* Copyright (c) 1989, 1993
* The Regents of the University of California. All rights reserved.
* This code is derived from software contributed to Berkeley by
* %sccs.include.redist.c%
static char copyright
[] =
"@(#) Copyright (c) 1989, 1993\n\
The Regents of the University of California. All rights reserved.\n";
static char sccsid
[] = "@(#)locate.code.c 8.1 (Berkeley) %G%";
* PURPOSE: sorted list compressor (works with a modified 'find'
* to encode/decode a filename database)
* USAGE: bigram < list > bigrams
* process bigrams (see updatedb) > common_bigrams
* code common_bigrams < list > squozen_list
* METHOD: Uses 'front compression' (see ";login:", Volume 8, Number 1
* February/March 1983, p. 8). Output format is, per line, an
* offset differential count byte followed by a partially bigram-
* encoded ascii residue. A bigram is a two-character sequence,
* the first 128 most common of which are encoded in one byte.
* EXAMPLE: For simple front compression with no bigram encoding,
* if the input is... then the output is...
* /usr/src/cmd/aardvark.c 8 /cmd/aardvark.c
* /usr/src/cmd/armadillo.c 14 armadillo.c
* 0-28 likeliest differential counts + offset to make nonnegative
* 30 switch code for out-of-range count to follow in next word
* 128-255 bigram codes (128 most common, as determined by 'updatedb')
* 32-127 single character (printable) ascii residue (ie, literal)
* SEE ALSO: updatedb.csh, bigram.c
* AUTHOR: James A. Woods, Informatics General Corp.,
* NASA Ames Research Center, 10/82
#define BGBUFSIZE (NBG * 2) /* size of bigram buffer */
char buf1
[MAXPATHLEN
] = " ";
char bigrams
[BGBUFSIZE
+ 1] = { 0 };
int bgindex
__P((char *));
register char *cp
, *oldpath
, *path
;
int ch
, code
, count
, diffcount
, oldcount
;
while ((ch
= getopt(argc
, argv
, "")) != EOF
)
if ((fp
= fopen(argv
[0], "r")) == NULL
)
/* First copy bigram array to stdout. */
(void)fgets(bigrams
, BGBUFSIZE
+ 1, fp
);
if (fwrite(bigrams
, 1, BGBUFSIZE
, stdout
) != BGBUFSIZE
)
while (fgets(path
, sizeof(buf2
), stdin
) != NULL
) {
cp
= path
+ strlen(path
) - 1;
if (cp
> path
&& *cp
== '\n')
/* Squelch characters that would botch the decoding. */
for (cp
= path
; *cp
!= NULL
; cp
++) {
if ((u_char
)*cp
>= PARITY
)
/* Skip longest common prefix. */
for (cp
= path
; *cp
== *oldpath
; cp
++, oldpath
++)
diffcount
= count
- oldcount
+ OFFSET
;
if (diffcount
< 0 || diffcount
> 2 * OFFSET
) {
if (putchar(SWITCH
) == EOF
||
putw(diffcount
, stdout
) == EOF
)
if (putchar(diffcount
) == EOF
)
if ((code
= bgindex(cp
)) < 0) {
if (putchar(*cp
++) == EOF
||
/* Found, so mark byte with parity bit. */
if (putchar((code
/ 2) | PARITY
) == EOF
)
if (path
== buf1
) { /* swap pointers */
/* Non-zero status if there were errors */
if (fflush(stdout
) != 0 || ferror(stdout
))
bgindex(bg
) /* Return location of bg in bigrams or -1. */
register char bg0
, bg1
, *p
;
for (p
= bigrams
; *p
!= NULL
; p
++)
if (*p
++ == bg0
&& *p
== bg1
)
return (*p
== NULL
? -1 : --p
- bigrams
);
"usage: locate.code common_bigrams < list > squozen_list\n");