strfile.c revision 132578 
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#if 0
38#ifndef lint
39static const char copyright[] =
40"@(#) Copyright (c) 1989, 1993\n\
41	The Regents of the University of California.  All rights reserved.\n";
42#endif /* not lint */
43
44#ifndef lint
45static const char sccsid[] = "@(#)strfile.c   8.1 (Berkeley) 5/31/93";
46#endif /* not lint */
47#endif
48#include <sys/cdefs.h>
49__FBSDID("$FreeBSD: head/games/fortune/strfile/strfile.c 132578 2004-07-23 11:36:35Z le $");
50
51# include	<sys/param.h>
52# include	<arpa/inet.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 separated 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 - Allow comments marked by a double delimiter at line's beginning
73 *	c - Change delimiting character from '%' to 'C'
74 *	s - Silent.  Give no summary of data processed at the end of
75 *	    the run.
76 *	o - order the strings in alphabetic order
77 *	i - if ordering, ignore case
78 *	r - randomize the order of the strings
79 *	x - set rotated bit
80 *
81 *		Ken Arnold	Sept. 7, 1978 --
82 *
83 *	Added ordering options.
84 */
85
86# define	TRUE	1
87# define	FALSE	0
88
89# define	STORING_PTRS	(Oflag || Rflag)
90# define	CHUNKSIZE	512
91
92# define        ALLOC(ptr,sz) { \
93			if (ptr == NULL) \
94				ptr = malloc((unsigned int) (CHUNKSIZE * sizeof *ptr)); \
95			else if (((sz) + 1) % CHUNKSIZE == 0) \
96				ptr = realloc((void *) ptr, ((unsigned int) ((sz) + CHUNKSIZE) * sizeof *ptr)); \
97			if (ptr == NULL) { \
98				fprintf(stderr, "out of space\n"); \
99				exit(1); \
100			} \
101		}
102
103#ifdef NO_VOID
104# define	void	char
105#endif
106
107typedef struct {
108	int	first;
109	long    pos;
110} STR;
111
112static char	*Infile		= NULL,		/* input file name */
113		Outfile[MAXPATHLEN] = "",	/* output file name */
114		Delimch		= '%';		/* delimiting character */
115
116static int	Cflag		= FALSE;	/* embedded comments */
117static int	Sflag		= FALSE;	/* silent run flag */
118static int	Oflag		= FALSE;	/* ordering flag */
119static int	Iflag		= FALSE;	/* ignore case flag */
120static int	Rflag		= FALSE;	/* randomize order flag */
121static int	Xflag		= FALSE;	/* set rotated bit */
122static long	Num_pts		= 0;		/* number of pointers/strings */
123
124static long    *Seekpts;
125
126static FILE	*Sort_1, *Sort_2;		/* pointers for sorting */
127
128static STRFILE	Tbl;				/* statistics table */
129
130static STR	*Firstch;			/* first chars of each string */
131
132static void	add_offset(FILE *, long);
133static int	cmp_str(const void *, const void *);
134static int      stable_collate_range_cmp(int, int);
135static void	do_order(void);
136static void	getargs(int, char **);
137static void	randomize(void);
138static void	usage(void);
139
140/*
141 * main:
142 *	Drive the sucker.  There are two main modes -- either we store
143 *	the seek pointers, if the table is to be sorted or randomized,
144 *	or we write the pointer directly to the file, if we are to stay
145 *	in file order.  If the former, we allocate and re-allocate in
146 *	CHUNKSIZE blocks; if the latter, we just write each pointer,
147 *	and then seek back to the beginning to write in the table.
148 */
149int main(int ac, char *av[])
150{
151	char		*sp, dc;
152	FILE		*inf, *outf;
153	long		last_off, pos, *p;
154	unsigned long	length;
155	int		first, cnt;
156	char		*nsp;
157	STR		*fp;
158	static char		string[257];
159
160	(void) setlocale(LC_ALL, "");
161
162	getargs(ac, av);		/* evalute arguments */
163	dc = Delimch;
164	if ((inf = fopen(Infile, "r")) == NULL) {
165		perror(Infile);
166		exit(1);
167	}
168
169	if ((outf = fopen(Outfile, "w")) == NULL) {
170		perror(Outfile);
171		exit(1);
172	}
173	if (!STORING_PTRS)
174		(void) fseek(outf, (long) sizeof Tbl, 0);
175
176	/*
177	 * Write the strings onto the file
178	 */
179
180	Tbl.str_longlen = 0;
181	Tbl.str_shortlen = ~((unsigned long) 0);
182	Tbl.str_delim = dc;
183	Tbl.str_version = VERSION;
184	first = Oflag;
185	add_offset(outf, ftell(inf));
186	last_off = 0;
187	do {
188		sp = fgets(string, 256, inf);
189		if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
190			pos = ftell(inf);
191			length = pos - last_off - (sp ? strlen(sp) : 0);
192			last_off = pos;
193			if (!length)
194				continue;
195			add_offset(outf, pos);
196			if (Tbl.str_longlen < length)
197				Tbl.str_longlen = length;
198			if (Tbl.str_shortlen > length)
199				Tbl.str_shortlen = length;
200			first = Oflag;
201		}
202		else if (first) {
203			for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
204				continue;
205			ALLOC(Firstch, Num_pts);
206			fp = &Firstch[Num_pts - 1];
207			if (Iflag && isupper((unsigned char)*nsp))
208				fp->first = tolower((unsigned char)*nsp);
209			else
210				fp->first = *nsp;
211			fp->pos = Seekpts[Num_pts - 1];
212			first = FALSE;
213		}
214	} while (sp != NULL);
215
216	/*
217	 * write the tables in
218	 */
219
220	(void) fclose(inf);
221	Tbl.str_numstr = Num_pts - 1;
222
223	if (Cflag)
224		Tbl.str_flags |= STR_COMMENTS;
225
226	if (Oflag)
227		do_order();
228	else if (Rflag)
229		randomize();
230
231	if (Xflag)
232		Tbl.str_flags |= STR_ROTATED;
233
234	if (!Sflag) {
235		printf("\"%s\" created\n", Outfile);
236		if (Num_pts == 2)
237			puts("There was 1 string");
238		else
239			printf("There were %ld strings\n", Num_pts - 1);
240		printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
241		       Tbl.str_longlen == 1 ? "" : "s");
242		printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
243		       Tbl.str_shortlen == 1 ? "" : "s");
244	}
245
246	rewind(outf);
247	Tbl.str_version = htonl(Tbl.str_version);
248	Tbl.str_numstr = htonl(Tbl.str_numstr);
249	Tbl.str_longlen = htonl(Tbl.str_longlen);
250	Tbl.str_shortlen = htonl(Tbl.str_shortlen);
251	Tbl.str_flags = htonl(Tbl.str_flags);
252	(void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
253	if (STORING_PTRS) {
254		for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
255			*p = htonl(*p);
256		(void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
257	}
258	(void) fclose(outf);
259	exit(0);
260}
261
262/*
263 *	This routine evaluates arguments from the command line
264 */
265void getargs(argc, argv)
266int	argc;
267char	**argv;
268{
269	int	ch;
270
271	while ((ch = getopt(argc, argv, "Cc:iorsx")) != EOF)
272		switch(ch) {
273		case 'C':			/* embedded comments */
274			Cflag++;
275			break;
276		case 'c':			/* new delimiting char */
277			Delimch = *optarg;
278			if (!isascii(Delimch)) {
279				printf("bad delimiting character: '\\%o\n'",
280				       (unsigned char)Delimch);
281			}
282			break;
283		case 'i':			/* ignore case in ordering */
284			Iflag++;
285			break;
286		case 'o':			/* order strings */
287			Oflag++;
288			break;
289		case 'r':			/* randomize pointers */
290			Rflag++;
291			break;
292		case 's':			/* silent */
293			Sflag++;
294			break;
295		case 'x':			/* set the rotated bit */
296			Xflag++;
297			break;
298		case '?':
299		default:
300			usage();
301		}
302	argv += optind;
303
304	if (*argv) {
305		Infile = *argv;
306		if (*++argv)
307			(void) strcpy(Outfile, *argv);
308	}
309	if (!Infile) {
310		puts("No input file name");
311		usage();
312	}
313	if (*Outfile == '\0') {
314		(void) strcpy(Outfile, Infile);
315		(void) strcat(Outfile, ".dat");
316	}
317}
318
319void usage()
320{
321	(void) fprintf(stderr,
322	    "strfile [-Ciorsx] [-c char] sourcefile [datafile]\n");
323	exit(1);
324}
325
326/*
327 * add_offset:
328 *	Add an offset to the list, or write it out, as appropriate.
329 */
330void add_offset(fp, off)
331FILE	*fp;
332long    off;
333{
334	long net;
335
336	if (!STORING_PTRS) {
337		net = htonl(off);
338		fwrite(&net, 1, sizeof net, fp);
339	} else {
340		ALLOC(Seekpts, Num_pts + 1);
341		Seekpts[Num_pts] = off;
342	}
343	Num_pts++;
344}
345
346/*
347 * do_order:
348 *	Order the strings alphabetically (possibly ignoring case).
349 */
350void do_order()
351{
352	int	i;
353	long   *lp;
354	STR	*fp;
355
356	Sort_1 = fopen(Infile, "r");
357	Sort_2 = fopen(Infile, "r");
358	qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str);
359	i = Tbl.str_numstr;
360	lp = Seekpts;
361	fp = Firstch;
362	while (i--)
363		*lp++ = fp++->pos;
364	(void) fclose(Sort_1);
365	(void) fclose(Sort_2);
366	Tbl.str_flags |= STR_ORDERED;
367}
368
369static int stable_collate_range_cmp(c1, c2)
370	int c1, c2;
371{
372	static char s1[2], s2[2];
373	int ret;
374
375	s1[0] = c1;
376	s2[0] = c2;
377	if ((ret = strcoll(s1, s2)) != 0)
378		return (ret);
379	return (c1 - c2);
380}
381
382/*
383 * cmp_str:
384 *	Compare two strings in the file
385 */
386int cmp_str(s1, s2)
387const void	*s1, *s2;
388{
389	const STR	*p1, *p2;
390	int	c1, c2;
391	int	n1, n2;
392	int r;
393
394# define	SET_N(nf,ch)	(nf = (ch == '\n'))
395# define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
396
397	p1 = (const STR *) s1;
398	p2 = (const STR *) s2;
399
400	c1 = (unsigned char) p1->first;
401	c2 = (unsigned char) p2->first;
402	if ((r = stable_collate_range_cmp(c1, c2)) != 0)
403		return (r);
404
405	(void) fseek(Sort_1, p1->pos, 0);
406	(void) fseek(Sort_2, p2->pos, 0);
407
408	n1 = FALSE;
409	n2 = FALSE;
410	while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
411		SET_N(n1, c1);
412	while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
413		SET_N(n2, c2);
414
415	while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
416		if (Iflag) {
417			if (isupper(c1))
418				c1 = tolower(c1);
419			if (isupper(c2))
420				c2 = tolower(c2);
421		}
422		if ((r = stable_collate_range_cmp(c1, c2)) != 0)
423			return (r);
424		SET_N(n1, c1);
425		SET_N(n2, c2);
426		c1 = getc(Sort_1);
427		c2 = getc(Sort_2);
428	}
429	if (IS_END(c1, n1))
430		c1 = 0;
431	if (IS_END(c2, n2))
432		c2 = 0;
433	return (stable_collate_range_cmp(c1, c2));
434}
435
436/*
437 * randomize:
438 *	Randomize the order of the string table.  We must be careful
439 *	not to randomize across delimiter boundaries.  All
440 *	randomization is done within each block.
441 */
442void randomize()
443{
444	int	cnt, i;
445	long   tmp;
446	long   *sp;
447
448	srandomdev();
449
450	Tbl.str_flags |= STR_RANDOM;
451	cnt = Tbl.str_numstr;
452
453	/*
454	 * move things around randomly
455	 */
456
457	for (sp = Seekpts; cnt > 0; cnt--, sp++) {
458		i = random() % cnt;
459		tmp = sp[0];
460		sp[0] = sp[i];
461		sp[i] = tmp;
462	}
463}
464