1 /* locate -- search databases for filenames that match patterns
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., 675 Mass Ave, Cambridge, MA 02139, USA. */
18 /* Usage: locate [options] pattern...
20 Scan a pathname list for the full pathname of a file, given only
21 a piece of the name (possibly containing shell globbing metacharacters).
22 The list has been processed with front-compression, which reduces
23 the list size by a factor of 4-5.
24 Recognizes two database formats, old and new. The old format is
25 bigram coded, which reduces space by a further 20-25% and uses the
26 following encoding of the database bytes:
28 0-28 likeliest differential counts + offset (14) to make nonnegative
29 30 escape code for out-of-range count to follow in next halfword
30 128-255 bigram codes (the 128 most common, as determined by `updatedb')
31 32-127 single character (printable) ASCII remainder
33 Uses a novel two-tiered string search technique:
35 First, match a metacharacter-free subpattern and a partial pathname
36 BACKWARDS to avoid full expansion of the pathname list.
37 The time savings is 40-50% over forward matching, which cannot efficiently
38 handle overlapped search patterns and compressed path remainders.
40 Then, match the actual shell glob-style regular expression (if in this form)
41 against the candidate pathnames using the slower shell filename
44 Described more fully in Usenix ;login:, Vol 8, No 1,
45 February/March, 1983, p. 8.
47 Written by James A. Woods <jwoods@adobe.com>.
48 Modified by David MacKenzie <djm@gnu.ai.mit.edu>. */
53 #include <sys/types.h>
62 #if defined(HAVE_STRING_H) || defined(STDC_HEADERS)
85 # define _(Text) gettext (Text)
88 #define textdomain(Domain)
89 #define bindtextdomain(Package, Directory)
92 # define N_(String) gettext_noop (String)
94 # define N_(String) (String)
100 /* Note that this evaluates C many times. */
102 # define TOUPPER(Ch) toupper (Ch)
103 # define TOLOWER(Ch) tolower (Ch)
105 # define TOUPPER(Ch) (islower (Ch) ? toupper (Ch) : (Ch))
106 # define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch))
109 typedef enum {false, true} boolean
;
111 /* Warn if a database is older than this. 8 days allows for a weekly
112 update that takes up to a day to perform. */
113 #define WARN_NUMBER_UNITS (8)
114 /* Printable name of units used in WARN_SECONDS */
115 static const char warn_name_units
[] = N_("days");
116 #define SECONDS_PER_UNIT (60 * 60 * 24)
118 #define WARN_SECONDS ((SECONDS_PER_UNIT) * (WARN_NUMBER_UNITS))
120 /* Check for existence of files before printing them out? */
121 int check_existence
= 0;
123 char *next_element ();
128 /* Read in a 16-bit int, high byte first (network byte order). */
138 x
|= (fgetc (fp
) & 0xff);
142 /* Return a pointer to the last character in a static copy of the last
143 glob-free subpattern in NAME,
144 with '\0' prepended for a fast backwards pre-match. */
147 last_literal_end (name
)
150 static char *globfree
= NULL
; /* A copy of the subpattern in NAME. */
151 static size_t gfalloc
= 0; /* Bytes allocated for `globfree'. */
152 register char *subp
; /* Return value. */
153 register char *p
; /* Search location in NAME. */
155 /* Find the end of the subpattern.
156 Skip trailing metacharacters and [] ranges. */
157 for (p
= name
+ strlen (name
) - 1; p
>= name
&& strchr ("*?]", *p
) != NULL
;
161 while (p
>= name
&& *p
!= '[')
167 if (p
- name
+ 3 > gfalloc
)
169 gfalloc
= p
- name
+ 3 + 64; /* Room to grow. */
170 globfree
= xrealloc (globfree
, gfalloc
);
175 /* If the pattern has only metacharacters, make every path match the
176 subpattern, so it gets checked the slow way. */
177 if (p
== name
&& strchr ("?*[]", *p
) != NULL
)
182 /* Find the start of the metacharacter-free subpattern. */
183 for (endmark
= p
; p
>= name
&& strchr ("]*?", *p
) == NULL
; p
--)
185 /* Copy the subpattern into globfree. */
186 for (++p
; p
<= endmark
; )
189 *subp
-- = '\0'; /* Null terminate, though it's not needed. */
194 /* Print the entries in DBFILE that match shell globbing pattern PATHPART.
195 Return the number of entries printed. */
198 locate (pathpart
, dbfile
, ignore_case
)
199 char *pathpart
, *dbfile
;
202 /* The pathname database. */
206 /* Number of bytes read from an entry. */
209 /* true if PATHPART contains globbing metacharacters. */
211 /* The end of the last glob-free subpattern in PATHPART. */
214 /* The current input database entry. */
216 /* Amount allocated for it. */
219 /* The length of the prefix shared with the previous database entry. */
221 /* Where in `path' to stop the backward search for the last character
222 in the subpattern. Set according to `count'. */
225 /* true if we found a fast match (of patend) on the previous path. */
226 boolean prev_fast_match
= false;
227 /* The return value. */
230 /* true if reading a bigram-encoded database. */
231 boolean old_format
= false;
232 /* For the old database format,
233 the first and second characters of the most common bigrams. */
234 char bigram1
[128], bigram2
[128];
236 /* To check the age of the database. */
240 if (stat (dbfile
, &st
) || (fp
= fopen (dbfile
, "r")) == NULL
)
242 error (0, errno
, "%s", dbfile
);
246 if (now
- st
.st_mtime
> WARN_SECONDS
)
249 warning: database `fred' is more than 8 days old */
250 error (0, 0, _("warning: database `%s' is more than %d %s old"),
251 dbfile
, WARN_NUMBER_UNITS
, _(warn_name_units
));
254 pathsize
= 1026; /* Increased as necessary by getstr. */
255 path
= xmalloc (pathsize
);
257 nread
= fread (path
, 1, sizeof (LOCATEDB_MAGIC
), fp
);
258 if (nread
!= sizeof (LOCATEDB_MAGIC
)
259 || memcmp (path
, LOCATEDB_MAGIC
, sizeof (LOCATEDB_MAGIC
)))
262 /* Read the list of the most common bigrams in the database. */
264 for (i
= 0; i
< 128; i
++)
266 bigram1
[i
] = getc (fp
);
267 bigram2
[i
] = getc (fp
);
272 /* If we ignore case,
273 convert it to lower first so we don't have to do it every time */
275 for (patend
=pathpart
;*patend
;++patend
){
276 *patend
=TOLOWER(*patend
);
281 globflag
= strchr (pathpart
, '*') || strchr (pathpart
, '?')
282 || strchr (pathpart
, '[');
284 patend
= last_literal_end (pathpart
);
289 register char *s
; /* Scan the path we read in. */
293 /* Get the offset in the path where this path info starts. */
294 if (c
== LOCATEDB_OLD_ESCAPE
)
295 count
+= getw (fp
) - LOCATEDB_OLD_OFFSET
;
297 count
+= c
- LOCATEDB_OLD_OFFSET
;
299 /* Overlay the old path with the remainder of the new. */
300 for (s
= path
+ count
; (c
= getc (fp
)) > LOCATEDB_OLD_ESCAPE
;)
302 *s
++ = c
; /* An ordinary character. */
305 /* Bigram markers have the high bit set. */
314 if (c
== LOCATEDB_ESCAPE
)
315 count
+= get_short (fp
);
321 /* Overlay the old path with the remainder of the new. */
322 nread
= getstr (&path
, &pathsize
, fp
, '\0', count
);
326 s
= path
+ count
+ nread
- 2; /* Move to the last char in path. */
327 assert (s
[0] != '\0');
328 assert (s
[1] == '\0'); /* Our terminator. */
329 assert (s
[2] == '\0'); /* Added by getstr. */
332 /* If the previous path matched, scan the whole path for the last
333 char in the subpattern. If not, the shared prefix doesn't match
334 the pattern, so don't scan it for the last char. */
335 cutoff
= prev_fast_match
? path
: path
+ count
;
337 /* Search backward starting at the end of the path we just read in,
338 for the character at the end of the last glob-free subpattern
342 for (prev_fast_match
= false; s
>= cutoff
; s
--)
343 /* Fast first char check. */
344 if (TOLOWER(*s
) == *patend
)
346 char *s2
; /* Scan the path we read in. */
347 register char *p2
; /* Scan `patend'. */
349 for (s2
= s
- 1, p2
= patend
- 1; *p2
!= '\0' && TOLOWER(*s2
) == *p2
;
354 /* Success on the fast match. Compare the whole pattern
355 if it contains globbing characters. */
356 prev_fast_match
= true;
357 if (globflag
== false || fnmatch (pathpart
, path
, FNM_CASEFOLD
) == 0)
359 if (!check_existence
|| stat(path
, &st
) == 0)
371 for (prev_fast_match
= false; s
>= cutoff
; s
--)
372 /* Fast first char check. */
375 char *s2
; /* Scan the path we read in. */
376 register char *p2
; /* Scan `patend'. */
378 for (s2
= s
- 1, p2
= patend
- 1; *p2
!= '\0' && *s2
== *p2
;
383 /* Success on the fast match. Compare the whole pattern
384 if it contains globbing characters. */
385 prev_fast_match
= true;
386 if (globflag
== false || fnmatch (pathpart
, path
,
389 if (!check_existence
|| stat(path
, &st
) == 0)
404 error (0, errno
, "%s", dbfile
);
407 if (fclose (fp
) == EOF
)
409 error (0, errno
, "%s", dbfile
);
416 extern char *version_string
;
418 /* The name this program was run with. */
422 usage (stream
, status
)
426 fprintf (stream
, _("\
427 Usage: %s [-d path | --database=path] [-e | --existing]\n\
428 [-i | --ignore-case] [--version] [--help] pattern...\n"),
430 fputs (_("\nReport bugs to <bug-findutils@gnu.org>."), stream
);
434 static struct option
const longopts
[] =
436 {"database", required_argument
, NULL
, 'd'},
437 {"existing", no_argument
, NULL
, 'e'},
438 {"ignore-case", no_argument
, NULL
, 'i'},
439 {"help", no_argument
, NULL
, 'h'},
440 {"version", no_argument
, NULL
, 'v'},
441 {NULL
, no_argument
, NULL
, 0}
450 int fnmatch_flags
= 0;
454 program_name
= argv
[0];
456 #ifdef HAVE_SETLOCALE
457 setlocale (LC_ALL
, "");
459 bindtextdomain (PACKAGE
, LOCALEDIR
);
460 textdomain (PACKAGE
);
462 dbpath
= getenv ("LOCATE_PATH");
468 while ((optc
= getopt_long (argc
, argv
, "d:ei", longopts
, (int *) 0)) != -1)
481 fnmatch_flags
|= FNM_CASEFOLD
;
488 printf (_("GNU locate version %s\n"), version_string
);
498 for (; optind
< argc
; optind
++)
501 next_element (dbpath
); /* Initialize. */
502 while ((e
= next_element ((char *) NULL
)) != NULL
)
503 found
|= locate (argv
[optind
], e
, ignore_case
);