2 * Copyright (c) 1989, 1993
3 * The Regents of the University of California. All rights reserved.
5 * This code is derived from software contributed to Berkeley by
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
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.
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
35 static const char copyright[] =
36 "@(#) Copyright (c) 1989, 1993\n\
37 The Regents of the University of California. All rights reserved.\n";
41 static const char sccsid[] = "@(#)strfile.c 8.1 (Berkeley) 5/31/93";
44 #include <sys/cdefs.h>
45 __FBSDID("$FreeBSD$");
47 #include <sys/param.h>
48 #include <sys/endian.h>
61 * This program takes a file composed of strings separated by
62 * lines starting with two consecutive delimiting character (default
63 * character is '%') and creates another file which consists of a table
64 * describing the file (structure from "strfile.h"), a table of seek
65 * pointers to the start of the strings, and the strings, each terminated
66 * by a null byte. Usage:
68 * % strfile [-iorsx] [ -cC ] sourcefile [ datafile ]
70 * C - Allow comments marked by a double delimiter at line's beginning
71 * c - Change delimiting character from '%' to 'C'
72 * s - Silent. Give no summary of data processed at the end of
74 * o - order the strings in alphabetic order
75 * i - if ordering, ignore case
76 * r - randomize the order of the strings
79 * Ken Arnold Sept. 7, 1978 --
81 * Added ordering options.
84 #define STORING_PTRS (Oflag || Rflag)
87 #define ALLOC(ptr, sz) do { \
89 ptr = malloc(CHUNKSIZE * sizeof(*ptr)); \
90 else if (((sz) + 1) % CHUNKSIZE == 0) \
91 ptr = realloc(ptr, ((sz) + CHUNKSIZE) * sizeof(*ptr)); \
93 fprintf(stderr, "out of space\n"); \
103 static char *Infile = NULL, /* input file name */
104 Outfile[MAXPATHLEN] = "", /* output file name */
105 Delimch = '%'; /* delimiting character */
107 static int Cflag = false; /* embedded comments */
108 static int Sflag = false; /* silent run flag */
109 static int Oflag = false; /* ordering flag */
110 static int Iflag = false; /* ignore case flag */
111 static int Rflag = false; /* randomize order flag */
112 static int Xflag = false; /* set rotated bit */
113 static uint32_t Num_pts = 0; /* number of pointers/strings */
115 static off_t *Seekpts;
117 static FILE *Sort_1, *Sort_2; /* pointers for sorting */
119 static STRFILE Tbl; /* statistics table */
121 static STR *Firstch; /* first chars of each string */
123 static void add_offset(FILE *, off_t);
124 static int cmp_str(const void *, const void *);
125 static int stable_collate_range_cmp(int, int);
126 static void do_order(void);
127 static void getargs(int, char **);
128 static void randomize(void);
129 static void usage(void);
133 * Drive the sucker. There are two main modes -- either we store
134 * the seek pointers, if the table is to be sorted or randomized,
135 * or we write the pointer directly to the file, if we are to stay
136 * in file order. If the former, we allocate and re-allocate in
137 * CHUNKSIZE blocks; if the latter, we just write each pointer,
138 * and then seek back to the beginning to write in the table.
141 main(int ac, char *av[])
145 off_t last_off, pos, *p;
150 static char string[257];
152 setlocale(LC_ALL, "");
154 getargs(ac, av); /* evalute arguments */
156 if ((inf = fopen(Infile, "r")) == NULL) {
161 if ((outf = fopen(Outfile, "w")) == NULL) {
166 fseek(outf, (long)sizeof(Tbl), SEEK_SET);
169 * Write the strings onto the file
173 Tbl.str_shortlen = 0xffffffff;
175 Tbl.str_version = VERSION;
177 add_offset(outf, ftello(inf));
180 sp = fgets(string, 256, inf);
181 if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
183 length = (size_t)(pos - last_off) -
184 (sp != NULL ? strlen(sp) : 0);
188 add_offset(outf, pos);
189 if ((size_t)Tbl.str_longlen < length)
190 Tbl.str_longlen = length;
191 if ((size_t)Tbl.str_shortlen > length)
192 Tbl.str_shortlen = length;
196 for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
198 ALLOC(Firstch, Num_pts);
199 fp = &Firstch[Num_pts - 1];
200 if (Iflag && isupper((unsigned char)*nsp))
201 fp->first = tolower((unsigned char)*nsp);
204 fp->pos = Seekpts[Num_pts - 1];
207 } while (sp != NULL);
210 * write the tables in
214 Tbl.str_numstr = Num_pts - 1;
217 Tbl.str_flags |= STR_COMMENTS;
225 Tbl.str_flags |= STR_ROTATED;
228 printf("\"%s\" created\n", Outfile);
230 puts("There was 1 string");
232 printf("There were %u strings\n", Num_pts - 1);
233 printf("Longest string: %u byte%s\n", Tbl.str_longlen,
234 Tbl.str_longlen == 1 ? "" : "s");
235 printf("Shortest string: %u byte%s\n", Tbl.str_shortlen,
236 Tbl.str_shortlen == 1 ? "" : "s");
240 Tbl.str_version = htobe32(Tbl.str_version);
241 Tbl.str_numstr = htobe32(Tbl.str_numstr);
242 Tbl.str_longlen = htobe32(Tbl.str_longlen);
243 Tbl.str_shortlen = htobe32(Tbl.str_shortlen);
244 Tbl.str_flags = htobe32(Tbl.str_flags);
245 fwrite((char *)&Tbl, sizeof(Tbl), 1, outf);
247 for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
249 fwrite(Seekpts, sizeof(*Seekpts), (size_t)Num_pts, outf);
256 * This routine evaluates arguments from the command line
259 getargs(int argc, char **argv)
263 while ((ch = getopt(argc, argv, "Cc:iorsx")) != -1)
265 case 'C': /* embedded comments */
268 case 'c': /* new delimiting char */
270 if (!isascii(Delimch)) {
271 printf("bad delimiting character: '\\%o\n'",
272 (unsigned char)Delimch);
275 case 'i': /* ignore case in ordering */
278 case 'o': /* order strings */
281 case 'r': /* randomize pointers */
284 case 's': /* silent */
287 case 'x': /* set the rotated bit */
299 strcpy(Outfile, *argv);
302 puts("No input file name");
305 if (*Outfile == '\0') {
306 strcpy(Outfile, Infile);
307 strcat(Outfile, ".dat");
315 "strfile [-Ciorsx] [-c char] source_file [output_file]\n");
321 * Add an offset to the list, or write it out, as appropriate.
324 add_offset(FILE *fp, off_t off)
329 beoff = htobe64(off);
330 fwrite(&beoff, 1, sizeof(beoff), fp);
332 ALLOC(Seekpts, Num_pts + 1);
333 Seekpts[Num_pts] = off;
340 * Order the strings alphabetically (possibly ignoring case).
349 Sort_1 = fopen(Infile, "r");
350 Sort_2 = fopen(Infile, "r");
351 qsort(Firstch, (size_t)Tbl.str_numstr, sizeof(*Firstch), cmp_str);
359 Tbl.str_flags |= STR_ORDERED;
363 stable_collate_range_cmp(int c1, int c2)
365 static char s1[2], s2[2];
370 if ((ret = strcoll(s1, s2)) != 0)
377 * Compare two strings in the file
380 cmp_str(const void *s1, const void *s2)
383 int c1, c2, n1, n2, r;
385 #define SET_N(nf,ch) (nf = (ch == '\n'))
386 #define IS_END(ch,nf) (ch == EOF || (ch == (unsigned char)Delimch && nf))
388 p1 = (const STR *)s1;
389 p2 = (const STR *)s2;
391 c1 = (unsigned char)p1->first;
392 c2 = (unsigned char)p2->first;
393 if ((r = stable_collate_range_cmp(c1, c2)) != 0)
396 fseeko(Sort_1, p1->pos, SEEK_SET);
397 fseeko(Sort_2, p2->pos, SEEK_SET);
401 while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
403 while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
406 while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
413 if ((r = stable_collate_range_cmp(c1, c2)) != 0)
425 return (stable_collate_range_cmp(c1, c2));
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.
441 #if __FreeBSD_version < 800041
445 Tbl.str_flags |= STR_RANDOM;
446 cnt = Tbl.str_numstr;
449 * move things around randomly
452 for (sp = Seekpts; cnt > 0; cnt--, sp++) {
453 #if __FreeBSD_version < 800041
456 i = arc4random_uniform(cnt);