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