1 /* code -- bigram- and front-encode filenames for locate
2 Copyright (C) 1994 Free Software Foundation, Inc.
4 This program is free software; you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation; either version 2, or (at your option)
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301,
20 /* Compress a sorted list.
21 Works with `find' to encode a filename database to save space
26 bigram < file_list > bigrams
27 process-bigrams > most_common_bigrams
28 code most_common_bigrams < file_list > squeezed_list
30 Uses `front compression' (see ";login:", March 1983, p. 8).
31 The output begins with the 128 most common bigrams.
32 After that, the output format is, for each line,
33 an offset (from the previous line) differential count byte
34 followed by a (partially bigram-encoded) ASCII remainder.
35 The output lines have no terminating byte; the start of the next line
36 is indicated by its first byte having a value <= 30.
38 The encoding of the output bytes is:
40 0-28 likeliest differential counts + offset (14) to make nonnegative
41 30 escape code for out-of-range count to follow in next halfword
42 128-255 bigram codes (the 128 most common, as determined by `updatedb')
43 32-127 single character (printable) ASCII remainder
45 Written by James A. Woods <jwoods@adobe.com>.
46 Modified by David MacKenzie <djm@gnu.org>. */
50 #include <sys/types.h>
52 #if defined(HAVE_STRING_H) || defined(STDC_HEADERS)
64 # define _(Text) gettext (Text)
67 #define textdomain(Domain)
68 #define bindtextdomain(Package, Directory)
71 # define N_(String) gettext_noop (String)
73 /* See locate.c for explanation as to why not use (String) */
74 # define N_(String) String
80 #include "gnulib-version.h"
82 char *xmalloc
PARAMS((size_t));
84 /* The name this program was run with. */
87 /* The 128 most common bigrams in the file list, padded with NULs
88 if there are fewer. */
89 static char bigrams
[257] = {0};
91 /* Return the offset of PATTERN in STRING, or -1 if not found. */
94 strindex (char *string
, char *pattern
)
98 for (s
= string
; *s
!= '\0'; s
++)
99 /* Fast first char check. */
102 register char *p2
= pattern
+ 1, *s2
= s
+ 1;
103 while (*p2
!= '\0' && *p2
== *s2
)
106 return s2
- strlen (pattern
) - string
;
111 /* Return the length of the longest common prefix of strings S1 and S2. */
114 prefix_length (char *s1
, char *s2
)
116 register char *start
;
118 for (start
= s1
; *s1
== *s2
&& *s1
!= '\0'; s1
++, s2
++)
123 extern char *version_string
;
128 fprintf (stream
, _("\
129 Usage: %s [--version | --help]\n\
130 or %s most_common_bigrams < file-list > locate-database\n"),
131 program_name
, program_name
);
132 fputs (_("\nReport bugs to <bug-findutils@gnu.org>.\n"), stream
);
137 main (int argc
, char **argv
)
139 char *path
; /* The current input entry. */
140 char *oldpath
; /* The previous input entry. */
141 size_t pathsize
, oldpathsize
; /* Amounts allocated for them. */
142 int count
, oldcount
, diffcount
; /* Their prefix lengths & the difference. */
143 char bigram
[3]; /* Bigram to search for in table. */
144 int code
; /* Index of `bigram' in bigrams table. */
145 FILE *fp
; /* Most common bigrams file. */
146 int line_len
; /* Length of input line. */
148 program_name
= argv
[0];
149 atexit (close_stdout
);
159 if (0 == strcmp(argv
[1], "--help"))
164 else if (0 == strcmp(argv
[1], "--version"))
166 printf (_("GNU findutils version %s\n"), version_string
);
167 printf (_("Built using GNU gnulib version %s\n"), gnulib_version
);
171 fp
= fopen (argv
[1], "r");
174 fprintf (stderr
, "%s: ", argv
[0]);
179 pathsize
= oldpathsize
= 1026; /* Increased as necessary by getline. */
180 path
= xmalloc (pathsize
);
181 oldpath
= xmalloc (oldpathsize
);
183 /* Set to anything not starting with a slash, to force the first
184 prefix count to 0. */
185 strcpy (oldpath
, " ");
188 /* Copy the list of most common bigrams to the output,
189 padding with NULs if there are <128 of them. */
190 fgets (bigrams
, 257, fp
);
191 fwrite (bigrams
, 1, 256, stdout
);
194 while ((line_len
= getline (&path
, &pathsize
, stdin
)) > 0)
198 path
[line_len
- 1] = '\0'; /* Remove newline. */
200 /* Squelch unprintable chars in path so as not to botch decoding. */
201 for (pp
= path
; *pp
!= '\0'; pp
++)
203 if (!(*pp
>= 040 && *pp
< 0177))
207 count
= prefix_length (oldpath
, path
);
208 diffcount
= count
- oldcount
;
210 /* If the difference is small, it fits in one byte;
211 otherwise, two bytes plus a marker noting that fact. */
212 if (diffcount
< -LOCATEDB_OLD_OFFSET
|| diffcount
> LOCATEDB_OLD_OFFSET
)
214 putc (LOCATEDB_OLD_ESCAPE
, stdout
);
215 putw (diffcount
+ LOCATEDB_OLD_OFFSET
, stdout
);
218 putc (diffcount
+ LOCATEDB_OLD_OFFSET
, stdout
);
220 /* Look for bigrams in the remainder of the path. */
221 for (pp
= path
+ count
; *pp
!= '\0'; pp
+= 2)
225 /* No bigram is possible; only one char is left. */
231 /* Linear search for specific bigram in string table. */
232 code
= strindex (bigrams
, bigram
);
234 putchar ((code
/ 2) | 0200); /* It's a common bigram. */
236 fputs (bigram
, stdout
); /* Write the text as printable ASCII. */
240 /* Swap path and oldpath and their sizes. */
241 char *tmppath
= oldpath
;
242 size_t tmppathsize
= oldpathsize
;
244 oldpathsize
= pathsize
;
246 pathsize
= tmppathsize
;