strfile.c revision 201175 
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 201175 2009-12-29 08:42:58Z ed $");
50
51# include	<sys/param.h>
52# include	<sys/endian.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(CHUNKSIZE * sizeof *ptr); \
95			else if (((sz) + 1) % CHUNKSIZE == 0) \
96				ptr = realloc(ptr, ((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	off_t	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 uint32_t	Num_pts		= 0;		/* number of pointers/strings */
123
124static off_t	*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 *, off_t);
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
150main(int ac, char *av[])
151{
152	char		*sp, dc;
153	FILE		*inf, *outf;
154	off_t		last_off, pos, *p;
155	size_t		length;
156	int		first;
157	uint32_t	cnt;
158	char		*nsp;
159	STR		*fp;
160	static char		string[257];
161
162	(void) setlocale(LC_ALL, "");
163
164	getargs(ac, av);		/* evalute arguments */
165	dc = Delimch;
166	if ((inf = fopen(Infile, "r")) == NULL) {
167		perror(Infile);
168		exit(1);
169	}
170
171	if ((outf = fopen(Outfile, "w")) == NULL) {
172		perror(Outfile);
173		exit(1);
174	}
175	if (!STORING_PTRS)
176		(void) fseek(outf, (long) sizeof Tbl, 0);
177
178	/*
179	 * Write the strings onto the file
180	 */
181
182	Tbl.str_longlen = 0;
183	Tbl.str_shortlen = 0xffffffff;
184	Tbl.str_delim = dc;
185	Tbl.str_version = VERSION;
186	first = Oflag;
187	add_offset(outf, ftello(inf));
188	last_off = 0;
189	do {
190		sp = fgets(string, 256, inf);
191		if (sp == NULL || (sp[0] == dc && sp[1] == '\n')) {
192			pos = ftello(inf);
193			length = (size_t)(pos - last_off) -
194			    (sp != NULL ? strlen(sp) : 0);
195			last_off = pos;
196			if (length == 0)
197				continue;
198			add_offset(outf, pos);
199			if ((size_t)Tbl.str_longlen < length)
200				Tbl.str_longlen = length;
201			if ((size_t)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 %u strings\n", Num_pts - 1);
243		printf("Longest string: %u byte%s\n", Tbl.str_longlen,
244		       Tbl.str_longlen == 1 ? "" : "s");
245		printf("Shortest string: %u byte%s\n", Tbl.str_shortlen,
246		       Tbl.str_shortlen == 1 ? "" : "s");
247	}
248
249	rewind(outf);
250	Tbl.str_version = htobe32(Tbl.str_version);
251	Tbl.str_numstr = htobe32(Tbl.str_numstr);
252	Tbl.str_longlen = htobe32(Tbl.str_longlen);
253	Tbl.str_shortlen = htobe32(Tbl.str_shortlen);
254	Tbl.str_flags = htobe32(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 = htobe64(*p);
259		(void) fwrite(Seekpts, sizeof *Seekpts, (size_t) 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
269getargs(int argc, char **argv)
270{
271	int	ch;
272
273	while ((ch = getopt(argc, argv, "Cc:iorsx")) != -1)
274		switch(ch) {
275		case 'C':			/* embedded comments */
276			Cflag++;
277			break;
278		case 'c':			/* new delimiting char */
279			Delimch = *optarg;
280			if (!isascii(Delimch)) {
281				printf("bad delimiting character: '\\%o\n'",
282				       (unsigned char)Delimch);
283			}
284			break;
285		case 'i':			/* ignore case in ordering */
286			Iflag++;
287			break;
288		case 'o':			/* order strings */
289			Oflag++;
290			break;
291		case 'r':			/* randomize pointers */
292			Rflag++;
293			break;
294		case 's':			/* silent */
295			Sflag++;
296			break;
297		case 'x':			/* set the rotated bit */
298			Xflag++;
299			break;
300		case '?':
301		default:
302			usage();
303		}
304	argv += optind;
305
306	if (*argv) {
307		Infile = *argv;
308		if (*++argv)
309			(void) strcpy(Outfile, *argv);
310	}
311	if (!Infile) {
312		puts("No input file name");
313		usage();
314	}
315	if (*Outfile == '\0') {
316		(void) strcpy(Outfile, Infile);
317		(void) strcat(Outfile, ".dat");
318	}
319}
320
321void
322usage(void)
323{
324	(void) fprintf(stderr,
325	    "strfile [-Ciorsx] [-c char] source_file [output_file]\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
334add_offset(FILE *fp, off_t off)
335{
336	off_t beoff;
337
338	if (!STORING_PTRS) {
339		beoff = htobe64(off);
340		fwrite(&beoff, 1, sizeof beoff, fp);
341	} else {
342		ALLOC(Seekpts, Num_pts + 1);
343		Seekpts[Num_pts] = off;
344	}
345	Num_pts++;
346}
347
348/*
349 * do_order:
350 *	Order the strings alphabetically (possibly ignoring case).
351 */
352void
353do_order(void)
354{
355	uint32_t i;
356	off_t	*lp;
357	STR	*fp;
358
359	Sort_1 = fopen(Infile, "r");
360	Sort_2 = fopen(Infile, "r");
361	qsort(Firstch, (size_t) 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
373stable_collate_range_cmp(int c1, int c2)
374{
375	static char s1[2], s2[2];
376	int ret;
377
378	s1[0] = c1;
379	s2[0] = c2;
380	if ((ret = strcoll(s1, s2)) != 0)
381		return (ret);
382	return (c1 - c2);
383}
384
385/*
386 * cmp_str:
387 *	Compare two strings in the file
388 */
389int
390cmp_str(const void *s1, const void *s2)
391{
392	const STR	*p1, *p2;
393	int	c1, c2;
394	int	n1, n2;
395	int r;
396
397# define	SET_N(nf,ch)	(nf = (ch == '\n'))
398# define        IS_END(ch,nf)   (ch == EOF || (ch == (unsigned char) Delimch && nf))
399
400	p1 = (const STR *) s1;
401	p2 = (const STR *) s2;
402
403	c1 = (unsigned char) p1->first;
404	c2 = (unsigned char) p2->first;
405	if ((r = stable_collate_range_cmp(c1, c2)) != 0)
406		return (r);
407
408	(void) fseeko(Sort_1, p1->pos, 0);
409	(void) fseeko(Sort_2, p2->pos, 0);
410
411	n1 = FALSE;
412	n2 = FALSE;
413	while (!isalnum(c1 = getc(Sort_1)) && c1 != '\0' && c1 != EOF)
414		SET_N(n1, c1);
415	while (!isalnum(c2 = getc(Sort_2)) && c2 != '\0' && c2 != EOF)
416		SET_N(n2, c2);
417
418	while (!IS_END(c1, n1) && !IS_END(c2, n2)) {
419		if (Iflag) {
420			if (isupper(c1))
421				c1 = tolower(c1);
422			if (isupper(c2))
423				c2 = tolower(c2);
424		}
425		if ((r = stable_collate_range_cmp(c1, c2)) != 0)
426			return (r);
427		SET_N(n1, c1);
428		SET_N(n2, c2);
429		c1 = getc(Sort_1);
430		c2 = getc(Sort_2);
431	}
432	if (IS_END(c1, n1))
433		c1 = 0;
434	if (IS_END(c2, n2))
435		c2 = 0;
436	return (stable_collate_range_cmp(c1, c2));
437}
438
439/*
440 * randomize:
441 *	Randomize the order of the string table.  We must be careful
442 *	not to randomize across delimiter boundaries.  All
443 *	randomization is done within each block.
444 */
445void
446randomize(void)
447{
448	uint32_t cnt, i;
449	off_t	tmp;
450	off_t	*sp;
451
452#if __FreeBSD_version < 800041
453	srandomdev();
454#endif
455
456	Tbl.str_flags |= STR_RANDOM;
457	cnt = Tbl.str_numstr;
458
459	/*
460	 * move things around randomly
461	 */
462
463	for (sp = Seekpts; cnt > 0; cnt--, sp++) {
464#if __FreeBSD_version < 800041
465		i = random() % cnt;
466#else
467		i = arc4random_uniform(cnt);
468#endif
469		tmp = sp[0];
470		sp[0] = sp[i];
471		sp[i] = tmp;
472	}
473}
474