1 | /**************************************************************************
|
---|
2 | *
|
---|
3 | * mgpp_passes.cpp -- Driver for the various passes
|
---|
4 | * Copyright (C) 1994 Neil Sharman
|
---|
5 | * Copyright (C) 1999 Rodger McNab
|
---|
6 | *
|
---|
7 | * This program is free software; you can redistribute it and/or modify
|
---|
8 | * it under the terms of the GNU General Public License as published by
|
---|
9 | * the Free Software Foundation; either version 2 of the License, or
|
---|
10 | * (at your option) any later version.
|
---|
11 | *
|
---|
12 | * This program is distributed in the hope that it will be useful,
|
---|
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
15 | * GNU General Public License for more details.
|
---|
16 | *
|
---|
17 | * You should have received a copy of the GNU General Public License
|
---|
18 | * along with this program; if not, write to the Free Software
|
---|
19 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
20 | *
|
---|
21 | **************************************************************************/
|
---|
22 |
|
---|
23 | #define _XOPEN_SOURCE 1
|
---|
24 | // This was added for Solaris, but it makes things worse on Solaris for me...
|
---|
25 | // #define _XOPEN_SOURCE_EXTENDED 1
|
---|
26 |
|
---|
27 | #if defined(WIN32) || defined(_WIN32) || defined(__WIN32__)
|
---|
28 | #pragma warning(disable:4786)
|
---|
29 | #endif
|
---|
30 |
|
---|
31 | // need this to avoid bizarre compiler problems under VC++ 6.0
|
---|
32 | #if defined (__WIN32__) && !defined (GSDL_USE_IOS_H)
|
---|
33 | # include <iostream>
|
---|
34 | #endif
|
---|
35 |
|
---|
36 | #include "sysfuncs.h"
|
---|
37 |
|
---|
38 | #ifdef HAVE_MALLINFO
|
---|
39 | # include <malloc.h>
|
---|
40 | #endif
|
---|
41 |
|
---|
42 | #if defined __WIN32__
|
---|
43 | # include <io.h>
|
---|
44 | # include "getopt_old.h"
|
---|
45 | # define close _close
|
---|
46 | # define open _open
|
---|
47 | #elif defined __CYGWIN__
|
---|
48 | #include "getopt_old.h"
|
---|
49 | #else
|
---|
50 | # include <unistd.h>
|
---|
51 | #endif
|
---|
52 |
|
---|
53 | #include "memlib.h"
|
---|
54 | #include "messages.h"
|
---|
55 | #include "longlong.h"
|
---|
56 | #include "mg_files.h"
|
---|
57 | #include "mg.h"
|
---|
58 | #include "build.h"
|
---|
59 | #include "text.h"
|
---|
60 | #include "stemmer.h"
|
---|
61 | #include "FileBuf.h"
|
---|
62 | #include "TextEl.h"
|
---|
63 | #include "TagInfo.h"
|
---|
64 | #include "words.h"
|
---|
65 | #define MAX_PASSES 5
|
---|
66 |
|
---|
67 | #define SPECIAL 1
|
---|
68 | #define TEXT_PASS_1 2
|
---|
69 | #define TEXT_PASS_2 4
|
---|
70 | #define IVF_PASS_1 8
|
---|
71 | #define IVF_PASS_2 16
|
---|
72 |
|
---|
73 | #define MIN_BUF 8192
|
---|
74 |
|
---|
75 |
|
---|
76 | unsigned long invf_buffer_size = 5 * 1024 * 1024; /* 5Mb */
|
---|
77 |
|
---|
78 | static char Passes = 0;
|
---|
79 | static char **files = NULL;
|
---|
80 | static int num_files = 0;
|
---|
81 |
|
---|
82 |
|
---|
83 | struct pass_data {
|
---|
84 | char *name;
|
---|
85 | int (*init) (const TagInfo &tagInfo, char *);
|
---|
86 | int (*process) (const TagInfo &tagInfo, const TextElArray &doc);
|
---|
87 | int (*done) (const TagInfo &tagInfo, char *);
|
---|
88 | };
|
---|
89 |
|
---|
90 |
|
---|
91 | static pass_data PassData[MAX_PASSES] = {
|
---|
92 | {"special", init_special, process_special, done_special},
|
---|
93 | {"text.pass1", init_text_1, process_text_1, done_text_1},
|
---|
94 | {"text.pass2", init_text_2, process_text_2, done_text_2},
|
---|
95 | {"ivf.pass1", init_ivf_1, process_ivf_1, done_ivf_1},
|
---|
96 | {"ivf.pass2", init_ivf_2, process_ivf_2, done_ivf_2},
|
---|
97 | };
|
---|
98 |
|
---|
99 | static char *usage_str = "\nUSAGE:\n"
|
---|
100 | " %s [-J doc-tag] [-K level-tag] [-L index-level]\n"
|
---|
101 | " %*s [-m invf-memory] [-T1] [-T2] [-I1] [-I2] [-S]\n"
|
---|
102 | " %*s [-C] [-h] [-d directory] [-M maxnumeric] -f name\n\n";
|
---|
103 |
|
---|
104 |
|
---|
105 |
|
---|
106 | static void Usage (char *err) {
|
---|
107 | if (err) Message (err);
|
---|
108 | fprintf (stderr, usage_str, msg_prefix, strlen (msg_prefix), "",
|
---|
109 | strlen (msg_prefix), "");
|
---|
110 | exit (1);
|
---|
111 | }
|
---|
112 |
|
---|
113 |
|
---|
114 |
|
---|
115 |
|
---|
116 | int OpenNextFile (int in_fd) {
|
---|
117 | if (in_fd > 0) close (in_fd);
|
---|
118 | if (num_files == 0) return (-1);
|
---|
119 | if ((in_fd = open (files[0], O_RDONLY)) == -1)
|
---|
120 | FatalError (1, "Cannot open %s", files[0]);
|
---|
121 | ++files;
|
---|
122 | --num_files;
|
---|
123 | return (in_fd);
|
---|
124 | }
|
---|
125 |
|
---|
126 |
|
---|
127 | static void Driver (int in_fd, char *file_name,
|
---|
128 | const TagInfo &tagInfo, bool compatMode) {
|
---|
129 | // cout << tagInfo;
|
---|
130 |
|
---|
131 | int pass;
|
---|
132 |
|
---|
133 | unsigned long numBytes = 0;
|
---|
134 | unsigned long numDocs = 0;
|
---|
135 |
|
---|
136 | // initialise all the passes
|
---|
137 | for (pass = 0; pass < MAX_PASSES; ++pass) {
|
---|
138 | if (Passes & (1 << pass)) {
|
---|
139 | if (PassData[pass].init (tagInfo, file_name) == COMPERROR)
|
---|
140 | FatalError (1, "Error during init of \"%s\"", PassData[pass].name);
|
---|
141 | }
|
---|
142 | }
|
---|
143 |
|
---|
144 |
|
---|
145 | // set up various variables
|
---|
146 | FileBuf buf;
|
---|
147 | TextElArray doc;
|
---|
148 | unsigned long docLen = 0;
|
---|
149 |
|
---|
150 | // read and process each file (start with an open file)
|
---|
151 | do {
|
---|
152 |
|
---|
153 | // read and process each document in this file
|
---|
154 | buf.SetFD (in_fd);
|
---|
155 | while (ReadDoc (buf, tagInfo.docTag, doc, docLen, compatMode)) {
|
---|
156 |
|
---|
157 | // give this document to each pass
|
---|
158 | for (pass = 0; pass < MAX_PASSES; ++pass) {
|
---|
159 | if (Passes & (1 << pass)) {
|
---|
160 | if (PassData[pass].process (tagInfo, doc) == COMPERROR)
|
---|
161 | FatalError(1,"Error during processing of \"%s\"",PassData[pass].name);
|
---|
162 | }
|
---|
163 | }
|
---|
164 |
|
---|
165 | // another document has been processed
|
---|
166 | numBytes += docLen;
|
---|
167 | ++numDocs;
|
---|
168 | }
|
---|
169 |
|
---|
170 | } while ((in_fd = OpenNextFile (in_fd)) > 0);
|
---|
171 |
|
---|
172 |
|
---|
173 | // do done for each pass
|
---|
174 | for (pass = 0; pass < MAX_PASSES; ++pass) {
|
---|
175 | if (Passes & (1 << pass)) {
|
---|
176 | if (PassData[pass].done (tagInfo, file_name) == COMPERROR)
|
---|
177 | FatalError (1, "Error during done of \"%s\"", PassData[pass].name);
|
---|
178 | }
|
---|
179 | }
|
---|
180 | }
|
---|
181 |
|
---|
182 | int main (int argc, char **argv) {
|
---|
183 | int ch, in_fd, maxnum;
|
---|
184 | char *filename = NULL;
|
---|
185 | bool compatMode = false;
|
---|
186 | TagInfo tagInfo;
|
---|
187 | tagInfo.SetDocTag ("Document");
|
---|
188 |
|
---|
189 | msg_prefix = argv[0];
|
---|
190 |
|
---|
191 | opterr = 0;
|
---|
192 | while ((ch=getopt(argc, argv, "J:K:L:M:f:d:m:I:T:SCh"))!=-1){
|
---|
193 | switch (ch) {
|
---|
194 | case 'J':
|
---|
195 | tagInfo.SetDocTag (optarg);
|
---|
196 | break;
|
---|
197 | case 'K':
|
---|
198 | tagInfo.AddLevelTag (optarg);
|
---|
199 | break;
|
---|
200 | case 'L':
|
---|
201 | tagInfo.SetIndexLevel (optarg);
|
---|
202 | break;
|
---|
203 | case 'M':
|
---|
204 | maxnum = atoi(optarg);
|
---|
205 | if (maxnum > 4 && maxnum < 512) {
|
---|
206 | MAXNUMERIC = maxnum;
|
---|
207 | }
|
---|
208 | break;
|
---|
209 | case 'f':
|
---|
210 | filename = optarg;
|
---|
211 | break;
|
---|
212 | case 'd':
|
---|
213 | set_basepath (optarg);
|
---|
214 | break;
|
---|
215 | case 'm':
|
---|
216 | invf_buffer_size = (int) (atof (optarg) * 1024 * 1024);
|
---|
217 | break;
|
---|
218 | case 'I':
|
---|
219 | if (*optarg == '1')
|
---|
220 | Passes |= IVF_PASS_1;
|
---|
221 | else if (*optarg == '2')
|
---|
222 | Passes |= IVF_PASS_2;
|
---|
223 | else
|
---|
224 | Usage ("Invalid pass number");
|
---|
225 | break;
|
---|
226 | case 'T':
|
---|
227 | if (*optarg == '1')
|
---|
228 | Passes |= TEXT_PASS_1;
|
---|
229 | else if (*optarg == '2')
|
---|
230 | Passes |= TEXT_PASS_2;
|
---|
231 | else
|
---|
232 | Usage ("Invalid pass number");
|
---|
233 | break;
|
---|
234 | case 'S':
|
---|
235 | Passes |= SPECIAL;
|
---|
236 | break;
|
---|
237 | case 'C':
|
---|
238 | compatMode = true;
|
---|
239 | break;
|
---|
240 | case 'h':
|
---|
241 | case '?':
|
---|
242 | Usage (NULL);
|
---|
243 | }
|
---|
244 | }
|
---|
245 |
|
---|
246 | if (!filename || *filename == '\0')
|
---|
247 | FatalError (1, "A document collection name must be specified.");
|
---|
248 |
|
---|
249 | if ((Passes & (IVF_PASS_1 | IVF_PASS_2)) == (IVF_PASS_1 | IVF_PASS_2))
|
---|
250 | FatalError (1, "I1 and I2 cannot be done simultaneously.");
|
---|
251 |
|
---|
252 | if ((Passes & (TEXT_PASS_1 | TEXT_PASS_2)) == (TEXT_PASS_1 | TEXT_PASS_2))
|
---|
253 | FatalError (1, "T1 and T2 cannot be done simultaneously.");
|
---|
254 |
|
---|
255 | if (!Passes)
|
---|
256 | FatalError (1, "S, T1, T2, I1 or I2 must be specified.");
|
---|
257 |
|
---|
258 | if (optind < argc) {
|
---|
259 | if ((in_fd = open (argv[optind], O_RDONLY)) == -1)
|
---|
260 | FatalError (1, "Cannot open %s", argv[optind]);
|
---|
261 | files = &argv[optind + 1];
|
---|
262 | num_files = argc - (optind + 1);
|
---|
263 |
|
---|
264 | } else in_fd = 0; // stdin
|
---|
265 |
|
---|
266 |
|
---|
267 | if (compatMode) tagInfo.SetDocTag ("Document");
|
---|
268 |
|
---|
269 | // a document tag is also a level tag
|
---|
270 | tagInfo.levelTags.insert (tagInfo.docTag);
|
---|
271 |
|
---|
272 | Driver (in_fd, filename, tagInfo, compatMode);
|
---|
273 |
|
---|
274 | return (0);
|
---|
275 | }
|
---|