1 | /**************************************************************************
|
---|
2 | *
|
---|
3 | * mg_invf_dump.c -- Program to dump uot an inverted fil
|
---|
4 | * Copyright (C) 1994 Neil Sharman
|
---|
5 | *
|
---|
6 | * This program is free software; you can redistribute it and/or modify
|
---|
7 | * it under the terms of the GNU General Public License as published by
|
---|
8 | * the Free Software Foundation; either version 2 of the License, or
|
---|
9 | * (at your option) any later version.
|
---|
10 | *
|
---|
11 | * This program is distributed in the hope that it will be useful,
|
---|
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
14 | * GNU General Public License for more details.
|
---|
15 | *
|
---|
16 | * You should have received a copy of the GNU General Public License
|
---|
17 | * along with this program; if not, write to the Free Software
|
---|
18 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
19 | *
|
---|
20 | * $Id: mg_invf_dump.c 439 1999-08-10 21:23:37Z sjboddie $
|
---|
21 | *
|
---|
22 | **************************************************************************/
|
---|
23 |
|
---|
24 | #include "sysfuncs.h"
|
---|
25 |
|
---|
26 | #include "messages.h"
|
---|
27 | #include "timing.h"
|
---|
28 | #include "bitio_m.h"
|
---|
29 | #include "bitio_m_stdio.h"
|
---|
30 | #include "bitio_gen.h"
|
---|
31 | #include "netorder.h" /* [RPAP - Jan 97: Endian Ordering] */
|
---|
32 |
|
---|
33 | #include "mg_files.h"
|
---|
34 | #include "locallib.h"
|
---|
35 | #include "words.h"
|
---|
36 | #include "invf.h"
|
---|
37 |
|
---|
38 | extern unsigned long S_btg;
|
---|
39 |
|
---|
40 | /*
|
---|
41 | $Log$
|
---|
42 | Revision 1.1 1999/08/10 21:18:09 sjboddie
|
---|
43 | renamed mg-1.3d directory mg
|
---|
44 |
|
---|
45 | Revision 1.2 1998/11/25 07:55:46 rjmcnab
|
---|
46 |
|
---|
47 | Modified mg to that you can specify the stemmer you want
|
---|
48 | to use via a command line option. You specify it to
|
---|
49 | mg_passes during the build process. The number of the
|
---|
50 | stemmer that you used is stored within the inverted
|
---|
51 | dictionary header and the stemmed dictionary header so
|
---|
52 | the correct stemmer is used in later stages of building
|
---|
53 | and querying.
|
---|
54 |
|
---|
55 | Revision 1.1 1998/11/17 09:35:05 rjmcnab
|
---|
56 | *** empty log message ***
|
---|
57 |
|
---|
58 | * Revision 1.3 1994/11/29 00:32:01 tes
|
---|
59 | * Committing the new merged files and changes.
|
---|
60 | *
|
---|
61 | * Revision 1.2 1994/09/20 04:41:50 tes
|
---|
62 | * For version 1.1
|
---|
63 | *
|
---|
64 | */
|
---|
65 |
|
---|
66 | static char *RCSID = "$Id: mg_invf_dump.c 439 1999-08-10 21:23:37Z sjboddie $";
|
---|
67 |
|
---|
68 |
|
---|
69 |
|
---|
70 | static void process_files (char *filename);
|
---|
71 |
|
---|
72 | int binary = 0;
|
---|
73 | int word_counts = 0;
|
---|
74 | int term_dump = 0;
|
---|
75 |
|
---|
76 |
|
---|
77 | int main (int argc, char **argv)
|
---|
78 | {
|
---|
79 | ProgTime start;
|
---|
80 | char *dir_name, *file_name = "";
|
---|
81 | int ch;
|
---|
82 | msg_prefix = argv[0];
|
---|
83 | dir_name = getenv ("MGDATA");
|
---|
84 | opterr = 0;
|
---|
85 | msg_prefix = argv[0];
|
---|
86 | while ((ch = getopt (argc, argv, "hbwtf:d:")) != -1)
|
---|
87 | switch (ch)
|
---|
88 | {
|
---|
89 | case 'f': /* input file */
|
---|
90 | file_name = optarg;
|
---|
91 | break;
|
---|
92 | case 'd':
|
---|
93 | set_basepath(optarg);
|
---|
94 | break;
|
---|
95 | case 'b':
|
---|
96 | binary = 1;
|
---|
97 | break;
|
---|
98 | case 'w':
|
---|
99 | word_counts = 1;
|
---|
100 | break;
|
---|
101 | case 't':
|
---|
102 | term_dump = 1;
|
---|
103 | break;
|
---|
104 | case 'h':
|
---|
105 | case '?':
|
---|
106 | fprintf (stderr, "usage: %s [-h] [-b] [-w] [-t] [-f input_file]"
|
---|
107 | "[-d data directory]\n", argv[0]);
|
---|
108 | exit (1);
|
---|
109 | }
|
---|
110 | GetTime (&start);
|
---|
111 | process_files (file_name);
|
---|
112 | Message ("%s\n", ElapsedTime (&start, NULL));
|
---|
113 | return 0;
|
---|
114 | }
|
---|
115 |
|
---|
116 | static void
|
---|
117 | process_files (char *name)
|
---|
118 | {
|
---|
119 | unsigned long N, Nstatic, k;
|
---|
120 | FILE *dict;
|
---|
121 | FILE *invf;
|
---|
122 | struct invf_dict_header idh;
|
---|
123 | struct invf_file_header ifh;
|
---|
124 | int i; /* [RPAP - Jan 97: Endian Ordering] */
|
---|
125 |
|
---|
126 | dict = open_file (name, INVF_DICT_SUFFIX, "rb", MAGIC_STEM_BUILD, MG_ABORT); /* [RPAP - Feb 97: WIN32 Port] */
|
---|
127 |
|
---|
128 | fread ((char *) &idh, sizeof (idh), 1, dict);
|
---|
129 |
|
---|
130 | /* [RPAP - Jan 97: Endian Ordering] */
|
---|
131 | NTOHUL(idh.lookback);
|
---|
132 | NTOHUL(idh.dict_size);
|
---|
133 | NTOHUL(idh.total_bytes);
|
---|
134 | NTOHUL(idh.index_string_bytes);
|
---|
135 | NTOHD(idh.input_bytes); /* [RJM 07/97: 4G limit] */
|
---|
136 | NTOHUL(idh.num_of_docs);
|
---|
137 | NTOHUL(idh.static_num_of_docs);
|
---|
138 | NTOHUL(idh.num_of_words);
|
---|
139 | NTOHUL(idh.stemmer_num);
|
---|
140 | NTOHUL(idh.stem_method);
|
---|
141 |
|
---|
142 | if (!(invf = open_file (name, INVF_SUFFIX ".ORG", "rb", MAGIC_INVF, MG_CONTINUE))) /* [RPAP - Feb 97: WIN32 Port] */
|
---|
143 | invf = open_file (name, INVF_SUFFIX, "rb", MAGIC_INVF, MG_ABORT); /* [RPAP - Feb 97: WIN32 Port] */
|
---|
144 |
|
---|
145 | fread ((char *) &ifh, sizeof (ifh), 1, invf);
|
---|
146 |
|
---|
147 | /* [RPAP - Jan 97: Endian Ordering] */
|
---|
148 | NTOHUL(ifh.no_of_words);
|
---|
149 | NTOHUL(ifh.no_of_ptrs);
|
---|
150 | NTOHUL(ifh.skip_mode);
|
---|
151 | for (i = 0; i < 16; i++)
|
---|
152 | NTOHUL(ifh.params[i]);
|
---|
153 | NTOHUL(ifh.InvfLevel);
|
---|
154 |
|
---|
155 | if (ifh.skip_mode != 0)
|
---|
156 | FatalError (1, "The invf file contains skips. Unable to dump.");
|
---|
157 |
|
---|
158 | DECODE_START (invf)
|
---|
159 | N = idh.num_of_docs;
|
---|
160 | Nstatic = idh.static_num_of_docs;
|
---|
161 | if (binary)
|
---|
162 | {
|
---|
163 | fwrite ((char *) &N, sizeof (N), 1, stdout);
|
---|
164 | fwrite ((char *) &ifh.no_of_words, sizeof (ifh.no_of_words), 1, stdout);
|
---|
165 | }
|
---|
166 | else
|
---|
167 | printf ("%ld %ld\n", N, ifh.no_of_words);
|
---|
168 | for (k = 0; k < ifh.no_of_words; k++)
|
---|
169 | {
|
---|
170 | int i, blk, doc;
|
---|
171 | register unsigned long suff, prefix;
|
---|
172 | unsigned long fcnt, wcnt;
|
---|
173 | char term[MAXSTEMLEN + 1];
|
---|
174 |
|
---|
175 | prefix = fgetc (dict);
|
---|
176 | suff = fgetc (dict);
|
---|
177 | fread (&term[prefix], sizeof (char), suff, dict);
|
---|
178 | fread ((char *) &fcnt, sizeof (fcnt), 1, dict);
|
---|
179 | fread ((char *) &wcnt, sizeof (wcnt), 1, dict);
|
---|
180 | term[prefix + suff] = '\0';
|
---|
181 |
|
---|
182 | /* [RPAP - Jan 97: Endian Ordering] */
|
---|
183 | NTOHUL(fcnt);
|
---|
184 | NTOHUL(wcnt);
|
---|
185 |
|
---|
186 | if (binary)
|
---|
187 | fwrite ((char *) &fcnt, sizeof (fcnt), 1, stdout);
|
---|
188 | else
|
---|
189 | {
|
---|
190 | if (term_dump)
|
---|
191 | printf ("%ld \"%s\"\n", fcnt, term);
|
---|
192 | else
|
---|
193 | printf ("%ld\n", fcnt);
|
---|
194 | }
|
---|
195 | blk = BIO_Bblock_Init (Nstatic, fcnt);
|
---|
196 | for (doc = i = 0; i < fcnt; i++)
|
---|
197 | {
|
---|
198 | int num;
|
---|
199 | BBLOCK_DECODE (num, blk);
|
---|
200 | doc += num;
|
---|
201 | if (binary)
|
---|
202 | fwrite ((char *) &doc, sizeof (doc), 1, stdout);
|
---|
203 | else
|
---|
204 | printf (" %d", doc);
|
---|
205 | if (ifh.InvfLevel >= 2)
|
---|
206 | {
|
---|
207 | int count;
|
---|
208 | GAMMA_DECODE (count);
|
---|
209 | if (word_counts)
|
---|
210 | if (binary)
|
---|
211 | fwrite ((char *) &count, sizeof (count), 1, stdout);
|
---|
212 | else
|
---|
213 | printf (" %d", count);
|
---|
214 | }
|
---|
215 | if (!binary)
|
---|
216 | putchar ('\n');
|
---|
217 | }
|
---|
218 | while (__btg)
|
---|
219 | DECODE_BIT;
|
---|
220 | }
|
---|
221 | DECODE_DONE
|
---|
222 | }
|
---|