1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * gdbmclass.cpp --
|
---|
4 | * Copyright (C) 1999-2008 The New Zealand Digital Library Project
|
---|
5 | *
|
---|
6 | * A component of the Greenstone digital library software
|
---|
7 | * from the New Zealand Digital Library Project at the
|
---|
8 | * University of Waikato, New Zealand.
|
---|
9 | *
|
---|
10 | * This program is free software; you can redistribute it and/or modify
|
---|
11 | * it under the terms of the GNU General Public License as published by
|
---|
12 | * the Free Software Foundation; either version 2 of the License, or
|
---|
13 | * (at your option) any later version.
|
---|
14 | *
|
---|
15 | * This program is distributed in the hope that it will be useful,
|
---|
16 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
17 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
18 | * GNU General Public License for more details.
|
---|
19 | *
|
---|
20 | * You should have received a copy of the GNU General Public License
|
---|
21 | * along with this program; if not, write to the Free Software
|
---|
22 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | *
|
---|
24 | *********************************************************************/
|
---|
25 |
|
---|
26 | #include "gdbmclass.h"
|
---|
27 | #include "unitool.h"
|
---|
28 | #include "gsdlunicode.h"
|
---|
29 | #include "OIDtools.h"
|
---|
30 |
|
---|
31 |
|
---|
32 | gdbmclass::~gdbmclass()
|
---|
33 | {
|
---|
34 | closedatabase();
|
---|
35 | }
|
---|
36 |
|
---|
37 |
|
---|
38 | // returns true if opened
|
---|
39 | bool gdbmclass::opendatabase (const text_t &filename, int mode, int num_retrys,
|
---|
40 | #ifdef __WIN32__
|
---|
41 | bool need_filelock
|
---|
42 | #else
|
---|
43 | bool
|
---|
44 | #endif
|
---|
45 | )
|
---|
46 | {
|
---|
47 | text_t data_location;
|
---|
48 | int block_size = 512;
|
---|
49 |
|
---|
50 | if (gdbmfile != NULL) {
|
---|
51 | if (openfile == filename) return true;
|
---|
52 | else closedatabase ();
|
---|
53 | }
|
---|
54 |
|
---|
55 | openfile = filename;
|
---|
56 |
|
---|
57 | // Map the DB mode values into GDBM mode values
|
---|
58 | int gdbm_mode = GDBM_READER;
|
---|
59 | if (mode == DB_WRITER)
|
---|
60 | {
|
---|
61 | gdbm_mode = GDBM_WRITER;
|
---|
62 | }
|
---|
63 | else if (mode == DB_WRITER_CREATE)
|
---|
64 | {
|
---|
65 | gdbm_mode = GDBM_WRCREAT;
|
---|
66 | }
|
---|
67 |
|
---|
68 | char *namebuffer = filename.getcstr();
|
---|
69 | do {
|
---|
70 | #ifdef __WIN32__
|
---|
71 | gdbmfile = gdbm_open (namebuffer, block_size, gdbm_mode, 00664, NULL, (need_filelock) ? 1 : 0);
|
---|
72 | #else
|
---|
73 | gdbmfile = gdbm_open (namebuffer, block_size, gdbm_mode, 00664, NULL);
|
---|
74 | #endif
|
---|
75 | --num_retrys;
|
---|
76 | } while (num_retrys>0 && gdbmfile==NULL &&
|
---|
77 | (gdbm_errno==GDBM_CANT_BE_READER || gdbm_errno==GDBM_CANT_BE_WRITER));
|
---|
78 | delete []namebuffer;
|
---|
79 |
|
---|
80 | if (gdbmfile == NULL && logout != NULL) {
|
---|
81 | outconvertclass text_t2ascii;
|
---|
82 | (*logout) << text_t2ascii << "database open failed on: " << filename << "\n";
|
---|
83 | }
|
---|
84 |
|
---|
85 | return (gdbmfile != NULL);
|
---|
86 | }
|
---|
87 |
|
---|
88 |
|
---|
89 | void gdbmclass::closedatabase ()
|
---|
90 | {
|
---|
91 | if (gdbmfile == NULL) return;
|
---|
92 |
|
---|
93 | gdbm_close (gdbmfile);
|
---|
94 | gdbmfile = NULL;
|
---|
95 | openfile.clear();
|
---|
96 | }
|
---|
97 |
|
---|
98 |
|
---|
99 | void gdbmclass::deletekey (const text_t &key)
|
---|
100 | {
|
---|
101 | if (gdbmfile == NULL) return;
|
---|
102 |
|
---|
103 | // get a utf-8 encoded c string of the unicode key
|
---|
104 | datum key_data;
|
---|
105 | key_data.dptr = (to_utf8(key)).getcstr();
|
---|
106 | if (key_data.dptr == NULL) return;
|
---|
107 | key_data.dsize = strlen (key_data.dptr);
|
---|
108 |
|
---|
109 | // delete the key
|
---|
110 | gdbm_delete (gdbmfile, key_data);
|
---|
111 |
|
---|
112 | // free up the key memory
|
---|
113 | delete []key_data.dptr;
|
---|
114 | }
|
---|
115 |
|
---|
116 |
|
---|
117 | // returns true if exists
|
---|
118 | bool gdbmclass::exists (const text_t& key)
|
---|
119 | {
|
---|
120 | text_t data;
|
---|
121 | return getkeydata (key, data);
|
---|
122 | }
|
---|
123 |
|
---|
124 |
|
---|
125 | // returns true on success
|
---|
126 | bool gdbmclass::getinfo (const text_t& key, infodbclass &info)
|
---|
127 | {
|
---|
128 | text_t data;
|
---|
129 |
|
---|
130 | if (!getkeydata (key, data)) return false;
|
---|
131 | text_t::iterator here = data.begin ();
|
---|
132 | text_t::iterator end = data.end ();
|
---|
133 |
|
---|
134 | text_t ikey, ivalue;
|
---|
135 | info.clear (); // reset info
|
---|
136 |
|
---|
137 | while (getinfoline (here, end, ikey, ivalue)) {
|
---|
138 | info.addinfo (ikey, ivalue);
|
---|
139 | }
|
---|
140 |
|
---|
141 | return true;
|
---|
142 | }
|
---|
143 |
|
---|
144 |
|
---|
145 | // returns true on success
|
---|
146 | bool gdbmclass::getkeydata (const text_t& key, text_t &data)
|
---|
147 | {
|
---|
148 | datum key_data;
|
---|
149 | datum return_data;
|
---|
150 |
|
---|
151 | if (gdbmfile == NULL) return false;
|
---|
152 |
|
---|
153 | // get a utf-8 encoded c string of the unicode key
|
---|
154 | key_data.dptr = (to_utf8(key)).getcstr();
|
---|
155 | if (key_data.dptr == NULL) {
|
---|
156 | if (logout != NULL) (*logout) << "gdbmclass: out of memory\n";
|
---|
157 | return false;
|
---|
158 | }
|
---|
159 | key_data.dsize = strlen (key_data.dptr);
|
---|
160 |
|
---|
161 | // fetch the result
|
---|
162 | return_data = gdbm_fetch (gdbmfile, key_data);
|
---|
163 | delete []key_data.dptr;
|
---|
164 |
|
---|
165 | if (return_data.dptr == NULL) return false;
|
---|
166 |
|
---|
167 | data.setcarr (return_data.dptr, return_data.dsize);
|
---|
168 | free (return_data.dptr);
|
---|
169 | data = to_uni(data); // convert to unicode
|
---|
170 |
|
---|
171 | return true;
|
---|
172 | }
|
---|
173 |
|
---|
174 |
|
---|
175 | // returns array of keys
|
---|
176 | text_tarray gdbmclass::getkeys ()
|
---|
177 | {
|
---|
178 | text_tarray keys;
|
---|
179 |
|
---|
180 | text_t key = getfirstkey();
|
---|
181 | while (!key.empty())
|
---|
182 | {
|
---|
183 | keys.push_back(key);
|
---|
184 | key = getnextkey(key);
|
---|
185 | }
|
---|
186 |
|
---|
187 | return keys;
|
---|
188 | }
|
---|
189 |
|
---|
190 |
|
---|
191 | // returns true on success
|
---|
192 | bool gdbmclass::setinfo (const text_t &key, const infodbclass &info)
|
---|
193 | {
|
---|
194 | if (gdbmfile == NULL) return false;
|
---|
195 |
|
---|
196 | text_t subkey;
|
---|
197 | text_t data;
|
---|
198 |
|
---|
199 | // get all the keys and values
|
---|
200 | infodbclass::const_iterator info_here = info.begin();
|
---|
201 | infodbclass::const_iterator info_end = info.end();
|
---|
202 | while (info_here != info_end) {
|
---|
203 | // add the key
|
---|
204 | subkey.clear();
|
---|
205 | subkey.push_back('<');
|
---|
206 | text_t::const_iterator subkey_here = (*info_here).first.begin();
|
---|
207 | text_t::const_iterator subkey_end = (*info_here).first.end();
|
---|
208 | while (subkey_here != subkey_end) {
|
---|
209 | if (*subkey_here == '>') {
|
---|
210 | subkey.push_back('\\'); subkey.push_back('>');
|
---|
211 | } else if (*subkey_here == '\n') {
|
---|
212 | subkey.push_back('\\'); subkey.push_back('n');
|
---|
213 | } else if (*subkey_here == '\r') {
|
---|
214 | subkey.push_back('\\'); subkey.push_back('r');
|
---|
215 | } else if (*subkey_here == '\\') {
|
---|
216 | subkey.push_back('\\'); subkey.push_back('\\');
|
---|
217 | } else {
|
---|
218 | subkey.push_back (*subkey_here);
|
---|
219 | }
|
---|
220 | ++subkey_here;
|
---|
221 | }
|
---|
222 | subkey.push_back('>');
|
---|
223 |
|
---|
224 | // add the values
|
---|
225 | text_tarray::const_iterator subvalue_here = (*info_here).second.begin();
|
---|
226 | text_tarray::const_iterator subvalue_end = (*info_here).second.end();
|
---|
227 | while (subvalue_here != subvalue_end) {
|
---|
228 | data += subkey;
|
---|
229 |
|
---|
230 | text_t::const_iterator thissubvalue_here = (*subvalue_here).begin();
|
---|
231 | text_t::const_iterator thissubvalue_end = (*subvalue_here).end();
|
---|
232 | while (thissubvalue_here != thissubvalue_end) {
|
---|
233 | if (*thissubvalue_here == '>') {
|
---|
234 | data.push_back('\\'); data.push_back('>');
|
---|
235 | } else if (*thissubvalue_here == '\n') {
|
---|
236 | data.push_back('\\'); data.push_back('n');
|
---|
237 | } else if (*thissubvalue_here == '\r') {
|
---|
238 | data.push_back('\\'); data.push_back('r');
|
---|
239 | } else if (*thissubvalue_here == '\\') {
|
---|
240 | data.push_back('\\'); data.push_back('\\');
|
---|
241 | } else {
|
---|
242 | data.push_back (*thissubvalue_here);
|
---|
243 | }
|
---|
244 |
|
---|
245 | ++thissubvalue_here;
|
---|
246 | }
|
---|
247 |
|
---|
248 | data.push_back('\n');
|
---|
249 | ++subvalue_here;
|
---|
250 | }
|
---|
251 |
|
---|
252 | ++info_here;
|
---|
253 | }
|
---|
254 |
|
---|
255 | return setkeydata(key, data);
|
---|
256 | }
|
---|
257 |
|
---|
258 |
|
---|
259 | // returns true on success
|
---|
260 | bool gdbmclass::setkeydata (const text_t &key, const text_t &data)
|
---|
261 | {
|
---|
262 | if (gdbmfile == NULL) return false;
|
---|
263 |
|
---|
264 | // store the value
|
---|
265 | datum key_data;
|
---|
266 | datum data_data;
|
---|
267 |
|
---|
268 | // get a utf-8 encoded c string of the unicode key
|
---|
269 | key_data.dptr = (to_utf8(key)).getcstr();
|
---|
270 | if (key_data.dptr == NULL) {
|
---|
271 | if (logout != NULL) (*logout) << "gdbmclass: out of memory\n";
|
---|
272 | return false;
|
---|
273 | }
|
---|
274 | key_data.dsize = strlen (key_data.dptr);
|
---|
275 |
|
---|
276 | data_data.dptr = (to_utf8(data)).getcstr();
|
---|
277 | if (data_data.dptr == NULL) {
|
---|
278 | if (logout != NULL) (*logout) << "gdbmclass: out of memory\n";
|
---|
279 | delete []key_data.dptr;
|
---|
280 | return false;
|
---|
281 | }
|
---|
282 | data_data.dsize = strlen (data_data.dptr);
|
---|
283 |
|
---|
284 | int ret = gdbm_store (gdbmfile, key_data, data_data, GDBM_REPLACE);
|
---|
285 | delete []key_data.dptr;
|
---|
286 | delete []data_data.dptr;
|
---|
287 |
|
---|
288 | return (ret == 0);
|
---|
289 | }
|
---|
290 |
|
---|
291 |
|
---|
292 | // ----------------------------------------------------------------------------------------
|
---|
293 | // GDBM-ONLY FUNCTIONS
|
---|
294 | // ----------------------------------------------------------------------------------------
|
---|
295 |
|
---|
296 | // getfirstkey and getnextkey are used for traversing the database
|
---|
297 | // no insertions or deletions should be carried out while traversing
|
---|
298 | // the database. when there are no keys left to visit in the database
|
---|
299 | // an empty string is returned.
|
---|
300 | text_t gdbmclass::getfirstkey ()
|
---|
301 | {
|
---|
302 | if (gdbmfile == NULL) return g_EmptyText;
|
---|
303 |
|
---|
304 | // get the first key
|
---|
305 | datum firstkey_data = gdbm_firstkey (gdbmfile);
|
---|
306 | if (firstkey_data.dptr == NULL) return g_EmptyText;
|
---|
307 |
|
---|
308 | // convert it to text_t
|
---|
309 | text_t firstkey;
|
---|
310 | firstkey.setcarr (firstkey_data.dptr, firstkey_data.dsize);
|
---|
311 | free (firstkey_data.dptr);
|
---|
312 | return to_uni(firstkey); // convert to unicode
|
---|
313 | }
|
---|
314 |
|
---|
315 |
|
---|
316 | text_t gdbmclass::getnextkey (const text_t &key)
|
---|
317 | {
|
---|
318 | if (gdbmfile == NULL || key.empty()) return g_EmptyText;
|
---|
319 |
|
---|
320 | // get a utf-8 encoded c string of the unicode key
|
---|
321 | datum key_data;
|
---|
322 | key_data.dptr = (to_utf8(key)).getcstr();
|
---|
323 | if (key_data.dptr == NULL) return g_EmptyText;
|
---|
324 | key_data.dsize = strlen (key_data.dptr);
|
---|
325 |
|
---|
326 | // get the next key
|
---|
327 | datum nextkey_data = gdbm_nextkey (gdbmfile, key_data);
|
---|
328 | if (nextkey_data.dptr == NULL) {
|
---|
329 | delete []key_data.dptr;
|
---|
330 | return g_EmptyText;
|
---|
331 | }
|
---|
332 |
|
---|
333 | // convert it to text_t
|
---|
334 | text_t nextkey;
|
---|
335 | nextkey.setcarr (nextkey_data.dptr, nextkey_data.dsize);
|
---|
336 | free (nextkey_data.dptr);
|
---|
337 | delete []key_data.dptr;
|
---|
338 | return to_uni(nextkey); // convert to unicode
|
---|
339 | }
|
---|
340 |
|
---|
341 |
|
---|
342 | // returns true on success
|
---|
343 | bool gdbmclass::getinfoline (text_t::iterator &here, text_t::iterator end,
|
---|
344 | text_t &key, text_t &value)
|
---|
345 | {
|
---|
346 | key.clear();
|
---|
347 | value.clear();
|
---|
348 |
|
---|
349 | // ignore white space
|
---|
350 | while (here != end && is_unicode_space (*here)) ++here;
|
---|
351 |
|
---|
352 | // get the '<'
|
---|
353 | if (here == end || *here != '<') return false;
|
---|
354 | ++here;
|
---|
355 |
|
---|
356 | // get the key
|
---|
357 | while (here != end && *here != '>') {
|
---|
358 | key.push_back(*here);
|
---|
359 | ++here;
|
---|
360 | }
|
---|
361 |
|
---|
362 | // get the '>'
|
---|
363 | if (here == end || *here != '>') return false;
|
---|
364 | ++here;
|
---|
365 |
|
---|
366 | // get the value
|
---|
367 | while (here != end && *here != '\n') {
|
---|
368 | if (*here == '\\') {
|
---|
369 | // found escape character
|
---|
370 | ++here;
|
---|
371 | if (here != end) {
|
---|
372 | if (*here == 'n') value.push_back ('\n');
|
---|
373 | else if (*here == 'r') value.push_back ('\r');
|
---|
374 | else value.push_back(*here);
|
---|
375 | }
|
---|
376 |
|
---|
377 | } else {
|
---|
378 | // a normal character
|
---|
379 | value.push_back(*here);
|
---|
380 | }
|
---|
381 |
|
---|
382 | ++here;
|
---|
383 | }
|
---|
384 |
|
---|
385 | return true;
|
---|
386 | }
|
---|