strfile.c revision 123243 
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 123243 2003-12-07 17:34:52Z des $");
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(ac, av)
150int	ac;
151char	**av;
152{
153	char		*sp, dc;
154	FILE		*inf, *outf;
155	long           last_off, length, pos, *p;
156	int		first, cnt;
157	char		*nsp;
158	STR		*fp;
159	static char		string[257];
160
161	(void) setlocale(LC_ALL, "");
162
163	getargs(ac, av);		/* evalute arguments */
164	dc = Delimch;
165	if ((inf = fopen(Infile, "r")) == NULL) {
166		perror(Infile);
167		exit(1);
168	}
169
170	if ((outf = fopen(Outfile, "w")) == NULL) {
171		perror(Outfile);
172		exit(1);
173	}
174	if (!STORING_PTRS)
175		(void) fseek(outf, (long) sizeof Tbl, 0);
176
177	/*
178	 * Write the strings onto the file
179	 */
180
181	Tbl.str_longlen = 0;
182	Tbl.str_shortlen = ~((unsigned long) 0);
183	Tbl.str_delim = dc;
184	Tbl.str_version = VERSION;
185	first = Oflag;
186	add_offset(outf, ftell(inf));
187	last_off = 0;
188	do {
189		sp = fgets(string, 256, inf);
190		if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
191			pos = ftell(inf);
192			length = pos - last_off - (sp ? strlen(sp) : 0);
193			last_off = pos;
194			if (!length)
195				continue;
196			add_offset(outf, pos);
197			if (Tbl.str_longlen < length)
198				Tbl.str_longlen = length;
199			if (Tbl.str_shortlen > length)
200				Tbl.str_shortlen = length;
201			first = Oflag;
202		}
203		else if (first) {
204			for (nsp = sp; !isalnum((unsigned char)*nsp); nsp++)
205				continue;
206			ALLOC(Firstch, Num_pts);
207			fp = &Firstch[Num_pts - 1];
208			if (Iflag && isupper((unsigned char)*nsp))
209				fp->first = tolower((unsigned char)*nsp);
210			else
211				fp->first = *nsp;
212			fp->pos = Seekpts[Num_pts - 1];
213			first = FALSE;
214		}
215	} while (sp != NULL);
216
217	/*
218	 * write the tables in
219	 */
220
221	(void) fclose(inf);
222	Tbl.str_numstr = Num_pts - 1;
223
224	if (Cflag)
225		Tbl.str_flags |= STR_COMMENTS;
226
227	if (Oflag)
228		do_order();
229	else if (Rflag)
230		randomize();
231
232	if (Xflag)
233		Tbl.str_flags |= STR_ROTATED;
234
235	if (!Sflag) {
236		printf("\"%s\" created\n", Outfile);
237		if (Num_pts == 2)
238			puts("There was 1 string");
239		else
240			printf("There were %ld strings\n", Num_pts - 1);
241		printf("Longest string: %lu byte%s\n", Tbl.str_longlen,
242		       Tbl.str_longlen == 1 ? "" : "s");
243		printf("Shortest string: %lu byte%s\n", Tbl.str_shortlen,
244		       Tbl.str_shortlen == 1 ? "" : "s");
245	}
246
247	rewind(outf);
248	Tbl.str_version = htonl(Tbl.str_version);
249	Tbl.str_numstr = htonl(Tbl.str_numstr);
250	Tbl.str_longlen = htonl(Tbl.str_longlen);
251	Tbl.str_shortlen = htonl(Tbl.str_shortlen);
252	Tbl.str_flags = htonl(Tbl.str_flags);
253	(void) fwrite((char *) &Tbl, sizeof Tbl, 1, outf);
254	if (STORING_PTRS) {
255		for (p = Seekpts, cnt = Num_pts; cnt--; ++p)
256			*p = htonl(*p);
257		(void) fwrite((char *) Seekpts, sizeof *Seekpts, (int) Num_pts, outf);
258	}
259	(void) fclose(outf);
260	exit(0);
261}
262
263/*
264 *	This routine evaluates arguments from the command line
265 */
266void getargs(argc, argv)
267int	argc;
268char	**argv;
269{
270	int	ch;
271
272	while ((ch = getopt(argc, argv, "Cc:iorsx")) != EOF)
273		switch(ch) {
274		case 'C':			/* embedded comments */
275			Cflag++;
276			break;
277		case 'c':			/* new delimiting char */
278			Delimch = *optarg;
279			if (!isascii(Delimch)) {
280				printf("bad delimiting character: '\\%o\n'",
281				       (unsigned char)Delimch);
282			}
283			break;
284		case 'i':			/* ignore case in ordering */
285			Iflag++;
286			break;
287		case 'o':			/* order strings */
288			Oflag++;
289			break;
290		case 'r':			/* randomize pointers */
291			Rflag++;
292			break;
293		case 's':			/* silent */
294			Sflag++;
295			break;
296		case 'x':			/* set the rotated bit */
297			Xflag++;
298			break;
299		case '?':
300		default:
301			usage();
302		}
303	argv += optind;
304
305	if (*argv) {
306		Infile = *argv;
307		if (*++argv)
308			(void) strcpy(Outfile, *argv);
309	}
310	if (!Infile) {
311		puts("No input file name");
312		usage();
313	}
314	if (*Outfile == '\0') {
315		(void) strcpy(Outfile, Infile);
316		(void) strcat(Outfile, ".dat");
317	}
318}
319
320void usage()
321{
322	(void) fprintf(stderr,
323	    "strfile [-Ciorsx] [-c char] sourcefile [datafile]\n");
324	exit(1);
325}
326
327/*
328 * add_offset:
329 *	Add an offset to the list, or write it out, as appropriate.
330 */
331void add_offset(fp, off)
332FILE	*fp;
333long    off;
334{
335	long net;
336
337	if (!STORING_PTRS) {
338		net = htonl(off);
339		fwrite(&net, 1, sizeof net, fp);
340	} else {
341		ALLOC(Seekpts, Num_pts + 1);
342		Seekpts[Num_pts] = off;
343	}
344	Num_pts++;
345}
346
347/*
348 * do_order:
349 *	Order the strings alphabetically (possibly ignoring case).
350 */
351void do_order()
352{
353	int	i;
354	long   *lp;
355	STR	*fp;
356
357	Sort_1 = fopen(Infile, "r");
358	Sort_2 = fopen(Infile, "r");
359	qsort((char *) Firstch, (int) Tbl.str_numstr, sizeof *Firstch, cmp_str);
360	i = Tbl.str_numstr;
361	lp = Seekpts;
362	fp = Firstch;
363	while (i--)
364		*lp++ = fp++->pos;
365	(void) fclose(Sort_1);
366	(void) fclose(Sort_2);
367	Tbl.str_flags |= STR_ORDERED;
368}
369
370static int stable_collate_range_cmp(c1, c2)
371	int c1, c2;
372{
373	static char s1[2], s2[2];
374	int ret;
375
376	s1[0] = c1;
377	s2[0] = c2;
378	if ((ret = strcoll(s1, s2)) != 0)
379		return (ret);
380	return (c1 - c2);
381}
382
383/*
384 * cmp_str:
385 *	Compare two strings in the file
386 */
387int cmp_str(s1, s2)
388const void	*s1, *s2;
389{
390	const STR	*p1, *p2;
391	int	c1, c2;
392	int	n1, n2;
393	int r;
394
395# define	SET_N(nf,ch)	(nf = (ch == '\n'))
396# define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
397
398	p1 = (const STR *) s1;
399	p2 = (const STR *) s2;
400
401	c1 = (unsigned char) p1->first;
402	c2 = (unsigned char) p2->first;
403	if ((r = stable_collate_range_cmp(c1, c2)) != 0)
404		return (r);
405
406	(void) fseek(Sort_1, p1->pos, 0);
407	(void) fseek(Sort_2, p2->pos, 0);
408
409	n1 = FALSE;
410	n2 = FALSE;
411	while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
412		SET_N(n1, c1);
413	while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
414		SET_N(n2, c2);
415
416	while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
417		if (Iflag) {
418			if (isupper(c1))
419				c1 = tolower(c1);
420			if (isupper(c2))
421				c2 = tolower(c2);
422		}
423		if ((r = stable_collate_range_cmp(c1, c2)) != 0)
424			return (r);
425		SET_N(n1, c1);
426		SET_N(n2, c2);
427		c1 = getc(Sort_1);
428		c2 = getc(Sort_2);
429	}
430	if (IS_END(c1, n1))
431		c1 = 0;
432	if (IS_END(c2, n2))
433		c2 = 0;
434	return (stable_collate_range_cmp(c1, c2));
435}
436
437/*
438 * randomize:
439 *	Randomize the order of the string table.  We must be careful
440 *	not to randomize across delimiter boundaries.  All
441 *	randomization is done within each block.
442 */
443void randomize()
444{
445	int	cnt, i;
446	long   tmp;
447	long   *sp;
448
449	srandomdev();
450
451	Tbl.str_flags |= STR_RANDOM;
452	cnt = Tbl.str_numstr;
453
454	/*
455	 * move things around randomly
456	 */
457
458	for (sp = Seekpts; cnt > 0; cnt--, sp++) {
459		i = random() % cnt;
460		tmp = sp[0];
461		sp[0] = sp[i];
462		sp[i] = tmp;
463	}
464}
465