1 /* frcode -- front-compress a sorted list
2 Copyright (C) 1994, 2005, 2006, 2007, 2010, 2011 Free Software
5 This program is free software: you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation, either version 3 of the License, or
8 (at your option) any later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>.
19 /* Usage: frcode < sorted-list > compressed-list
21 Uses front compression (also known as incremental encoding);
22 see ";login:", March 1983, p. 8.
24 The input is a sorted list of NUL-terminated strings (or
25 newline-terminated if the -0 option is not given).
27 The output entries are in the same order as the input; each entry
28 consists of a signed offset-differential count byte (the additional
29 number of characters of prefix of the preceding entry to use beyond
30 the number that the preceding entry is using of its predecessor),
31 followed by a null-terminated ASCII remainder.
33 If the offset-differential count is larger than can be stored
34 in a byte (+/-127), the byte has the value LOCATEDB_ESCAPE
35 and the count follows in a 2-byte word, with the high byte first
40 Input, with NULs changed to newlines:
42 /usr/src/cmd/aardvark.c
43 /usr/src/cmd/armadillo.c
46 Length of the longest prefix of the preceding entry to share:
52 Output, with NULs changed to newlines and count bytes made printable:
59 (6 = 14 - 8, and -9 = 5 - 14)
61 Written by James A. Woods <jwoods@adobe.com>.
62 Modified by David MacKenzie <djm@gnu.org>.
63 Modified by James Youngman <jay@gnu.org>.
66 /* config.h must be included first. */
78 #include <sys/types.h>
88 #include "findutils-version.h"
93 # define _(Text) gettext (Text)
96 #define textdomain(Domain)
97 #define bindtextdomain(Package, Directory)
100 # define N_(String) gettext_noop (String)
102 /* We used to use (String) instead of just String, but apparently ISO C
103 * doesn't allow this (at least, that's what HP said when someone reported
104 * this as a compiler bug). This is HP case number 1205608192. See
105 * also http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11250 (which references
106 * ANSI 3.5.7p14-15). The Intel icc compiler also rejects constructs
107 * like: static const char buf[] = ("string");
109 # define N_(String) String
113 /* Write out a 16-bit int, high byte first (network byte order).
114 * Return true iff all went well.
117 put_short (int c, FILE *fp)
119 /* XXX: The value of c may be negative. ANSI C 1989 (section 6.3.7)
120 * indicates that the result of shifting a negative value right is
121 * implementation defined.
123 assert (c <= SHRT_MAX);
124 assert (c >= SHRT_MIN);
125 return (putc (c >> 8, fp) != EOF) && (putc (c, fp) != EOF);
128 /* Return the length of the longest common prefix of strings S1 and S2. */
131 prefix_length (char *s1, char *s2)
133 register char *start;
135 for (start = s1; *s1 == *s2 && *s1 != '\0'; s1++, s2++)
137 /* Don't emit a prefix length that will not fit into
146 static struct option const longopts[] =
148 {"help", no_argument, NULL, 'h'},
149 {"version", no_argument, NULL, 'v'},
150 {"null", no_argument, NULL, '0'},
151 {NULL, no_argument, NULL, 0}
154 extern char *version_string;
160 _("Usage: %s [-0 | --null] [--version] [--help]\n"),
162 fputs (_("\nReport bugs to <bug-findutils@gnu.org>.\n"), stream);
166 get_seclevel (char *s)
171 /* Reset errno in oreder to be able to distinguish LONG_MAX/LONG_MIN
172 * from values whichare actually out of range
176 result = strtol (s, &p, 10);
177 if ((0==result) && (p == optarg))
179 error (EXIT_FAILURE, 0,
180 _("You need to specify a security level as a decimal integer."));
184 else if ((LONG_MIN==result || LONG_MAX==result) && errno)
187 error (EXIT_FAILURE, 0,
188 _("Security level %s is outside the convertible range."), s);
194 /* Some suffix exists */
195 error (EXIT_FAILURE, 0,
196 _("Security level %s has unexpected suffix %s."), s, p);
209 /* Issue the same error message as closeout () would. */
210 error (EXIT_FAILURE, errno, _("write error"));
214 main (int argc, char **argv)
216 char *path; /* The current input entry. */
217 char *oldpath; /* The previous input entry. */
218 size_t pathsize, oldpathsize; /* Amounts allocated for them. */
219 int count, oldcount, diffcount; /* Their prefix lengths & the difference. */
220 int line_len; /* Length of input line. */
221 int delimiter = '\n';
223 int slocate_compat = 0;
224 long slocate_seclevel = 0L;
227 set_program_name (argv[0]);
229 set_program_name ("frcode");
231 if (atexit (close_stdout))
233 error (EXIT_FAILURE, errno, _("The atexit library function failed"));
236 pathsize = oldpathsize = 1026; /* Increased as necessary by getline. */
237 path = xmalloc (pathsize);
238 oldpath = xmalloc (oldpathsize);
244 while ((optc = getopt_long (argc, argv, "hv0S:", longopts, (int *) 0)) != -1)
253 slocate_seclevel = get_seclevel (optarg);
254 if (slocate_seclevel < 0 || slocate_seclevel > 1)
256 error (EXIT_FAILURE, 0,
257 _("slocate security level %ld is unsupported."),
267 display_findutils_version ("frcode");
275 /* We expect to have no arguments. */
285 fputc (slocate_seclevel ? '1' : '0', stdout);
291 /* GNU LOCATE02 format */
292 if (fwrite (LOCATEDB_MAGIC, 1, sizeof (LOCATEDB_MAGIC), stdout)
293 != sizeof (LOCATEDB_MAGIC))
295 error (EXIT_FAILURE, errno, _("Failed to write to standard output"));
300 while ((line_len = getdelim (&path, &pathsize, delimiter, stdin)) > 0)
302 path[line_len - 1] = '\0'; /* FIXME temporary: nuke the newline. */
304 count = prefix_length (oldpath, path);
305 diffcount = count - oldcount;
306 if ( (diffcount > SHRT_MAX) || (diffcount < SHRT_MIN) )
308 /* We do this to prevent overflow of the value we
309 * write with put_short ()
312 diffcount = (-oldcount);
318 /* Emit no count for the first pathname. */
323 /* If the difference is small, it fits in one byte;
324 otherwise, two bytes plus a marker noting that fact. */
325 if (diffcount < LOCATEDB_ONEBYTE_MIN
326 || diffcount > LOCATEDB_ONEBYTE_MAX)
328 if (EOF == putc (LOCATEDB_ESCAPE, stdout))
330 if (!put_short (diffcount, stdout))
335 if (EOF == putc (diffcount, stdout))
340 if ( (EOF == fputs (path + count, stdout))
341 || (EOF == putc ('\0', stdout)))
348 /* Swap path and oldpath and their sizes. */
349 char *tmppath = oldpath;
350 size_t tmppathsize = oldpathsize;
352 oldpathsize = pathsize;
354 pathsize = tmppathsize;