look.c revision 27623 
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 const 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
44#if 0
45static char sccsid[] = "@(#)look.c	8.2 (Berkeley) 5/4/95";
46#endif
47static const char rcsid[] =
48	"$Id$";
49#endif /* not lint */
50
51/*
52 * look -- find lines in a sorted list.
53 *
54 * The man page said that TABs and SPACEs participate in -d comparisons.
55 * In fact, they were ignored.  This implements historic practice, not
56 * the manual page.
57 */
58
59#include <sys/types.h>
60#include <sys/mman.h>
61#include <sys/stat.h>
62
63#include <ctype.h>
64#include <err.h>
65#include <errno.h>
66#include <fcntl.h>
67#include <limits.h>
68#include <locale.h>
69#include <stdio.h>
70#include <stdlib.h>
71#include <string.h>
72#include <unistd.h>
73
74#include "pathnames.h"
75
76/*
77 * FOLD and DICT convert characters to a normal form for comparison,
78 * according to the user specified flags.
79 *
80 * DICT expects integers because it uses a non-character value to
81 * indicate a character which should not participate in comparisons.
82 */
83#define	EQUAL		0
84#define	GREATER		1
85#define	LESS		(-1)
86#define NO_COMPARE	(-2)
87
88#define FOLD(c) (isupper(c) ? tolower(c) : (unsigned char) (c))
89#define DICT(c) (isalnum(c) ? (c) & 0xFF /* int */ : NO_COMPARE)
90
91int dflag, fflag;
92
93char    *binary_search __P((unsigned char *, unsigned char *, unsigned char *));
94int      compare __P((unsigned char *, unsigned char *, unsigned char *));
95char    *linear_search __P((unsigned char *, unsigned char *, unsigned char *));
96int      look __P((unsigned char *, unsigned char *, unsigned char *));
97void     print_from __P((unsigned char *, unsigned char *, unsigned char *));
98
99static void usage __P((void));
100
101int
102main(argc, argv)
103	int argc;
104	char *argv[];
105{
106	struct stat sb;
107	int ch, fd, termchar;
108	unsigned char *back, *file, *front, *string, *p;
109
110	(void) setlocale(LC_CTYPE, "");
111
112	file = _PATH_WORDS;
113	termchar = '\0';
114	while ((ch = getopt(argc, argv, "dft:")) != -1)
115		switch(ch) {
116		case 'd':
117			dflag = 1;
118			break;
119		case 'f':
120			fflag = 1;
121			break;
122		case 't':
123			termchar = *optarg;
124			break;
125		case '?':
126		default:
127			usage();
128		}
129	argc -= optind;
130	argv += optind;
131
132	switch (argc) {
133	case 2:				/* Don't set -df for user. */
134		string = *argv++;
135		file = *argv;
136		break;
137	case 1:				/* But set -df by default. */
138		dflag = fflag = 1;
139		string = *argv;
140		break;
141	default:
142		usage();
143	}
144
145	if (termchar != '\0' && (p = strchr(string, termchar)) != NULL)
146		*++p = '\0';
147
148	if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
149		err(2, "%s", file);
150	if (sb.st_size > SIZE_T_MAX)
151		errx(2, "%s: %s", file, strerror(EFBIG));
152	if ((front = mmap(NULL,
153	    (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED)
154		err(2, "%s", file);
155	back = front + sb.st_size;
156	exit(look(string, front, back));
157}
158
159int
160look(string, front, back)
161	unsigned char *string, *front, *back;
162{
163	register int ch;
164	register unsigned char *readp, *writep;
165
166	/* Reformat string string to avoid doing it multiple times later. */
167	for (readp = writep = string; ch = *readp++;) {
168		if (fflag)
169			ch = FOLD(ch);
170		if (dflag)
171			ch = DICT(ch);
172		if (ch != NO_COMPARE)
173			*(writep++) = ch;
174	}
175	*writep = '\0';
176
177	front = binary_search(string, front, back);
178	front = linear_search(string, front, back);
179
180	if (front)
181		print_from(string, front, back);
182	return (front ? 0 : 1);
183}
184
185
186/*
187 * Binary search for "string" in memory between "front" and "back".
188 *
189 * This routine is expected to return a pointer to the start of a line at
190 * *or before* the first word matching "string".  Relaxing the constraint
191 * this way simplifies the algorithm.
192 *
193 * Invariants:
194 * 	front points to the beginning of a line at or before the first
195 *	matching string.
196 *
197 * 	back points to the beginning of a line at or after the first
198 *	matching line.
199 *
200 * Base of the Invariants.
201 * 	front = NULL;
202 *	back = EOF;
203 *
204 * Advancing the Invariants:
205 *
206 * 	p = first newline after halfway point from front to back.
207 *
208 * 	If the string at "p" is not greater than the string to match,
209 *	p is the new front.  Otherwise it is the new back.
210 *
211 * Termination:
212 *
213 * 	The definition of the routine allows it return at any point,
214 *	since front is always at or before the line to print.
215 *
216 * 	In fact, it returns when the chosen "p" equals "back".  This
217 *	implies that there exists a string is least half as long as
218 *	(back - front), which in turn implies that a linear search will
219 *	be no more expensive than the cost of simply printing a string or two.
220 *
221 * 	Trying to continue with binary search at this point would be
222 *	more trouble than it's worth.
223 */
224#define	SKIP_PAST_NEWLINE(p, back) \
225	while (p < back && *p++ != '\n');
226
227char *
228binary_search(string, front, back)
229	register unsigned char *string, *front, *back;
230{
231	register unsigned char *p;
232
233	p = front + (back - front) / 2;
234	SKIP_PAST_NEWLINE(p, back);
235
236	/*
237	 * If the file changes underneath us, make sure we don't
238	 * infinitely loop.
239	 */
240	while (p < back && back > front) {
241		if (compare(string, p, back) == GREATER)
242			front = p;
243		else
244			back = p;
245		p = front + (back - front) / 2;
246		SKIP_PAST_NEWLINE(p, back);
247	}
248	return (front);
249}
250
251/*
252 * Find the first line that starts with string, linearly searching from front
253 * to back.
254 *
255 * Return NULL for no such line.
256 *
257 * This routine assumes:
258 *
259 * 	o front points at the first character in a line.
260 *	o front is before or at the first line to be printed.
261 */
262char *
263linear_search(string, front, back)
264	unsigned char *string, *front, *back;
265{
266	while (front < back) {
267		switch (compare(string, front, back)) {
268		case EQUAL:		/* Found it. */
269			return (front);
270			break;
271		case LESS:		/* No such string. */
272			return (NULL);
273			break;
274		case GREATER:		/* Keep going. */
275			break;
276		}
277		SKIP_PAST_NEWLINE(front, back);
278	}
279	return (NULL);
280}
281
282/*
283 * Print as many lines as match string, starting at front.
284 */
285void
286print_from(string, front, back)
287	register unsigned char *string, *front, *back;
288{
289	for (; front < back && compare(string, front, back) == EQUAL; ++front) {
290		for (; front < back && *front != '\n'; ++front)
291			if (putchar(*front) == EOF)
292				err(2, "stdout");
293		if (putchar('\n') == EOF)
294			err(2, "stdout");
295	}
296}
297
298/*
299 * Return LESS, GREATER, or EQUAL depending on how the string1 compares with
300 * string2 (s1 ??? s2).
301 *
302 * 	o Matches up to len(s1) are EQUAL.
303 *	o Matches up to len(s2) are GREATER.
304 *
305 * Compare understands about the -f and -d flags, and treats comparisons
306 * appropriately.
307 *
308 * The string "s1" is null terminated.  The string s2 is '\n' terminated (or
309 * "back" terminated).
310 */
311int
312compare(s1, s2, back)
313	register unsigned char *s1, *s2, *back;
314{
315	register int ch;
316
317	for (; *s1 && s2 < back && *s2 != '\n'; ++s1, ++s2) {
318		ch = *s2;
319		if (fflag)
320			ch = FOLD(ch);
321		if (dflag)
322			ch = DICT(ch);
323
324		if (ch == NO_COMPARE) {
325			++s2;		/* Ignore character in comparison. */
326			continue;
327		}
328		if (*s1 != ch)
329			return (*s1 < ch ? LESS : GREATER);
330	}
331	return (*s1 ? GREATER : EQUAL);
332}
333
334static void
335usage()
336{
337	(void)fprintf(stderr, "usage: look [-df] [-t char] string [file]\n");
338	exit(2);
339}
340