games/fortune/strfile/strfile.c

   1 /*-
   2  * Copyright (c) 1989, 1993
   3  *      The Regents of the University of California.  All rights reserved.
   4  *
   5  * This code is derived from software contributed to Berkeley by
   6  * Ken Arnold.
   7  *
   8  * Redistribution and use in source and binary forms, with or without
   9  * modification, are permitted provided that the following conditions
  10  * are met:
  11  * 1. Redistributions of source code must retain the above copyright
  12  *    notice, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  * 3. All advertising materials mentioning features or use of this software
  17  *    must display the following acknowledgement:
  18  *      This product includes software developed by the University of
  19  *      California, Berkeley and its contributors.
  20  * 4. Neither the name of the University nor the names of its contributors
  21  *    may be used to endorse or promote products derived from this software
  22  *    without specific prior written permission.
  23  *
  24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  34  * SUCH DAMAGE.
  35  *
  36  * $FreeBSD: src/games/fortune/strfile/strfile.c,v 1.15.2.2 2001/03/05 11:52:37 kris Exp $
  37  * $DragonFly: src/games/fortune/strfile/strfile.c,v 1.6 2008/07/10 18:29:51 swildner Exp $
  38  *
  39  * @(#) Copyright (c) 1989, 1993 The Regents of the University of California.  All rights reserved.
  40  * @(#)strfile.c   8.1 (Berkeley) 5/31/93
  41  * $FreeBSD: src/games/fortune/strfile/strfile.c,v 1.15.2.2 2001/03/05 11:52:37 kris Exp $
  42  */
  43
  44 # include       <sys/param.h>
  45 # include       <stdbool.h>
  46 # include       <stdio.h>
  47 # include       <stdlib.h>
  48 # include       <ctype.h>
  49 # include       <string.h>
  50 # include       <time.h>
  51 # include       <locale.h>
  52 # include       <unistd.h>
  53 # include       "strfile.h"
  54
  55 /*
  56  *      This program takes a file composed of strings separated by
  57  * lines starting with two consecutive delimiting character (default
  58  * character is '%') and creates another file which consists of a table
  59  * describing the file (structure from "strfile.h"), a table of seek
  60  * pointers to the start of the strings, and the strings, each terminated
  61  * by a null byte.  Usage:
  62  *
  63  *      % strfile [-iorsx] [ -cC ] sourcefile [ datafile ]
  64  *
  65  *      C - Allow comments marked by a double delimiter at line's beginning
  66  *      c - Change delimiting character from '%' to 'C'
  67  *      s - Silent.  Give no summary of data processed at the end of
  68  *          the run.
  69  *      o - order the strings in alphabetic order
  70  *      i - if ordering, ignore case
  71  *      r - randomize the order of the strings
  72  *      x - set rotated bit
  73  *
  74  *              Ken Arnold      Sept. 7, 1978 --
  75  *
  76  *      Added ordering options.
  77  */
  78
  79 # define        STORING_PTRS    (Oflag || Rflag)
  80 # define        CHUNKSIZE       512
  81
  82 # define        ALLOC(ptr,sz) { \
  83                         if (ptr == NULL) \
  84                                 ptr = malloc((unsigned int) (CHUNKSIZE * sizeof *ptr)); \
  85                         else if (((sz) + 1) % CHUNKSIZE == 0) \
  86                                 ptr = realloc((void *) ptr, ((unsigned int) ((sz) + CHUNKSIZE) * sizeof *ptr)); \
  87                         if (ptr == NULL) { \
  88                                 fprintf(stderr, "out of space\n"); \
  89                                 exit(1); \
  90                         } \
  91                 }
  92
  93 typedef struct {
  94         char    first;
  95         long    pos;
  96 } STR;
  97
  98 char    *Infile         = NULL,         /* input file name */
  99         Outfile[MAXPATHLEN] = "",       /* output file name */
 100         Delimch         = '%';          /* delimiting character */
 101
 102 int     Cflag           = false;        /* embedded comments */
 103 int     Sflag           = false;        /* silent run flag */
 104 int     Oflag           = false;        /* ordering flag */
 105 int     Iflag           = false;        /* ignore case flag */
 106 int     Rflag           = false;        /* randomize order flag */
 107 int     Xflag           = false;        /* set rotated bit */
 108 long    Num_pts         = 0;            /* number of pointers/strings */
 109
 110 long    *Seekpts;
 111
 112 FILE    *Sort_1, *Sort_2;               /* pointers for sorting */
 113
 114 STRFILE Tbl;                            /* statistics table */
 115
 116 STR     *Firstch;                       /* first chars of each string */
 117
 118 void    add_offset (FILE *, long);
 119 int     cmp_str (const void *, const void *);
 120 static int      collate_range_cmp  (int, int);
 121 void    do_order (void);
 122 void    getargs (int, char **);
 123 void    randomize (void);
 124 void    usage (void);
 125
 126 /*
 127  * main:
 128  *      Drive the sucker.  There are two main modes -- either we store
 129  *      the seek pointers, if the table is to be sorted or randomized,
 130  *      or we write the pointer directly to the file, if we are to stay
 131  *      in file order.  If the former, we allocate and re-allocate in
 132  *      CHUNKSIZE blocks; if the latter, we just write each pointer,
 133  *      and then seek back to the beginning to write in the table.
 134  */
 135 int
 136 main(int ac, char **av)
 137 {
 138         char            *sp, dc;
 139         FILE            *inf, *outf;
 140         long           last_off, length, pos, *p;
 141         int             first, cnt;
 142         char            *nsp;
 143         STR             *fp;
 144         static char             string[257];
 145
 146         (void) setlocale(LC_ALL, "");
 147
 148         getargs(ac, av);                /* evalute arguments */
 149         dc = Delimch;
 150         if ((inf = fopen(Infile, "r")) == NULL) {
 151                 perror(Infile);
 152                 exit(1);
 153         }
 154
 155         if ((outf = fopen(Outfile, "w")) == NULL) {
 156                 perror(Outfile);
 157                 exit(1);
 158         }
 159         if (!STORING_PTRS)
 160                 (void) fseek(outf, (long) sizeof Tbl, 0);
 161
 162         /*
 163          * Write the strings onto the file
 164          */
 165
 166         Tbl.str_longlen = 0;
 167         Tbl.str_shortlen = ~((unsigned long) 0);
 168         Tbl.str_delim = dc;
 169         Tbl.str_version = VERSION;
 170         first = Oflag;
 171         add_offset(outf, ftell(inf));
 172         last_off = 0;
 173         do {
 174                 sp = fgets(string, 256, inf);
 175                 if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
 176                         pos = ftell(inf);
 177                         length = pos - last_off - (sp ? strlen(sp) : 0);
 178                         last_off = pos;
 179                         if (!length)
 180                                 continue;
 181                         add_offset(outf, pos);
 182                         if (Tbl.str_longlen < (unsigned long)length)
 183                                 Tbl.str_longlen = length;
 184                         if (Tbl.str_shortlen > (unsigned long)length)
 185                                 Tbl.str_shortlen = length;
 186                         first = Oflag;
 187                 }
 188                 else if (first) {
 189                         for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
 190                                 continue;
 191                         ALLOC(Firstch, Num_pts);
 192                         fp = &Firstch[Num_pts - 1];
 193                         if (Iflag && isupper((unsigned char)*nsp))
 194                                 fp->first = tolower((unsigned char)*nsp);
 195                         else
 196                                 fp->first = *nsp;
 197                         fp->pos = Seekpts[Num_pts - 1];
 198                         first = false;
 199                 }
 200         } while (sp != NULL);
 201
 202         /*
 203          * write the tables in
 204          */
 205
 206         (void) fclose(inf);
 207         Tbl.str_numstr = Num_pts - 1;
 208
 209         if (Cflag)
 210                 Tbl.str_flags |= STR_COMMENTS;
 211
 212         if (Oflag)
 213                 do_order();
 214         else if (Rflag)
 215                 randomize();
 216
 217         if (Xflag)
 218                 Tbl.str_flags |= STR_ROTATED;
 219
 220         if (!Sflag) {
 221                 printf("\"%s\" created\n", Outfile);
 222                 if (Num_pts == 2)
 223                         puts("There was 1 string");
 224                 else
 225                         printf("There were %ld strings\n", Num_pts - 1);
 226                 printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
 227                        Tbl.str_longlen == 1 ? "" : "s");
 228                 printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
 229                        Tbl.str_shortlen == 1 ? "" : "s");
 230         }
 231
 232         rewind(outf);
 233         Tbl.str_version = htonl(Tbl.str_version);
 234         Tbl.str_numstr = htonl(Tbl.str_numstr);
 235         Tbl.str_longlen = htonl(Tbl.str_longlen);
 236         Tbl.str_shortlen = htonl(Tbl.str_shortlen);
 237         Tbl.str_flags = htonl(Tbl.str_flags);
 238         (void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
 239         if (STORING_PTRS) {
 240                 for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
 241                         *p = htonl(*p);
 242                 (void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
 243         }
 244         (void) fclose(outf);
 245         exit(0);
 246 }
 247
 248 /*
 249  *      This routine evaluates arguments from the command line
 250  */
 251 void
 252 getargs(int argc, char **argv)
 253 {
 254         int     ch;
 255
 256         while ((ch = getopt(argc, argv, "Cc:iorsx")) != -1)
 257                 switch(ch) {
 258                 case 'C':                       /* embedded comments */
 259                         Cflag++;
 260                         break;
 261                 case 'c':                       /* new delimiting char */
 262                         Delimch = *optarg;
 263                         if (!isascii(Delimch)) {
 264                                 printf("bad delimiting character: '\\%o\n'",
 265                                        (unsigned char)Delimch);
 266                         }
 267                         break;
 268                 case 'i':                       /* ignore case in ordering */
 269                         Iflag++;
 270                         break;
 271                 case 'o':                       /* order strings */
 272                         Oflag++;
 273                         break;
 274                 case 'r':                       /* randomize pointers */
 275                         Rflag++;
 276                         break;
 277                 case 's':                       /* silent */
 278                         Sflag++;
 279                         break;
 280                 case 'x':                       /* set the rotated bit */
 281                         Xflag++;
 282                         break;
 283                 case '?':
 284                 default:
 285                         usage();
 286                 }
 287         argv += optind;
 288
 289         if (*argv) {
 290                 Infile = *argv;
 291                 if (*++argv)
 292                         (void) strcpy(Outfile, *argv);
 293         }
 294         if (!Infile) {
 295                 puts("No input file name");
 296                 usage();
 297         }
 298         if (*Outfile == '\0') {
 299                 (void) strcpy(Outfile, Infile);
 300                 (void) strcat(Outfile, ".dat");
 301         }
 302 }
 303
 304 void
 305 usage(void)
 306 {
 307         (void) fprintf(stderr,
 308             "strfile [-Ciorsx] [-c char] sourcefile [datafile]\n");
 309         exit(1);
 310 }
 311
 312 /*
 313  * add_offset:
 314  *      Add an offset to the list, or write it out, as appropriate.
 315  */
 316 void
 317 add_offset(FILE *fp, long off)
 318 {
 319         long net;
 320
 321         if (!STORING_PTRS) {
 322                 net = htonl(off);
 323                 fwrite(&net, 1, sizeof net, fp);
 324         } else {
 325                 ALLOC(Seekpts, Num_pts + 1);
 326                 Seekpts[Num_pts] = off;
 327         }
 328         Num_pts++;
 329 }
 330
 331 /*
 332  * do_order:
 333  *      Order the strings alphabetically (possibly ignoring case).
 334  */
 335 void
 336 do_order(void)
 337 {
 338         int     i;
 339         long   *lp;
 340         STR     *fp;
 341
 342         Sort_1 = fopen(Infile, "r");
 343         Sort_2 = fopen(Infile, "r");
 344         qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str);
 345         i = Tbl.str_numstr;
 346         lp = Seekpts;
 347         fp = Firstch;
 348         while (i--)
 349                 *lp++ = fp++->pos;
 350         (void) fclose(Sort_1);
 351         (void) fclose(Sort_2);
 352         Tbl.str_flags |= STR_ORDERED;
 353 }
 354
 355 static int
 356 collate_range_cmp (int c1, int c2)
 357 {
 358         static char s1[2], s2[2];
 359         int ret;
 360
 361         c1 &= UCHAR_MAX;
 362         c2 &= UCHAR_MAX;
 363         if (c1 == c2)
 364                 return (0);
 365         s1[0] = c1;
 366         s2[0] = c2;
 367         if ((ret = strcoll(s1, s2)) != 0)
 368                 return (ret);
 369         return (c1 - c2);
 370 }
 371
 372 /*
 373  * cmp_str:
 374  *      Compare two strings in the file
 375  */
 376 int
 377 cmp_str(const void *s1, const void  *s2)
 378 {
 379         const STR       *p1, *p2;
 380         int     c1, c2;
 381         int     n1, n2;
 382         int r;
 383
 384 # define        SET_N(nf,ch)    (nf = (ch == '\n'))
 385 # define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
 386
 387         p1 = (const STR *) s1;
 388         p2 = (const STR *) s2;
 389
 390         c1 = (unsigned char) p1->first;
 391         c2 = (unsigned char) p2->first;
 392         if ((r = collate_range_cmp(c1, c2)) != 0)
 393                 return r;
 394
 395         (void) fseek(Sort_1, p1->pos, 0);
 396         (void) fseek(Sort_2, p2->pos, 0);
 397
 398         n1 = false;
 399         n2 = false;
 400         while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
 401                 SET_N(n1, c1);
 402         while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
 403                 SET_N(n2, c2);
 404
 405         while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
 406                 if (Iflag) {
 407                         if (isupper(c1))
 408                                 c1 = tolower(c1);
 409                         if (isupper(c2))
 410                                 c2 = tolower(c2);
 411                 }
 412                 if ((r = collate_range_cmp(c1, c2)) != 0)
 413                         return r;
 414                 SET_N(n1, c1);
 415                 SET_N(n2, c2);
 416                 c1 = getc(Sort_1);
 417                 c2 = getc(Sort_2);
 418         }
 419         if (IS_END(c1, n1))
 420                 c1 = 0;
 421         if (IS_END(c2, n2))
 422                 c2 = 0;
 423         return collate_range_cmp(c1, c2);
 424 }
 425
 426 /*
 427  * randomize:
 428  *      Randomize the order of the string table.  We must be careful
 429  *      not to randomize across delimiter boundaries.  All
 430  *      randomization is done within each block.
 431  */
 432 void
 433 randomize(void)
 434 {
 435         int     cnt, i;
 436         long   tmp;
 437         long   *sp;
 438
 439         srandomdev();
 440
 441         Tbl.str_flags |= STR_RANDOM;
 442         cnt = Tbl.str_numstr;
 443
 444         /*
 445          * move things around randomly
 446          */
 447
 448         for (sp = Seekpts; cnt > 0; cnt--, sp++) {
 449                 i = random() % cnt;
 450                 tmp = sp[0];
 451                 sp[0] = sp[i];
 452                 sp[i] = tmp;
 453         }
 454 }