strfile.c revision 51287 
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
37#ifndef lint
38static const char copyright[] =
39"@(#) Copyright (c) 1989, 1993\n\
40	The Regents of the University of California.  All rights reserved.\n";
41#endif /* not lint */
42
43#ifndef lint
44#if 0
45static const char sccsid[] = "@(#)strfile.c   8.1 (Berkeley) 5/31/93";
46#else
47static const char rcsid[] =
48  "$FreeBSD: head/games/fortune/strfile/strfile.c 51287 1999-09-15 01:58:44Z peter $";
49#endif
50#endif /* not lint */
51
52# include	<sys/param.h>
53# include	<stdio.h>
54# include       <stdlib.h>
55# include	<ctype.h>
56# include       <string.h>
57# include       <time.h>
58# include       <locale.h>
59# include       <unistd.h>
60# include	"strfile.h"
61
62/*
63 *	This program takes a file composed of strings seperated by
64 * lines starting with two consecutive delimiting character (default
65 * character is '%') and creates another file which consists of a table
66 * describing the file (structure from "strfile.h"), a table of seek
67 * pointers to the start of the strings, and the strings, each terminated
68 * by a null byte.  Usage:
69 *
70 *	% strfile [-iorsx] [ -cC ] sourcefile [ datafile ]
71 *
72 *	c - Change delimiting character from '%' to 'C'
73 *	s - Silent.  Give no summary of data processed at the end of
74 *	    the run.
75 *	o - order the strings in alphabetic order
76 *	i - if ordering, ignore case
77 *	r - randomize the order of the strings
78 *	x - set rotated bit
79 *
80 *		Ken Arnold	Sept. 7, 1978 --
81 *
82 *	Added ordering options.
83 */
84
85# define	TRUE	1
86# define	FALSE	0
87
88# define	STORING_PTRS	(Oflag || Rflag)
89# define	CHUNKSIZE	512
90
91# define        ALLOC(ptr,sz) { \
92			if (ptr == NULL) \
93				ptr = malloc((unsigned int) (CHUNKSIZE * sizeof *ptr)); \
94			else if (((sz) + 1) % CHUNKSIZE == 0) \
95				ptr = realloc((void *) ptr, ((unsigned int) ((sz) + CHUNKSIZE) * sizeof *ptr)); \
96			if (ptr == NULL) { \
97				fprintf(stderr, "out of space\n"); \
98				exit(1); \
99			} \
100		}
101
102#ifdef NO_VOID
103# define	void	char
104#endif
105
106typedef struct {
107	char	first;
108	long    pos;
109} STR;
110
111char	*Infile		= NULL,		/* input file name */
112	Outfile[MAXPATHLEN] = "",	/* output file name */
113	Delimch		= '%';		/* delimiting character */
114
115int	Sflag		= FALSE;	/* silent run flag */
116int	Oflag		= FALSE;	/* ordering flag */
117int	Iflag		= FALSE;	/* ignore case flag */
118int	Rflag		= FALSE;	/* randomize order flag */
119int	Xflag		= FALSE;	/* set rotated bit */
120long	Num_pts		= 0;		/* number of pointers/strings */
121
122long    *Seekpts;
123
124FILE	*Sort_1, *Sort_2;		/* pointers for sorting */
125
126STRFILE	Tbl;				/* statistics table */
127
128STR	*Firstch;			/* first chars of each string */
129
130void getargs(), add_offset(), do_order(), randomize(), usage();
131int cmp_str();
132
133/*
134 * main:
135 *	Drive the sucker.  There are two main modes -- either we store
136 *	the seek pointers, if the table is to be sorted or randomized,
137 *	or we write the pointer directly to the file, if we are to stay
138 *	in file order.  If the former, we allocate and re-allocate in
139 *	CHUNKSIZE blocks; if the latter, we just write each pointer,
140 *	and then seek back to the beginning to write in the table.
141 */
142int main(ac, av)
143int	ac;
144char	**av;
145{
146	register char		*sp, dc;
147	register FILE		*inf, *outf;
148	register long           last_off, length, pos, *p;
149	register int		first, cnt;
150	register char		*nsp;
151	register STR		*fp;
152	static char		string[257];
153
154	(void) setlocale(LC_ALL, "");
155
156	getargs(ac, av);		/* evalute arguments */
157	dc = Delimch;
158	if ((inf = fopen(Infile, "r")) == NULL) {
159		perror(Infile);
160		exit(1);
161	}
162
163	if ((outf = fopen(Outfile, "w")) == NULL) {
164		perror(Outfile);
165		exit(1);
166	}
167	if (!STORING_PTRS)
168		(void) fseek(outf, (long) sizeof Tbl, 0);
169
170	/*
171	 * Write the strings onto the file
172	 */
173
174	Tbl.str_longlen = 0;
175	Tbl.str_shortlen = ~((unsigned long) 0);
176	Tbl.str_delim = dc;
177	Tbl.str_version = VERSION;
178	first = Oflag;
179	add_offset(outf, ftell(inf));
180	last_off = 0;
181	do {
182		sp = fgets(string, 256, inf);
183		if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
184			pos = ftell(inf);
185			length = pos - last_off - (sp ? strlen(sp) : 0);
186			last_off = pos;
187			if (!length)
188				continue;
189			add_offset(outf, pos);
190			if (Tbl.str_longlen < length)
191				Tbl.str_longlen = length;
192			if (Tbl.str_shortlen > length)
193				Tbl.str_shortlen = length;
194			first = Oflag;
195		}
196		else if (first) {
197			for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
198				continue;
199			ALLOC(Firstch, Num_pts);
200			fp = &Firstch[Num_pts - 1];
201			if (Iflag && isupper((unsigned char)*nsp))
202				fp->first = tolower((unsigned char)*nsp);
203			else
204				fp->first = *nsp;
205			fp->pos = Seekpts[Num_pts - 1];
206			first = FALSE;
207		}
208	} while (sp != NULL);
209
210	/*
211	 * write the tables in
212	 */
213
214	(void) fclose(inf);
215	Tbl.str_numstr = Num_pts - 1;
216
217	if (Oflag)
218		do_order();
219	else if (Rflag)
220		randomize();
221
222	if (Xflag)
223		Tbl.str_flags |= STR_ROTATED;
224
225	if (!Sflag) {
226		printf("\"%s\" created\n", Outfile);
227		if (Num_pts == 2)
228			puts("There was 1 string");
229		else
230			printf("There were %ld strings\n", Num_pts - 1);
231		printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
232		       Tbl.str_longlen == 1 ? "" : "s");
233		printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
234		       Tbl.str_shortlen == 1 ? "" : "s");
235	}
236
237	rewind(outf);
238	Tbl.str_version = htonl(Tbl.str_version);
239	Tbl.str_numstr = htonl(Tbl.str_numstr);
240	Tbl.str_longlen = htonl(Tbl.str_longlen);
241	Tbl.str_shortlen = htonl(Tbl.str_shortlen);
242	Tbl.str_flags = htonl(Tbl.str_flags);
243	(void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
244	if (STORING_PTRS) {
245		for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
246			*p = htonl(*p);
247		(void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
248	}
249	(void) fclose(outf);
250	exit(0);
251}
252
253/*
254 *	This routine evaluates arguments from the command line
255 */
256void getargs(argc, argv)
257int	argc;
258char	**argv;
259{
260	extern char	*optarg;
261	extern int	optind;
262	int	ch;
263
264	while ((ch = getopt(argc, argv, "c:iorsx")) != EOF)
265		switch(ch) {
266		case 'c':			/* new delimiting char */
267			Delimch = *optarg;
268			if (!isascii(Delimch)) {
269				printf("bad delimiting character: '\\%o\n'",
270				       (unsigned char)Delimch);
271			}
272			break;
273		case 'i':			/* ignore case in ordering */
274			Iflag++;
275			break;
276		case 'o':			/* order strings */
277			Oflag++;
278			break;
279		case 'r':			/* randomize pointers */
280			Rflag++;
281			break;
282		case 's':			/* silent */
283			Sflag++;
284			break;
285		case 'x':			/* set the rotated bit */
286			Xflag++;
287			break;
288		case '?':
289		default:
290			usage();
291		}
292	argv += optind;
293
294	if (*argv) {
295		Infile = *argv;
296		if (*++argv)
297			(void) strcpy(Outfile, *argv);
298	}
299	if (!Infile) {
300		puts("No input file name");
301		usage();
302	}
303	if (*Outfile == '\0') {
304		(void) strcpy(Outfile, Infile);
305		(void) strcat(Outfile, ".dat");
306	}
307}
308
309void usage()
310{
311	(void) fprintf(stderr,
312	    "strfile [-iorsx] [-c char] sourcefile [datafile]\n");
313	exit(1);
314}
315
316/*
317 * add_offset:
318 *	Add an offset to the list, or write it out, as appropriate.
319 */
320void add_offset(fp, off)
321FILE	*fp;
322long    off;
323{
324	long net;
325
326	if (!STORING_PTRS) {
327		net = htonl(off);
328		fwrite(&net, 1, sizeof net, fp);
329	} else {
330		ALLOC(Seekpts, Num_pts + 1);
331		Seekpts[Num_pts] = off;
332	}
333	Num_pts++;
334}
335
336/*
337 * do_order:
338 *	Order the strings alphabetically (possibly ignoring case).
339 */
340void do_order()
341{
342	register int	i;
343	register long   *lp;
344	register STR	*fp;
345
346	Sort_1 = fopen(Infile, "r");
347	Sort_2 = fopen(Infile, "r");
348	qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str);
349	i = Tbl.str_numstr;
350	lp = Seekpts;
351	fp = Firstch;
352	while (i--)
353		*lp++ = fp++->pos;
354	(void) fclose(Sort_1);
355	(void) fclose(Sort_2);
356	Tbl.str_flags |= STR_ORDERED;
357}
358
359static int collate_range_cmp (c1, c2)
360	int c1, c2;
361{
362	static char s1[2], s2[2];
363	int ret;
364
365	c1 &= UCHAR_MAX;
366	c2 &= UCHAR_MAX;
367	if (c1 == c2)
368		return (0);
369	s1[0] = c1;
370	s2[0] = c2;
371	if ((ret = strcoll(s1, s2)) != 0)
372		return (ret);
373	return (c1 - c2);
374}
375
376/*
377 * cmp_str:
378 *	Compare two strings in the file
379 */
380int cmp_str(p1, p2)
381STR	*p1, *p2;
382{
383	register int	c1, c2;
384	register int	n1, n2;
385	int r;
386
387# define	SET_N(nf,ch)	(nf = (ch == '\n'))
388# define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
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 */
432void randomize()
433{
434	register int	cnt, i;
435	register long   tmp;
436	register long   *sp;
437
438	srandomdev();
439
440	Tbl.str_flags |= STR_RANDOM;
441	cnt = Tbl.str_numstr;
442
443	/*
444	 * move things around randomly
445	 */
446
447	for (sp = Seekpts; cnt > 0; cnt--, sp++) {
448		i = random() % cnt;
449		tmp = sp[0];
450		sp[0] = sp[i];
451		sp[i] = tmp;
452	}
453}
454