rubysl-sdbm 2.0.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +17 -0
- data/.travis.yml +7 -0
- data/Gemfile +4 -0
- data/LICENSE +25 -0
- data/README.md +29 -0
- data/Rakefile +1 -0
- data/ext/rubysl/sdbm/_sdbm.c +929 -0
- data/ext/rubysl/sdbm/extconf.rb +4 -0
- data/ext/rubysl/sdbm/init.c +713 -0
- data/ext/rubysl/sdbm/sdbm.h +86 -0
- data/lib/rubysl/sdbm.rb +2 -0
- data/lib/rubysl/sdbm/version.rb +5 -0
- data/lib/sdbm.rb +1 -0
- data/rubysl-sdbm.gemspec +23 -0
- metadata +101 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: f749712537e38df1d36f99b235803d31ee3e731e
|
4
|
+
data.tar.gz: c916491362e0edc9004efa8e6f0e52e7e03e4f4e
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 7d703fd20c9b29936ba70cd4b54e8f7e0609554a826c81d35a655b7a0d430e7255f9f6b87074b5f48e78d9b60a983921c3e4e743315240e39b2f9470264704c7
|
7
|
+
data.tar.gz: f7fe2567012ab5e2db635e2c0c6954be738c5b6891c4f6055f72bb4c1c9292f0d855798030cd3533d0e0af906e173c9273d6c4fd97a4e6b3db0843f7f336444b
|
data/.gitignore
ADDED
data/.travis.yml
ADDED
data/Gemfile
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,25 @@
|
|
1
|
+
Copyright (c) 2013, Brian Shirai
|
2
|
+
All rights reserved.
|
3
|
+
|
4
|
+
Redistribution and use in source and binary forms, with or without
|
5
|
+
modification, are permitted provided that the following conditions are met:
|
6
|
+
|
7
|
+
1. Redistributions of source code must retain the above copyright notice, this
|
8
|
+
list of conditions and the following disclaimer.
|
9
|
+
2. Redistributions in binary form must reproduce the above copyright notice,
|
10
|
+
this list of conditions and the following disclaimer in the documentation
|
11
|
+
and/or other materials provided with the distribution.
|
12
|
+
3. Neither the name of the library nor the names of its contributors may be
|
13
|
+
used to endorse or promote products derived from this software without
|
14
|
+
specific prior written permission.
|
15
|
+
|
16
|
+
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
17
|
+
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
18
|
+
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
|
19
|
+
DISCLAIMED. IN NO EVENT SHALL <COPYRIGHT HOLDER> BE LIABLE FOR ANY DIRECT,
|
20
|
+
INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
|
21
|
+
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
22
|
+
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
23
|
+
OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
|
24
|
+
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
|
25
|
+
EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
data/README.md
ADDED
@@ -0,0 +1,29 @@
|
|
1
|
+
# Rubysl::Sdbm
|
2
|
+
|
3
|
+
TODO: Write a gem description
|
4
|
+
|
5
|
+
## Installation
|
6
|
+
|
7
|
+
Add this line to your application's Gemfile:
|
8
|
+
|
9
|
+
gem 'rubysl-sdbm'
|
10
|
+
|
11
|
+
And then execute:
|
12
|
+
|
13
|
+
$ bundle
|
14
|
+
|
15
|
+
Or install it yourself as:
|
16
|
+
|
17
|
+
$ gem install rubysl-sdbm
|
18
|
+
|
19
|
+
## Usage
|
20
|
+
|
21
|
+
TODO: Write usage instructions here
|
22
|
+
|
23
|
+
## Contributing
|
24
|
+
|
25
|
+
1. Fork it
|
26
|
+
2. Create your feature branch (`git checkout -b my-new-feature`)
|
27
|
+
3. Commit your changes (`git commit -am 'Add some feature'`)
|
28
|
+
4. Push to the branch (`git push origin my-new-feature`)
|
29
|
+
5. Create new Pull Request
|
data/Rakefile
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
require "bundler/gem_tasks"
|
@@ -0,0 +1,929 @@
|
|
1
|
+
/*
|
2
|
+
* sdbm - ndbm work-alike hashed database library
|
3
|
+
* based on Per-Aake Larson's Dynamic Hashing algorithms. BIT 18 (1978).
|
4
|
+
* author: oz@nexus.yorku.ca
|
5
|
+
* status: public domain.
|
6
|
+
*
|
7
|
+
* core routines
|
8
|
+
*/
|
9
|
+
|
10
|
+
#ifndef lint
|
11
|
+
/*char sdbm_rcsid[] = "$Id$";*/
|
12
|
+
#endif
|
13
|
+
|
14
|
+
#include "ruby/config.h"
|
15
|
+
#include "ruby/defines.h"
|
16
|
+
|
17
|
+
#ifdef HAVE_UNISTD_H
|
18
|
+
#include <unistd.h>
|
19
|
+
#endif
|
20
|
+
|
21
|
+
#include "sdbm.h"
|
22
|
+
|
23
|
+
/*
|
24
|
+
* sdbm - ndbm work-alike hashed database library
|
25
|
+
* tuning and portability constructs [not nearly enough]
|
26
|
+
* author: oz@nexus.yorku.ca
|
27
|
+
*/
|
28
|
+
|
29
|
+
#define BYTESIZ 8
|
30
|
+
|
31
|
+
#ifdef BSD42
|
32
|
+
#define SEEK_SET L_SET
|
33
|
+
#define memset(s,c,n) bzero((s), (n)) /* only when c is zero */
|
34
|
+
#define memcpy(s1,s2,n) bcopy((s2), (s1), (n))
|
35
|
+
#define memcmp(s1,s2,n) bcmp((s1),(s2),(n))
|
36
|
+
#endif
|
37
|
+
|
38
|
+
/*
|
39
|
+
* important tuning parms (hah)
|
40
|
+
*/
|
41
|
+
|
42
|
+
#ifndef SEEDUPS
|
43
|
+
#define SEEDUPS 1 /* always detect duplicates */
|
44
|
+
#endif
|
45
|
+
#ifndef BADMESS
|
46
|
+
#define BADMESS 1 /* generate a message for worst case:
|
47
|
+
cannot make room after SPLTMAX splits */
|
48
|
+
#endif
|
49
|
+
|
50
|
+
/*
|
51
|
+
* misc
|
52
|
+
*/
|
53
|
+
#ifdef DEBUG
|
54
|
+
#define debug(x) printf x
|
55
|
+
#else
|
56
|
+
#define debug(x)
|
57
|
+
#endif
|
58
|
+
|
59
|
+
#ifdef BIG_E
|
60
|
+
#define GET_SHORT(p, i) (((unsigned)((unsigned char *)(p))[(i)*2] << 8) + (((unsigned char *)(p))[(i)*2 + 1]))
|
61
|
+
#define PUT_SHORT(p, i, s) (((unsigned char *)(p))[(i)*2] = (unsigned char)((s) >> 8), ((unsigned char *)(p))[(i)*2 + 1] = (unsigned char)(s))
|
62
|
+
#else
|
63
|
+
#define GET_SHORT(p, i) ((p)[(i)])
|
64
|
+
#define PUT_SHORT(p, i, s) ((p)[(i)] = (s))
|
65
|
+
#endif
|
66
|
+
|
67
|
+
/*#include "pair.h"*/
|
68
|
+
static int fitpair proto((char *, int));
|
69
|
+
static void putpair proto((char *, datum, datum));
|
70
|
+
static datum getpair proto((char *, datum));
|
71
|
+
static int delpair proto((char *, datum));
|
72
|
+
static int chkpage proto((char *));
|
73
|
+
static datum getnkey proto((char *, int));
|
74
|
+
static void splpage proto((char *, char *, long));
|
75
|
+
#if SEEDUPS
|
76
|
+
static int duppair proto((char *, datum));
|
77
|
+
#endif
|
78
|
+
|
79
|
+
#include <stdio.h>
|
80
|
+
#include <stdlib.h>
|
81
|
+
#ifdef DOSISH
|
82
|
+
#include <io.h>
|
83
|
+
#endif
|
84
|
+
#include <sys/types.h>
|
85
|
+
#include <sys/stat.h>
|
86
|
+
#ifdef BSD42
|
87
|
+
#include <sys/file.h>
|
88
|
+
#else
|
89
|
+
#include <fcntl.h>
|
90
|
+
/*#include <memory.h>*/
|
91
|
+
#endif
|
92
|
+
#ifndef O_BINARY
|
93
|
+
#define O_BINARY 0
|
94
|
+
#endif
|
95
|
+
|
96
|
+
#include <errno.h>
|
97
|
+
#ifndef EPERM
|
98
|
+
#define EPERM EACCES
|
99
|
+
#endif
|
100
|
+
#include <string.h>
|
101
|
+
|
102
|
+
#ifdef __STDC__
|
103
|
+
#include <stddef.h>
|
104
|
+
#endif
|
105
|
+
|
106
|
+
#ifndef NULL
|
107
|
+
#define NULL 0
|
108
|
+
#endif
|
109
|
+
|
110
|
+
/*
|
111
|
+
* externals
|
112
|
+
*/
|
113
|
+
#if !defined sun && !defined _WIN32 && !defined __CYGWIN__ && !defined(errno)
|
114
|
+
extern int errno;
|
115
|
+
#endif
|
116
|
+
|
117
|
+
/*
|
118
|
+
* forward
|
119
|
+
*/
|
120
|
+
static int getdbit proto((DBM *, long));
|
121
|
+
static int setdbit proto((DBM *, long));
|
122
|
+
static int getpage proto((DBM *, long));
|
123
|
+
static datum getnext proto((DBM *));
|
124
|
+
static int makroom proto((DBM *, long, int));
|
125
|
+
|
126
|
+
/*
|
127
|
+
* useful macros
|
128
|
+
*/
|
129
|
+
#define bad(x) ((x).dptr == NULL || (x).dsize < 0)
|
130
|
+
#define exhash(item) sdbm_hash((item).dptr, (item).dsize)
|
131
|
+
#define ioerr(db) ((db)->flags |= DBM_IOERR)
|
132
|
+
|
133
|
+
#define OFF_PAG(off) (long) (off) * PBLKSIZ
|
134
|
+
#define OFF_DIR(off) (long) (off) * DBLKSIZ
|
135
|
+
|
136
|
+
static long masks[] = {
|
137
|
+
000000000000L, 000000000001L, 000000000003L,
|
138
|
+
000000000007L, 000000000017L, 000000000037L,
|
139
|
+
000000000077L, 000000000177L, 000000000377L,
|
140
|
+
000000000777L, 000000001777L, 000000003777L,
|
141
|
+
000000007777L, 000000017777L, 000000037777L,
|
142
|
+
000000077777L, 000000177777L, 000000377777L,
|
143
|
+
000000777777L, 000001777777L, 000003777777L,
|
144
|
+
000007777777L, 000017777777L, 000037777777L,
|
145
|
+
000077777777L, 000177777777L, 000377777777L,
|
146
|
+
000777777777L, 001777777777L, 003777777777L,
|
147
|
+
007777777777L, 017777777777L
|
148
|
+
};
|
149
|
+
|
150
|
+
datum nullitem = {NULL, 0};
|
151
|
+
|
152
|
+
DBM *
|
153
|
+
sdbm_open(register char *file, register int flags, register int mode)
|
154
|
+
{
|
155
|
+
register DBM *db;
|
156
|
+
register char *dirname;
|
157
|
+
register char *pagname;
|
158
|
+
register size_t n;
|
159
|
+
|
160
|
+
if (file == NULL || !*file)
|
161
|
+
return errno = EINVAL, (DBM *) NULL;
|
162
|
+
/*
|
163
|
+
* need space for two seperate filenames
|
164
|
+
*/
|
165
|
+
n = strlen(file) * 2 + strlen(DIRFEXT) + strlen(PAGFEXT) + 2;
|
166
|
+
|
167
|
+
if ((dirname = malloc(n)) == NULL)
|
168
|
+
return errno = ENOMEM, (DBM *) NULL;
|
169
|
+
/*
|
170
|
+
* build the file names
|
171
|
+
*/
|
172
|
+
dirname = strcat(strcpy(dirname, file), DIRFEXT);
|
173
|
+
pagname = strcpy(dirname + strlen(dirname) + 1, file);
|
174
|
+
pagname = strcat(pagname, PAGFEXT);
|
175
|
+
|
176
|
+
db = sdbm_prep(dirname, pagname, flags, mode);
|
177
|
+
free((char *) dirname);
|
178
|
+
return db;
|
179
|
+
}
|
180
|
+
|
181
|
+
DBM *
|
182
|
+
sdbm_prep(char *dirname, char *pagname, int flags, int mode)
|
183
|
+
{
|
184
|
+
register DBM *db;
|
185
|
+
struct stat dstat;
|
186
|
+
|
187
|
+
if ((db = (DBM *) malloc(sizeof(DBM))) == NULL)
|
188
|
+
return errno = ENOMEM, (DBM *) NULL;
|
189
|
+
|
190
|
+
db->flags = 0;
|
191
|
+
db->hmask = 0;
|
192
|
+
db->blkptr = 0;
|
193
|
+
db->keyptr = 0;
|
194
|
+
/*
|
195
|
+
* adjust user flags so that WRONLY becomes RDWR,
|
196
|
+
* as required by this package. Also set our internal
|
197
|
+
* flag for RDONLY.
|
198
|
+
*/
|
199
|
+
if (flags & O_WRONLY)
|
200
|
+
flags = (flags & ~O_WRONLY) | O_RDWR;
|
201
|
+
if (flags & O_RDONLY)
|
202
|
+
db->flags = DBM_RDONLY;
|
203
|
+
/*
|
204
|
+
* open the files in sequence, and stat the dirfile.
|
205
|
+
* If we fail anywhere, undo everything, return NULL.
|
206
|
+
*/
|
207
|
+
flags |= O_BINARY;
|
208
|
+
if ((db->pagf = open(pagname, flags, mode)) > -1) {
|
209
|
+
if ((db->dirf = open(dirname, flags, mode)) > -1) {
|
210
|
+
/*
|
211
|
+
* need the dirfile size to establish max bit number.
|
212
|
+
*/
|
213
|
+
if (fstat(db->dirf, &dstat) == 0) {
|
214
|
+
/*
|
215
|
+
* zero size: either a fresh database, or one with a single,
|
216
|
+
* unsplit data page: dirpage is all zeros.
|
217
|
+
*/
|
218
|
+
db->dirbno = (!dstat.st_size) ? 0 : -1;
|
219
|
+
db->pagbno = -1;
|
220
|
+
db->maxbno = dstat.st_size * (long) BYTESIZ;
|
221
|
+
|
222
|
+
(void) memset(db->pagbuf, 0, PBLKSIZ);
|
223
|
+
(void) memset(db->dirbuf, 0, DBLKSIZ);
|
224
|
+
/*
|
225
|
+
* success
|
226
|
+
*/
|
227
|
+
return db;
|
228
|
+
}
|
229
|
+
(void) close(db->dirf);
|
230
|
+
}
|
231
|
+
(void) close(db->pagf);
|
232
|
+
}
|
233
|
+
free((char *) db);
|
234
|
+
return (DBM *) NULL;
|
235
|
+
}
|
236
|
+
|
237
|
+
void
|
238
|
+
sdbm_close(register DBM *db)
|
239
|
+
{
|
240
|
+
if (db == NULL)
|
241
|
+
errno = EINVAL;
|
242
|
+
else {
|
243
|
+
(void) close(db->dirf);
|
244
|
+
(void) close(db->pagf);
|
245
|
+
free((char *) db);
|
246
|
+
}
|
247
|
+
}
|
248
|
+
|
249
|
+
datum
|
250
|
+
sdbm_fetch(register DBM *db, datum key)
|
251
|
+
{
|
252
|
+
if (db == NULL || bad(key))
|
253
|
+
return errno = EINVAL, nullitem;
|
254
|
+
|
255
|
+
if (getpage(db, exhash(key)))
|
256
|
+
return getpair(db->pagbuf, key);
|
257
|
+
|
258
|
+
return ioerr(db), nullitem;
|
259
|
+
}
|
260
|
+
|
261
|
+
int
|
262
|
+
sdbm_delete(register DBM *db, datum key)
|
263
|
+
{
|
264
|
+
if (db == NULL || bad(key))
|
265
|
+
return errno = EINVAL, -1;
|
266
|
+
if (sdbm_rdonly(db))
|
267
|
+
return errno = EPERM, -1;
|
268
|
+
|
269
|
+
if (getpage(db, exhash(key))) {
|
270
|
+
if (!delpair(db->pagbuf, key))
|
271
|
+
return -1;
|
272
|
+
/*
|
273
|
+
* update the page file
|
274
|
+
*/
|
275
|
+
if (lseek(db->pagf, OFF_PAG(db->pagbno), SEEK_SET) < 0
|
276
|
+
|| write(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
277
|
+
return ioerr(db), -1;
|
278
|
+
|
279
|
+
return 0;
|
280
|
+
}
|
281
|
+
|
282
|
+
return ioerr(db), -1;
|
283
|
+
}
|
284
|
+
|
285
|
+
int
|
286
|
+
sdbm_store(register DBM *db, datum key, datum val, int flags)
|
287
|
+
{
|
288
|
+
int need;
|
289
|
+
register long hash;
|
290
|
+
|
291
|
+
if (db == NULL || bad(key))
|
292
|
+
return errno = EINVAL, -1;
|
293
|
+
if (sdbm_rdonly(db))
|
294
|
+
return errno = EPERM, -1;
|
295
|
+
|
296
|
+
need = key.dsize + val.dsize;
|
297
|
+
/*
|
298
|
+
* is the pair too big (or too small) for this database ??
|
299
|
+
*/
|
300
|
+
if (need < 0 || need > PAIRMAX)
|
301
|
+
return errno = EINVAL, -1;
|
302
|
+
|
303
|
+
if (getpage(db, (hash = exhash(key)))) {
|
304
|
+
/*
|
305
|
+
* if we need to replace, delete the key/data pair
|
306
|
+
* first. If it is not there, ignore.
|
307
|
+
*/
|
308
|
+
if (flags == DBM_REPLACE)
|
309
|
+
(void) delpair(db->pagbuf, key);
|
310
|
+
#if SEEDUPS
|
311
|
+
else if (duppair(db->pagbuf, key))
|
312
|
+
return 1;
|
313
|
+
#endif
|
314
|
+
/*
|
315
|
+
* if we do not have enough room, we have to split.
|
316
|
+
*/
|
317
|
+
if (!fitpair(db->pagbuf, need))
|
318
|
+
if (!makroom(db, hash, need))
|
319
|
+
return ioerr(db), -1;
|
320
|
+
/*
|
321
|
+
* we have enough room or split is successful. insert the key,
|
322
|
+
* and update the page file.
|
323
|
+
*/
|
324
|
+
(void) putpair(db->pagbuf, key, val);
|
325
|
+
|
326
|
+
if (lseek(db->pagf, OFF_PAG(db->pagbno), SEEK_SET) < 0
|
327
|
+
|| write(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
328
|
+
return ioerr(db), -1;
|
329
|
+
/*
|
330
|
+
* success
|
331
|
+
*/
|
332
|
+
return 0;
|
333
|
+
}
|
334
|
+
|
335
|
+
return ioerr(db), -1;
|
336
|
+
}
|
337
|
+
|
338
|
+
/*
|
339
|
+
* makroom - make room by splitting the overfull page
|
340
|
+
* this routine will attempt to make room for SPLTMAX times before
|
341
|
+
* giving up.
|
342
|
+
*/
|
343
|
+
static int
|
344
|
+
makroom(register DBM *db, long int hash, int need)
|
345
|
+
{
|
346
|
+
long newp;
|
347
|
+
char twin[PBLKSIZ];
|
348
|
+
#if defined _WIN32 && !defined __CYGWIN__
|
349
|
+
char zer[PBLKSIZ];
|
350
|
+
long oldtail;
|
351
|
+
#endif
|
352
|
+
char *pag = db->pagbuf;
|
353
|
+
char *new = twin;
|
354
|
+
register int smax = SPLTMAX;
|
355
|
+
|
356
|
+
do {
|
357
|
+
/*
|
358
|
+
* split the current page
|
359
|
+
*/
|
360
|
+
(void) splpage(pag, new, db->hmask + 1);
|
361
|
+
/*
|
362
|
+
* address of the new page
|
363
|
+
*/
|
364
|
+
newp = (hash & db->hmask) | (db->hmask + 1);
|
365
|
+
debug(("newp: %ld\n", newp));
|
366
|
+
/*
|
367
|
+
* write delay, read avoidence/cache shuffle:
|
368
|
+
* select the page for incoming pair: if key is to go to the new page,
|
369
|
+
* write out the previous one, and copy the new one over, thus making
|
370
|
+
* it the current page. If not, simply write the new page, and we are
|
371
|
+
* still looking at the page of interest. current page is not updated
|
372
|
+
* here, as sdbm_store will do so, after it inserts the incoming pair.
|
373
|
+
*/
|
374
|
+
|
375
|
+
#if defined _WIN32 && !defined __CYGWIN__
|
376
|
+
/*
|
377
|
+
* Fill hole with 0 if made it.
|
378
|
+
* (hole is NOT read as 0)
|
379
|
+
*/
|
380
|
+
oldtail = lseek(db->pagf, 0L, SEEK_END);
|
381
|
+
memset(zer, 0, PBLKSIZ);
|
382
|
+
while (OFF_PAG(newp) > oldtail) {
|
383
|
+
if (lseek(db->pagf, 0L, SEEK_END) < 0 ||
|
384
|
+
write(db->pagf, zer, PBLKSIZ) < 0) {
|
385
|
+
|
386
|
+
return 0;
|
387
|
+
}
|
388
|
+
oldtail += PBLKSIZ;
|
389
|
+
}
|
390
|
+
#endif
|
391
|
+
|
392
|
+
if (hash & (db->hmask + 1)) {
|
393
|
+
if (lseek(db->pagf, OFF_PAG(db->pagbno), SEEK_SET) < 0
|
394
|
+
|| write(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
395
|
+
return 0;
|
396
|
+
db->pagbno = newp;
|
397
|
+
(void) memcpy(pag, new, PBLKSIZ);
|
398
|
+
}
|
399
|
+
else if (lseek(db->pagf, OFF_PAG(newp), SEEK_SET) < 0
|
400
|
+
|| write(db->pagf, new, PBLKSIZ) < 0)
|
401
|
+
return 0;
|
402
|
+
|
403
|
+
if (!setdbit(db, db->curbit))
|
404
|
+
return 0;
|
405
|
+
/*
|
406
|
+
* see if we have enough room now
|
407
|
+
*/
|
408
|
+
if (fitpair(pag, need))
|
409
|
+
return 1;
|
410
|
+
/*
|
411
|
+
* try again... update curbit and hmask as getpage would have
|
412
|
+
* done. because of our update of the current page, we do not
|
413
|
+
* need to read in anything. BUT we have to write the current
|
414
|
+
* [deferred] page out, as the window of failure is too great.
|
415
|
+
*/
|
416
|
+
db->curbit = 2 * db->curbit +
|
417
|
+
((hash & (db->hmask + 1)) ? 2 : 1);
|
418
|
+
db->hmask |= (db->hmask + 1);
|
419
|
+
|
420
|
+
if (lseek(db->pagf, OFF_PAG(db->pagbno), SEEK_SET) < 0
|
421
|
+
|| write(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
422
|
+
return 0;
|
423
|
+
|
424
|
+
} while (--smax);
|
425
|
+
/*
|
426
|
+
* if we are here, this is real bad news. After SPLTMAX splits,
|
427
|
+
* we still cannot fit the key. say goodnight.
|
428
|
+
*/
|
429
|
+
#if BADMESS
|
430
|
+
(void) (write(2, "sdbm: cannot insert after SPLTMAX attempts.\n", 44) < 0);
|
431
|
+
#endif
|
432
|
+
return 0;
|
433
|
+
|
434
|
+
}
|
435
|
+
|
436
|
+
/*
|
437
|
+
* the following two routines will break if
|
438
|
+
* deletions aren't taken into account. (ndbm bug)
|
439
|
+
*/
|
440
|
+
datum
|
441
|
+
sdbm_firstkey(register DBM *db)
|
442
|
+
{
|
443
|
+
if (db == NULL)
|
444
|
+
return errno = EINVAL, nullitem;
|
445
|
+
/*
|
446
|
+
* start at page 0
|
447
|
+
*/
|
448
|
+
(void) memset(db->pagbuf, 0, PBLKSIZ);
|
449
|
+
if (lseek(db->pagf, OFF_PAG(0), SEEK_SET) < 0
|
450
|
+
|| read(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
451
|
+
return ioerr(db), nullitem;
|
452
|
+
db->pagbno = 0;
|
453
|
+
db->blkptr = 0;
|
454
|
+
db->keyptr = 0;
|
455
|
+
|
456
|
+
return getnext(db);
|
457
|
+
}
|
458
|
+
|
459
|
+
datum
|
460
|
+
sdbm_nextkey(register DBM *db)
|
461
|
+
{
|
462
|
+
if (db == NULL)
|
463
|
+
return errno = EINVAL, nullitem;
|
464
|
+
return getnext(db);
|
465
|
+
}
|
466
|
+
|
467
|
+
/*
|
468
|
+
* all important binary trie traversal
|
469
|
+
*/
|
470
|
+
static int
|
471
|
+
getpage(register DBM *db, register long int hash)
|
472
|
+
{
|
473
|
+
register int hbit;
|
474
|
+
register long dbit;
|
475
|
+
register long pagb;
|
476
|
+
|
477
|
+
dbit = 0;
|
478
|
+
hbit = 0;
|
479
|
+
while (dbit < db->maxbno && getdbit(db, dbit))
|
480
|
+
dbit = 2 * dbit + ((hash & ((long) 1 << hbit++)) ? 2 : 1);
|
481
|
+
|
482
|
+
debug(("dbit: %d...", dbit));
|
483
|
+
|
484
|
+
db->curbit = dbit;
|
485
|
+
db->hmask = masks[hbit];
|
486
|
+
|
487
|
+
pagb = hash & db->hmask;
|
488
|
+
/*
|
489
|
+
* see if the block we need is already in memory.
|
490
|
+
* note: this lookaside cache has about 10% hit rate.
|
491
|
+
*/
|
492
|
+
if (pagb != db->pagbno) {
|
493
|
+
/*
|
494
|
+
* note: here, we assume a "hole" is read as 0s.
|
495
|
+
* if not, must zero pagbuf first.
|
496
|
+
*/
|
497
|
+
(void) memset(db->pagbuf, 0, PBLKSIZ);
|
498
|
+
|
499
|
+
if (lseek(db->pagf, OFF_PAG(pagb), SEEK_SET) < 0
|
500
|
+
|| read(db->pagf, db->pagbuf, PBLKSIZ) < 0)
|
501
|
+
return 0;
|
502
|
+
if (!chkpage(db->pagbuf)) {
|
503
|
+
return 0;
|
504
|
+
}
|
505
|
+
db->pagbno = pagb;
|
506
|
+
|
507
|
+
debug(("pag read: %d\n", pagb));
|
508
|
+
}
|
509
|
+
return 1;
|
510
|
+
}
|
511
|
+
|
512
|
+
static int
|
513
|
+
getdbit(register DBM *db, register long int dbit)
|
514
|
+
{
|
515
|
+
register long c;
|
516
|
+
register long dirb;
|
517
|
+
|
518
|
+
c = dbit / BYTESIZ;
|
519
|
+
dirb = c / DBLKSIZ;
|
520
|
+
|
521
|
+
if (dirb != db->dirbno) {
|
522
|
+
if (lseek(db->dirf, OFF_DIR(dirb), SEEK_SET) < 0
|
523
|
+
|| read(db->dirf, db->dirbuf, DBLKSIZ) < 0)
|
524
|
+
return 0;
|
525
|
+
db->dirbno = dirb;
|
526
|
+
|
527
|
+
debug(("dir read: %d\n", dirb));
|
528
|
+
}
|
529
|
+
|
530
|
+
return db->dirbuf[c % DBLKSIZ] & (1 << (dbit % BYTESIZ));
|
531
|
+
}
|
532
|
+
|
533
|
+
static int
|
534
|
+
setdbit(register DBM *db, register long int dbit)
|
535
|
+
{
|
536
|
+
register long c;
|
537
|
+
register long dirb;
|
538
|
+
|
539
|
+
c = dbit / BYTESIZ;
|
540
|
+
dirb = c / DBLKSIZ;
|
541
|
+
|
542
|
+
if (dirb != db->dirbno) {
|
543
|
+
if (lseek(db->dirf, OFF_DIR(dirb), SEEK_SET) < 0
|
544
|
+
|| read(db->dirf, db->dirbuf, DBLKSIZ) < 0)
|
545
|
+
return 0;
|
546
|
+
db->dirbno = dirb;
|
547
|
+
|
548
|
+
debug(("dir read: %d\n", dirb));
|
549
|
+
}
|
550
|
+
|
551
|
+
db->dirbuf[c % DBLKSIZ] |= (1 << (dbit % BYTESIZ));
|
552
|
+
|
553
|
+
if (dbit >= db->maxbno)
|
554
|
+
db->maxbno += (long) DBLKSIZ * BYTESIZ;
|
555
|
+
|
556
|
+
if (lseek(db->dirf, OFF_DIR(dirb), SEEK_SET) < 0
|
557
|
+
|| write(db->dirf, db->dirbuf, DBLKSIZ) < 0)
|
558
|
+
return 0;
|
559
|
+
|
560
|
+
return 1;
|
561
|
+
}
|
562
|
+
|
563
|
+
/*
|
564
|
+
* getnext - get the next key in the page, and if done with
|
565
|
+
* the page, try the next page in sequence
|
566
|
+
*/
|
567
|
+
static datum
|
568
|
+
getnext(register DBM *db)
|
569
|
+
{
|
570
|
+
datum key;
|
571
|
+
|
572
|
+
for (;;) {
|
573
|
+
db->keyptr++;
|
574
|
+
key = getnkey(db->pagbuf, db->keyptr);
|
575
|
+
if (key.dptr != NULL)
|
576
|
+
return key;
|
577
|
+
/*
|
578
|
+
* we either run out, or there is nothing on this page..
|
579
|
+
* try the next one... If we lost our position on the
|
580
|
+
* file, we will have to seek.
|
581
|
+
*/
|
582
|
+
db->keyptr = 0;
|
583
|
+
if (db->pagbno != db->blkptr++)
|
584
|
+
if (lseek(db->pagf, OFF_PAG(db->blkptr), SEEK_SET) < 0)
|
585
|
+
break;
|
586
|
+
db->pagbno = db->blkptr;
|
587
|
+
if (read(db->pagf, db->pagbuf, PBLKSIZ) <= 0)
|
588
|
+
break;
|
589
|
+
if (!chkpage(db->pagbuf)) {
|
590
|
+
break;
|
591
|
+
}
|
592
|
+
}
|
593
|
+
|
594
|
+
return ioerr(db), nullitem;
|
595
|
+
}
|
596
|
+
|
597
|
+
/* pair.c */
|
598
|
+
/*
|
599
|
+
* sdbm - ndbm work-alike hashed database library
|
600
|
+
* based on Per-Aake Larson's Dynamic Hashing algorithms. BIT 18 (1978).
|
601
|
+
* author: oz@nexus.yorku.ca
|
602
|
+
* status: public domain.
|
603
|
+
*
|
604
|
+
* page-level routines
|
605
|
+
*/
|
606
|
+
|
607
|
+
#ifndef lint
|
608
|
+
/*char pair_rcsid[] = "$Id$";*/
|
609
|
+
#endif
|
610
|
+
|
611
|
+
#ifndef BSD42
|
612
|
+
/*#include <memory.h>*/
|
613
|
+
#endif
|
614
|
+
|
615
|
+
#define exhash(item) sdbm_hash((item).dptr, (item).dsize)
|
616
|
+
|
617
|
+
/*
|
618
|
+
* forward
|
619
|
+
*/
|
620
|
+
static int seepair proto((char *, int, char *, int));
|
621
|
+
|
622
|
+
/*
|
623
|
+
* page format:
|
624
|
+
* +------------------------------+
|
625
|
+
* ino | n | keyoff | datoff | keyoff |
|
626
|
+
* +------------+--------+--------+
|
627
|
+
* | datoff | - - - ----> |
|
628
|
+
* +--------+---------------------+
|
629
|
+
* | F R E E A R E A |
|
630
|
+
* +--------------+---------------+
|
631
|
+
* | <---- - - - | data |
|
632
|
+
* +--------+-----+----+----------+
|
633
|
+
* | key | data | key |
|
634
|
+
* +--------+----------+----------+
|
635
|
+
*
|
636
|
+
* calculating the offsets for free area: if the number
|
637
|
+
* of entries (ino[0]) is zero, the offset to the END of
|
638
|
+
* the free area is the block size. Otherwise, it is the
|
639
|
+
* nth (ino[ino[0]]) entry's offset.
|
640
|
+
*/
|
641
|
+
|
642
|
+
static int
|
643
|
+
fitpair(char *pag, int need)
|
644
|
+
{
|
645
|
+
register int n;
|
646
|
+
register int off;
|
647
|
+
register int free;
|
648
|
+
register short *ino = (short *) pag;
|
649
|
+
|
650
|
+
off = ((n = GET_SHORT(ino,0)) > 0) ? GET_SHORT(ino,n) : PBLKSIZ;
|
651
|
+
free = off - (n + 1) * sizeof(short);
|
652
|
+
need += 2 * sizeof(short);
|
653
|
+
|
654
|
+
debug(("free %d need %d\n", free, need));
|
655
|
+
|
656
|
+
return need <= free;
|
657
|
+
}
|
658
|
+
|
659
|
+
static void
|
660
|
+
putpair(char *pag, datum key, datum val)
|
661
|
+
{
|
662
|
+
register int n;
|
663
|
+
register int off;
|
664
|
+
register short *ino = (short *) pag;
|
665
|
+
|
666
|
+
off = ((n = GET_SHORT(ino,0)) > 0) ? GET_SHORT(ino,n) : PBLKSIZ;
|
667
|
+
/*
|
668
|
+
* enter the key first
|
669
|
+
*/
|
670
|
+
off -= key.dsize;
|
671
|
+
if (key.dsize)
|
672
|
+
(void) memcpy(pag + off, key.dptr, key.dsize);
|
673
|
+
PUT_SHORT(ino,n + 1,off);
|
674
|
+
/*
|
675
|
+
* now the data
|
676
|
+
*/
|
677
|
+
off -= val.dsize;
|
678
|
+
if (val.dsize)
|
679
|
+
(void) memcpy(pag + off, val.dptr, val.dsize);
|
680
|
+
PUT_SHORT(ino,n + 2,off);
|
681
|
+
/*
|
682
|
+
* adjust item count
|
683
|
+
*/
|
684
|
+
PUT_SHORT(ino,0,GET_SHORT(ino,0) + 2);
|
685
|
+
}
|
686
|
+
|
687
|
+
static datum
|
688
|
+
getpair(char *pag, datum key)
|
689
|
+
{
|
690
|
+
register int i;
|
691
|
+
register int n;
|
692
|
+
datum val;
|
693
|
+
register short *ino = (short *) pag;
|
694
|
+
|
695
|
+
if ((n = GET_SHORT(ino,0)) == 0)
|
696
|
+
return nullitem;
|
697
|
+
|
698
|
+
if ((i = seepair(pag, n, key.dptr, key.dsize)) == 0)
|
699
|
+
return nullitem;
|
700
|
+
|
701
|
+
val.dptr = pag + GET_SHORT(ino,i + 1);
|
702
|
+
val.dsize = GET_SHORT(ino,i) - GET_SHORT(ino,i + 1);
|
703
|
+
return val;
|
704
|
+
}
|
705
|
+
|
706
|
+
#if SEEDUPS
|
707
|
+
static int
|
708
|
+
duppair(char *pag, datum key)
|
709
|
+
{
|
710
|
+
register short *ino = (short *) pag;
|
711
|
+
return GET_SHORT(ino,0) > 0 &&
|
712
|
+
seepair(pag, GET_SHORT(ino,0), key.dptr, key.dsize) > 0;
|
713
|
+
}
|
714
|
+
#endif
|
715
|
+
|
716
|
+
static datum
|
717
|
+
getnkey(char *pag, int num)
|
718
|
+
{
|
719
|
+
datum key;
|
720
|
+
register int off;
|
721
|
+
register short *ino = (short *) pag;
|
722
|
+
|
723
|
+
num = num * 2 - 1;
|
724
|
+
if (GET_SHORT(ino,0) == 0 || num > GET_SHORT(ino,0))
|
725
|
+
return nullitem;
|
726
|
+
|
727
|
+
off = (num > 1) ? GET_SHORT(ino,num - 1) : PBLKSIZ;
|
728
|
+
|
729
|
+
key.dptr = pag + GET_SHORT(ino,num);
|
730
|
+
key.dsize = off - GET_SHORT(ino,num);
|
731
|
+
|
732
|
+
return key;
|
733
|
+
}
|
734
|
+
|
735
|
+
static int
|
736
|
+
delpair(char *pag, datum key)
|
737
|
+
{
|
738
|
+
register int n;
|
739
|
+
register int i;
|
740
|
+
register short *ino = (short *) pag;
|
741
|
+
|
742
|
+
if ((n = GET_SHORT(ino,0)) == 0)
|
743
|
+
return 0;
|
744
|
+
|
745
|
+
if ((i = seepair(pag, n, key.dptr, key.dsize)) == 0)
|
746
|
+
return 0;
|
747
|
+
/*
|
748
|
+
* found the key. if it is the last entry
|
749
|
+
* [i.e. i == n - 1] we just adjust the entry count.
|
750
|
+
* hard case: move all data down onto the deleted pair,
|
751
|
+
* shift offsets onto deleted offsets, and adjust them.
|
752
|
+
* [note: 0 < i < n]
|
753
|
+
*/
|
754
|
+
if (i < n - 1) {
|
755
|
+
register int m;
|
756
|
+
register char *dst = pag + (i == 1 ? PBLKSIZ : GET_SHORT(ino,i - 1));
|
757
|
+
register char *src = pag + GET_SHORT(ino,i + 1);
|
758
|
+
register ptrdiff_t zoo = dst - src;
|
759
|
+
|
760
|
+
debug(("free-up %"PRIdPTRDIFF" ", zoo));
|
761
|
+
/*
|
762
|
+
* shift data/keys down
|
763
|
+
*/
|
764
|
+
m = GET_SHORT(ino,i + 1) - GET_SHORT(ino,n);
|
765
|
+
#ifdef DUFF
|
766
|
+
#define MOVB *--dst = *--src
|
767
|
+
|
768
|
+
if (m > 0) {
|
769
|
+
register int loop = (m + 8 - 1) >> 3;
|
770
|
+
|
771
|
+
switch (m & (8 - 1)) {
|
772
|
+
case 0: do {
|
773
|
+
MOVB; case 7: MOVB;
|
774
|
+
case 6: MOVB; case 5: MOVB;
|
775
|
+
case 4: MOVB; case 3: MOVB;
|
776
|
+
case 2: MOVB; case 1: MOVB;
|
777
|
+
} while (--loop);
|
778
|
+
}
|
779
|
+
}
|
780
|
+
#else
|
781
|
+
#ifdef MEMMOVE
|
782
|
+
memmove(dst, src, m);
|
783
|
+
#else
|
784
|
+
while (m--)
|
785
|
+
*--dst = *--src;
|
786
|
+
#endif
|
787
|
+
#endif
|
788
|
+
/*
|
789
|
+
* adjust offset index up
|
790
|
+
*/
|
791
|
+
while (i < n - 1) {
|
792
|
+
PUT_SHORT(ino,i, GET_SHORT(ino,i + 2) + zoo);
|
793
|
+
i++;
|
794
|
+
}
|
795
|
+
}
|
796
|
+
PUT_SHORT(ino, 0, GET_SHORT(ino, 0) - 2);
|
797
|
+
return 1;
|
798
|
+
}
|
799
|
+
|
800
|
+
/*
|
801
|
+
* search for the key in the page.
|
802
|
+
* return offset index in the range 0 < i < n.
|
803
|
+
* return 0 if not found.
|
804
|
+
*/
|
805
|
+
static int
|
806
|
+
seepair(char *pag, register int n, register char *key, register int siz)
|
807
|
+
{
|
808
|
+
register int i;
|
809
|
+
register int off = PBLKSIZ;
|
810
|
+
register short *ino = (short *) pag;
|
811
|
+
|
812
|
+
for (i = 1; i < n; i += 2) {
|
813
|
+
if (siz == off - GET_SHORT(ino,i) &&
|
814
|
+
memcmp(key, pag + GET_SHORT(ino,i), siz) == 0)
|
815
|
+
return i;
|
816
|
+
off = GET_SHORT(ino,i + 1);
|
817
|
+
}
|
818
|
+
return 0;
|
819
|
+
}
|
820
|
+
|
821
|
+
static void
|
822
|
+
splpage(char *pag, char *new, long int sbit)
|
823
|
+
{
|
824
|
+
datum key;
|
825
|
+
datum val;
|
826
|
+
|
827
|
+
register int n;
|
828
|
+
register int off = PBLKSIZ;
|
829
|
+
char cur[PBLKSIZ];
|
830
|
+
register short *ino = (short *) cur;
|
831
|
+
|
832
|
+
(void) memcpy(cur, pag, PBLKSIZ);
|
833
|
+
(void) memset(pag, 0, PBLKSIZ);
|
834
|
+
(void) memset(new, 0, PBLKSIZ);
|
835
|
+
|
836
|
+
n = GET_SHORT(ino,0);
|
837
|
+
for (ino++; n > 0; ino += 2) {
|
838
|
+
key.dptr = cur + GET_SHORT(ino,0);
|
839
|
+
key.dsize = off - GET_SHORT(ino,0);
|
840
|
+
val.dptr = cur + GET_SHORT(ino,1);
|
841
|
+
val.dsize = GET_SHORT(ino,0) - GET_SHORT(ino,1);
|
842
|
+
/*
|
843
|
+
* select the page pointer (by looking at sbit) and insert
|
844
|
+
*/
|
845
|
+
(void) putpair((exhash(key) & sbit) ? new : pag, key, val);
|
846
|
+
|
847
|
+
off = GET_SHORT(ino,1);
|
848
|
+
n -= 2;
|
849
|
+
}
|
850
|
+
|
851
|
+
debug(("%d split %d/%d\n", ((short *) cur)[0] / 2,
|
852
|
+
((short *) new)[0] / 2,
|
853
|
+
((short *) pag)[0] / 2));
|
854
|
+
}
|
855
|
+
|
856
|
+
/*
|
857
|
+
* check page sanity:
|
858
|
+
* number of entries should be something
|
859
|
+
* reasonable, and all offsets in the index should be in order.
|
860
|
+
* this could be made more rigorous.
|
861
|
+
*/
|
862
|
+
static int
|
863
|
+
chkpage(char *pag)
|
864
|
+
{
|
865
|
+
register int n;
|
866
|
+
register int off;
|
867
|
+
register short *ino = (short *) pag;
|
868
|
+
|
869
|
+
if ((n = GET_SHORT(ino,0)) < 0 || n > PBLKSIZ / (int)sizeof(short))
|
870
|
+
return 0;
|
871
|
+
|
872
|
+
if (n > 0) {
|
873
|
+
off = PBLKSIZ;
|
874
|
+
for (ino++; n > 0; ino += 2) {
|
875
|
+
if (GET_SHORT(ino,0) > off || GET_SHORT(ino,1) > off ||
|
876
|
+
GET_SHORT(ino,1) > GET_SHORT(ino,0))
|
877
|
+
return 0;
|
878
|
+
off = GET_SHORT(ino,1);
|
879
|
+
n -= 2;
|
880
|
+
}
|
881
|
+
}
|
882
|
+
return 1;
|
883
|
+
}
|
884
|
+
|
885
|
+
/* hash.c */
|
886
|
+
/*
|
887
|
+
* sdbm - ndbm work-alike hashed database library
|
888
|
+
* based on Per-Aake Larson's Dynamic Hashing algorithms. BIT 18 (1978).
|
889
|
+
* author: oz@nexus.yorku.ca
|
890
|
+
* status: public domain. keep it that way.
|
891
|
+
*
|
892
|
+
* hashing routine
|
893
|
+
*/
|
894
|
+
|
895
|
+
/*
|
896
|
+
* polynomial conversion ignoring overflows
|
897
|
+
* [this seems to work remarkably well, in fact better
|
898
|
+
* then the ndbm hash function. Replace at your own risk]
|
899
|
+
* use: 65599 nice.
|
900
|
+
* 65587 even better.
|
901
|
+
*/
|
902
|
+
long
|
903
|
+
sdbm_hash(register char *str, register int len)
|
904
|
+
{
|
905
|
+
register unsigned long n = 0;
|
906
|
+
|
907
|
+
#ifdef DUFF
|
908
|
+
|
909
|
+
#define HASHC n = *str++ + 65599 * n
|
910
|
+
|
911
|
+
if (len > 0) {
|
912
|
+
register int loop = (len + 8 - 1) >> 3;
|
913
|
+
|
914
|
+
switch(len & (8 - 1)) {
|
915
|
+
case 0: do {
|
916
|
+
HASHC; case 7: HASHC;
|
917
|
+
case 6: HASHC; case 5: HASHC;
|
918
|
+
case 4: HASHC; case 3: HASHC;
|
919
|
+
case 2: HASHC; case 1: HASHC;
|
920
|
+
} while (--loop);
|
921
|
+
}
|
922
|
+
|
923
|
+
}
|
924
|
+
#else
|
925
|
+
while (len--)
|
926
|
+
n = ((*str++) & 255) + 65587L * n;
|
927
|
+
#endif
|
928
|
+
return n;
|
929
|
+
}
|