strfile.c revision 62671 
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: head/games/fortune/strfile/strfile.c 62671 2000-07-06 05:46:37Z billf $
37 */
38
39#ifndef lint
40static const char copyright[] =
41"@(#) Copyright (c) 1989, 1993\n\
42	The Regents of the University of California.  All rights reserved.\n";
43#endif /* not lint */
44
45#ifndef lint
46#if 0
47static const char sccsid[] = "@(#)strfile.c   8.1 (Berkeley) 5/31/93";
48#else
49static const char rcsid[] =
50  "$FreeBSD: head/games/fortune/strfile/strfile.c 62671 2000-07-06 05:46:37Z billf $";
51#endif
52#endif /* not lint */
53
54# include	<sys/param.h>
55# include	<stdio.h>
56# include       <stdlib.h>
57# include	<ctype.h>
58# include       <string.h>
59# include       <time.h>
60# include       <locale.h>
61# include       <unistd.h>
62# include	"strfile.h"
63
64/*
65 *	This program takes a file composed of strings seperated by
66 * lines starting with two consecutive delimiting character (default
67 * character is '%') and creates another file which consists of a table
68 * describing the file (structure from "strfile.h"), a table of seek
69 * pointers to the start of the strings, and the strings, each terminated
70 * by a null byte.  Usage:
71 *
72 *	% strfile [-iorsx] [ -cC ] sourcefile [ datafile ]
73 *
74 *	C - Allow comments marked by a double delimiter at line's beginning
75 *	c - Change delimiting character from '%' to 'C'
76 *	s - Silent.  Give no summary of data processed at the end of
77 *	    the run.
78 *	o - order the strings in alphabetic order
79 *	i - if ordering, ignore case
80 *	r - randomize the order of the strings
81 *	x - set rotated bit
82 *
83 *		Ken Arnold	Sept. 7, 1978 --
84 *
85 *	Added ordering options.
86 */
87
88# define	TRUE	1
89# define	FALSE	0
90
91# define	STORING_PTRS	(Oflag || Rflag)
92# define	CHUNKSIZE	512
93
94# define        ALLOC(ptr,sz) { \
95			if (ptr == NULL) \
96				ptr = malloc((unsigned int) (CHUNKSIZE * sizeof *ptr)); \
97			else if (((sz) + 1) % CHUNKSIZE == 0) \
98				ptr = realloc((void *) ptr, ((unsigned int) ((sz) + CHUNKSIZE) * sizeof *ptr)); \
99			if (ptr == NULL) { \
100				fprintf(stderr, "out of space\n"); \
101				exit(1); \
102			} \
103		}
104
105#ifdef NO_VOID
106# define	void	char
107#endif
108
109typedef struct {
110	char	first;
111	long    pos;
112} STR;
113
114char	*Infile		= NULL,		/* input file name */
115	Outfile[MAXPATHLEN] = "",	/* output file name */
116	Delimch		= '%';		/* delimiting character */
117
118int	Cflag		= FALSE;	/* embedded comments */
119int	Sflag		= FALSE;	/* silent run flag */
120int	Oflag		= FALSE;	/* ordering flag */
121int	Iflag		= FALSE;	/* ignore case flag */
122int	Rflag		= FALSE;	/* randomize order flag */
123int	Xflag		= FALSE;	/* set rotated bit */
124long	Num_pts		= 0;		/* number of pointers/strings */
125
126long    *Seekpts;
127
128FILE	*Sort_1, *Sort_2;		/* pointers for sorting */
129
130STRFILE	Tbl;				/* statistics table */
131
132STR	*Firstch;			/* first chars of each string */
133
134void	add_offset __P((FILE *, long));
135int	cmp_str __P((const void *, const void *));
136static int	collate_range_cmp  __P((int, int));
137void	do_order __P((void));
138void	getargs __P((int, char **));
139void	randomize __P((void));
140void	usage __P((void));
141
142/*
143 * main:
144 *	Drive the sucker.  There are two main modes -- either we store
145 *	the seek pointers, if the table is to be sorted or randomized,
146 *	or we write the pointer directly to the file, if we are to stay
147 *	in file order.  If the former, we allocate and re-allocate in
148 *	CHUNKSIZE blocks; if the latter, we just write each pointer,
149 *	and then seek back to the beginning to write in the table.
150 */
151int main(ac, av)
152int	ac;
153char	**av;
154{
155	char		*sp, dc;
156	FILE		*inf, *outf;
157	long           last_off, length, pos, *p;
158	int		first, cnt;
159	char		*nsp;
160	STR		*fp;
161	static char		string[257];
162
163	(void) setlocale(LC_ALL, "");
164
165	getargs(ac, av);		/* evalute arguments */
166	dc = Delimch;
167	if ((inf = fopen(Infile, "r")) == NULL) {
168		perror(Infile);
169		exit(1);
170	}
171
172	if ((outf = fopen(Outfile, "w")) == NULL) {
173		perror(Outfile);
174		exit(1);
175	}
176	if (!STORING_PTRS)
177		(void) fseek(outf, (long) sizeof Tbl, 0);
178
179	/*
180	 * Write the strings onto the file
181	 */
182
183	Tbl.str_longlen = 0;
184	Tbl.str_shortlen = ~((unsigned long) 0);
185	Tbl.str_delim = dc;
186	Tbl.str_version = VERSION;
187	first = Oflag;
188	add_offset(outf, ftell(inf));
189	last_off = 0;
190	do {
191		sp = fgets(string, 256, inf);
192		if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
193			pos = ftell(inf);
194			length = pos - last_off - (sp ? strlen(sp) : 0);
195			last_off = pos;
196			if (!length)
197				continue;
198			add_offset(outf, pos);
199			if (Tbl.str_longlen < length)
200				Tbl.str_longlen = length;
201			if (Tbl.str_shortlen > length)
202				Tbl.str_shortlen = length;
203			first = Oflag;
204		}
205		else if (first) {
206			for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
207				continue;
208			ALLOC(Firstch, Num_pts);
209			fp = &Firstch[Num_pts - 1];
210			if (Iflag && isupper((unsigned char)*nsp))
211				fp->first = tolower((unsigned char)*nsp);
212			else
213				fp->first = *nsp;
214			fp->pos = Seekpts[Num_pts - 1];
215			first = FALSE;
216		}
217	} while (sp != NULL);
218
219	/*
220	 * write the tables in
221	 */
222
223	(void) fclose(inf);
224	Tbl.str_numstr = Num_pts - 1;
225
226	if (Cflag)
227		Tbl.str_flags |= STR_COMMENTS;
228
229	if (Oflag)
230		do_order();
231	else if (Rflag)
232		randomize();
233
234	if (Xflag)
235		Tbl.str_flags |= STR_ROTATED;
236
237	if (!Sflag) {
238		printf("\"%s\" created\n", Outfile);
239		if (Num_pts == 2)
240			puts("There was 1 string");
241		else
242			printf("There were %ld strings\n", Num_pts - 1);
243		printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
244		       Tbl.str_longlen == 1 ? "" : "s");
245		printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
246		       Tbl.str_shortlen == 1 ? "" : "s");
247	}
248
249	rewind(outf);
250	Tbl.str_version = htonl(Tbl.str_version);
251	Tbl.str_numstr = htonl(Tbl.str_numstr);
252	Tbl.str_longlen = htonl(Tbl.str_longlen);
253	Tbl.str_shortlen = htonl(Tbl.str_shortlen);
254	Tbl.str_flags = htonl(Tbl.str_flags);
255	(void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
256	if (STORING_PTRS) {
257		for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
258			*p = htonl(*p);
259		(void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
260	}
261	(void) fclose(outf);
262	exit(0);
263}
264
265/*
266 *	This routine evaluates arguments from the command line
267 */
268void getargs(argc, argv)
269int	argc;
270char	**argv;
271{
272	int	ch;
273
274	while ((ch = getopt(argc, argv, "Cc:iorsx")) != EOF)
275		switch(ch) {
276		case 'C':			/* embedded comments */
277			Cflag++;
278			break;
279		case 'c':			/* new delimiting char */
280			Delimch = *optarg;
281			if (!isascii(Delimch)) {
282				printf("bad delimiting character: '\\%o\n'",
283				       (unsigned char)Delimch);
284			}
285			break;
286		case 'i':			/* ignore case in ordering */
287			Iflag++;
288			break;
289		case 'o':			/* order strings */
290			Oflag++;
291			break;
292		case 'r':			/* randomize pointers */
293			Rflag++;
294			break;
295		case 's':			/* silent */
296			Sflag++;
297			break;
298		case 'x':			/* set the rotated bit */
299			Xflag++;
300			break;
301		case '?':
302		default:
303			usage();
304		}
305	argv += optind;
306
307	if (*argv) {
308		Infile = *argv;
309		if (*++argv)
310			(void) strcpy(Outfile, *argv);
311	}
312	if (!Infile) {
313		puts("No input file name");
314		usage();
315	}
316	if (*Outfile == '\0') {
317		(void) strcpy(Outfile, Infile);
318		(void) strcat(Outfile, ".dat");
319	}
320}
321
322void usage()
323{
324	(void) fprintf(stderr,
325	    "strfile [-Ciorsx] [-c char] sourcefile [datafile]\n");
326	exit(1);
327}
328
329/*
330 * add_offset:
331 *	Add an offset to the list, or write it out, as appropriate.
332 */
333void add_offset(fp, off)
334FILE	*fp;
335long    off;
336{
337	long net;
338
339	if (!STORING_PTRS) {
340		net = htonl(off);
341		fwrite(&net, 1, sizeof net, fp);
342	} else {
343		ALLOC(Seekpts, Num_pts + 1);
344		Seekpts[Num_pts] = off;
345	}
346	Num_pts++;
347}
348
349/*
350 * do_order:
351 *	Order the strings alphabetically (possibly ignoring case).
352 */
353void do_order()
354{
355	int	i;
356	long   *lp;
357	STR	*fp;
358
359	Sort_1 = fopen(Infile, "r");
360	Sort_2 = fopen(Infile, "r");
361	qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str);
362	i = Tbl.str_numstr;
363	lp = Seekpts;
364	fp = Firstch;
365	while (i--)
366		*lp++ = fp++->pos;
367	(void) fclose(Sort_1);
368	(void) fclose(Sort_2);
369	Tbl.str_flags |= STR_ORDERED;
370}
371
372static int collate_range_cmp (c1, c2)
373	int c1, c2;
374{
375	static char s1[2], s2[2];
376	int ret;
377
378	c1 &= UCHAR_MAX;
379	c2 &= UCHAR_MAX;
380	if (c1 == c2)
381		return (0);
382	s1[0] = c1;
383	s2[0] = c2;
384	if ((ret = strcoll(s1, s2)) != 0)
385		return (ret);
386	return (c1 - c2);
387}
388
389/*
390 * cmp_str:
391 *	Compare two strings in the file
392 */
393int cmp_str(s1, s2)
394const void	*s1, *s2;
395{
396	const STR	*p1, *p2;
397	int	c1, c2;
398	int	n1, n2;
399	int r;
400
401# define	SET_N(nf,ch)	(nf = (ch == '\n'))
402# define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
403
404	p1 = (const STR *) s1;
405	p2 = (const STR *) s2;
406
407	c1 = (unsigned char) p1->first;
408	c2 = (unsigned char) p2->first;
409	if ((r = collate_range_cmp(c1, c2)) != 0)
410		return r;
411
412	(void) fseek(Sort_1, p1->pos, 0);
413	(void) fseek(Sort_2, p2->pos, 0);
414
415	n1 = FALSE;
416	n2 = FALSE;
417	while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
418		SET_N(n1, c1);
419	while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
420		SET_N(n2, c2);
421
422	while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
423		if (Iflag) {
424			if (isupper(c1))
425				c1 = tolower(c1);
426			if (isupper(c2))
427				c2 = tolower(c2);
428		}
429		if ((r = collate_range_cmp(c1, c2)) != 0)
430			return r;
431		SET_N(n1, c1);
432		SET_N(n2, c2);
433		c1 = getc(Sort_1);
434		c2 = getc(Sort_2);
435	}
436	if (IS_END(c1, n1))
437		c1 = 0;
438	if (IS_END(c2, n2))
439		c2 = 0;
440	return collate_range_cmp(c1, c2);
441}
442
443/*
444 * randomize:
445 *	Randomize the order of the string table.  We must be careful
446 *	not to randomize across delimiter boundaries.  All
447 *	randomization is done within each block.
448 */
449void randomize()
450{
451	int	cnt, i;
452	long   tmp;
453	long   *sp;
454
455	srandomdev();
456
457	Tbl.str_flags |= STR_RANDOM;
458	cnt = Tbl.str_numstr;
459
460	/*
461	 * move things around randomly
462	 */
463
464	for (sp = Seekpts; cnt > 0; cnt--, sp++) {
465		i = random() % cnt;
466		tmp = sp[0];
467		sp[0] = sp[i];
468		sp[i] = tmp;
469	}
470}
471