look.c revision 23693 
1/*-
2 * Copyright (c) 1991, 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 * David Hitz of Auspex Systems, Inc.
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 char copyright[] =
39"@(#) Copyright (c) 1991, 1993\n\
40	The Regents of the University of California.  All rights reserved.\n";
41#endif /* not lint */
42
43#ifndef lint
44static char sccsid[] = "@(#)look.c	8.2 (Berkeley) 5/4/95";
45#endif /* not lint */
46
47/*
48 * look -- find lines in a sorted list.
49 *
50 * The man page said that TABs and SPACEs participate in -d comparisons.
51 * In fact, they were ignored.  This implements historic practice, not
52 * the manual page.
53 */
54
55#include <sys/types.h>
56#include <sys/mman.h>
57#include <sys/stat.h>
58
59#include <limits.h>
60#include <locale.h>
61#include <ctype.h>
62#include <errno.h>
63#include <fcntl.h>
64#include <limits.h>
65#include <stdio.h>
66#include <stdlib.h>
67#include <string.h>
68#include <unistd.h>
69
70#include "pathnames.h"
71
72/*
73 * FOLD and DICT convert characters to a normal form for comparison,
74 * according to the user specified flags.
75 *
76 * DICT expects integers because it uses a non-character value to
77 * indicate a character which should not participate in comparisons.
78 */
79#define	EQUAL		0
80#define	GREATER		1
81#define	LESS		(-1)
82#define NO_COMPARE	(-2)
83
84#define FOLD(c) (isupper(c) ? tolower(c) : (unsigned char) (c))
85#define DICT(c) (isalnum(c) ? (c) & 0xFF /* int */ : NO_COMPARE)
86
87int dflag, fflag;
88
89char    *binary_search __P((unsigned char *, unsigned char *, unsigned char *));
90int      compare __P((unsigned char *, unsigned char *, unsigned char *));
91void	 err __P((const char *fmt, ...));
92char    *linear_search __P((unsigned char *, unsigned char *, unsigned char *));
93int      look __P((unsigned char *, unsigned char *, unsigned char *));
94void     print_from __P((unsigned char *, unsigned char *, unsigned char *));
95
96static void usage __P((void));
97
98main(argc, argv)
99	int argc;
100	char *argv[];
101{
102	struct stat sb;
103	int ch, fd, termchar;
104	unsigned char *back, *file, *front, *string, *p;
105
106	(void) setlocale(LC_CTYPE, "");
107
108	file = _PATH_WORDS;
109	termchar = '\0';
110	while ((ch = getopt(argc, argv, "dft:")) != EOF)
111		switch(ch) {
112		case 'd':
113			dflag = 1;
114			break;
115		case 'f':
116			fflag = 1;
117			break;
118		case 't':
119			termchar = *optarg;
120			break;
121		case '?':
122		default:
123			usage();
124		}
125	argc -= optind;
126	argv += optind;
127
128	switch (argc) {
129	case 2:				/* Don't set -df for user. */
130		string = *argv++;
131		file = *argv;
132		break;
133	case 1:				/* But set -df by default. */
134		dflag = fflag = 1;
135		string = *argv;
136		break;
137	default:
138		usage();
139	}
140
141	if (termchar != '\0' && (p = strchr(string, termchar)) != NULL)
142		*++p = '\0';
143
144	if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
145		err("%s: %s", file, strerror(errno));
146	if (sb.st_size > SIZE_T_MAX)
147		err("%s: %s", file, strerror(EFBIG));
148	if ((front = mmap(NULL,
149	    (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED)
150		err("%s: %s", file, strerror(errno));
151	back = front + sb.st_size;
152	exit(look(string, front, back));
153}
154
155look(string, front, back)
156	unsigned char *string, *front, *back;
157{
158	register int ch;
159	register unsigned char *readp, *writep;
160
161	/* Reformat string string to avoid doing it multiple times later. */
162	for (readp = writep = string; ch = *readp++;) {
163		if (fflag)
164			ch = FOLD(ch);
165		if (dflag)
166			ch = DICT(ch);
167		if (ch != NO_COMPARE)
168			*(writep++) = ch;
169	}
170	*writep = '\0';
171
172	front = binary_search(string, front, back);
173	front = linear_search(string, front, back);
174
175	if (front)
176		print_from(string, front, back);
177	return (front ? 0 : 1);
178}
179
180
181/*
182 * Binary search for "string" in memory between "front" and "back".
183 *
184 * This routine is expected to return a pointer to the start of a line at
185 * *or before* the first word matching "string".  Relaxing the constraint
186 * this way simplifies the algorithm.
187 *
188 * Invariants:
189 * 	front points to the beginning of a line at or before the first
190 *	matching string.
191 *
192 * 	back points to the beginning of a line at or after the first
193 *	matching line.
194 *
195 * Base of the Invariants.
196 * 	front = NULL;
197 *	back = EOF;
198 *
199 * Advancing the Invariants:
200 *
201 * 	p = first newline after halfway point from front to back.
202 *
203 * 	If the string at "p" is not greater than the string to match,
204 *	p is the new front.  Otherwise it is the new back.
205 *
206 * Termination:
207 *
208 * 	The definition of the routine allows it return at any point,
209 *	since front is always at or before the line to print.
210 *
211 * 	In fact, it returns when the chosen "p" equals "back".  This
212 *	implies that there exists a string is least half as long as
213 *	(back - front), which in turn implies that a linear search will
214 *	be no more expensive than the cost of simply printing a string or two.
215 *
216 * 	Trying to continue with binary search at this point would be
217 *	more trouble than it's worth.
218 */
219#define	SKIP_PAST_NEWLINE(p, back) \
220	while (p < back && *p++ != '\n');
221
222char *
223binary_search(string, front, back)
224	register unsigned char *string, *front, *back;
225{
226	register unsigned char *p;
227
228	p = front + (back - front) / 2;
229	SKIP_PAST_NEWLINE(p, back);
230
231	/*
232	 * If the file changes underneath us, make sure we don't
233	 * infinitely loop.
234	 */
235	while (p < back && back > front) {
236		if (compare(string, p, back) == GREATER)
237			front = p;
238		else
239			back = p;
240		p = front + (back - front) / 2;
241		SKIP_PAST_NEWLINE(p, back);
242	}
243	return (front);
244}
245
246/*
247 * Find the first line that starts with string, linearly searching from front
248 * to back.
249 *
250 * Return NULL for no such line.
251 *
252 * This routine assumes:
253 *
254 * 	o front points at the first character in a line.
255 *	o front is before or at the first line to be printed.
256 */
257char *
258linear_search(string, front, back)
259	unsigned char *string, *front, *back;
260{
261	while (front < back) {
262		switch (compare(string, front, back)) {
263		case EQUAL:		/* Found it. */
264			return (front);
265			break;
266		case LESS:		/* No such string. */
267			return (NULL);
268			break;
269		case GREATER:		/* Keep going. */
270			break;
271		}
272		SKIP_PAST_NEWLINE(front, back);
273	}
274	return (NULL);
275}
276
277/*
278 * Print as many lines as match string, starting at front.
279 */
280void
281print_from(string, front, back)
282	register unsigned char *string, *front, *back;
283{
284	for (; front < back && compare(string, front, back) == EQUAL; ++front) {
285		for (; front < back && *front != '\n'; ++front)
286			if (putchar(*front) == EOF)
287				err("stdout: %s", strerror(errno));
288		if (putchar('\n') == EOF)
289			err("stdout: %s", strerror(errno));
290	}
291}
292
293/*
294 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with
295 * string2 (s1 ??? s2).
296 *
297 * 	o Matches up to len(s1) are EQUAL.
298 *	o Matches up to len(s2) are GREATER.
299 *
300 * Compare understands about the -f and -d flags, and treats comparisons
301 * appropriately.
302 *
303 * The string "s1" is null terminated.  The string s2 is '\n' terminated (or
304 * "back" terminated).
305 */
306int
307compare(s1, s2, back)
308	register unsigned char *s1, *s2, *back;
309{
310	register int ch;
311
312	for (; *s1 && s2 < back && *s2 != '\n'; ++s1, ++s2) {
313		ch = *s2;
314		if (fflag)
315			ch = FOLD(ch);
316		if (dflag)
317			ch = DICT(ch);
318
319		if (ch == NO_COMPARE) {
320			++s2;		/* Ignore character in comparison. */
321			continue;
322		}
323		if (*s1 != ch)
324			return (*s1 < ch ? LESS : GREATER);
325	}
326	return (*s1 ? GREATER : EQUAL);
327}
328
329static void
330usage()
331{
332	(void)fprintf(stderr, "usage: look [-df] [-t char] string [file]\n");
333	exit(2);
334}
335
336#if __STDC__
337#include <stdarg.h>
338#else
339#include <varargs.h>
340#endif
341
342void
343#if __STDC__
344err(const char *fmt, ...)
345#else
346err(fmt, va_alist)
347	char *fmt;
348	va_dcl
349#endif
350{
351	va_list ap;
352#if __STDC__
353	va_start(ap, fmt);
354#else
355	va_start(ap);
356#endif
357	(void)fprintf(stderr, "look: ");
358	(void)vfprintf(stderr, fmt, ap);
359	va_end(ap);
360	(void)fprintf(stderr, "\n");
361	exit(2);
362	/* NOTREACHED */
363}
364