Home | History | Annotate | Line # | Download | only in code
locate.code.c revision 1.1.1.3
      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  * James A. Woods.
      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
     38 static char copyright[] =
     39 "@(#) Copyright (c) 1989, 1993\n\
     40 	The Regents of the University of California.  All rights reserved.\n";
     41 #endif /* not lint */
     42 
     43 #ifndef lint
     44 static char sccsid[] = "@(#)locate.code.c	8.4 (Berkeley) 5/4/95";
     45 #endif /* not lint */
     46 
     47 /*
     48  * PURPOSE:	sorted list compressor (works with a modified 'find'
     49  *		to encode/decode a filename database)
     50  *
     51  * USAGE:	bigram < list > bigrams
     52  *		process bigrams (see updatedb) > common_bigrams
     53  *		code common_bigrams < list > squozen_list
     54  *
     55  * METHOD:	Uses 'front compression' (see ";login:", Volume 8, Number 1
     56  *		February/March 1983, p. 8).  Output format is, per line, an
     57  *		offset differential count byte followed by a partially bigram-
     58  *		encoded ascii residue.  A bigram is a two-character sequence,
     59  *		the first 128 most common of which are encoded in one byte.
     60  *
     61  * EXAMPLE:	For simple front compression with no bigram encoding,
     62  *		if the input is...		then the output is...
     63  *
     64  *		/usr/src			 0 /usr/src
     65  *		/usr/src/cmd/aardvark.c		 8 /cmd/aardvark.c
     66  *		/usr/src/cmd/armadillo.c	14 armadillo.c
     67  *		/usr/tmp/zoo			 5 tmp/zoo
     68  *
     69  *	The codes are:
     70  *
     71  *	0-28	likeliest differential counts + offset to make nonnegative
     72  *	30	switch code for out-of-range count to follow in next word
     73  *	128-255 bigram codes (128 most common, as determined by 'updatedb')
     74  *	32-127  single character (printable) ascii residue (ie, literal)
     75  *
     76  * SEE ALSO:	updatedb.csh, bigram.c
     77  *
     78  * AUTHOR:	James A. Woods, Informatics General Corp.,
     79  *		NASA Ames Research Center, 10/82
     80  */
     81 
     82 #include <sys/param.h>
     83 
     84 #include <err.h>
     85 #include <errno.h>
     86 #include <stdio.h>
     87 #include <stdlib.h>
     88 #include <string.h>
     89 #include <unistd.h>
     90 
     91 #include "locate.h"
     92 
     93 #define	BGBUFSIZE	(NBG * 2)	/* size of bigram buffer */
     94 
     95 char buf1[MAXPATHLEN] = " ";
     96 char buf2[MAXPATHLEN];
     97 char bigrams[BGBUFSIZE + 1] = { 0 };
     98 
     99 int	bgindex __P((char *));
    100 void	usage __P((void));
    101 
    102 int
    103 main(argc, argv)
    104 	int argc;
    105 	char *argv[];
    106 {
    107 	register char *cp, *oldpath, *path;
    108 	int ch, code, count, diffcount, oldcount;
    109 	FILE *fp;
    110 
    111 	while ((ch = getopt(argc, argv, "")) != EOF)
    112 		switch(ch) {
    113 		case '?':
    114 		default:
    115 			usage();
    116 		}
    117 	argc -= optind;
    118 	argv += optind;
    119 
    120 	if (argc != 1)
    121 		usage();
    122 
    123 	if ((fp = fopen(argv[0], "r")) == NULL)
    124 		err(1, "%s", argv[0]);
    125 
    126 	/* First copy bigram array to stdout. */
    127 	(void)fgets(bigrams, BGBUFSIZE + 1, fp);
    128 	if (fwrite(bigrams, 1, BGBUFSIZE, stdout) != BGBUFSIZE)
    129 		err(1, "stdout");
    130 	(void)fclose(fp);
    131 
    132 	oldpath = buf1;
    133 	path = buf2;
    134 	oldcount = 0;
    135 	while (fgets(path, sizeof(buf2), stdin) != NULL) {
    136 		/* Truncate newline. */
    137 		cp = path + strlen(path) - 1;
    138 		if (cp > path && *cp == '\n')
    139 			*cp = '\0';
    140 
    141 		/* Squelch characters that would botch the decoding. */
    142 		for (cp = path; *cp != '\0'; cp++) {
    143 			if ((u_char)*cp >= PARITY || *cp <= SWITCH)
    144 				*cp = '?';
    145 		}
    146 
    147 		/* Skip longest common prefix. */
    148 		for (cp = path; *cp == *oldpath; cp++, oldpath++)
    149 			if (*oldpath == '\0')
    150 				break;
    151 		count = cp - path;
    152 		diffcount = count - oldcount + OFFSET;
    153 		oldcount = count;
    154 		if (diffcount < 0 || diffcount > 2 * OFFSET) {
    155 			if (putchar(SWITCH) == EOF ||
    156 			    putw(diffcount, stdout) == EOF)
    157 				err(1, "stdout");
    158 		} else
    159 			if (putchar(diffcount) == EOF)
    160 				err(1, "stdout");
    161 
    162 		while (*cp != '\0') {
    163 			if (*(cp + 1) == '\0') {
    164 				if (putchar(*cp) == EOF)
    165 					err(1, "stdout");
    166 				break;
    167 			}
    168 			if ((code = bgindex(cp)) < 0) {
    169 				if (putchar(*cp++) == EOF ||
    170 				    putchar(*cp++) == EOF)
    171 					err(1, "stdout");
    172 			} else {
    173 				/* Found, so mark byte with parity bit. */
    174 				if (putchar((code / 2) | PARITY) == EOF)
    175 					err(1, "stdout");
    176 				cp += 2;
    177 			}
    178 		}
    179 		if (path == buf1) {		/* swap pointers */
    180 			path = buf2;
    181 			oldpath = buf1;
    182 		} else {
    183 			path = buf1;
    184 			oldpath = buf2;
    185 		}
    186 	}
    187 	/* Non-zero status if there were errors */
    188 	if (fflush(stdout) != 0 || ferror(stdout))
    189 		exit(1);
    190 	exit(0);
    191 }
    192 
    193 int
    194 bgindex(bg)			/* Return location of bg in bigrams or -1. */
    195 	char *bg;
    196 {
    197 	register char bg0, bg1, *p;
    198 
    199 	bg0 = bg[0];
    200 	bg1 = bg[1];
    201 	for (p = bigrams; *p != '\0'; p++)
    202 		if (*p++ == bg0 && *p == bg1)
    203 			break;
    204 	return (*p == '\0' ? -1 : --p - bigrams);
    205 }
    206 
    207 void
    208 usage()
    209 {
    210 	(void)fprintf(stderr,
    211 	    "usage: locate.code common_bigrams < list > squozen_list\n");
    212 	exit(1);
    213 }
    214