Commit | Line | Data |
a0d0e21e |
1 | # DB_File.pm -- Perl 5 interface to Berkeley DB |
2 | # |
6ca2e664 |
3 | # written by Paul Marquess (Paul.Marquess@btinternet.com) |
cad2e5aa |
4 | # last modified 6th June 1999 |
5 | # version 1.67 |
36477c24 |
6 | # |
cad2e5aa |
7 | # Copyright (c) 1995-1999 Paul Marquess. All rights reserved. |
36477c24 |
8 | # This program is free software; you can redistribute it and/or |
9 | # modify it under the same terms as Perl itself. |
10 | |
8e07c86e |
11 | |
12 | package DB_File::HASHINFO ; |
785da04d |
13 | |
610ab055 |
14 | require 5.003 ; |
15 | |
785da04d |
16 | use strict; |
8e07c86e |
17 | use Carp; |
88108326 |
18 | require Tie::Hash; |
19 | @DB_File::HASHINFO::ISA = qw(Tie::Hash); |
8e07c86e |
20 | |
88108326 |
21 | sub new |
8e07c86e |
22 | { |
88108326 |
23 | my $pkg = shift ; |
24 | my %x ; |
25 | tie %x, $pkg ; |
26 | bless \%x, $pkg ; |
8e07c86e |
27 | } |
28 | |
610ab055 |
29 | |
88108326 |
30 | sub TIEHASH |
31 | { |
32 | my $pkg = shift ; |
33 | |
36477c24 |
34 | bless { VALID => { map {$_, 1} |
35 | qw( bsize ffactor nelem cachesize hash lorder) |
36 | }, |
37 | GOT => {} |
38 | }, $pkg ; |
88108326 |
39 | } |
8e07c86e |
40 | |
610ab055 |
41 | |
8e07c86e |
42 | sub FETCH |
43 | { |
88108326 |
44 | my $self = shift ; |
45 | my $key = shift ; |
8e07c86e |
46 | |
36477c24 |
47 | return $self->{GOT}{$key} if exists $self->{VALID}{$key} ; |
88108326 |
48 | |
49 | my $pkg = ref $self ; |
50 | croak "${pkg}::FETCH - Unknown element '$key'" ; |
8e07c86e |
51 | } |
52 | |
53 | |
54 | sub STORE |
55 | { |
88108326 |
56 | my $self = shift ; |
57 | my $key = shift ; |
58 | my $value = shift ; |
59 | |
36477c24 |
60 | if ( exists $self->{VALID}{$key} ) |
8e07c86e |
61 | { |
36477c24 |
62 | $self->{GOT}{$key} = $value ; |
8e07c86e |
63 | return ; |
64 | } |
65 | |
88108326 |
66 | my $pkg = ref $self ; |
67 | croak "${pkg}::STORE - Unknown element '$key'" ; |
8e07c86e |
68 | } |
69 | |
70 | sub DELETE |
71 | { |
88108326 |
72 | my $self = shift ; |
73 | my $key = shift ; |
74 | |
36477c24 |
75 | if ( exists $self->{VALID}{$key} ) |
8e07c86e |
76 | { |
36477c24 |
77 | delete $self->{GOT}{$key} ; |
8e07c86e |
78 | return ; |
79 | } |
80 | |
88108326 |
81 | my $pkg = ref $self ; |
82 | croak "DB_File::HASHINFO::DELETE - Unknown element '$key'" ; |
8e07c86e |
83 | } |
84 | |
88108326 |
85 | sub EXISTS |
8e07c86e |
86 | { |
88108326 |
87 | my $self = shift ; |
88 | my $key = shift ; |
8e07c86e |
89 | |
36477c24 |
90 | exists $self->{VALID}{$key} ; |
8e07c86e |
91 | } |
92 | |
88108326 |
93 | sub NotHere |
8e07c86e |
94 | { |
18d2dc8c |
95 | my $self = shift ; |
88108326 |
96 | my $method = shift ; |
8e07c86e |
97 | |
18d2dc8c |
98 | croak ref($self) . " does not define the method ${method}" ; |
8e07c86e |
99 | } |
100 | |
18d2dc8c |
101 | sub FIRSTKEY { my $self = shift ; $self->NotHere("FIRSTKEY") } |
102 | sub NEXTKEY { my $self = shift ; $self->NotHere("NEXTKEY") } |
103 | sub CLEAR { my $self = shift ; $self->NotHere("CLEAR") } |
8e07c86e |
104 | |
105 | package DB_File::RECNOINFO ; |
785da04d |
106 | |
88108326 |
107 | use strict ; |
108 | |
045291aa |
109 | @DB_File::RECNOINFO::ISA = qw(DB_File::HASHINFO) ; |
8e07c86e |
110 | |
111 | sub TIEHASH |
112 | { |
88108326 |
113 | my $pkg = shift ; |
114 | |
36477c24 |
115 | bless { VALID => { map {$_, 1} |
116 | qw( bval cachesize psize flags lorder reclen bfname ) |
117 | }, |
118 | GOT => {}, |
119 | }, $pkg ; |
8e07c86e |
120 | } |
121 | |
88108326 |
122 | package DB_File::BTREEINFO ; |
8e07c86e |
123 | |
88108326 |
124 | use strict ; |
8e07c86e |
125 | |
88108326 |
126 | @DB_File::BTREEINFO::ISA = qw(DB_File::HASHINFO) ; |
8e07c86e |
127 | |
88108326 |
128 | sub TIEHASH |
8e07c86e |
129 | { |
88108326 |
130 | my $pkg = shift ; |
131 | |
36477c24 |
132 | bless { VALID => { map {$_, 1} |
133 | qw( flags cachesize maxkeypage minkeypage psize |
134 | compare prefix lorder ) |
135 | }, |
136 | GOT => {}, |
137 | }, $pkg ; |
8e07c86e |
138 | } |
139 | |
140 | |
8e07c86e |
141 | package DB_File ; |
785da04d |
142 | |
143 | use strict; |
1f70e1ea |
144 | use vars qw($VERSION @ISA @EXPORT $AUTOLOAD $DB_BTREE $DB_HASH $DB_RECNO $db_version) ; |
8e07c86e |
145 | use Carp; |
146 | |
785da04d |
147 | |
cad2e5aa |
148 | $VERSION = "1.67" ; |
8e07c86e |
149 | |
150 | #typedef enum { DB_BTREE, DB_HASH, DB_RECNO } DBTYPE; |
88108326 |
151 | $DB_BTREE = new DB_File::BTREEINFO ; |
152 | $DB_HASH = new DB_File::HASHINFO ; |
153 | $DB_RECNO = new DB_File::RECNOINFO ; |
8e07c86e |
154 | |
785da04d |
155 | require Tie::Hash; |
8e07c86e |
156 | require Exporter; |
157 | use AutoLoader; |
158 | require DynaLoader; |
785da04d |
159 | @ISA = qw(Tie::Hash Exporter DynaLoader); |
8e07c86e |
160 | @EXPORT = qw( |
161 | $DB_BTREE $DB_HASH $DB_RECNO |
88108326 |
162 | |
8e07c86e |
163 | BTREEMAGIC |
164 | BTREEVERSION |
165 | DB_LOCK |
166 | DB_SHMEM |
167 | DB_TXN |
168 | HASHMAGIC |
169 | HASHVERSION |
170 | MAX_PAGE_NUMBER |
171 | MAX_PAGE_OFFSET |
172 | MAX_REC_NUMBER |
173 | RET_ERROR |
174 | RET_SPECIAL |
175 | RET_SUCCESS |
176 | R_CURSOR |
177 | R_DUP |
178 | R_FIRST |
179 | R_FIXEDLEN |
180 | R_IAFTER |
181 | R_IBEFORE |
182 | R_LAST |
183 | R_NEXT |
184 | R_NOKEY |
185 | R_NOOVERWRITE |
186 | R_PREV |
187 | R_RECNOSYNC |
188 | R_SETCURSOR |
189 | R_SNAPSHOT |
190 | __R_UNUSED |
88108326 |
191 | |
045291aa |
192 | ); |
8e07c86e |
193 | |
194 | sub AUTOLOAD { |
785da04d |
195 | my($constname); |
8e07c86e |
196 | ($constname = $AUTOLOAD) =~ s/.*:://; |
785da04d |
197 | my $val = constant($constname, @_ ? $_[0] : 0); |
8e07c86e |
198 | if ($! != 0) { |
199 | if ($! =~ /Invalid/) { |
200 | $AutoLoader::AUTOLOAD = $AUTOLOAD; |
201 | goto &AutoLoader::AUTOLOAD; |
202 | } |
203 | else { |
785da04d |
204 | my($pack,$file,$line) = caller; |
8e07c86e |
205 | croak "Your vendor has not defined DB macro $constname, used at $file line $line. |
206 | "; |
207 | } |
208 | } |
209 | eval "sub $AUTOLOAD { $val }"; |
210 | goto &$AUTOLOAD; |
211 | } |
212 | |
f6b705ef |
213 | |
a6ed719b |
214 | eval { |
1f70e1ea |
215 | # Make all Fcntl O_XXX constants available for importing |
216 | require Fcntl; |
217 | my @O = grep /^O_/, @Fcntl::EXPORT; |
218 | Fcntl->import(@O); # first we import what we want to export |
219 | push(@EXPORT, @O); |
a6ed719b |
220 | }; |
f6b705ef |
221 | |
1f70e1ea |
222 | ## import borrowed from IO::File |
223 | ## exports Fcntl constants if available. |
224 | #sub import { |
225 | # my $pkg = shift; |
226 | # my $callpkg = caller; |
227 | # Exporter::export $pkg, $callpkg, @_; |
228 | # eval { |
229 | # require Fcntl; |
230 | # Exporter::export 'Fcntl', $callpkg, '/^O_/'; |
231 | # }; |
232 | #} |
233 | |
785da04d |
234 | bootstrap DB_File $VERSION; |
8e07c86e |
235 | |
236 | # Preloaded methods go here. Autoload methods go after __END__, and are |
237 | # processed by the autosplit program. |
238 | |
05475680 |
239 | sub tie_hash_or_array |
610ab055 |
240 | { |
241 | my (@arg) = @_ ; |
05475680 |
242 | my $tieHASH = ( (caller(1))[3] =~ /TIEHASH/ ) ; |
610ab055 |
243 | |
244 | $arg[4] = tied %{ $arg[4] } |
245 | if @arg >= 5 && ref $arg[4] && $arg[4] =~ /=HASH/ && tied %{ $arg[4] } ; |
246 | |
1f70e1ea |
247 | # make recno in Berkeley DB version 2 work like recno in version 1. |
248 | if ($db_version > 1 and defined $arg[4] and $arg[4] =~ /RECNO/ and |
249 | $arg[1] and ! -e $arg[1]) { |
250 | open(FH, ">$arg[1]") or return undef ; |
251 | close FH ; |
252 | chmod $arg[3] ? $arg[3] : 0666 , $arg[1] ; |
253 | } |
254 | |
05475680 |
255 | DoTie_($tieHASH, @arg) ; |
610ab055 |
256 | } |
257 | |
05475680 |
258 | sub TIEHASH |
259 | { |
260 | tie_hash_or_array(@_) ; |
261 | } |
262 | |
263 | sub TIEARRAY |
264 | { |
265 | tie_hash_or_array(@_) ; |
266 | } |
88108326 |
267 | |
045291aa |
268 | sub CLEAR |
269 | { |
1f70e1ea |
270 | my $self = shift; |
271 | my $key = "" ; |
272 | my $value = "" ; |
273 | my $status = $self->seq($key, $value, R_FIRST()); |
274 | my @keys; |
275 | |
276 | while ($status == 0) { |
277 | push @keys, $key; |
278 | $status = $self->seq($key, $value, R_NEXT()); |
279 | } |
280 | foreach $key (reverse @keys) { |
281 | my $s = $self->del($key); |
282 | } |
283 | } |
284 | |
045291aa |
285 | sub EXTEND { } |
286 | |
287 | sub STORESIZE |
288 | { |
289 | my $self = shift; |
290 | my $length = shift ; |
291 | my $current_length = $self->length() ; |
292 | |
293 | if ($length < $current_length) { |
294 | my $key ; |
295 | for ($key = $current_length - 1 ; $key >= $length ; -- $key) |
296 | { $self->del($key) } |
297 | } |
a9fd575d |
298 | elsif ($length > $current_length) { |
299 | $self->put($length-1, "") ; |
300 | } |
045291aa |
301 | } |
302 | |
6ca2e664 |
303 | sub find_dup |
304 | { |
305 | croak "Usage: \$db->find_dup(key,value)\n" |
306 | unless @_ == 3 ; |
307 | |
308 | my $db = shift ; |
309 | my ($origkey, $value_wanted) = @_ ; |
310 | my ($key, $value) = ($origkey, 0); |
311 | my ($status) = 0 ; |
312 | |
313 | for ($status = $db->seq($key, $value, R_CURSOR() ) ; |
314 | $status == 0 ; |
315 | $status = $db->seq($key, $value, R_NEXT() ) ) { |
316 | |
317 | return 0 if $key eq $origkey and $value eq $value_wanted ; |
318 | } |
319 | |
320 | return $status ; |
321 | } |
322 | |
323 | sub del_dup |
324 | { |
325 | croak "Usage: \$db->del_dup(key,value)\n" |
326 | unless @_ == 3 ; |
327 | |
328 | my $db = shift ; |
329 | my ($key, $value) = @_ ; |
330 | my ($status) = $db->find_dup($key, $value) ; |
331 | return $status if $status != 0 ; |
332 | |
333 | $status = $db->del($key, R_CURSOR() ) ; |
334 | return $status ; |
335 | } |
336 | |
88108326 |
337 | sub get_dup |
338 | { |
339 | croak "Usage: \$db->get_dup(key [,flag])\n" |
340 | unless @_ == 2 or @_ == 3 ; |
341 | |
342 | my $db = shift ; |
343 | my $key = shift ; |
344 | my $flag = shift ; |
f6b705ef |
345 | my $value = 0 ; |
88108326 |
346 | my $origkey = $key ; |
347 | my $wantarray = wantarray ; |
f6b705ef |
348 | my %values = () ; |
88108326 |
349 | my @values = () ; |
350 | my $counter = 0 ; |
f6b705ef |
351 | my $status = 0 ; |
88108326 |
352 | |
f6b705ef |
353 | # iterate through the database until either EOF ($status == 0) |
354 | # or a different key is encountered ($key ne $origkey). |
355 | for ($status = $db->seq($key, $value, R_CURSOR()) ; |
356 | $status == 0 and $key eq $origkey ; |
357 | $status = $db->seq($key, $value, R_NEXT()) ) { |
88108326 |
358 | |
f6b705ef |
359 | # save the value or count number of matches |
360 | if ($wantarray) { |
361 | if ($flag) |
362 | { ++ $values{$value} } |
363 | else |
364 | { push (@values, $value) } |
365 | } |
366 | else |
367 | { ++ $counter } |
88108326 |
368 | |
88108326 |
369 | } |
370 | |
f6b705ef |
371 | return ($wantarray ? ($flag ? %values : @values) : $counter) ; |
88108326 |
372 | } |
373 | |
374 | |
8e07c86e |
375 | 1; |
376 | __END__ |
377 | |
3b35bae3 |
378 | =head1 NAME |
379 | |
1f70e1ea |
380 | DB_File - Perl5 access to Berkeley DB version 1.x |
3b35bae3 |
381 | |
382 | =head1 SYNOPSIS |
383 | |
384 | use DB_File ; |
88108326 |
385 | |
386 | [$X =] tie %hash, 'DB_File', [$filename, $flags, $mode, $DB_HASH] ; |
387 | [$X =] tie %hash, 'DB_File', $filename, $flags, $mode, $DB_BTREE ; |
388 | [$X =] tie @array, 'DB_File', $filename, $flags, $mode, $DB_RECNO ; |
760ac839 |
389 | |
3b35bae3 |
390 | $status = $X->del($key [, $flags]) ; |
391 | $status = $X->put($key, $value [, $flags]) ; |
392 | $status = $X->get($key, $value [, $flags]) ; |
760ac839 |
393 | $status = $X->seq($key, $value, $flags) ; |
3b35bae3 |
394 | $status = $X->sync([$flags]) ; |
395 | $status = $X->fd ; |
760ac839 |
396 | |
f6b705ef |
397 | # BTREE only |
88108326 |
398 | $count = $X->get_dup($key) ; |
399 | @list = $X->get_dup($key) ; |
400 | %list = $X->get_dup($key, 1) ; |
6ca2e664 |
401 | $status = $X->find_dup($key, $value) ; |
402 | $status = $X->del_dup($key, $value) ; |
88108326 |
403 | |
f6b705ef |
404 | # RECNO only |
405 | $a = $X->length; |
406 | $a = $X->pop ; |
407 | $X->push(list); |
408 | $a = $X->shift; |
409 | $X->unshift(list); |
410 | |
cad2e5aa |
411 | # DBM Filters |
412 | $old_filter = $db->filter_store_key ( sub { ... } ) ; |
413 | $old_filter = $db->filter_store_value( sub { ... } ) ; |
414 | $old_filter = $db->filter_fetch_key ( sub { ... } ) ; |
415 | $old_filter = $db->filter_fetch_value( sub { ... } ) ; |
416 | |
3b35bae3 |
417 | untie %hash ; |
418 | untie @array ; |
419 | |
420 | =head1 DESCRIPTION |
421 | |
8e07c86e |
422 | B<DB_File> is a module which allows Perl programs to make use of the |
1f70e1ea |
423 | facilities provided by Berkeley DB version 1.x (if you have a newer |
424 | version of DB, see L<Using DB_File with Berkeley DB version 2>). It is |
425 | assumed that you have a copy of the Berkeley DB manual pages at hand |
426 | when reading this documentation. The interface defined here mirrors the |
427 | Berkeley DB interface closely. |
68dc0745 |
428 | |
8e07c86e |
429 | Berkeley DB is a C library which provides a consistent interface to a |
430 | number of database formats. B<DB_File> provides an interface to all |
431 | three of the database types currently supported by Berkeley DB. |
3b35bae3 |
432 | |
433 | The file types are: |
434 | |
435 | =over 5 |
436 | |
88108326 |
437 | =item B<DB_HASH> |
3b35bae3 |
438 | |
88108326 |
439 | This database type allows arbitrary key/value pairs to be stored in data |
8e07c86e |
440 | files. This is equivalent to the functionality provided by other |
441 | hashing packages like DBM, NDBM, ODBM, GDBM, and SDBM. Remember though, |
442 | the files created using DB_HASH are not compatible with any of the |
443 | other packages mentioned. |
3b35bae3 |
444 | |
8e07c86e |
445 | A default hashing algorithm, which will be adequate for most |
446 | applications, is built into Berkeley DB. If you do need to use your own |
447 | hashing algorithm it is possible to write your own in Perl and have |
448 | B<DB_File> use it instead. |
3b35bae3 |
449 | |
88108326 |
450 | =item B<DB_BTREE> |
451 | |
452 | The btree format allows arbitrary key/value pairs to be stored in a |
8e07c86e |
453 | sorted, balanced binary tree. |
3b35bae3 |
454 | |
8e07c86e |
455 | As with the DB_HASH format, it is possible to provide a user defined |
456 | Perl routine to perform the comparison of keys. By default, though, the |
457 | keys are stored in lexical order. |
3b35bae3 |
458 | |
88108326 |
459 | =item B<DB_RECNO> |
3b35bae3 |
460 | |
8e07c86e |
461 | DB_RECNO allows both fixed-length and variable-length flat text files |
462 | to be manipulated using the same key/value pair interface as in DB_HASH |
463 | and DB_BTREE. In this case the key will consist of a record (line) |
464 | number. |
3b35bae3 |
465 | |
466 | =back |
467 | |
1f70e1ea |
468 | =head2 Using DB_File with Berkeley DB version 2 |
469 | |
470 | Although B<DB_File> is intended to be used with Berkeley DB version 1, |
471 | it can also be used with version 2. In this case the interface is |
472 | limited to the functionality provided by Berkeley DB 1.x. Anywhere the |
473 | version 2 interface differs, B<DB_File> arranges for it to work like |
474 | version 1. This feature allows B<DB_File> scripts that were built with |
475 | version 1 to be migrated to version 2 without any changes. |
476 | |
477 | If you want to make use of the new features available in Berkeley DB |
478 | 2.x, use the Perl module B<BerkeleyDB> instead. |
479 | |
480 | At the time of writing this document the B<BerkeleyDB> module is still |
481 | alpha quality (the version number is < 1.0), and so unsuitable for use |
482 | in any serious development work. Once its version number is >= 1.0, it |
483 | is considered stable enough for real work. |
484 | |
485 | B<Note:> The database file format has changed in Berkeley DB version 2. |
486 | If you cannot recreate your databases, you must dump any existing |
487 | databases with the C<db_dump185> utility that comes with Berkeley DB. |
6ca2e664 |
488 | Once you have rebuilt DB_File to use Berkeley DB version 2, your |
1f70e1ea |
489 | databases can be recreated using C<db_load>. Refer to the Berkeley DB |
490 | documentation for further details. |
491 | |
6ca2e664 |
492 | Please read L<"COPYRIGHT"> before using version 2.x of Berkeley DB with |
1f70e1ea |
493 | DB_File. |
494 | |
68dc0745 |
495 | =head2 Interface to Berkeley DB |
3b35bae3 |
496 | |
497 | B<DB_File> allows access to Berkeley DB files using the tie() mechanism |
8e07c86e |
498 | in Perl 5 (for full details, see L<perlfunc/tie()>). This facility |
499 | allows B<DB_File> to access Berkeley DB files using either an |
500 | associative array (for DB_HASH & DB_BTREE file types) or an ordinary |
501 | array (for the DB_RECNO file type). |
3b35bae3 |
502 | |
88108326 |
503 | In addition to the tie() interface, it is also possible to access most |
504 | of the functions provided in the Berkeley DB API directly. |
f6b705ef |
505 | See L<THE API INTERFACE>. |
3b35bae3 |
506 | |
88108326 |
507 | =head2 Opening a Berkeley DB Database File |
3b35bae3 |
508 | |
8e07c86e |
509 | Berkeley DB uses the function dbopen() to open or create a database. |
f6b705ef |
510 | Here is the C prototype for dbopen(): |
3b35bae3 |
511 | |
512 | DB* |
513 | dbopen (const char * file, int flags, int mode, |
514 | DBTYPE type, const void * openinfo) |
515 | |
516 | The parameter C<type> is an enumeration which specifies which of the 3 |
517 | interface methods (DB_HASH, DB_BTREE or DB_RECNO) is to be used. |
518 | Depending on which of these is actually chosen, the final parameter, |
519 | I<openinfo> points to a data structure which allows tailoring of the |
520 | specific interface method. |
521 | |
8e07c86e |
522 | This interface is handled slightly differently in B<DB_File>. Here is |
88108326 |
523 | an equivalent call using B<DB_File>: |
3b35bae3 |
524 | |
88108326 |
525 | tie %array, 'DB_File', $filename, $flags, $mode, $DB_HASH ; |
3b35bae3 |
526 | |
8e07c86e |
527 | The C<filename>, C<flags> and C<mode> parameters are the direct |
528 | equivalent of their dbopen() counterparts. The final parameter $DB_HASH |
529 | performs the function of both the C<type> and C<openinfo> parameters in |
530 | dbopen(). |
3b35bae3 |
531 | |
88108326 |
532 | In the example above $DB_HASH is actually a pre-defined reference to a |
533 | hash object. B<DB_File> has three of these pre-defined references. |
534 | Apart from $DB_HASH, there is also $DB_BTREE and $DB_RECNO. |
3b35bae3 |
535 | |
8e07c86e |
536 | The keys allowed in each of these pre-defined references is limited to |
537 | the names used in the equivalent C structure. So, for example, the |
538 | $DB_HASH reference will only allow keys called C<bsize>, C<cachesize>, |
88108326 |
539 | C<ffactor>, C<hash>, C<lorder> and C<nelem>. |
540 | |
541 | To change one of these elements, just assign to it like this: |
542 | |
543 | $DB_HASH->{'cachesize'} = 10000 ; |
544 | |
545 | The three predefined variables $DB_HASH, $DB_BTREE and $DB_RECNO are |
546 | usually adequate for most applications. If you do need to create extra |
547 | instances of these objects, constructors are available for each file |
548 | type. |
549 | |
550 | Here are examples of the constructors and the valid options available |
551 | for DB_HASH, DB_BTREE and DB_RECNO respectively. |
552 | |
553 | $a = new DB_File::HASHINFO ; |
554 | $a->{'bsize'} ; |
555 | $a->{'cachesize'} ; |
556 | $a->{'ffactor'}; |
557 | $a->{'hash'} ; |
558 | $a->{'lorder'} ; |
559 | $a->{'nelem'} ; |
560 | |
561 | $b = new DB_File::BTREEINFO ; |
562 | $b->{'flags'} ; |
563 | $b->{'cachesize'} ; |
564 | $b->{'maxkeypage'} ; |
565 | $b->{'minkeypage'} ; |
566 | $b->{'psize'} ; |
567 | $b->{'compare'} ; |
568 | $b->{'prefix'} ; |
569 | $b->{'lorder'} ; |
570 | |
571 | $c = new DB_File::RECNOINFO ; |
572 | $c->{'bval'} ; |
573 | $c->{'cachesize'} ; |
574 | $c->{'psize'} ; |
575 | $c->{'flags'} ; |
576 | $c->{'lorder'} ; |
577 | $c->{'reclen'} ; |
578 | $c->{'bfname'} ; |
579 | |
580 | The values stored in the hashes above are mostly the direct equivalent |
581 | of their C counterpart. Like their C counterparts, all are set to a |
f6b705ef |
582 | default values - that means you don't have to set I<all> of the |
88108326 |
583 | values when you only want to change one. Here is an example: |
584 | |
585 | $a = new DB_File::HASHINFO ; |
586 | $a->{'cachesize'} = 12345 ; |
587 | tie %y, 'DB_File', "filename", $flags, 0777, $a ; |
588 | |
36477c24 |
589 | A few of the options need extra discussion here. When used, the C |
88108326 |
590 | equivalent of the keys C<hash>, C<compare> and C<prefix> store pointers |
591 | to C functions. In B<DB_File> these keys are used to store references |
592 | to Perl subs. Below are templates for each of the subs: |
593 | |
594 | sub hash |
595 | { |
596 | my ($data) = @_ ; |
597 | ... |
598 | # return the hash value for $data |
599 | return $hash ; |
600 | } |
3b35bae3 |
601 | |
88108326 |
602 | sub compare |
603 | { |
604 | my ($key, $key2) = @_ ; |
605 | ... |
606 | # return 0 if $key1 eq $key2 |
607 | # -1 if $key1 lt $key2 |
608 | # 1 if $key1 gt $key2 |
609 | return (-1 , 0 or 1) ; |
610 | } |
3b35bae3 |
611 | |
88108326 |
612 | sub prefix |
613 | { |
614 | my ($key, $key2) = @_ ; |
615 | ... |
616 | # return number of bytes of $key2 which are |
617 | # necessary to determine that it is greater than $key1 |
618 | return $bytes ; |
619 | } |
3b35bae3 |
620 | |
f6b705ef |
621 | See L<Changing the BTREE sort order> for an example of using the |
622 | C<compare> template. |
88108326 |
623 | |
36477c24 |
624 | If you are using the DB_RECNO interface and you intend making use of |
9a2c4ce3 |
625 | C<bval>, you should check out L<The 'bval' Option>. |
36477c24 |
626 | |
88108326 |
627 | =head2 Default Parameters |
628 | |
629 | It is possible to omit some or all of the final 4 parameters in the |
630 | call to C<tie> and let them take default values. As DB_HASH is the most |
631 | common file format used, the call: |
632 | |
633 | tie %A, "DB_File", "filename" ; |
634 | |
635 | is equivalent to: |
636 | |
18d2dc8c |
637 | tie %A, "DB_File", "filename", O_CREAT|O_RDWR, 0666, $DB_HASH ; |
88108326 |
638 | |
639 | It is also possible to omit the filename parameter as well, so the |
640 | call: |
641 | |
642 | tie %A, "DB_File" ; |
643 | |
644 | is equivalent to: |
645 | |
18d2dc8c |
646 | tie %A, "DB_File", undef, O_CREAT|O_RDWR, 0666, $DB_HASH ; |
88108326 |
647 | |
f6b705ef |
648 | See L<In Memory Databases> for a discussion on the use of C<undef> |
88108326 |
649 | in place of a filename. |
650 | |
f6b705ef |
651 | =head2 In Memory Databases |
652 | |
653 | Berkeley DB allows the creation of in-memory databases by using NULL |
654 | (that is, a C<(char *)0> in C) in place of the filename. B<DB_File> |
655 | uses C<undef> instead of NULL to provide this functionality. |
656 | |
657 | =head1 DB_HASH |
658 | |
659 | The DB_HASH file format is probably the most commonly used of the three |
660 | file formats that B<DB_File> supports. It is also very straightforward |
661 | to use. |
662 | |
68dc0745 |
663 | =head2 A Simple Example |
f6b705ef |
664 | |
665 | This example shows how to create a database, add key/value pairs to the |
666 | database, delete keys/value pairs and finally how to enumerate the |
667 | contents of the database. |
668 | |
610ab055 |
669 | use strict ; |
f6b705ef |
670 | use DB_File ; |
610ab055 |
671 | use vars qw( %h $k $v ) ; |
f6b705ef |
672 | |
673 | tie %h, "DB_File", "fruit", O_RDWR|O_CREAT, 0640, $DB_HASH |
674 | or die "Cannot open file 'fruit': $!\n"; |
675 | |
676 | # Add a few key/value pairs to the file |
677 | $h{"apple"} = "red" ; |
678 | $h{"orange"} = "orange" ; |
679 | $h{"banana"} = "yellow" ; |
680 | $h{"tomato"} = "red" ; |
681 | |
682 | # Check for existence of a key |
683 | print "Banana Exists\n\n" if $h{"banana"} ; |
684 | |
685 | # Delete a key/value pair. |
686 | delete $h{"apple"} ; |
687 | |
688 | # print the contents of the file |
689 | while (($k, $v) = each %h) |
690 | { print "$k -> $v\n" } |
691 | |
692 | untie %h ; |
693 | |
694 | here is the output: |
695 | |
696 | Banana Exists |
697 | |
698 | orange -> orange |
699 | tomato -> red |
700 | banana -> yellow |
701 | |
702 | Note that the like ordinary associative arrays, the order of the keys |
703 | retrieved is in an apparently random order. |
704 | |
705 | =head1 DB_BTREE |
706 | |
707 | The DB_BTREE format is useful when you want to store data in a given |
708 | order. By default the keys will be stored in lexical order, but as you |
709 | will see from the example shown in the next section, it is very easy to |
710 | define your own sorting function. |
711 | |
712 | =head2 Changing the BTREE sort order |
713 | |
714 | This script shows how to override the default sorting algorithm that |
715 | BTREE uses. Instead of using the normal lexical ordering, a case |
716 | insensitive compare function will be used. |
88108326 |
717 | |
610ab055 |
718 | use strict ; |
f6b705ef |
719 | use DB_File ; |
610ab055 |
720 | |
721 | my %h ; |
f6b705ef |
722 | |
723 | sub Compare |
724 | { |
725 | my ($key1, $key2) = @_ ; |
726 | "\L$key1" cmp "\L$key2" ; |
727 | } |
728 | |
729 | # specify the Perl sub that will do the comparison |
730 | $DB_BTREE->{'compare'} = \&Compare ; |
731 | |
732 | tie %h, "DB_File", "tree", O_RDWR|O_CREAT, 0640, $DB_BTREE |
733 | or die "Cannot open file 'tree': $!\n" ; |
734 | |
735 | # Add a key/value pair to the file |
736 | $h{'Wall'} = 'Larry' ; |
737 | $h{'Smith'} = 'John' ; |
738 | $h{'mouse'} = 'mickey' ; |
739 | $h{'duck'} = 'donald' ; |
740 | |
741 | # Delete |
742 | delete $h{"duck"} ; |
743 | |
744 | # Cycle through the keys printing them in order. |
745 | # Note it is not necessary to sort the keys as |
746 | # the btree will have kept them in order automatically. |
747 | foreach (keys %h) |
748 | { print "$_\n" } |
749 | |
750 | untie %h ; |
751 | |
752 | Here is the output from the code above. |
753 | |
754 | mouse |
755 | Smith |
756 | Wall |
757 | |
758 | There are a few point to bear in mind if you want to change the |
759 | ordering in a BTREE database: |
760 | |
761 | =over 5 |
762 | |
763 | =item 1. |
764 | |
765 | The new compare function must be specified when you create the database. |
766 | |
767 | =item 2. |
768 | |
769 | You cannot change the ordering once the database has been created. Thus |
770 | you must use the same compare function every time you access the |
88108326 |
771 | database. |
772 | |
f6b705ef |
773 | =back |
774 | |
68dc0745 |
775 | =head2 Handling Duplicate Keys |
f6b705ef |
776 | |
777 | The BTREE file type optionally allows a single key to be associated |
778 | with an arbitrary number of values. This option is enabled by setting |
779 | the flags element of C<$DB_BTREE> to R_DUP when creating the database. |
780 | |
88108326 |
781 | There are some difficulties in using the tied hash interface if you |
782 | want to manipulate a BTREE database with duplicate keys. Consider this |
783 | code: |
784 | |
610ab055 |
785 | use strict ; |
88108326 |
786 | use DB_File ; |
610ab055 |
787 | |
788 | use vars qw($filename %h ) ; |
789 | |
88108326 |
790 | $filename = "tree" ; |
791 | unlink $filename ; |
792 | |
793 | # Enable duplicate records |
794 | $DB_BTREE->{'flags'} = R_DUP ; |
795 | |
796 | tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0640, $DB_BTREE |
797 | or die "Cannot open $filename: $!\n"; |
798 | |
799 | # Add some key/value pairs to the file |
800 | $h{'Wall'} = 'Larry' ; |
801 | $h{'Wall'} = 'Brick' ; # Note the duplicate key |
f6b705ef |
802 | $h{'Wall'} = 'Brick' ; # Note the duplicate key and value |
88108326 |
803 | $h{'Smith'} = 'John' ; |
804 | $h{'mouse'} = 'mickey' ; |
805 | |
806 | # iterate through the associative array |
807 | # and print each key/value pair. |
808 | foreach (keys %h) |
809 | { print "$_ -> $h{$_}\n" } |
810 | |
f6b705ef |
811 | untie %h ; |
812 | |
88108326 |
813 | Here is the output: |
814 | |
815 | Smith -> John |
816 | Wall -> Larry |
817 | Wall -> Larry |
f6b705ef |
818 | Wall -> Larry |
88108326 |
819 | mouse -> mickey |
820 | |
f6b705ef |
821 | As you can see 3 records have been successfully created with key C<Wall> |
88108326 |
822 | - the only thing is, when they are retrieved from the database they |
f6b705ef |
823 | I<seem> to have the same value, namely C<Larry>. The problem is caused |
824 | by the way that the associative array interface works. Basically, when |
825 | the associative array interface is used to fetch the value associated |
826 | with a given key, it will only ever retrieve the first value. |
88108326 |
827 | |
828 | Although it may not be immediately obvious from the code above, the |
829 | associative array interface can be used to write values with duplicate |
830 | keys, but it cannot be used to read them back from the database. |
831 | |
832 | The way to get around this problem is to use the Berkeley DB API method |
833 | called C<seq>. This method allows sequential access to key/value |
f6b705ef |
834 | pairs. See L<THE API INTERFACE> for details of both the C<seq> method |
835 | and the API in general. |
88108326 |
836 | |
837 | Here is the script above rewritten using the C<seq> API method. |
838 | |
610ab055 |
839 | use strict ; |
88108326 |
840 | use DB_File ; |
88108326 |
841 | |
610ab055 |
842 | use vars qw($filename $x %h $status $key $value) ; |
843 | |
88108326 |
844 | $filename = "tree" ; |
845 | unlink $filename ; |
846 | |
847 | # Enable duplicate records |
848 | $DB_BTREE->{'flags'} = R_DUP ; |
849 | |
850 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0640, $DB_BTREE |
851 | or die "Cannot open $filename: $!\n"; |
852 | |
853 | # Add some key/value pairs to the file |
854 | $h{'Wall'} = 'Larry' ; |
855 | $h{'Wall'} = 'Brick' ; # Note the duplicate key |
f6b705ef |
856 | $h{'Wall'} = 'Brick' ; # Note the duplicate key and value |
88108326 |
857 | $h{'Smith'} = 'John' ; |
858 | $h{'mouse'} = 'mickey' ; |
859 | |
f6b705ef |
860 | # iterate through the btree using seq |
88108326 |
861 | # and print each key/value pair. |
610ab055 |
862 | $key = $value = 0 ; |
f6b705ef |
863 | for ($status = $x->seq($key, $value, R_FIRST) ; |
864 | $status == 0 ; |
865 | $status = $x->seq($key, $value, R_NEXT) ) |
88108326 |
866 | { print "$key -> $value\n" } |
867 | |
868 | undef $x ; |
869 | untie %h ; |
870 | |
871 | that prints: |
872 | |
873 | Smith -> John |
874 | Wall -> Brick |
f6b705ef |
875 | Wall -> Brick |
88108326 |
876 | Wall -> Larry |
877 | mouse -> mickey |
878 | |
f6b705ef |
879 | This time we have got all the key/value pairs, including the multiple |
88108326 |
880 | values associated with the key C<Wall>. |
881 | |
6ca2e664 |
882 | To make life easier when dealing with duplicate keys, B<DB_File> comes with |
883 | a few utility methods. |
884 | |
68dc0745 |
885 | =head2 The get_dup() Method |
f6b705ef |
886 | |
6ca2e664 |
887 | The C<get_dup> method assists in |
88108326 |
888 | reading duplicate values from BTREE databases. The method can take the |
889 | following forms: |
890 | |
891 | $count = $x->get_dup($key) ; |
892 | @list = $x->get_dup($key) ; |
893 | %list = $x->get_dup($key, 1) ; |
894 | |
895 | In a scalar context the method returns the number of values associated |
896 | with the key, C<$key>. |
897 | |
898 | In list context, it returns all the values which match C<$key>. Note |
f6b705ef |
899 | that the values will be returned in an apparently random order. |
88108326 |
900 | |
7a2e2cd6 |
901 | In list context, if the second parameter is present and evaluates |
902 | TRUE, the method returns an associative array. The keys of the |
903 | associative array correspond to the values that matched in the BTREE |
904 | and the values of the array are a count of the number of times that |
905 | particular value occurred in the BTREE. |
88108326 |
906 | |
f6b705ef |
907 | So assuming the database created above, we can use C<get_dup> like |
88108326 |
908 | this: |
909 | |
610ab055 |
910 | my $cnt = $x->get_dup("Wall") ; |
88108326 |
911 | print "Wall occurred $cnt times\n" ; |
912 | |
610ab055 |
913 | my %hash = $x->get_dup("Wall", 1) ; |
88108326 |
914 | print "Larry is there\n" if $hash{'Larry'} ; |
f6b705ef |
915 | print "There are $hash{'Brick'} Brick Walls\n" ; |
88108326 |
916 | |
610ab055 |
917 | my @list = $x->get_dup("Wall") ; |
88108326 |
918 | print "Wall => [@list]\n" ; |
919 | |
f6b705ef |
920 | @list = $x->get_dup("Smith") ; |
88108326 |
921 | print "Smith => [@list]\n" ; |
922 | |
f6b705ef |
923 | @list = $x->get_dup("Dog") ; |
88108326 |
924 | print "Dog => [@list]\n" ; |
925 | |
926 | |
927 | and it will print: |
928 | |
f6b705ef |
929 | Wall occurred 3 times |
88108326 |
930 | Larry is there |
f6b705ef |
931 | There are 2 Brick Walls |
932 | Wall => [Brick Brick Larry] |
88108326 |
933 | Smith => [John] |
934 | Dog => [] |
3b35bae3 |
935 | |
6ca2e664 |
936 | =head2 The find_dup() Method |
937 | |
938 | $status = $X->find_dup($key, $value) ; |
939 | |
940 | This method checks for the existance of a specific key/value pair. If the |
941 | pair exists, the cursor is left pointing to the pair and the method |
942 | returns 0. Otherwise the method returns a non-zero value. |
943 | |
944 | Assuming the database from the previous example: |
945 | |
946 | use strict ; |
947 | use DB_File ; |
948 | |
949 | use vars qw($filename $x %h $found) ; |
950 | |
951 | my $filename = "tree" ; |
952 | |
953 | # Enable duplicate records |
954 | $DB_BTREE->{'flags'} = R_DUP ; |
955 | |
956 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0640, $DB_BTREE |
957 | or die "Cannot open $filename: $!\n"; |
958 | |
959 | $found = ( $x->find_dup("Wall", "Larry") == 0 ? "" : "not") ; |
960 | print "Larry Wall is $found there\n" ; |
961 | |
962 | $found = ( $x->find_dup("Wall", "Harry") == 0 ? "" : "not") ; |
963 | print "Harry Wall is $found there\n" ; |
964 | |
965 | undef $x ; |
966 | untie %h ; |
967 | |
968 | prints this |
969 | |
970 | Larry Wall is there |
971 | Harry Wall is not there |
972 | |
973 | |
974 | =head2 The del_dup() Method |
975 | |
976 | $status = $X->del_dup($key, $value) ; |
977 | |
978 | This method deletes a specific key/value pair. It returns |
979 | 0 if they exist and have been deleted successfully. |
980 | Otherwise the method returns a non-zero value. |
981 | |
982 | Again assuming the existance of the C<tree> database |
983 | |
984 | use strict ; |
985 | use DB_File ; |
986 | |
987 | use vars qw($filename $x %h $found) ; |
988 | |
989 | my $filename = "tree" ; |
990 | |
991 | # Enable duplicate records |
992 | $DB_BTREE->{'flags'} = R_DUP ; |
993 | |
994 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0640, $DB_BTREE |
995 | or die "Cannot open $filename: $!\n"; |
996 | |
997 | $x->del_dup("Wall", "Larry") ; |
998 | |
999 | $found = ( $x->find_dup("Wall", "Larry") == 0 ? "" : "not") ; |
1000 | print "Larry Wall is $found there\n" ; |
1001 | |
1002 | undef $x ; |
1003 | untie %h ; |
1004 | |
1005 | prints this |
1006 | |
1007 | Larry Wall is not there |
1008 | |
f6b705ef |
1009 | =head2 Matching Partial Keys |
1010 | |
1011 | The BTREE interface has a feature which allows partial keys to be |
1012 | matched. This functionality is I<only> available when the C<seq> method |
1013 | is used along with the R_CURSOR flag. |
1014 | |
1015 | $x->seq($key, $value, R_CURSOR) ; |
1016 | |
1017 | Here is the relevant quote from the dbopen man page where it defines |
1018 | the use of the R_CURSOR flag with seq: |
1019 | |
f6b705ef |
1020 | Note, for the DB_BTREE access method, the returned key is not |
1021 | necessarily an exact match for the specified key. The returned key |
1022 | is the smallest key greater than or equal to the specified key, |
1023 | permitting partial key matches and range searches. |
1024 | |
f6b705ef |
1025 | In the example script below, the C<match> sub uses this feature to find |
1026 | and print the first matching key/value pair given a partial key. |
1027 | |
610ab055 |
1028 | use strict ; |
f6b705ef |
1029 | use DB_File ; |
1030 | use Fcntl ; |
610ab055 |
1031 | |
1032 | use vars qw($filename $x %h $st $key $value) ; |
f6b705ef |
1033 | |
1034 | sub match |
1035 | { |
1036 | my $key = shift ; |
610ab055 |
1037 | my $value = 0; |
f6b705ef |
1038 | my $orig_key = $key ; |
1039 | $x->seq($key, $value, R_CURSOR) ; |
1040 | print "$orig_key\t-> $key\t-> $value\n" ; |
1041 | } |
1042 | |
1043 | $filename = "tree" ; |
1044 | unlink $filename ; |
1045 | |
1046 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0640, $DB_BTREE |
1047 | or die "Cannot open $filename: $!\n"; |
1048 | |
1049 | # Add some key/value pairs to the file |
1050 | $h{'mouse'} = 'mickey' ; |
1051 | $h{'Wall'} = 'Larry' ; |
1052 | $h{'Walls'} = 'Brick' ; |
1053 | $h{'Smith'} = 'John' ; |
1054 | |
1055 | |
610ab055 |
1056 | $key = $value = 0 ; |
f6b705ef |
1057 | print "IN ORDER\n" ; |
1058 | for ($st = $x->seq($key, $value, R_FIRST) ; |
1059 | $st == 0 ; |
1060 | $st = $x->seq($key, $value, R_NEXT) ) |
1061 | |
1062 | { print "$key -> $value\n" } |
1063 | |
1064 | print "\nPARTIAL MATCH\n" ; |
1065 | |
1066 | match "Wa" ; |
1067 | match "A" ; |
1068 | match "a" ; |
1069 | |
1070 | undef $x ; |
1071 | untie %h ; |
1072 | |
1073 | Here is the output: |
1074 | |
1075 | IN ORDER |
1076 | Smith -> John |
1077 | Wall -> Larry |
1078 | Walls -> Brick |
1079 | mouse -> mickey |
1080 | |
1081 | PARTIAL MATCH |
1082 | Wa -> Wall -> Larry |
1083 | A -> Smith -> John |
1084 | a -> mouse -> mickey |
1085 | |
1086 | =head1 DB_RECNO |
1087 | |
1088 | DB_RECNO provides an interface to flat text files. Both variable and |
1089 | fixed length records are supported. |
3b35bae3 |
1090 | |
6ca2e664 |
1091 | In order to make RECNO more compatible with Perl, the array offset for |
88108326 |
1092 | all RECNO arrays begins at 0 rather than 1 as in Berkeley DB. |
3b35bae3 |
1093 | |
88108326 |
1094 | As with normal Perl arrays, a RECNO array can be accessed using |
1095 | negative indexes. The index -1 refers to the last element of the array, |
1096 | -2 the second last, and so on. Attempting to access an element before |
1097 | the start of the array will raise a fatal run-time error. |
3b35bae3 |
1098 | |
68dc0745 |
1099 | =head2 The 'bval' Option |
36477c24 |
1100 | |
1101 | The operation of the bval option warrants some discussion. Here is the |
1102 | definition of bval from the Berkeley DB 1.85 recno manual page: |
1103 | |
1104 | The delimiting byte to be used to mark the end of a |
1105 | record for variable-length records, and the pad charac- |
1106 | ter for fixed-length records. If no value is speci- |
1107 | fied, newlines (``\n'') are used to mark the end of |
1108 | variable-length records and fixed-length records are |
1109 | padded with spaces. |
1110 | |
1111 | The second sentence is wrong. In actual fact bval will only default to |
1112 | C<"\n"> when the openinfo parameter in dbopen is NULL. If a non-NULL |
1113 | openinfo parameter is used at all, the value that happens to be in bval |
1114 | will be used. That means you always have to specify bval when making |
1115 | use of any of the options in the openinfo parameter. This documentation |
1116 | error will be fixed in the next release of Berkeley DB. |
1117 | |
1118 | That clarifies the situation with regards Berkeley DB itself. What |
1119 | about B<DB_File>? Well, the behavior defined in the quote above is |
6ca2e664 |
1120 | quite useful, so B<DB_File> conforms to it. |
36477c24 |
1121 | |
1122 | That means that you can specify other options (e.g. cachesize) and |
1123 | still have bval default to C<"\n"> for variable length records, and |
1124 | space for fixed length records. |
1125 | |
f6b705ef |
1126 | =head2 A Simple Example |
3b35bae3 |
1127 | |
6ca2e664 |
1128 | Here is a simple example that uses RECNO (if you are using a version |
1129 | of Perl earlier than 5.004_57 this example won't work -- see |
1130 | L<Extra RECNO Methods> for a workaround). |
f6b705ef |
1131 | |
610ab055 |
1132 | use strict ; |
f6b705ef |
1133 | use DB_File ; |
f6b705ef |
1134 | |
610ab055 |
1135 | my @h ; |
f6b705ef |
1136 | tie @h, "DB_File", "text", O_RDWR|O_CREAT, 0640, $DB_RECNO |
1137 | or die "Cannot open file 'text': $!\n" ; |
1138 | |
1139 | # Add a few key/value pairs to the file |
1140 | $h[0] = "orange" ; |
1141 | $h[1] = "blue" ; |
1142 | $h[2] = "yellow" ; |
1143 | |
6ca2e664 |
1144 | push @h, "green", "black" ; |
1145 | |
1146 | my $elements = scalar @h ; |
1147 | print "The array contains $elements entries\n" ; |
1148 | |
1149 | my $last = pop @h ; |
1150 | print "popped $last\n" ; |
1151 | |
1152 | unshift @h, "white" ; |
1153 | my $first = shift @h ; |
1154 | print "shifted $first\n" ; |
1155 | |
f6b705ef |
1156 | # Check for existence of a key |
1157 | print "Element 1 Exists with value $h[1]\n" if $h[1] ; |
1158 | |
1159 | # use a negative index |
1160 | print "The last element is $h[-1]\n" ; |
1161 | print "The 2nd last element is $h[-2]\n" ; |
1162 | |
1163 | untie @h ; |
3b35bae3 |
1164 | |
f6b705ef |
1165 | Here is the output from the script: |
1166 | |
6ca2e664 |
1167 | The array contains 5 entries |
1168 | popped black |
1169 | unshifted white |
f6b705ef |
1170 | Element 1 Exists with value blue |
6ca2e664 |
1171 | The last element is green |
1172 | The 2nd last element is yellow |
f6b705ef |
1173 | |
6ca2e664 |
1174 | =head2 Extra RECNO Methods |
f6b705ef |
1175 | |
045291aa |
1176 | If you are using a version of Perl earlier than 5.004_57, the tied |
6ca2e664 |
1177 | array interface is quite limited. In the example script above |
1178 | C<push>, C<pop>, C<shift>, C<unshift> |
1179 | or determining the array length will not work with a tied array. |
045291aa |
1180 | |
1181 | To make the interface more useful for older versions of Perl, a number |
1182 | of methods are supplied with B<DB_File> to simulate the missing array |
1183 | operations. All these methods are accessed via the object returned from |
1184 | the tie call. |
f6b705ef |
1185 | |
1186 | Here are the methods: |
1187 | |
1188 | =over 5 |
3b35bae3 |
1189 | |
f6b705ef |
1190 | =item B<$X-E<gt>push(list) ;> |
1191 | |
1192 | Pushes the elements of C<list> to the end of the array. |
1193 | |
1194 | =item B<$value = $X-E<gt>pop ;> |
1195 | |
1196 | Removes and returns the last element of the array. |
1197 | |
1198 | =item B<$X-E<gt>shift> |
1199 | |
1200 | Removes and returns the first element of the array. |
1201 | |
1202 | =item B<$X-E<gt>unshift(list) ;> |
1203 | |
1204 | Pushes the elements of C<list> to the start of the array. |
1205 | |
1206 | =item B<$X-E<gt>length> |
1207 | |
1208 | Returns the number of elements in the array. |
1209 | |
1210 | =back |
1211 | |
1212 | =head2 Another Example |
1213 | |
1214 | Here is a more complete example that makes use of some of the methods |
1215 | described above. It also makes use of the API interface directly (see |
1216 | L<THE API INTERFACE>). |
1217 | |
1218 | use strict ; |
1219 | use vars qw(@h $H $file $i) ; |
1220 | use DB_File ; |
1221 | use Fcntl ; |
1222 | |
1223 | $file = "text" ; |
1224 | |
1225 | unlink $file ; |
1226 | |
1227 | $H = tie @h, "DB_File", $file, O_RDWR|O_CREAT, 0640, $DB_RECNO |
1228 | or die "Cannot open file $file: $!\n" ; |
1229 | |
1230 | # first create a text file to play with |
1231 | $h[0] = "zero" ; |
1232 | $h[1] = "one" ; |
1233 | $h[2] = "two" ; |
1234 | $h[3] = "three" ; |
1235 | $h[4] = "four" ; |
1236 | |
1237 | |
1238 | # Print the records in order. |
1239 | # |
1240 | # The length method is needed here because evaluating a tied |
1241 | # array in a scalar context does not return the number of |
1242 | # elements in the array. |
1243 | |
1244 | print "\nORIGINAL\n" ; |
1245 | foreach $i (0 .. $H->length - 1) { |
1246 | print "$i: $h[$i]\n" ; |
1247 | } |
1248 | |
1249 | # use the push & pop methods |
1250 | $a = $H->pop ; |
1251 | $H->push("last") ; |
1252 | print "\nThe last record was [$a]\n" ; |
1253 | |
1254 | # and the shift & unshift methods |
1255 | $a = $H->shift ; |
1256 | $H->unshift("first") ; |
1257 | print "The first record was [$a]\n" ; |
1258 | |
1259 | # Use the API to add a new record after record 2. |
1260 | $i = 2 ; |
1261 | $H->put($i, "Newbie", R_IAFTER) ; |
1262 | |
1263 | # and a new record before record 1. |
1264 | $i = 1 ; |
1265 | $H->put($i, "New One", R_IBEFORE) ; |
1266 | |
1267 | # delete record 3 |
1268 | $H->del(3) ; |
1269 | |
1270 | # now print the records in reverse order |
1271 | print "\nREVERSE\n" ; |
1272 | for ($i = $H->length - 1 ; $i >= 0 ; -- $i) |
1273 | { print "$i: $h[$i]\n" } |
1274 | |
1275 | # same again, but use the API functions instead |
1276 | print "\nREVERSE again\n" ; |
610ab055 |
1277 | my ($s, $k, $v) = (0, 0, 0) ; |
f6b705ef |
1278 | for ($s = $H->seq($k, $v, R_LAST) ; |
1279 | $s == 0 ; |
1280 | $s = $H->seq($k, $v, R_PREV)) |
1281 | { print "$k: $v\n" } |
1282 | |
1283 | undef $H ; |
1284 | untie @h ; |
1285 | |
1286 | and this is what it outputs: |
1287 | |
1288 | ORIGINAL |
1289 | 0: zero |
1290 | 1: one |
1291 | 2: two |
1292 | 3: three |
1293 | 4: four |
1294 | |
1295 | The last record was [four] |
1296 | The first record was [zero] |
1297 | |
1298 | REVERSE |
1299 | 5: last |
1300 | 4: three |
1301 | 3: Newbie |
1302 | 2: one |
1303 | 1: New One |
1304 | 0: first |
1305 | |
1306 | REVERSE again |
1307 | 5: last |
1308 | 4: three |
1309 | 3: Newbie |
1310 | 2: one |
1311 | 1: New One |
1312 | 0: first |
1313 | |
1314 | Notes: |
1315 | |
1316 | =over 5 |
1317 | |
1318 | =item 1. |
1319 | |
1320 | Rather than iterating through the array, C<@h> like this: |
1321 | |
1322 | foreach $i (@h) |
1323 | |
1324 | it is necessary to use either this: |
1325 | |
1326 | foreach $i (0 .. $H->length - 1) |
1327 | |
1328 | or this: |
1329 | |
1330 | for ($a = $H->get($k, $v, R_FIRST) ; |
1331 | $a == 0 ; |
1332 | $a = $H->get($k, $v, R_NEXT) ) |
1333 | |
1334 | =item 2. |
1335 | |
1336 | Notice that both times the C<put> method was used the record index was |
1337 | specified using a variable, C<$i>, rather than the literal value |
1338 | itself. This is because C<put> will return the record number of the |
1339 | inserted line via that parameter. |
1340 | |
1341 | =back |
1342 | |
1343 | =head1 THE API INTERFACE |
3b35bae3 |
1344 | |
1345 | As well as accessing Berkeley DB using a tied hash or array, it is also |
88108326 |
1346 | possible to make direct use of most of the API functions defined in the |
8e07c86e |
1347 | Berkeley DB documentation. |
3b35bae3 |
1348 | |
88108326 |
1349 | To do this you need to store a copy of the object returned from the tie. |
3b35bae3 |
1350 | |
88108326 |
1351 | $db = tie %hash, "DB_File", "filename" ; |
3b35bae3 |
1352 | |
8e07c86e |
1353 | Once you have done that, you can access the Berkeley DB API functions |
88108326 |
1354 | as B<DB_File> methods directly like this: |
3b35bae3 |
1355 | |
1356 | $db->put($key, $value, R_NOOVERWRITE) ; |
1357 | |
88108326 |
1358 | B<Important:> If you have saved a copy of the object returned from |
1359 | C<tie>, the underlying database file will I<not> be closed until both |
1360 | the tied variable is untied and all copies of the saved object are |
610ab055 |
1361 | destroyed. |
88108326 |
1362 | |
1363 | use DB_File ; |
1364 | $db = tie %hash, "DB_File", "filename" |
1365 | or die "Cannot tie filename: $!" ; |
1366 | ... |
1367 | undef $db ; |
1368 | untie %hash ; |
1369 | |
9a2c4ce3 |
1370 | See L<The untie() Gotcha> for more details. |
778183f3 |
1371 | |
88108326 |
1372 | All the functions defined in L<dbopen> are available except for |
1373 | close() and dbopen() itself. The B<DB_File> method interface to the |
1374 | supported functions have been implemented to mirror the way Berkeley DB |
1375 | works whenever possible. In particular note that: |
1376 | |
1377 | =over 5 |
1378 | |
1379 | =item * |
1380 | |
1381 | The methods return a status value. All return 0 on success. |
1382 | All return -1 to signify an error and set C<$!> to the exact |
1383 | error code. The return code 1 generally (but not always) means that the |
1384 | key specified did not exist in the database. |
1385 | |
1386 | Other return codes are defined. See below and in the Berkeley DB |
1387 | documentation for details. The Berkeley DB documentation should be used |
1388 | as the definitive source. |
1389 | |
1390 | =item * |
3b35bae3 |
1391 | |
88108326 |
1392 | Whenever a Berkeley DB function returns data via one of its parameters, |
1393 | the equivalent B<DB_File> method does exactly the same. |
3b35bae3 |
1394 | |
88108326 |
1395 | =item * |
1396 | |
1397 | If you are careful, it is possible to mix API calls with the tied |
1398 | hash/array interface in the same piece of code. Although only a few of |
1399 | the methods used to implement the tied interface currently make use of |
1400 | the cursor, you should always assume that the cursor has been changed |
1401 | any time the tied hash/array interface is used. As an example, this |
1402 | code will probably not do what you expect: |
1403 | |
1404 | $X = tie %x, 'DB_File', $filename, O_RDWR|O_CREAT, 0777, $DB_BTREE |
1405 | or die "Cannot tie $filename: $!" ; |
1406 | |
1407 | # Get the first key/value pair and set the cursor |
1408 | $X->seq($key, $value, R_FIRST) ; |
1409 | |
1410 | # this line will modify the cursor |
1411 | $count = scalar keys %x ; |
1412 | |
1413 | # Get the second key/value pair. |
1414 | # oops, it didn't, it got the last key/value pair! |
1415 | $X->seq($key, $value, R_NEXT) ; |
1416 | |
1417 | The code above can be rearranged to get around the problem, like this: |
1418 | |
1419 | $X = tie %x, 'DB_File', $filename, O_RDWR|O_CREAT, 0777, $DB_BTREE |
1420 | or die "Cannot tie $filename: $!" ; |
1421 | |
1422 | # this line will modify the cursor |
1423 | $count = scalar keys %x ; |
1424 | |
1425 | # Get the first key/value pair and set the cursor |
1426 | $X->seq($key, $value, R_FIRST) ; |
1427 | |
1428 | # Get the second key/value pair. |
1429 | # worked this time. |
1430 | $X->seq($key, $value, R_NEXT) ; |
1431 | |
1432 | =back |
1433 | |
1434 | All the constants defined in L<dbopen> for use in the flags parameters |
1435 | in the methods defined below are also available. Refer to the Berkeley |
1436 | DB documentation for the precise meaning of the flags values. |
1437 | |
1438 | Below is a list of the methods available. |
3b35bae3 |
1439 | |
1440 | =over 5 |
1441 | |
f6b705ef |
1442 | =item B<$status = $X-E<gt>get($key, $value [, $flags]) ;> |
88108326 |
1443 | |
1444 | Given a key (C<$key>) this method reads the value associated with it |
1445 | from the database. The value read from the database is returned in the |
1446 | C<$value> parameter. |
3b35bae3 |
1447 | |
88108326 |
1448 | If the key does not exist the method returns 1. |
3b35bae3 |
1449 | |
88108326 |
1450 | No flags are currently defined for this method. |
3b35bae3 |
1451 | |
f6b705ef |
1452 | =item B<$status = $X-E<gt>put($key, $value [, $flags]) ;> |
3b35bae3 |
1453 | |
88108326 |
1454 | Stores the key/value pair in the database. |
1455 | |
1456 | If you use either the R_IAFTER or R_IBEFORE flags, the C<$key> parameter |
8e07c86e |
1457 | will have the record number of the inserted key/value pair set. |
3b35bae3 |
1458 | |
88108326 |
1459 | Valid flags are R_CURSOR, R_IAFTER, R_IBEFORE, R_NOOVERWRITE and |
1460 | R_SETCURSOR. |
1461 | |
f6b705ef |
1462 | =item B<$status = $X-E<gt>del($key [, $flags]) ;> |
3b35bae3 |
1463 | |
88108326 |
1464 | Removes all key/value pairs with key C<$key> from the database. |
3b35bae3 |
1465 | |
88108326 |
1466 | A return code of 1 means that the requested key was not in the |
1467 | database. |
3b35bae3 |
1468 | |
88108326 |
1469 | R_CURSOR is the only valid flag at present. |
3b35bae3 |
1470 | |
f6b705ef |
1471 | =item B<$status = $X-E<gt>fd ;> |
3b35bae3 |
1472 | |
88108326 |
1473 | Returns the file descriptor for the underlying database. |
3b35bae3 |
1474 | |
f6b705ef |
1475 | See L<Locking Databases> for an example of how to make use of the |
88108326 |
1476 | C<fd> method to lock your database. |
3b35bae3 |
1477 | |
f6b705ef |
1478 | =item B<$status = $X-E<gt>seq($key, $value, $flags) ;> |
3b35bae3 |
1479 | |
88108326 |
1480 | This interface allows sequential retrieval from the database. See |
1481 | L<dbopen> for full details. |
1482 | |
1483 | Both the C<$key> and C<$value> parameters will be set to the key/value |
1484 | pair read from the database. |
1485 | |
1486 | The flags parameter is mandatory. The valid flag values are R_CURSOR, |
1487 | R_FIRST, R_LAST, R_NEXT and R_PREV. |
1488 | |
f6b705ef |
1489 | =item B<$status = $X-E<gt>sync([$flags]) ;> |
88108326 |
1490 | |
1491 | Flushes any cached buffers to disk. |
1492 | |
1493 | R_RECNOSYNC is the only valid flag at present. |
3b35bae3 |
1494 | |
1495 | =back |
1496 | |
cad2e5aa |
1497 | =head1 DBM FILTERS |
1498 | |
1499 | A DBM Filter is a piece of code that is be used when you I<always> |
1500 | want to make the same transformation to all keys and/or values in a |
1501 | DBM database. |
1502 | |
1503 | There are four methods associated with DBM Filters. All work identically, |
1504 | and each is used to install (or uninstall) a single DBM Filter. Each |
1505 | expects a single parameter, namely a reference to a sub. The only |
1506 | difference between them is the place that the filter is installed. |
1507 | |
1508 | To summarise: |
1509 | |
1510 | =over 5 |
1511 | |
1512 | =item B<filter_store_key> |
1513 | |
1514 | If a filter has been installed with this method, it will be invoked |
1515 | every time you write a key to a DBM database. |
1516 | |
1517 | =item B<filter_store_value> |
1518 | |
1519 | If a filter has been installed with this method, it will be invoked |
1520 | every time you write a value to a DBM database. |
1521 | |
1522 | |
1523 | =item B<filter_fetch_key> |
1524 | |
1525 | If a filter has been installed with this method, it will be invoked |
1526 | every time you read a key from a DBM database. |
1527 | |
1528 | =item B<filter_fetch_value> |
1529 | |
1530 | If a filter has been installed with this method, it will be invoked |
1531 | every time you read a value from a DBM database. |
1532 | |
1533 | =back |
1534 | |
1535 | You can use any combination of the methods, from none, to all four. |
1536 | |
1537 | All filter methods return the existing filter, if present, or C<undef> |
1538 | in not. |
1539 | |
1540 | To delete a filter pass C<undef> to it. |
1541 | |
1542 | =head2 The Filter |
1543 | |
1544 | When each filter is called by Perl, a local copy of C<$_> will contain |
1545 | the key or value to be filtered. Filtering is achieved by modifying |
1546 | the contents of C<$_>. The return code from the filter is ignored. |
1547 | |
1548 | =head2 An Example -- the NULL termination problem. |
1549 | |
1550 | Consider the following scenario. You have a DBM database |
1551 | that you need to share with a third-party C application. The C application |
1552 | assumes that I<all> keys and values are NULL terminated. Unfortunately |
1553 | when Perl writes to DBM databases it doesn't use NULL termination, so |
1554 | your Perl application will have to manage NULL termination itself. When |
1555 | you write to the database you will have to use something like this: |
1556 | |
1557 | $hash{"$key\0"} = "$value\0" ; |
1558 | |
1559 | Similarly the NULL needs to be taken into account when you are considering |
1560 | the length of existing keys/values. |
1561 | |
1562 | It would be much better if you could ignore the NULL terminations issue |
1563 | in the main application code and have a mechanism that automatically |
1564 | added the terminating NULL to all keys and values whenever you write to |
1565 | the database and have them removed when you read from the database. As I'm |
1566 | sure you have already guessed, this is a problem that DBM Filters can |
1567 | fix very easily. |
1568 | |
1569 | use strict ; |
1570 | use DB_File ; |
1571 | |
1572 | my %hash ; |
1573 | my $filename = "/tmp/filt" ; |
1574 | unlink $filename ; |
1575 | |
1576 | my $db = tie %hash, 'DB_File', $filename, O_CREAT|O_RDWR, 0666, $DB_HASH |
1577 | or die "Cannot open $filename: $!\n" ; |
1578 | |
1579 | # Install DBM Filters |
1580 | $db->filter_fetch_key ( sub { s/\0$// } ) ; |
1581 | $db->filter_store_key ( sub { $_ .= "\0" } ) ; |
1582 | $db->filter_fetch_value( sub { s/\0$// } ) ; |
1583 | $db->filter_store_value( sub { $_ .= "\0" } ) ; |
1584 | |
1585 | $hash{"abc"} = "def" ; |
1586 | my $a = $hash{"ABC"} ; |
1587 | # ... |
1588 | undef $db ; |
1589 | untie %hash ; |
1590 | |
1591 | Hopefully the contents of each of the filters should be |
1592 | self-explanatory. Both "fetch" filters remove the terminating NULL, |
1593 | and both "store" filters add a terminating NULL. |
1594 | |
1595 | |
1596 | =head2 Another Example -- Key is a C int. |
1597 | |
1598 | Here is another real-life example. By default, whenever Perl writes to |
1599 | a DBM database it always writes the key and value as strings. So when |
1600 | you use this: |
1601 | |
1602 | $hash{12345} = "soemthing" ; |
1603 | |
1604 | the key 12345 will get stored in the DBM database as the 5 byte string |
1605 | "12345". If you actually want the key to be stored in the DBM database |
1606 | as a C int, you will have to use C<pack> when writing, and C<unpack> |
1607 | when reading. |
1608 | |
1609 | Here is a DBM Filter that does it: |
1610 | |
1611 | use strict ; |
1612 | use DB_File ; |
1613 | my %hash ; |
1614 | my $filename = "/tmp/filt" ; |
1615 | unlink $filename ; |
1616 | |
1617 | |
1618 | my $db = tie %hash, 'DB_File', $filename, O_CREAT|O_RDWR, 0666, $DB_HASH |
1619 | or die "Cannot open $filename: $!\n" ; |
1620 | |
1621 | $db->filter_fetch_key ( sub { $_ = unpack("i", $_) } ) ; |
1622 | $db->filter_store_key ( sub { $_ = pack ("i", $_) } ) ; |
1623 | $hash{123} = "def" ; |
1624 | # ... |
1625 | undef $db ; |
1626 | untie %hash ; |
1627 | |
1628 | This time only two filters have been used -- we only need to manipulate |
1629 | the contents of the key, so it wasn't necessary to install any value |
1630 | filters. |
1631 | |
f6b705ef |
1632 | =head1 HINTS AND TIPS |
3b35bae3 |
1633 | |
3b35bae3 |
1634 | |
cb1a09d0 |
1635 | =head2 Locking Databases |
3b35bae3 |
1636 | |
cb1a09d0 |
1637 | Concurrent access of a read-write database by several parties requires |
1638 | them all to use some kind of locking. Here's an example of Tom's that |
1639 | uses the I<fd> method to get the file descriptor, and then a careful |
1640 | open() to give something Perl will flock() for you. Run this repeatedly |
1641 | in the background to watch the locks granted in proper order. |
3b35bae3 |
1642 | |
cb1a09d0 |
1643 | use DB_File; |
1644 | |
1645 | use strict; |
1646 | |
1647 | sub LOCK_SH { 1 } |
1648 | sub LOCK_EX { 2 } |
1649 | sub LOCK_NB { 4 } |
1650 | sub LOCK_UN { 8 } |
1651 | |
1652 | my($oldval, $fd, $db, %db, $value, $key); |
1653 | |
1654 | $key = shift || 'default'; |
1655 | $value = shift || 'magic'; |
1656 | |
1657 | $value .= " $$"; |
1658 | |
1659 | $db = tie(%db, 'DB_File', '/tmp/foo.db', O_CREAT|O_RDWR, 0644) |
1660 | || die "dbcreat /tmp/foo.db $!"; |
1661 | $fd = $db->fd; |
1662 | print "$$: db fd is $fd\n"; |
1663 | open(DB_FH, "+<&=$fd") || die "dup $!"; |
1664 | |
1665 | |
1666 | unless (flock (DB_FH, LOCK_SH | LOCK_NB)) { |
1667 | print "$$: CONTENTION; can't read during write update! |
1668 | Waiting for read lock ($!) ...."; |
1669 | unless (flock (DB_FH, LOCK_SH)) { die "flock: $!" } |
1670 | } |
1671 | print "$$: Read lock granted\n"; |
1672 | |
1673 | $oldval = $db{$key}; |
1674 | print "$$: Old value was $oldval\n"; |
1675 | flock(DB_FH, LOCK_UN); |
1676 | |
1677 | unless (flock (DB_FH, LOCK_EX | LOCK_NB)) { |
1678 | print "$$: CONTENTION; must have exclusive lock! |
1679 | Waiting for write lock ($!) ...."; |
1680 | unless (flock (DB_FH, LOCK_EX)) { die "flock: $!" } |
1681 | } |
1682 | |
1683 | print "$$: Write lock granted\n"; |
1684 | $db{$key} = $value; |
610ab055 |
1685 | $db->sync; # to flush |
cb1a09d0 |
1686 | sleep 10; |
1687 | |
1688 | flock(DB_FH, LOCK_UN); |
88108326 |
1689 | undef $db; |
cb1a09d0 |
1690 | untie %db; |
1691 | close(DB_FH); |
1692 | print "$$: Updated db to $key=$value\n"; |
1693 | |
68dc0745 |
1694 | =head2 Sharing Databases With C Applications |
f6b705ef |
1695 | |
1696 | There is no technical reason why a Berkeley DB database cannot be |
1697 | shared by both a Perl and a C application. |
1698 | |
1699 | The vast majority of problems that are reported in this area boil down |
1700 | to the fact that C strings are NULL terminated, whilst Perl strings are |
cad2e5aa |
1701 | not. See L<DBM FILTERS> for a generic way to work around this problem. |
f6b705ef |
1702 | |
1703 | Here is a real example. Netscape 2.0 keeps a record of the locations you |
1704 | visit along with the time you last visited them in a DB_HASH database. |
1705 | This is usually stored in the file F<~/.netscape/history.db>. The key |
1706 | field in the database is the location string and the value field is the |
1707 | time the location was last visited stored as a 4 byte binary value. |
1708 | |
1709 | If you haven't already guessed, the location string is stored with a |
1710 | terminating NULL. This means you need to be careful when accessing the |
1711 | database. |
1712 | |
1713 | Here is a snippet of code that is loosely based on Tom Christiansen's |
1714 | I<ggh> script (available from your nearest CPAN archive in |
1715 | F<authors/id/TOMC/scripts/nshist.gz>). |
1716 | |
610ab055 |
1717 | use strict ; |
f6b705ef |
1718 | use DB_File ; |
1719 | use Fcntl ; |
f6b705ef |
1720 | |
610ab055 |
1721 | use vars qw( $dotdir $HISTORY %hist_db $href $binary_time $date ) ; |
f6b705ef |
1722 | $dotdir = $ENV{HOME} || $ENV{LOGNAME}; |
1723 | |
1724 | $HISTORY = "$dotdir/.netscape/history.db"; |
1725 | |
1726 | tie %hist_db, 'DB_File', $HISTORY |
1727 | or die "Cannot open $HISTORY: $!\n" ;; |
1728 | |
1729 | # Dump the complete database |
1730 | while ( ($href, $binary_time) = each %hist_db ) { |
1731 | |
1732 | # remove the terminating NULL |
1733 | $href =~ s/\x00$// ; |
1734 | |
1735 | # convert the binary time into a user friendly string |
1736 | $date = localtime unpack("V", $binary_time); |
1737 | print "$date $href\n" ; |
1738 | } |
1739 | |
1740 | # check for the existence of a specific key |
1741 | # remember to add the NULL |
1742 | if ( $binary_time = $hist_db{"http://mox.perl.com/\x00"} ) { |
1743 | $date = localtime unpack("V", $binary_time) ; |
1744 | print "Last visited mox.perl.com on $date\n" ; |
1745 | } |
1746 | else { |
1747 | print "Never visited mox.perl.com\n" |
1748 | } |
1749 | |
1750 | untie %hist_db ; |
1751 | |
68dc0745 |
1752 | =head2 The untie() Gotcha |
778183f3 |
1753 | |
7a2e2cd6 |
1754 | If you make use of the Berkeley DB API, it is I<very> strongly |
68dc0745 |
1755 | recommended that you read L<perltie/The untie Gotcha>. |
778183f3 |
1756 | |
1757 | Even if you don't currently make use of the API interface, it is still |
1758 | worth reading it. |
1759 | |
1760 | Here is an example which illustrates the problem from a B<DB_File> |
1761 | perspective: |
1762 | |
1763 | use DB_File ; |
1764 | use Fcntl ; |
1765 | |
1766 | my %x ; |
1767 | my $X ; |
1768 | |
1769 | $X = tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_TRUNC |
1770 | or die "Cannot tie first time: $!" ; |
1771 | |
1772 | $x{123} = 456 ; |
1773 | |
1774 | untie %x ; |
1775 | |
1776 | tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_CREAT |
1777 | or die "Cannot tie second time: $!" ; |
1778 | |
1779 | untie %x ; |
1780 | |
1781 | When run, the script will produce this error message: |
1782 | |
1783 | Cannot tie second time: Invalid argument at bad.file line 14. |
1784 | |
1785 | Although the error message above refers to the second tie() statement |
1786 | in the script, the source of the problem is really with the untie() |
1787 | statement that precedes it. |
1788 | |
1789 | Having read L<perltie> you will probably have already guessed that the |
1790 | error is caused by the extra copy of the tied object stored in C<$X>. |
1791 | If you haven't, then the problem boils down to the fact that the |
1792 | B<DB_File> destructor, DESTROY, will not be called until I<all> |
1793 | references to the tied object are destroyed. Both the tied variable, |
1794 | C<%x>, and C<$X> above hold a reference to the object. The call to |
1795 | untie() will destroy the first, but C<$X> still holds a valid |
1796 | reference, so the destructor will not get called and the database file |
1797 | F<tst.fil> will remain open. The fact that Berkeley DB then reports the |
1798 | attempt to open a database that is alreday open via the catch-all |
1799 | "Invalid argument" doesn't help. |
1800 | |
1801 | If you run the script with the C<-w> flag the error message becomes: |
1802 | |
1803 | untie attempted while 1 inner references still exist at bad.file line 12. |
1804 | Cannot tie second time: Invalid argument at bad.file line 14. |
1805 | |
1806 | which pinpoints the real problem. Finally the script can now be |
1807 | modified to fix the original problem by destroying the API object |
1808 | before the untie: |
1809 | |
1810 | ... |
1811 | $x{123} = 456 ; |
1812 | |
1813 | undef $X ; |
1814 | untie %x ; |
1815 | |
1816 | $X = tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_CREAT |
1817 | ... |
1818 | |
f6b705ef |
1819 | |
1820 | =head1 COMMON QUESTIONS |
1821 | |
1822 | =head2 Why is there Perl source in my database? |
1823 | |
1824 | If you look at the contents of a database file created by DB_File, |
1825 | there can sometimes be part of a Perl script included in it. |
1826 | |
1827 | This happens because Berkeley DB uses dynamic memory to allocate |
1828 | buffers which will subsequently be written to the database file. Being |
1829 | dynamic, the memory could have been used for anything before DB |
1830 | malloced it. As Berkeley DB doesn't clear the memory once it has been |
1831 | allocated, the unused portions will contain random junk. In the case |
1832 | where a Perl script gets written to the database, the random junk will |
1833 | correspond to an area of dynamic memory that happened to be used during |
1834 | the compilation of the script. |
1835 | |
1836 | Unless you don't like the possibility of there being part of your Perl |
1837 | scripts embedded in a database file, this is nothing to worry about. |
1838 | |
1839 | =head2 How do I store complex data structures with DB_File? |
1840 | |
1841 | Although B<DB_File> cannot do this directly, there is a module which |
1842 | can layer transparently over B<DB_File> to accomplish this feat. |
1843 | |
1844 | Check out the MLDBM module, available on CPAN in the directory |
1845 | F<modules/by-module/MLDBM>. |
1846 | |
1847 | =head2 What does "Invalid Argument" mean? |
1848 | |
1849 | You will get this error message when one of the parameters in the |
1850 | C<tie> call is wrong. Unfortunately there are quite a few parameters to |
1851 | get wrong, so it can be difficult to figure out which one it is. |
1852 | |
1853 | Here are a couple of possibilities: |
1854 | |
1855 | =over 5 |
1856 | |
1857 | =item 1. |
1858 | |
610ab055 |
1859 | Attempting to reopen a database without closing it. |
f6b705ef |
1860 | |
1861 | =item 2. |
1862 | |
1863 | Using the O_WRONLY flag. |
1864 | |
1865 | =back |
1866 | |
1867 | =head2 What does "Bareword 'DB_File' not allowed" mean? |
1868 | |
1869 | You will encounter this particular error message when you have the |
1870 | C<strict 'subs'> pragma (or the full strict pragma) in your script. |
1871 | Consider this script: |
1872 | |
1873 | use strict ; |
1874 | use DB_File ; |
1875 | use vars qw(%x) ; |
1876 | tie %x, DB_File, "filename" ; |
1877 | |
1878 | Running it produces the error in question: |
1879 | |
1880 | Bareword "DB_File" not allowed while "strict subs" in use |
1881 | |
1882 | To get around the error, place the word C<DB_File> in either single or |
1883 | double quotes, like this: |
1884 | |
1885 | tie %x, "DB_File", "filename" ; |
1886 | |
1887 | Although it might seem like a real pain, it is really worth the effort |
1888 | of having a C<use strict> in all your scripts. |
1889 | |
cad2e5aa |
1890 | =head1 REFERENCES |
1891 | |
1892 | Articles that are either about B<DB_File> or make use of it. |
1893 | |
1894 | =over 5 |
1895 | |
1896 | =item 1. |
1897 | |
1898 | I<Full-Text Searching in Perl>, Tim Kientzle (tkientzle@ddj.com), |
1899 | Dr. Dobb's Journal, Issue 295, January 1999, pp 34-41 |
1900 | |
1901 | =back |
1902 | |
cb1a09d0 |
1903 | =head1 HISTORY |
1904 | |
1f70e1ea |
1905 | Moved to the Changes file. |
610ab055 |
1906 | |
1f70e1ea |
1907 | =head1 BUGS |
05475680 |
1908 | |
1f70e1ea |
1909 | Some older versions of Berkeley DB had problems with fixed length |
1910 | records using the RECNO file format. This problem has been fixed since |
1911 | version 1.85 of Berkeley DB. |
e858de61 |
1912 | |
1f70e1ea |
1913 | I am sure there are bugs in the code. If you do find any, or can |
1914 | suggest any enhancements, I would welcome your comments. |
a6ed719b |
1915 | |
1f70e1ea |
1916 | =head1 AVAILABILITY |
a6ed719b |
1917 | |
1f70e1ea |
1918 | B<DB_File> comes with the standard Perl source distribution. Look in |
1919 | the directory F<ext/DB_File>. Given the amount of time between releases |
1920 | of Perl the version that ships with Perl is quite likely to be out of |
1921 | date, so the most recent version can always be found on CPAN (see |
1922 | L<perlmod/CPAN> for details), in the directory |
1923 | F<modules/by-module/DB_File>. |
a6ed719b |
1924 | |
1f70e1ea |
1925 | This version of B<DB_File> will work with either version 1.x or 2.x of |
1926 | Berkeley DB, but is limited to the functionality provided by version 1. |
a6ed719b |
1927 | |
cad2e5aa |
1928 | The official web site for Berkeley DB is F<http://www.sleepycat.com>. |
1929 | Both versions 1 and 2 of Berkeley DB are available there. |
93af7a87 |
1930 | |
1f70e1ea |
1931 | Alternatively, Berkeley DB version 1 is available at your nearest CPAN |
1932 | archive in F<src/misc/db.1.85.tar.gz>. |
e858de61 |
1933 | |
1f70e1ea |
1934 | If you are running IRIX, then get Berkeley DB version 1 from |
1935 | F<http://reality.sgi.com/ariel>. It has the patches necessary to |
1936 | compile properly on IRIX 5.3. |
610ab055 |
1937 | |
1f70e1ea |
1938 | =head1 COPYRIGHT |
3b35bae3 |
1939 | |
cad2e5aa |
1940 | Copyright (c) 1995-1999 Paul Marquess. All rights reserved. This program |
a9fd575d |
1941 | is free software; you can redistribute it and/or modify it under the |
1942 | same terms as Perl itself. |
3b35bae3 |
1943 | |
1f70e1ea |
1944 | Although B<DB_File> is covered by the Perl license, the library it |
1945 | makes use of, namely Berkeley DB, is not. Berkeley DB has its own |
1946 | copyright and its own license. Please take the time to read it. |
3b35bae3 |
1947 | |
a9fd575d |
1948 | Here are are few words taken from the Berkeley DB FAQ (at |
1949 | http://www.sleepycat.com) regarding the license: |
68dc0745 |
1950 | |
a9fd575d |
1951 | Do I have to license DB to use it in Perl scripts? |
3b35bae3 |
1952 | |
a9fd575d |
1953 | No. The Berkeley DB license requires that software that uses |
1954 | Berkeley DB be freely redistributable. In the case of Perl, that |
1955 | software is Perl, and not your scripts. Any Perl scripts that you |
1956 | write are your property, including scripts that make use of |
1957 | Berkeley DB. Neither the Perl license nor the Berkeley DB license |
1958 | place any restriction on what you may do with them. |
88108326 |
1959 | |
1f70e1ea |
1960 | If you are in any doubt about the license situation, contact either the |
1961 | Berkeley DB authors or the author of DB_File. See L<"AUTHOR"> for details. |
a0b8c8c1 |
1962 | |
1963 | |
3b35bae3 |
1964 | =head1 SEE ALSO |
1965 | |
9fe6733a |
1966 | L<perl(1)>, L<dbopen(3)>, L<hash(3)>, L<recno(3)>, L<btree(3)>, |
1967 | L<dbmfilter> |
3b35bae3 |
1968 | |
3b35bae3 |
1969 | =head1 AUTHOR |
1970 | |
8e07c86e |
1971 | The DB_File interface was written by Paul Marquess |
6ca2e664 |
1972 | E<lt>Paul.Marquess@btinternet.comE<gt>. |
d3ef3b8a |
1973 | Questions about the DB system itself may be addressed to |
1974 | E<lt>db@sleepycat.com<gt>. |
3b35bae3 |
1975 | |
1976 | =cut |