1, sizeof (*hp))) == NULL)
 422                 err(1, NULL);
 423         hp->hpt = mesgpt;
 424         if ((hp->hstr = strdup(str)) == NULL)
 425                 err(1, NULL);
 426         mesgpt += strlen(hp->hstr) + 1;
 427         hp->hnext = hp0->hnext;
 428         hp->hnew = new;
 429         hp0->hnext = hp;
 430         return (hp->hpt);
 431 }
 432
 433 static void
 434 flushsh(void)
 435 {
 436         int i;
 437         struct hash *hp;
 438         FILE *mesgwrit;
 439         int old = 0, new = 0;
 440
 441         for (i = 0; i < BUCKETS; i++)
 442                 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext)
 443                         if (hp->hnew)
 444                                 new++;
 445                         else
 446                                 old++;
 447         if (new == 0 && old != 0)
 448                 return;
 449         mesgwrit = fopen(strings, old ? "r+" : "w");
 450         if (mesgwrit == NULL)
 451                 err(1, "Cannot open `%s'", strings);
 452         for (i = 0; i < BUCKETS; i++)
 453                 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext) {
 454                         found(hp->hnew, hp->hpt, hp->hstr);
 455                         if (hp->hnew) {
 456                                 (void)fseek(mesgwrit, hp->hpt, 0);
 457                                 (void)fwrite(hp->hstr, strlen(hp->hstr) + 1, 1,
 458                                     mesgwrit);
 459                                 if (ferror(mesgwrit))
 460                                         err(1, "Error writing `%s'", strings);
 461                         }
 462                 }
 463         if (fclose(mesgwrit) == EOF)
 464                 err(1, "Error closing `%s'", strings);
 465 }
 466
 467 static void
 468 found(int new, off_t off, const char *str)
 469 {
 470         if (vflg == 0)
 471                 return;
 472         if (!new)
 473                 (void)fprintf(stderr, "found at %d:", (int) off);
 474         else
 475                 (void)fprintf(stderr, "new at %d:", (int) off);
 476         prstr(str);
 477         (void)fprintf(stderr, "\n");
 478 }
 479
 480 static void
 481 prstr(const char *cp)
 482 {
 483         int c;
 484
 485         while ((c = (*cp++ & 0377)) != '\0')
 486                 if (c < ' ')
 487                         (void)fprintf(stderr, "^%c", c + '`');
 488                 else if (c == 0177)
 489                         (void)fprintf(stderr, "^?");
 490                 else if (c > 0200)
 491                         (void)fprintf(stderr, "\\%03o", c);
 492                 else
 493                         (void)fprintf(stderr, "%c", c);
 494 }
 495
 496 static void
 497 xsdotc(void)
 498 {
 499         FILE *strf = fopen(strings, "r");
 500         FILE *xdotcf;
 501
 502         if (strf == NULL)
 503                 err(1, "Cannot open `%s'", strings);
 504         xdotcf = fopen("xs.c", "w");
 505         if (xdotcf == NULL)
 506                 err(1, "Cannot open `%s'", "xs.c");
 507         (void)fprintf(xdotcf, "char\t%s[] = {\n", array);
 508         for (;;) {
 509                 int i, c;
 510
 511                 for (i = 0; i < 8; i++) {
 512                         c = getc(strf);
 513                         if (ferror(strf)) {
 514                                 warn("Error reading `%s'", strings);
 515                                 onintr(1);
 516                         }
 517                         if (feof(strf)) {
 518                                 (void)fprintf(xdotcf, "\n");
 519                                 goto out;
 520                         }
 521                         (void)fprintf(xdotcf, "0x%02x,", c);
 522                 }
 523                 (void)fprintf(xdotcf, "\n");
 524         }
 525 out:
 526         (void)fprintf(xdotcf, "};\n");
 527         (void)fclose(xdotcf);
 528         (void)fclose(strf);
 529 }
 530
 531 static char
 532 lastchr(const char *cp)
 533 {
 534
 535         while (cp[0] && cp[1])
 536                 cp++;
 537         return (*cp);
 538 }
 539
 540 static int
 541 istail(const char *str, const char *of)
 542 {
 543         int d = strlen(of) - strlen(str);
 544
 545         if (d < 0 || strcmp(&of[d], str) != 0)
 546                 return (-1);
 547         return (d);
 548 }
 549
 550 static void
 551 onintr(int dummy)
 552 {
 553
 554         (void)signal(SIGINT, SIG_IGN);
 555         if (strings[0] == '/')
 556                 (void)unlink(strings);
 557         (void)unlink("x.c");
 558         (void)unlink("xs.c");
 559         exit(dummy);
 560 }
 561
 562 static void
 563 usage(void)
 564 {
 565
 566         (void)fprintf(stderr, "usage: %s [-cv] [-l array] [-] [<name> ...]\n",
 567             getprogname());
 568         exit(1);
 569 }