Commit | Line | Data |
---|---|---|
a0d0e21e LW |
1 | # DB_File.pm -- Perl 5 interface to Berkeley DB |
2 | # | |
6d02d21f | 3 | # written by Paul Marquess (pmqs@cpan.org) |
dcdb9d1f RGS |
4 | # last modified 12th March 2005 |
5 | # version 1.811 | |
36477c24 | 6 | # |
dcdb9d1f | 7 | # Copyright (c) 1995-2005 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 AD |
11 | |
12 | package DB_File::HASHINFO ; | |
785da04d | 13 | |
e5021521 | 14 | require 5.00404; |
610ab055 | 15 | |
3245f058 | 16 | use warnings; |
785da04d | 17 | use strict; |
8e07c86e | 18 | use Carp; |
88108326 | 19 | require Tie::Hash; |
20 | @DB_File::HASHINFO::ISA = qw(Tie::Hash); | |
8e07c86e | 21 | |
88108326 | 22 | sub new |
8e07c86e | 23 | { |
88108326 | 24 | my $pkg = shift ; |
25 | my %x ; | |
26 | tie %x, $pkg ; | |
27 | bless \%x, $pkg ; | |
8e07c86e AD |
28 | } |
29 | ||
610ab055 | 30 | |
88108326 | 31 | sub TIEHASH |
32 | { | |
33 | my $pkg = shift ; | |
34 | ||
efc79c7d PM |
35 | bless { VALID => { |
36 | bsize => 1, | |
37 | ffactor => 1, | |
38 | nelem => 1, | |
39 | cachesize => 1, | |
40 | hash => 2, | |
41 | lorder => 1, | |
36477c24 | 42 | }, |
43 | GOT => {} | |
44 | }, $pkg ; | |
88108326 | 45 | } |
8e07c86e | 46 | |
610ab055 | 47 | |
8e07c86e AD |
48 | sub FETCH |
49 | { | |
88108326 | 50 | my $self = shift ; |
51 | my $key = shift ; | |
8e07c86e | 52 | |
36477c24 | 53 | return $self->{GOT}{$key} if exists $self->{VALID}{$key} ; |
88108326 | 54 | |
55 | my $pkg = ref $self ; | |
56 | croak "${pkg}::FETCH - Unknown element '$key'" ; | |
8e07c86e AD |
57 | } |
58 | ||
59 | ||
60 | sub STORE | |
61 | { | |
88108326 | 62 | my $self = shift ; |
63 | my $key = shift ; | |
64 | my $value = shift ; | |
65 | ||
efc79c7d PM |
66 | my $type = $self->{VALID}{$key}; |
67 | ||
68 | if ( $type ) | |
8e07c86e | 69 | { |
efc79c7d PM |
70 | croak "Key '$key' not associated with a code reference" |
71 | if $type == 2 && !ref $value && ref $value ne 'CODE'; | |
36477c24 | 72 | $self->{GOT}{$key} = $value ; |
8e07c86e AD |
73 | return ; |
74 | } | |
75 | ||
88108326 | 76 | my $pkg = ref $self ; |
77 | croak "${pkg}::STORE - Unknown element '$key'" ; | |
8e07c86e AD |
78 | } |
79 | ||
80 | sub DELETE | |
81 | { | |
88108326 | 82 | my $self = shift ; |
83 | my $key = shift ; | |
84 | ||
36477c24 | 85 | if ( exists $self->{VALID}{$key} ) |
8e07c86e | 86 | { |
36477c24 | 87 | delete $self->{GOT}{$key} ; |
8e07c86e AD |
88 | return ; |
89 | } | |
90 | ||
88108326 | 91 | my $pkg = ref $self ; |
92 | croak "DB_File::HASHINFO::DELETE - Unknown element '$key'" ; | |
8e07c86e AD |
93 | } |
94 | ||
88108326 | 95 | sub EXISTS |
8e07c86e | 96 | { |
88108326 | 97 | my $self = shift ; |
98 | my $key = shift ; | |
8e07c86e | 99 | |
36477c24 | 100 | exists $self->{VALID}{$key} ; |
8e07c86e AD |
101 | } |
102 | ||
88108326 | 103 | sub NotHere |
8e07c86e | 104 | { |
18d2dc8c | 105 | my $self = shift ; |
88108326 | 106 | my $method = shift ; |
8e07c86e | 107 | |
18d2dc8c | 108 | croak ref($self) . " does not define the method ${method}" ; |
8e07c86e AD |
109 | } |
110 | ||
18d2dc8c PM |
111 | sub FIRSTKEY { my $self = shift ; $self->NotHere("FIRSTKEY") } |
112 | sub NEXTKEY { my $self = shift ; $self->NotHere("NEXTKEY") } | |
113 | sub CLEAR { my $self = shift ; $self->NotHere("CLEAR") } | |
8e07c86e AD |
114 | |
115 | package DB_File::RECNOINFO ; | |
785da04d | 116 | |
3245f058 | 117 | use warnings; |
88108326 | 118 | use strict ; |
119 | ||
045291aa | 120 | @DB_File::RECNOINFO::ISA = qw(DB_File::HASHINFO) ; |
8e07c86e AD |
121 | |
122 | sub TIEHASH | |
123 | { | |
88108326 | 124 | my $pkg = shift ; |
125 | ||
36477c24 | 126 | bless { VALID => { map {$_, 1} |
127 | qw( bval cachesize psize flags lorder reclen bfname ) | |
128 | }, | |
129 | GOT => {}, | |
130 | }, $pkg ; | |
8e07c86e AD |
131 | } |
132 | ||
88108326 | 133 | package DB_File::BTREEINFO ; |
8e07c86e | 134 | |
3245f058 | 135 | use warnings; |
88108326 | 136 | use strict ; |
8e07c86e | 137 | |
88108326 | 138 | @DB_File::BTREEINFO::ISA = qw(DB_File::HASHINFO) ; |
8e07c86e | 139 | |
88108326 | 140 | sub TIEHASH |
8e07c86e | 141 | { |
88108326 | 142 | my $pkg = shift ; |
143 | ||
efc79c7d PM |
144 | bless { VALID => { |
145 | flags => 1, | |
146 | cachesize => 1, | |
147 | maxkeypage => 1, | |
148 | minkeypage => 1, | |
149 | psize => 1, | |
150 | compare => 2, | |
151 | prefix => 2, | |
152 | lorder => 1, | |
36477c24 | 153 | }, |
154 | GOT => {}, | |
155 | }, $pkg ; | |
8e07c86e AD |
156 | } |
157 | ||
158 | ||
8e07c86e | 159 | package DB_File ; |
785da04d | 160 | |
3245f058 | 161 | use warnings; |
785da04d | 162 | use strict; |
07200f1b | 163 | our ($VERSION, @ISA, @EXPORT, $AUTOLOAD, $DB_BTREE, $DB_HASH, $DB_RECNO); |
d85a743d | 164 | our ($db_version, $use_XSLoader, $splice_end_array); |
8e07c86e AD |
165 | use Carp; |
166 | ||
785da04d | 167 | |
dcdb9d1f | 168 | $VERSION = "1.811" ; |
d85a743d PM |
169 | |
170 | { | |
171 | local $SIG{__WARN__} = sub {$splice_end_array = "@_";}; | |
172 | my @a =(1); splice(@a, 3); | |
173 | $splice_end_array = | |
174 | ($splice_end_array =~ /^splice\(\) offset past end of array at /); | |
175 | } | |
8e07c86e AD |
176 | |
177 | #typedef enum { DB_BTREE, DB_HASH, DB_RECNO } DBTYPE; | |
88108326 | 178 | $DB_BTREE = new DB_File::BTREEINFO ; |
179 | $DB_HASH = new DB_File::HASHINFO ; | |
180 | $DB_RECNO = new DB_File::RECNOINFO ; | |
8e07c86e | 181 | |
785da04d | 182 | require Tie::Hash; |
8e07c86e AD |
183 | require Exporter; |
184 | use AutoLoader; | |
b90e71be GS |
185 | BEGIN { |
186 | $use_XSLoader = 1 ; | |
e5021521 | 187 | { local $SIG{__DIE__} ; eval { require XSLoader } ; } |
b90e71be GS |
188 | |
189 | if ($@) { | |
190 | $use_XSLoader = 0 ; | |
191 | require DynaLoader; | |
192 | @ISA = qw(DynaLoader); | |
193 | } | |
194 | } | |
195 | ||
196 | push @ISA, qw(Tie::Hash Exporter); | |
8e07c86e AD |
197 | @EXPORT = qw( |
198 | $DB_BTREE $DB_HASH $DB_RECNO | |
88108326 | 199 | |
8e07c86e AD |
200 | BTREEMAGIC |
201 | BTREEVERSION | |
202 | DB_LOCK | |
203 | DB_SHMEM | |
204 | DB_TXN | |
205 | HASHMAGIC | |
206 | HASHVERSION | |
207 | MAX_PAGE_NUMBER | |
208 | MAX_PAGE_OFFSET | |
209 | MAX_REC_NUMBER | |
210 | RET_ERROR | |
211 | RET_SPECIAL | |
212 | RET_SUCCESS | |
213 | R_CURSOR | |
214 | R_DUP | |
215 | R_FIRST | |
216 | R_FIXEDLEN | |
217 | R_IAFTER | |
218 | R_IBEFORE | |
219 | R_LAST | |
220 | R_NEXT | |
221 | R_NOKEY | |
222 | R_NOOVERWRITE | |
223 | R_PREV | |
224 | R_RECNOSYNC | |
225 | R_SETCURSOR | |
226 | R_SNAPSHOT | |
227 | __R_UNUSED | |
88108326 | 228 | |
045291aa | 229 | ); |
8e07c86e AD |
230 | |
231 | sub AUTOLOAD { | |
785da04d | 232 | my($constname); |
8e07c86e | 233 | ($constname = $AUTOLOAD) =~ s/.*:://; |
07200f1b PM |
234 | my ($error, $val) = constant($constname); |
235 | Carp::croak $error if $error; | |
57c77851 JS |
236 | no strict 'refs'; |
237 | *{$AUTOLOAD} = sub { $val }; | |
238 | goto &{$AUTOLOAD}; | |
07200f1b | 239 | } |
8e07c86e | 240 | |
f6b705ef | 241 | |
a6ed719b | 242 | eval { |
1f70e1ea PM |
243 | # Make all Fcntl O_XXX constants available for importing |
244 | require Fcntl; | |
245 | my @O = grep /^O_/, @Fcntl::EXPORT; | |
246 | Fcntl->import(@O); # first we import what we want to export | |
247 | push(@EXPORT, @O); | |
a6ed719b | 248 | }; |
f6b705ef | 249 | |
b90e71be GS |
250 | if ($use_XSLoader) |
251 | { XSLoader::load("DB_File", $VERSION)} | |
252 | else | |
253 | { bootstrap DB_File $VERSION } | |
8e07c86e AD |
254 | |
255 | # Preloaded methods go here. Autoload methods go after __END__, and are | |
256 | # processed by the autosplit program. | |
257 | ||
05475680 | 258 | sub tie_hash_or_array |
610ab055 PM |
259 | { |
260 | my (@arg) = @_ ; | |
05475680 | 261 | my $tieHASH = ( (caller(1))[3] =~ /TIEHASH/ ) ; |
610ab055 PM |
262 | |
263 | $arg[4] = tied %{ $arg[4] } | |
264 | if @arg >= 5 && ref $arg[4] && $arg[4] =~ /=HASH/ && tied %{ $arg[4] } ; | |
265 | ||
efc79c7d PM |
266 | $arg[2] = O_CREAT()|O_RDWR() if @arg >=3 && ! defined $arg[2]; |
267 | $arg[3] = 0666 if @arg >=4 && ! defined $arg[3]; | |
268 | ||
9c095db2 PM |
269 | # make recno in Berkeley DB version 2 (or better) work like |
270 | # recno in version 1. | |
1f70e1ea PM |
271 | if ($db_version > 1 and defined $arg[4] and $arg[4] =~ /RECNO/ and |
272 | $arg[1] and ! -e $arg[1]) { | |
273 | open(FH, ">$arg[1]") or return undef ; | |
274 | close FH ; | |
275 | chmod $arg[3] ? $arg[3] : 0666 , $arg[1] ; | |
276 | } | |
277 | ||
05475680 | 278 | DoTie_($tieHASH, @arg) ; |
610ab055 PM |
279 | } |
280 | ||
05475680 PM |
281 | sub TIEHASH |
282 | { | |
283 | tie_hash_or_array(@_) ; | |
284 | } | |
285 | ||
286 | sub TIEARRAY | |
287 | { | |
288 | tie_hash_or_array(@_) ; | |
289 | } | |
88108326 | 290 | |
045291aa PM |
291 | sub CLEAR |
292 | { | |
1f70e1ea | 293 | my $self = shift; |
3245f058 | 294 | my $key = 0 ; |
1f70e1ea PM |
295 | my $value = "" ; |
296 | my $status = $self->seq($key, $value, R_FIRST()); | |
297 | my @keys; | |
298 | ||
299 | while ($status == 0) { | |
300 | push @keys, $key; | |
301 | $status = $self->seq($key, $value, R_NEXT()); | |
302 | } | |
303 | foreach $key (reverse @keys) { | |
304 | my $s = $self->del($key); | |
305 | } | |
306 | } | |
307 | ||
045291aa PM |
308 | sub EXTEND { } |
309 | ||
310 | sub STORESIZE | |
311 | { | |
312 | my $self = shift; | |
313 | my $length = shift ; | |
314 | my $current_length = $self->length() ; | |
315 | ||
316 | if ($length < $current_length) { | |
317 | my $key ; | |
318 | for ($key = $current_length - 1 ; $key >= $length ; -- $key) | |
319 | { $self->del($key) } | |
320 | } | |
a9fd575d PM |
321 | elsif ($length > $current_length) { |
322 | $self->put($length-1, "") ; | |
323 | } | |
045291aa PM |
324 | } |
325 | ||
c5da4faf PM |
326 | |
327 | sub SPLICE | |
328 | { | |
329 | my $self = shift; | |
330 | my $offset = shift; | |
331 | if (not defined $offset) { | |
d85a743d | 332 | warnings::warnif('uninitialized', 'Use of uninitialized value in splice'); |
c5da4faf PM |
333 | $offset = 0; |
334 | } | |
335 | ||
336 | my $length = @_ ? shift : 0; | |
337 | # Carping about definedness comes _after_ the OFFSET sanity check. | |
338 | # This is so we get the same error messages as Perl's splice(). | |
339 | # | |
340 | ||
341 | my @list = @_; | |
342 | ||
343 | my $size = $self->FETCHSIZE(); | |
344 | ||
345 | # 'If OFFSET is negative then it start that far from the end of | |
346 | # the array.' | |
347 | # | |
348 | if ($offset < 0) { | |
349 | my $new_offset = $size + $offset; | |
350 | if ($new_offset < 0) { | |
351 | die "Modification of non-creatable array value attempted, " | |
352 | . "subscript $offset"; | |
353 | } | |
354 | $offset = $new_offset; | |
355 | } | |
356 | ||
c5da4faf | 357 | if (not defined $length) { |
d85a743d | 358 | warnings::warnif('uninitialized', 'Use of uninitialized value in splice'); |
c5da4faf PM |
359 | $length = 0; |
360 | } | |
361 | ||
d85a743d PM |
362 | if ($offset > $size) { |
363 | $offset = $size; | |
364 | warnings::warnif('misc', 'splice() offset past end of array') | |
365 | if $splice_end_array; | |
366 | } | |
367 | ||
c5da4faf PM |
368 | # 'If LENGTH is omitted, removes everything from OFFSET onward.' |
369 | if (not defined $length) { | |
370 | $length = $size - $offset; | |
371 | } | |
372 | ||
373 | # 'If LENGTH is negative, leave that many elements off the end of | |
374 | # the array.' | |
375 | # | |
376 | if ($length < 0) { | |
377 | $length = $size - $offset + $length; | |
378 | ||
379 | if ($length < 0) { | |
380 | # The user must have specified a length bigger than the | |
381 | # length of the array passed in. But perl's splice() | |
382 | # doesn't catch this, it just behaves as for length=0. | |
383 | # | |
384 | $length = 0; | |
385 | } | |
386 | } | |
387 | ||
388 | if ($length > $size - $offset) { | |
389 | $length = $size - $offset; | |
390 | } | |
391 | ||
392 | # $num_elems holds the current number of elements in the database. | |
393 | my $num_elems = $size; | |
394 | ||
395 | # 'Removes the elements designated by OFFSET and LENGTH from an | |
396 | # array,'... | |
397 | # | |
398 | my @removed = (); | |
399 | foreach (0 .. $length - 1) { | |
400 | my $old; | |
401 | my $status = $self->get($offset, $old); | |
402 | if ($status != 0) { | |
403 | my $msg = "error from Berkeley DB on get($offset, \$old)"; | |
404 | if ($status == 1) { | |
405 | $msg .= ' (no such element?)'; | |
406 | } | |
407 | else { | |
408 | $msg .= ": error status $status"; | |
409 | if (defined $! and $! ne '') { | |
410 | $msg .= ", message $!"; | |
411 | } | |
412 | } | |
413 | die $msg; | |
414 | } | |
415 | push @removed, $old; | |
416 | ||
417 | $status = $self->del($offset); | |
418 | if ($status != 0) { | |
419 | my $msg = "error from Berkeley DB on del($offset)"; | |
420 | if ($status == 1) { | |
421 | $msg .= ' (no such element?)'; | |
422 | } | |
423 | else { | |
424 | $msg .= ": error status $status"; | |
425 | if (defined $! and $! ne '') { | |
426 | $msg .= ", message $!"; | |
427 | } | |
428 | } | |
429 | die $msg; | |
430 | } | |
431 | ||
432 | -- $num_elems; | |
433 | } | |
434 | ||
435 | # ...'and replaces them with the elements of LIST, if any.' | |
436 | my $pos = $offset; | |
437 | while (defined (my $elem = shift @list)) { | |
438 | my $old_pos = $pos; | |
439 | my $status; | |
440 | if ($pos >= $num_elems) { | |
441 | $status = $self->put($pos, $elem); | |
442 | } | |
443 | else { | |
444 | $status = $self->put($pos, $elem, $self->R_IBEFORE); | |
445 | } | |
446 | ||
447 | if ($status != 0) { | |
448 | my $msg = "error from Berkeley DB on put($pos, $elem, ...)"; | |
449 | if ($status == 1) { | |
450 | $msg .= ' (no such element?)'; | |
451 | } | |
452 | else { | |
453 | $msg .= ", error status $status"; | |
454 | if (defined $! and $! ne '') { | |
455 | $msg .= ", message $!"; | |
456 | } | |
457 | } | |
458 | die $msg; | |
459 | } | |
460 | ||
461 | die "pos unexpectedly changed from $old_pos to $pos with R_IBEFORE" | |
462 | if $old_pos != $pos; | |
463 | ||
464 | ++ $pos; | |
465 | ++ $num_elems; | |
466 | } | |
467 | ||
468 | if (wantarray) { | |
469 | # 'In list context, returns the elements removed from the | |
470 | # array.' | |
471 | # | |
472 | return @removed; | |
473 | } | |
474 | elsif (defined wantarray and not wantarray) { | |
475 | # 'In scalar context, returns the last element removed, or | |
476 | # undef if no elements are removed.' | |
477 | # | |
478 | if (@removed) { | |
479 | my $last = pop @removed; | |
480 | return "$last"; | |
481 | } | |
482 | else { | |
483 | return undef; | |
484 | } | |
485 | } | |
486 | elsif (not defined wantarray) { | |
487 | # Void context | |
488 | } | |
489 | else { die } | |
490 | } | |
491 | sub ::DB_File::splice { &SPLICE } | |
492 | ||
6ca2e664 PM |
493 | sub find_dup |
494 | { | |
495 | croak "Usage: \$db->find_dup(key,value)\n" | |
496 | unless @_ == 3 ; | |
497 | ||
498 | my $db = shift ; | |
499 | my ($origkey, $value_wanted) = @_ ; | |
500 | my ($key, $value) = ($origkey, 0); | |
501 | my ($status) = 0 ; | |
502 | ||
503 | for ($status = $db->seq($key, $value, R_CURSOR() ) ; | |
504 | $status == 0 ; | |
505 | $status = $db->seq($key, $value, R_NEXT() ) ) { | |
506 | ||
507 | return 0 if $key eq $origkey and $value eq $value_wanted ; | |
508 | } | |
509 | ||
510 | return $status ; | |
511 | } | |
512 | ||
513 | sub del_dup | |
514 | { | |
515 | croak "Usage: \$db->del_dup(key,value)\n" | |
516 | unless @_ == 3 ; | |
517 | ||
518 | my $db = shift ; | |
519 | my ($key, $value) = @_ ; | |
520 | my ($status) = $db->find_dup($key, $value) ; | |
521 | return $status if $status != 0 ; | |
522 | ||
523 | $status = $db->del($key, R_CURSOR() ) ; | |
524 | return $status ; | |
525 | } | |
526 | ||
88108326 | 527 | sub get_dup |
528 | { | |
529 | croak "Usage: \$db->get_dup(key [,flag])\n" | |
530 | unless @_ == 2 or @_ == 3 ; | |
531 | ||
532 | my $db = shift ; | |
533 | my $key = shift ; | |
534 | my $flag = shift ; | |
f6b705ef | 535 | my $value = 0 ; |
88108326 | 536 | my $origkey = $key ; |
537 | my $wantarray = wantarray ; | |
f6b705ef | 538 | my %values = () ; |
88108326 | 539 | my @values = () ; |
540 | my $counter = 0 ; | |
f6b705ef | 541 | my $status = 0 ; |
88108326 | 542 | |
f6b705ef | 543 | # iterate through the database until either EOF ($status == 0) |
544 | # or a different key is encountered ($key ne $origkey). | |
545 | for ($status = $db->seq($key, $value, R_CURSOR()) ; | |
546 | $status == 0 and $key eq $origkey ; | |
547 | $status = $db->seq($key, $value, R_NEXT()) ) { | |
88108326 | 548 | |
f6b705ef | 549 | # save the value or count number of matches |
550 | if ($wantarray) { | |
551 | if ($flag) | |
552 | { ++ $values{$value} } | |
553 | else | |
554 | { push (@values, $value) } | |
555 | } | |
556 | else | |
557 | { ++ $counter } | |
88108326 | 558 | |
88108326 | 559 | } |
560 | ||
f6b705ef | 561 | return ($wantarray ? ($flag ? %values : @values) : $counter) ; |
88108326 | 562 | } |
563 | ||
564 | ||
8e07c86e AD |
565 | 1; |
566 | __END__ | |
567 | ||
3b35bae3 AD |
568 | =head1 NAME |
569 | ||
1f70e1ea | 570 | DB_File - Perl5 access to Berkeley DB version 1.x |
3b35bae3 AD |
571 | |
572 | =head1 SYNOPSIS | |
573 | ||
bbc7dcd2 MS |
574 | use DB_File; |
575 | ||
88108326 | 576 | [$X =] tie %hash, 'DB_File', [$filename, $flags, $mode, $DB_HASH] ; |
577 | [$X =] tie %hash, 'DB_File', $filename, $flags, $mode, $DB_BTREE ; | |
578 | [$X =] tie @array, 'DB_File', $filename, $flags, $mode, $DB_RECNO ; | |
760ac839 | 579 | |
3b35bae3 AD |
580 | $status = $X->del($key [, $flags]) ; |
581 | $status = $X->put($key, $value [, $flags]) ; | |
582 | $status = $X->get($key, $value [, $flags]) ; | |
760ac839 | 583 | $status = $X->seq($key, $value, $flags) ; |
3b35bae3 AD |
584 | $status = $X->sync([$flags]) ; |
585 | $status = $X->fd ; | |
760ac839 | 586 | |
f6b705ef | 587 | # BTREE only |
88108326 | 588 | $count = $X->get_dup($key) ; |
589 | @list = $X->get_dup($key) ; | |
590 | %list = $X->get_dup($key, 1) ; | |
6ca2e664 PM |
591 | $status = $X->find_dup($key, $value) ; |
592 | $status = $X->del_dup($key, $value) ; | |
88108326 | 593 | |
f6b705ef | 594 | # RECNO only |
595 | $a = $X->length; | |
596 | $a = $X->pop ; | |
597 | $X->push(list); | |
598 | $a = $X->shift; | |
599 | $X->unshift(list); | |
c5da4faf | 600 | @r = $X->splice(offset, length, elements); |
f6b705ef | 601 | |
cad2e5aa JH |
602 | # DBM Filters |
603 | $old_filter = $db->filter_store_key ( sub { ... } ) ; | |
604 | $old_filter = $db->filter_store_value( sub { ... } ) ; | |
605 | $old_filter = $db->filter_fetch_key ( sub { ... } ) ; | |
606 | $old_filter = $db->filter_fetch_value( sub { ... } ) ; | |
607 | ||
3b35bae3 AD |
608 | untie %hash ; |
609 | untie @array ; | |
610 | ||
611 | =head1 DESCRIPTION | |
612 | ||
8e07c86e | 613 | B<DB_File> is a module which allows Perl programs to make use of the |
1f70e1ea | 614 | facilities provided by Berkeley DB version 1.x (if you have a newer |
0d735f06 | 615 | version of DB, see L<Using DB_File with Berkeley DB version 2 or greater>). |
039d031f PM |
616 | It is assumed that you have a copy of the Berkeley DB manual pages at |
617 | hand when reading this documentation. The interface defined here | |
618 | mirrors the Berkeley DB interface closely. | |
68dc0745 | 619 | |
8e07c86e AD |
620 | Berkeley DB is a C library which provides a consistent interface to a |
621 | number of database formats. B<DB_File> provides an interface to all | |
622 | three of the database types currently supported by Berkeley DB. | |
3b35bae3 AD |
623 | |
624 | The file types are: | |
625 | ||
626 | =over 5 | |
627 | ||
88108326 | 628 | =item B<DB_HASH> |
3b35bae3 | 629 | |
88108326 | 630 | This database type allows arbitrary key/value pairs to be stored in data |
8e07c86e AD |
631 | files. This is equivalent to the functionality provided by other |
632 | hashing packages like DBM, NDBM, ODBM, GDBM, and SDBM. Remember though, | |
633 | the files created using DB_HASH are not compatible with any of the | |
634 | other packages mentioned. | |
3b35bae3 | 635 | |
8e07c86e AD |
636 | A default hashing algorithm, which will be adequate for most |
637 | applications, is built into Berkeley DB. If you do need to use your own | |
638 | hashing algorithm it is possible to write your own in Perl and have | |
639 | B<DB_File> use it instead. | |
3b35bae3 | 640 | |
88108326 | 641 | =item B<DB_BTREE> |
642 | ||
643 | The btree format allows arbitrary key/value pairs to be stored in a | |
8e07c86e | 644 | sorted, balanced binary tree. |
3b35bae3 | 645 | |
8e07c86e AD |
646 | As with the DB_HASH format, it is possible to provide a user defined |
647 | Perl routine to perform the comparison of keys. By default, though, the | |
648 | keys are stored in lexical order. | |
3b35bae3 | 649 | |
88108326 | 650 | =item B<DB_RECNO> |
3b35bae3 | 651 | |
8e07c86e AD |
652 | DB_RECNO allows both fixed-length and variable-length flat text files |
653 | to be manipulated using the same key/value pair interface as in DB_HASH | |
654 | and DB_BTREE. In this case the key will consist of a record (line) | |
655 | number. | |
3b35bae3 AD |
656 | |
657 | =back | |
658 | ||
e5021521 | 659 | =head2 Using DB_File with Berkeley DB version 2 or greater |
1f70e1ea PM |
660 | |
661 | Although B<DB_File> is intended to be used with Berkeley DB version 1, | |
e5021521 | 662 | it can also be used with version 2, 3 or 4. In this case the interface is |
1f70e1ea | 663 | limited to the functionality provided by Berkeley DB 1.x. Anywhere the |
e5021521 | 664 | version 2 or greater interface differs, B<DB_File> arranges for it to work |
039d031f | 665 | like version 1. This feature allows B<DB_File> scripts that were built |
e5021521 | 666 | with version 1 to be migrated to version 2 or greater without any changes. |
1f70e1ea PM |
667 | |
668 | If you want to make use of the new features available in Berkeley DB | |
b90e71be | 669 | 2.x or greater, use the Perl module B<BerkeleyDB> instead. |
1f70e1ea | 670 | |
e5021521 JH |
671 | B<Note:> The database file format has changed multiple times in Berkeley |
672 | DB version 2, 3 and 4. If you cannot recreate your databases, you | |
673 | must dump any existing databases with either the C<db_dump> or the | |
674 | C<db_dump185> utility that comes with Berkeley DB. | |
675 | Once you have rebuilt DB_File to use Berkeley DB version 2 or greater, | |
676 | your databases can be recreated using C<db_load>. Refer to the Berkeley DB | |
1f70e1ea PM |
677 | documentation for further details. |
678 | ||
e5021521 | 679 | Please read L<"COPYRIGHT"> before using version 2.x or greater of Berkeley |
039d031f | 680 | DB with DB_File. |
1f70e1ea | 681 | |
68dc0745 | 682 | =head2 Interface to Berkeley DB |
3b35bae3 AD |
683 | |
684 | B<DB_File> allows access to Berkeley DB files using the tie() mechanism | |
8e07c86e AD |
685 | in Perl 5 (for full details, see L<perlfunc/tie()>). This facility |
686 | allows B<DB_File> to access Berkeley DB files using either an | |
687 | associative array (for DB_HASH & DB_BTREE file types) or an ordinary | |
688 | array (for the DB_RECNO file type). | |
3b35bae3 | 689 | |
88108326 | 690 | In addition to the tie() interface, it is also possible to access most |
691 | of the functions provided in the Berkeley DB API directly. | |
f6b705ef | 692 | See L<THE API INTERFACE>. |
3b35bae3 | 693 | |
88108326 | 694 | =head2 Opening a Berkeley DB Database File |
3b35bae3 | 695 | |
8e07c86e | 696 | Berkeley DB uses the function dbopen() to open or create a database. |
f6b705ef | 697 | Here is the C prototype for dbopen(): |
3b35bae3 AD |
698 | |
699 | DB* | |
700 | dbopen (const char * file, int flags, int mode, | |
701 | DBTYPE type, const void * openinfo) | |
702 | ||
703 | The parameter C<type> is an enumeration which specifies which of the 3 | |
704 | interface methods (DB_HASH, DB_BTREE or DB_RECNO) is to be used. | |
705 | Depending on which of these is actually chosen, the final parameter, | |
706 | I<openinfo> points to a data structure which allows tailoring of the | |
707 | specific interface method. | |
708 | ||
8e07c86e | 709 | This interface is handled slightly differently in B<DB_File>. Here is |
88108326 | 710 | an equivalent call using B<DB_File>: |
3b35bae3 | 711 | |
88108326 | 712 | tie %array, 'DB_File', $filename, $flags, $mode, $DB_HASH ; |
3b35bae3 | 713 | |
8e07c86e AD |
714 | The C<filename>, C<flags> and C<mode> parameters are the direct |
715 | equivalent of their dbopen() counterparts. The final parameter $DB_HASH | |
716 | performs the function of both the C<type> and C<openinfo> parameters in | |
717 | dbopen(). | |
3b35bae3 | 718 | |
88108326 | 719 | In the example above $DB_HASH is actually a pre-defined reference to a |
720 | hash object. B<DB_File> has three of these pre-defined references. | |
721 | Apart from $DB_HASH, there is also $DB_BTREE and $DB_RECNO. | |
3b35bae3 | 722 | |
8e07c86e AD |
723 | The keys allowed in each of these pre-defined references is limited to |
724 | the names used in the equivalent C structure. So, for example, the | |
725 | $DB_HASH reference will only allow keys called C<bsize>, C<cachesize>, | |
88108326 | 726 | C<ffactor>, C<hash>, C<lorder> and C<nelem>. |
727 | ||
728 | To change one of these elements, just assign to it like this: | |
729 | ||
730 | $DB_HASH->{'cachesize'} = 10000 ; | |
731 | ||
732 | The three predefined variables $DB_HASH, $DB_BTREE and $DB_RECNO are | |
733 | usually adequate for most applications. If you do need to create extra | |
734 | instances of these objects, constructors are available for each file | |
735 | type. | |
736 | ||
737 | Here are examples of the constructors and the valid options available | |
738 | for DB_HASH, DB_BTREE and DB_RECNO respectively. | |
739 | ||
740 | $a = new DB_File::HASHINFO ; | |
741 | $a->{'bsize'} ; | |
742 | $a->{'cachesize'} ; | |
743 | $a->{'ffactor'}; | |
744 | $a->{'hash'} ; | |
745 | $a->{'lorder'} ; | |
746 | $a->{'nelem'} ; | |
747 | ||
748 | $b = new DB_File::BTREEINFO ; | |
749 | $b->{'flags'} ; | |
750 | $b->{'cachesize'} ; | |
751 | $b->{'maxkeypage'} ; | |
752 | $b->{'minkeypage'} ; | |
753 | $b->{'psize'} ; | |
754 | $b->{'compare'} ; | |
755 | $b->{'prefix'} ; | |
756 | $b->{'lorder'} ; | |
757 | ||
758 | $c = new DB_File::RECNOINFO ; | |
759 | $c->{'bval'} ; | |
760 | $c->{'cachesize'} ; | |
761 | $c->{'psize'} ; | |
762 | $c->{'flags'} ; | |
763 | $c->{'lorder'} ; | |
764 | $c->{'reclen'} ; | |
765 | $c->{'bfname'} ; | |
766 | ||
767 | The values stored in the hashes above are mostly the direct equivalent | |
768 | of their C counterpart. Like their C counterparts, all are set to a | |
f6b705ef | 769 | default values - that means you don't have to set I<all> of the |
88108326 | 770 | values when you only want to change one. Here is an example: |
771 | ||
772 | $a = new DB_File::HASHINFO ; | |
773 | $a->{'cachesize'} = 12345 ; | |
774 | tie %y, 'DB_File', "filename", $flags, 0777, $a ; | |
775 | ||
36477c24 | 776 | A few of the options need extra discussion here. When used, the C |
88108326 | 777 | equivalent of the keys C<hash>, C<compare> and C<prefix> store pointers |
778 | to C functions. In B<DB_File> these keys are used to store references | |
779 | to Perl subs. Below are templates for each of the subs: | |
780 | ||
781 | sub hash | |
782 | { | |
783 | my ($data) = @_ ; | |
784 | ... | |
785 | # return the hash value for $data | |
786 | return $hash ; | |
787 | } | |
3b35bae3 | 788 | |
88108326 | 789 | sub compare |
790 | { | |
791 | my ($key, $key2) = @_ ; | |
792 | ... | |
793 | # return 0 if $key1 eq $key2 | |
794 | # -1 if $key1 lt $key2 | |
795 | # 1 if $key1 gt $key2 | |
796 | return (-1 , 0 or 1) ; | |
797 | } | |
3b35bae3 | 798 | |
88108326 | 799 | sub prefix |
800 | { | |
801 | my ($key, $key2) = @_ ; | |
802 | ... | |
803 | # return number of bytes of $key2 which are | |
804 | # necessary to determine that it is greater than $key1 | |
805 | return $bytes ; | |
806 | } | |
3b35bae3 | 807 | |
f6b705ef | 808 | See L<Changing the BTREE sort order> for an example of using the |
809 | C<compare> template. | |
88108326 | 810 | |
36477c24 | 811 | If you are using the DB_RECNO interface and you intend making use of |
9a2c4ce3 | 812 | C<bval>, you should check out L<The 'bval' Option>. |
36477c24 | 813 | |
88108326 | 814 | =head2 Default Parameters |
815 | ||
816 | It is possible to omit some or all of the final 4 parameters in the | |
817 | call to C<tie> and let them take default values. As DB_HASH is the most | |
818 | common file format used, the call: | |
819 | ||
820 | tie %A, "DB_File", "filename" ; | |
821 | ||
822 | is equivalent to: | |
823 | ||
18d2dc8c | 824 | tie %A, "DB_File", "filename", O_CREAT|O_RDWR, 0666, $DB_HASH ; |
88108326 | 825 | |
826 | It is also possible to omit the filename parameter as well, so the | |
827 | call: | |
828 | ||
829 | tie %A, "DB_File" ; | |
830 | ||
831 | is equivalent to: | |
832 | ||
18d2dc8c | 833 | tie %A, "DB_File", undef, O_CREAT|O_RDWR, 0666, $DB_HASH ; |
88108326 | 834 | |
f6b705ef | 835 | See L<In Memory Databases> for a discussion on the use of C<undef> |
88108326 | 836 | in place of a filename. |
837 | ||
f6b705ef | 838 | =head2 In Memory Databases |
839 | ||
840 | Berkeley DB allows the creation of in-memory databases by using NULL | |
841 | (that is, a C<(char *)0> in C) in place of the filename. B<DB_File> | |
842 | uses C<undef> instead of NULL to provide this functionality. | |
843 | ||
844 | =head1 DB_HASH | |
845 | ||
846 | The DB_HASH file format is probably the most commonly used of the three | |
847 | file formats that B<DB_File> supports. It is also very straightforward | |
848 | to use. | |
849 | ||
68dc0745 | 850 | =head2 A Simple Example |
f6b705ef | 851 | |
852 | This example shows how to create a database, add key/value pairs to the | |
853 | database, delete keys/value pairs and finally how to enumerate the | |
854 | contents of the database. | |
855 | ||
3245f058 | 856 | use warnings ; |
610ab055 | 857 | use strict ; |
f6b705ef | 858 | use DB_File ; |
07200f1b | 859 | our (%h, $k, $v) ; |
f6b705ef | 860 | |
2c2d71f5 | 861 | unlink "fruit" ; |
45a340cb | 862 | tie %h, "DB_File", "fruit", O_RDWR|O_CREAT, 0666, $DB_HASH |
f6b705ef | 863 | or die "Cannot open file 'fruit': $!\n"; |
864 | ||
865 | # Add a few key/value pairs to the file | |
866 | $h{"apple"} = "red" ; | |
867 | $h{"orange"} = "orange" ; | |
868 | $h{"banana"} = "yellow" ; | |
869 | $h{"tomato"} = "red" ; | |
870 | ||
871 | # Check for existence of a key | |
872 | print "Banana Exists\n\n" if $h{"banana"} ; | |
873 | ||
874 | # Delete a key/value pair. | |
875 | delete $h{"apple"} ; | |
876 | ||
877 | # print the contents of the file | |
878 | while (($k, $v) = each %h) | |
879 | { print "$k -> $v\n" } | |
880 | ||
881 | untie %h ; | |
882 | ||
883 | here is the output: | |
884 | ||
885 | Banana Exists | |
bbc7dcd2 | 886 | |
f6b705ef | 887 | orange -> orange |
888 | tomato -> red | |
889 | banana -> yellow | |
890 | ||
891 | Note that the like ordinary associative arrays, the order of the keys | |
892 | retrieved is in an apparently random order. | |
893 | ||
894 | =head1 DB_BTREE | |
895 | ||
896 | The DB_BTREE format is useful when you want to store data in a given | |
897 | order. By default the keys will be stored in lexical order, but as you | |
898 | will see from the example shown in the next section, it is very easy to | |
899 | define your own sorting function. | |
900 | ||
901 | =head2 Changing the BTREE sort order | |
902 | ||
903 | This script shows how to override the default sorting algorithm that | |
904 | BTREE uses. Instead of using the normal lexical ordering, a case | |
905 | insensitive compare function will be used. | |
88108326 | 906 | |
3245f058 | 907 | use warnings ; |
610ab055 | 908 | use strict ; |
f6b705ef | 909 | use DB_File ; |
610ab055 PM |
910 | |
911 | my %h ; | |
f6b705ef | 912 | |
913 | sub Compare | |
914 | { | |
915 | my ($key1, $key2) = @_ ; | |
916 | "\L$key1" cmp "\L$key2" ; | |
917 | } | |
918 | ||
919 | # specify the Perl sub that will do the comparison | |
920 | $DB_BTREE->{'compare'} = \&Compare ; | |
921 | ||
2c2d71f5 | 922 | unlink "tree" ; |
45a340cb | 923 | tie %h, "DB_File", "tree", O_RDWR|O_CREAT, 0666, $DB_BTREE |
f6b705ef | 924 | or die "Cannot open file 'tree': $!\n" ; |
925 | ||
926 | # Add a key/value pair to the file | |
927 | $h{'Wall'} = 'Larry' ; | |
928 | $h{'Smith'} = 'John' ; | |
929 | $h{'mouse'} = 'mickey' ; | |
930 | $h{'duck'} = 'donald' ; | |
931 | ||
932 | # Delete | |
933 | delete $h{"duck"} ; | |
934 | ||
935 | # Cycle through the keys printing them in order. | |
936 | # Note it is not necessary to sort the keys as | |
937 | # the btree will have kept them in order automatically. | |
938 | foreach (keys %h) | |
939 | { print "$_\n" } | |
940 | ||
941 | untie %h ; | |
942 | ||
943 | Here is the output from the code above. | |
944 | ||
945 | mouse | |
946 | Smith | |
947 | Wall | |
948 | ||
949 | There are a few point to bear in mind if you want to change the | |
950 | ordering in a BTREE database: | |
951 | ||
952 | =over 5 | |
953 | ||
954 | =item 1. | |
955 | ||
956 | The new compare function must be specified when you create the database. | |
957 | ||
958 | =item 2. | |
959 | ||
960 | You cannot change the ordering once the database has been created. Thus | |
961 | you must use the same compare function every time you access the | |
88108326 | 962 | database. |
963 | ||
39793c41 PM |
964 | =item 3 |
965 | ||
966 | Duplicate keys are entirely defined by the comparison function. | |
967 | In the case-insensitive example above, the keys: 'KEY' and 'key' | |
968 | would be considered duplicates, and assigning to the second one | |
52ffee89 | 969 | would overwrite the first. If duplicates are allowed for (with the |
59e51af5 | 970 | R_DUP flag discussed below), only a single copy of duplicate keys |
39793c41 PM |
971 | is stored in the database --- so (again with example above) assigning |
972 | three values to the keys: 'KEY', 'Key', and 'key' would leave just | |
973 | the first key: 'KEY' in the database with three values. For some | |
974 | situations this results in information loss, so care should be taken | |
975 | to provide fully qualified comparison functions when necessary. | |
976 | For example, the above comparison routine could be modified to | |
977 | additionally compare case-sensitively if two keys are equal in the | |
978 | case insensitive comparison: | |
979 | ||
980 | sub compare { | |
981 | my($key1, $key2) = @_; | |
982 | lc $key1 cmp lc $key2 || | |
983 | $key1 cmp $key2; | |
984 | } | |
985 | ||
986 | And now you will only have duplicates when the keys themselves | |
987 | are truly the same. (note: in versions of the db library prior to | |
988 | about November 1996, such duplicate keys were retained so it was | |
989 | possible to recover the original keys in sets of keys that | |
990 | compared as equal). | |
991 | ||
992 | ||
f6b705ef | 993 | =back |
994 | ||
68dc0745 | 995 | =head2 Handling Duplicate Keys |
f6b705ef | 996 | |
997 | The BTREE file type optionally allows a single key to be associated | |
998 | with an arbitrary number of values. This option is enabled by setting | |
999 | the flags element of C<$DB_BTREE> to R_DUP when creating the database. | |
1000 | ||
88108326 | 1001 | There are some difficulties in using the tied hash interface if you |
1002 | want to manipulate a BTREE database with duplicate keys. Consider this | |
1003 | code: | |
1004 | ||
3245f058 | 1005 | use warnings ; |
610ab055 | 1006 | use strict ; |
88108326 | 1007 | use DB_File ; |
610ab055 | 1008 | |
962cee9f | 1009 | my ($filename, %h) ; |
610ab055 | 1010 | |
88108326 | 1011 | $filename = "tree" ; |
1012 | unlink $filename ; | |
bbc7dcd2 | 1013 | |
88108326 | 1014 | # Enable duplicate records |
1015 | $DB_BTREE->{'flags'} = R_DUP ; | |
bbc7dcd2 | 1016 | |
45a340cb | 1017 | tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
88108326 | 1018 | or die "Cannot open $filename: $!\n"; |
bbc7dcd2 | 1019 | |
88108326 | 1020 | # Add some key/value pairs to the file |
1021 | $h{'Wall'} = 'Larry' ; | |
1022 | $h{'Wall'} = 'Brick' ; # Note the duplicate key | |
f6b705ef | 1023 | $h{'Wall'} = 'Brick' ; # Note the duplicate key and value |
88108326 | 1024 | $h{'Smith'} = 'John' ; |
1025 | $h{'mouse'} = 'mickey' ; | |
1026 | ||
1027 | # iterate through the associative array | |
1028 | # and print each key/value pair. | |
2c2d71f5 | 1029 | foreach (sort keys %h) |
88108326 | 1030 | { print "$_ -> $h{$_}\n" } |
1031 | ||
f6b705ef | 1032 | untie %h ; |
1033 | ||
88108326 | 1034 | Here is the output: |
1035 | ||
1036 | Smith -> John | |
1037 | Wall -> Larry | |
1038 | Wall -> Larry | |
f6b705ef | 1039 | Wall -> Larry |
88108326 | 1040 | mouse -> mickey |
1041 | ||
f6b705ef | 1042 | As you can see 3 records have been successfully created with key C<Wall> |
88108326 | 1043 | - the only thing is, when they are retrieved from the database they |
f6b705ef | 1044 | I<seem> to have the same value, namely C<Larry>. The problem is caused |
1045 | by the way that the associative array interface works. Basically, when | |
1046 | the associative array interface is used to fetch the value associated | |
1047 | with a given key, it will only ever retrieve the first value. | |
88108326 | 1048 | |
1049 | Although it may not be immediately obvious from the code above, the | |
1050 | associative array interface can be used to write values with duplicate | |
1051 | keys, but it cannot be used to read them back from the database. | |
1052 | ||
1053 | The way to get around this problem is to use the Berkeley DB API method | |
1054 | called C<seq>. This method allows sequential access to key/value | |
f6b705ef | 1055 | pairs. See L<THE API INTERFACE> for details of both the C<seq> method |
1056 | and the API in general. | |
88108326 | 1057 | |
1058 | Here is the script above rewritten using the C<seq> API method. | |
1059 | ||
3245f058 | 1060 | use warnings ; |
610ab055 | 1061 | use strict ; |
88108326 | 1062 | use DB_File ; |
bbc7dcd2 | 1063 | |
962cee9f | 1064 | my ($filename, $x, %h, $status, $key, $value) ; |
610ab055 | 1065 | |
88108326 | 1066 | $filename = "tree" ; |
1067 | unlink $filename ; | |
bbc7dcd2 | 1068 | |
88108326 | 1069 | # Enable duplicate records |
1070 | $DB_BTREE->{'flags'} = R_DUP ; | |
bbc7dcd2 | 1071 | |
45a340cb | 1072 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
88108326 | 1073 | or die "Cannot open $filename: $!\n"; |
bbc7dcd2 | 1074 | |
88108326 | 1075 | # Add some key/value pairs to the file |
1076 | $h{'Wall'} = 'Larry' ; | |
1077 | $h{'Wall'} = 'Brick' ; # Note the duplicate key | |
f6b705ef | 1078 | $h{'Wall'} = 'Brick' ; # Note the duplicate key and value |
88108326 | 1079 | $h{'Smith'} = 'John' ; |
1080 | $h{'mouse'} = 'mickey' ; | |
bbc7dcd2 | 1081 | |
f6b705ef | 1082 | # iterate through the btree using seq |
88108326 | 1083 | # and print each key/value pair. |
610ab055 | 1084 | $key = $value = 0 ; |
f6b705ef | 1085 | for ($status = $x->seq($key, $value, R_FIRST) ; |
1086 | $status == 0 ; | |
1087 | $status = $x->seq($key, $value, R_NEXT) ) | |
88108326 | 1088 | { print "$key -> $value\n" } |
bbc7dcd2 | 1089 | |
88108326 | 1090 | undef $x ; |
1091 | untie %h ; | |
1092 | ||
1093 | that prints: | |
1094 | ||
1095 | Smith -> John | |
1096 | Wall -> Brick | |
f6b705ef | 1097 | Wall -> Brick |
88108326 | 1098 | Wall -> Larry |
1099 | mouse -> mickey | |
1100 | ||
f6b705ef | 1101 | This time we have got all the key/value pairs, including the multiple |
88108326 | 1102 | values associated with the key C<Wall>. |
1103 | ||
6ca2e664 PM |
1104 | To make life easier when dealing with duplicate keys, B<DB_File> comes with |
1105 | a few utility methods. | |
1106 | ||
68dc0745 | 1107 | =head2 The get_dup() Method |
f6b705ef | 1108 | |
6ca2e664 | 1109 | The C<get_dup> method assists in |
88108326 | 1110 | reading duplicate values from BTREE databases. The method can take the |
1111 | following forms: | |
1112 | ||
1113 | $count = $x->get_dup($key) ; | |
1114 | @list = $x->get_dup($key) ; | |
1115 | %list = $x->get_dup($key, 1) ; | |
1116 | ||
1117 | In a scalar context the method returns the number of values associated | |
1118 | with the key, C<$key>. | |
1119 | ||
1120 | In list context, it returns all the values which match C<$key>. Note | |
f6b705ef | 1121 | that the values will be returned in an apparently random order. |
88108326 | 1122 | |
7a2e2cd6 | 1123 | In list context, if the second parameter is present and evaluates |
1124 | TRUE, the method returns an associative array. The keys of the | |
1125 | associative array correspond to the values that matched in the BTREE | |
1126 | and the values of the array are a count of the number of times that | |
1127 | particular value occurred in the BTREE. | |
88108326 | 1128 | |
f6b705ef | 1129 | So assuming the database created above, we can use C<get_dup> like |
88108326 | 1130 | this: |
1131 | ||
3245f058 | 1132 | use warnings ; |
2c2d71f5 JH |
1133 | use strict ; |
1134 | use DB_File ; | |
bbc7dcd2 | 1135 | |
962cee9f | 1136 | my ($filename, $x, %h) ; |
2c2d71f5 JH |
1137 | |
1138 | $filename = "tree" ; | |
bbc7dcd2 | 1139 | |
2c2d71f5 JH |
1140 | # Enable duplicate records |
1141 | $DB_BTREE->{'flags'} = R_DUP ; | |
bbc7dcd2 | 1142 | |
45a340cb | 1143 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
2c2d71f5 JH |
1144 | or die "Cannot open $filename: $!\n"; |
1145 | ||
610ab055 | 1146 | my $cnt = $x->get_dup("Wall") ; |
88108326 | 1147 | print "Wall occurred $cnt times\n" ; |
1148 | ||
610ab055 | 1149 | my %hash = $x->get_dup("Wall", 1) ; |
88108326 | 1150 | print "Larry is there\n" if $hash{'Larry'} ; |
f6b705ef | 1151 | print "There are $hash{'Brick'} Brick Walls\n" ; |
88108326 | 1152 | |
2c2d71f5 | 1153 | my @list = sort $x->get_dup("Wall") ; |
88108326 | 1154 | print "Wall => [@list]\n" ; |
1155 | ||
f6b705ef | 1156 | @list = $x->get_dup("Smith") ; |
88108326 | 1157 | print "Smith => [@list]\n" ; |
bbc7dcd2 | 1158 | |
f6b705ef | 1159 | @list = $x->get_dup("Dog") ; |
88108326 | 1160 | print "Dog => [@list]\n" ; |
1161 | ||
1162 | ||
1163 | and it will print: | |
1164 | ||
f6b705ef | 1165 | Wall occurred 3 times |
88108326 | 1166 | Larry is there |
f6b705ef | 1167 | There are 2 Brick Walls |
1168 | Wall => [Brick Brick Larry] | |
88108326 | 1169 | Smith => [John] |
1170 | Dog => [] | |
3b35bae3 | 1171 | |
6ca2e664 PM |
1172 | =head2 The find_dup() Method |
1173 | ||
1174 | $status = $X->find_dup($key, $value) ; | |
1175 | ||
b90e71be | 1176 | This method checks for the existence of a specific key/value pair. If the |
6ca2e664 PM |
1177 | pair exists, the cursor is left pointing to the pair and the method |
1178 | returns 0. Otherwise the method returns a non-zero value. | |
1179 | ||
1180 | Assuming the database from the previous example: | |
1181 | ||
3245f058 | 1182 | use warnings ; |
6ca2e664 PM |
1183 | use strict ; |
1184 | use DB_File ; | |
bbc7dcd2 | 1185 | |
962cee9f | 1186 | my ($filename, $x, %h, $found) ; |
6ca2e664 | 1187 | |
07200f1b | 1188 | $filename = "tree" ; |
bbc7dcd2 | 1189 | |
6ca2e664 PM |
1190 | # Enable duplicate records |
1191 | $DB_BTREE->{'flags'} = R_DUP ; | |
bbc7dcd2 | 1192 | |
45a340cb | 1193 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
6ca2e664 PM |
1194 | or die "Cannot open $filename: $!\n"; |
1195 | ||
1196 | $found = ( $x->find_dup("Wall", "Larry") == 0 ? "" : "not") ; | |
1197 | print "Larry Wall is $found there\n" ; | |
bbc7dcd2 | 1198 | |
6ca2e664 PM |
1199 | $found = ( $x->find_dup("Wall", "Harry") == 0 ? "" : "not") ; |
1200 | print "Harry Wall is $found there\n" ; | |
bbc7dcd2 | 1201 | |
6ca2e664 PM |
1202 | undef $x ; |
1203 | untie %h ; | |
1204 | ||
1205 | prints this | |
1206 | ||
2c2d71f5 | 1207 | Larry Wall is there |
6ca2e664 PM |
1208 | Harry Wall is not there |
1209 | ||
1210 | ||
1211 | =head2 The del_dup() Method | |
1212 | ||
1213 | $status = $X->del_dup($key, $value) ; | |
1214 | ||
1215 | This method deletes a specific key/value pair. It returns | |
1216 | 0 if they exist and have been deleted successfully. | |
1217 | Otherwise the method returns a non-zero value. | |
1218 | ||
b90e71be | 1219 | Again assuming the existence of the C<tree> database |
6ca2e664 | 1220 | |
3245f058 | 1221 | use warnings ; |
6ca2e664 PM |
1222 | use strict ; |
1223 | use DB_File ; | |
bbc7dcd2 | 1224 | |
962cee9f | 1225 | my ($filename, $x, %h, $found) ; |
6ca2e664 | 1226 | |
07200f1b | 1227 | $filename = "tree" ; |
bbc7dcd2 | 1228 | |
6ca2e664 PM |
1229 | # Enable duplicate records |
1230 | $DB_BTREE->{'flags'} = R_DUP ; | |
bbc7dcd2 | 1231 | |
45a340cb | 1232 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
6ca2e664 PM |
1233 | or die "Cannot open $filename: $!\n"; |
1234 | ||
1235 | $x->del_dup("Wall", "Larry") ; | |
1236 | ||
1237 | $found = ( $x->find_dup("Wall", "Larry") == 0 ? "" : "not") ; | |
1238 | print "Larry Wall is $found there\n" ; | |
bbc7dcd2 | 1239 | |
6ca2e664 PM |
1240 | undef $x ; |
1241 | untie %h ; | |
1242 | ||
1243 | prints this | |
1244 | ||
1245 | Larry Wall is not there | |
1246 | ||
f6b705ef | 1247 | =head2 Matching Partial Keys |
1248 | ||
1249 | The BTREE interface has a feature which allows partial keys to be | |
1250 | matched. This functionality is I<only> available when the C<seq> method | |
1251 | is used along with the R_CURSOR flag. | |
1252 | ||
1253 | $x->seq($key, $value, R_CURSOR) ; | |
1254 | ||
1255 | Here is the relevant quote from the dbopen man page where it defines | |
1256 | the use of the R_CURSOR flag with seq: | |
1257 | ||
f6b705ef | 1258 | Note, for the DB_BTREE access method, the returned key is not |
1259 | necessarily an exact match for the specified key. The returned key | |
1260 | is the smallest key greater than or equal to the specified key, | |
1261 | permitting partial key matches and range searches. | |
1262 | ||
f6b705ef | 1263 | In the example script below, the C<match> sub uses this feature to find |
1264 | and print the first matching key/value pair given a partial key. | |
1265 | ||
3245f058 | 1266 | use warnings ; |
610ab055 | 1267 | use strict ; |
f6b705ef | 1268 | use DB_File ; |
1269 | use Fcntl ; | |
610ab055 | 1270 | |
962cee9f | 1271 | my ($filename, $x, %h, $st, $key, $value) ; |
f6b705ef | 1272 | |
1273 | sub match | |
1274 | { | |
1275 | my $key = shift ; | |
610ab055 | 1276 | my $value = 0; |
f6b705ef | 1277 | my $orig_key = $key ; |
1278 | $x->seq($key, $value, R_CURSOR) ; | |
1279 | print "$orig_key\t-> $key\t-> $value\n" ; | |
1280 | } | |
1281 | ||
1282 | $filename = "tree" ; | |
1283 | unlink $filename ; | |
1284 | ||
45a340cb | 1285 | $x = tie %h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_BTREE |
f6b705ef | 1286 | or die "Cannot open $filename: $!\n"; |
bbc7dcd2 | 1287 | |
f6b705ef | 1288 | # Add some key/value pairs to the file |
1289 | $h{'mouse'} = 'mickey' ; | |
1290 | $h{'Wall'} = 'Larry' ; | |
1291 | $h{'Walls'} = 'Brick' ; | |
1292 | $h{'Smith'} = 'John' ; | |
bbc7dcd2 | 1293 | |
f6b705ef | 1294 | |
610ab055 | 1295 | $key = $value = 0 ; |
f6b705ef | 1296 | print "IN ORDER\n" ; |
1297 | for ($st = $x->seq($key, $value, R_FIRST) ; | |
1298 | $st == 0 ; | |
1299 | $st = $x->seq($key, $value, R_NEXT) ) | |
bbc7dcd2 | 1300 | |
2c2d71f5 | 1301 | { print "$key -> $value\n" } |
bbc7dcd2 | 1302 | |
f6b705ef | 1303 | print "\nPARTIAL MATCH\n" ; |
1304 | ||
1305 | match "Wa" ; | |
1306 | match "A" ; | |
1307 | match "a" ; | |
1308 | ||
1309 | undef $x ; | |
1310 | untie %h ; | |
1311 | ||
1312 | Here is the output: | |
1313 | ||
1314 | IN ORDER | |
1315 | Smith -> John | |
1316 | Wall -> Larry | |
1317 | Walls -> Brick | |
1318 | mouse -> mickey | |
1319 | ||
1320 | PARTIAL MATCH | |
1321 | Wa -> Wall -> Larry | |
1322 | A -> Smith -> John | |
1323 | a -> mouse -> mickey | |
1324 | ||
1325 | =head1 DB_RECNO | |
1326 | ||
1327 | DB_RECNO provides an interface to flat text files. Both variable and | |
1328 | fixed length records are supported. | |
3b35bae3 | 1329 | |
6ca2e664 | 1330 | In order to make RECNO more compatible with Perl, the array offset for |
88108326 | 1331 | all RECNO arrays begins at 0 rather than 1 as in Berkeley DB. |
3b35bae3 | 1332 | |
88108326 | 1333 | As with normal Perl arrays, a RECNO array can be accessed using |
1334 | negative indexes. The index -1 refers to the last element of the array, | |
1335 | -2 the second last, and so on. Attempting to access an element before | |
1336 | the start of the array will raise a fatal run-time error. | |
3b35bae3 | 1337 | |
68dc0745 | 1338 | =head2 The 'bval' Option |
36477c24 | 1339 | |
1340 | The operation of the bval option warrants some discussion. Here is the | |
1341 | definition of bval from the Berkeley DB 1.85 recno manual page: | |
1342 | ||
1343 | The delimiting byte to be used to mark the end of a | |
1344 | record for variable-length records, and the pad charac- | |
1345 | ter for fixed-length records. If no value is speci- | |
1346 | fied, newlines (``\n'') are used to mark the end of | |
1347 | variable-length records and fixed-length records are | |
1348 | padded with spaces. | |
1349 | ||
1350 | The second sentence is wrong. In actual fact bval will only default to | |
1351 | C<"\n"> when the openinfo parameter in dbopen is NULL. If a non-NULL | |
1352 | openinfo parameter is used at all, the value that happens to be in bval | |
1353 | will be used. That means you always have to specify bval when making | |
1354 | use of any of the options in the openinfo parameter. This documentation | |
1355 | error will be fixed in the next release of Berkeley DB. | |
1356 | ||
1357 | That clarifies the situation with regards Berkeley DB itself. What | |
1358 | about B<DB_File>? Well, the behavior defined in the quote above is | |
6ca2e664 | 1359 | quite useful, so B<DB_File> conforms to it. |
36477c24 | 1360 | |
1361 | That means that you can specify other options (e.g. cachesize) and | |
1362 | still have bval default to C<"\n"> for variable length records, and | |
1363 | space for fixed length records. | |
1364 | ||
c5da4faf | 1365 | Also note that the bval option only allows you to specify a single byte |
a6d6498e | 1366 | as a delimiter. |
c5da4faf | 1367 | |
f6b705ef | 1368 | =head2 A Simple Example |
3b35bae3 | 1369 | |
6ca2e664 PM |
1370 | Here is a simple example that uses RECNO (if you are using a version |
1371 | of Perl earlier than 5.004_57 this example won't work -- see | |
1372 | L<Extra RECNO Methods> for a workaround). | |
f6b705ef | 1373 | |
3245f058 | 1374 | use warnings ; |
610ab055 | 1375 | use strict ; |
f6b705ef | 1376 | use DB_File ; |
f6b705ef | 1377 | |
2c2d71f5 JH |
1378 | my $filename = "text" ; |
1379 | unlink $filename ; | |
1380 | ||
610ab055 | 1381 | my @h ; |
45a340cb | 1382 | tie @h, "DB_File", $filename, O_RDWR|O_CREAT, 0666, $DB_RECNO |
f6b705ef | 1383 | or die "Cannot open file 'text': $!\n" ; |
1384 | ||
1385 | # Add a few key/value pairs to the file | |
1386 | $h[0] = "orange" ; | |
1387 | $h[1] = "blue" ; | |
1388 | $h[2] = "yellow" ; | |
1389 | ||
6ca2e664 PM |
1390 | push @h, "green", "black" ; |
1391 | ||
1392 | my $elements = scalar @h ; | |
1393 | print "The array contains $elements entries\n" ; | |
1394 | ||
1395 | my $last = pop @h ; | |
1396 | print "popped $last\n" ; | |
1397 | ||
1398 | unshift @h, "white" ; | |
1399 | my $first = shift @h ; | |
1400 | print "shifted $first\n" ; | |
1401 | ||
f6b705ef | 1402 | # Check for existence of a key |
1403 | print "Element 1 Exists with value $h[1]\n" if $h[1] ; | |
1404 | ||
1405 | # use a negative index | |
1406 | print "The last element is $h[-1]\n" ; | |
1407 | print "The 2nd last element is $h[-2]\n" ; | |
1408 | ||
1409 | untie @h ; | |
3b35bae3 | 1410 | |
f6b705ef | 1411 | Here is the output from the script: |
1412 | ||
6ca2e664 PM |
1413 | The array contains 5 entries |
1414 | popped black | |
2c2d71f5 | 1415 | shifted white |
f6b705ef | 1416 | Element 1 Exists with value blue |
6ca2e664 PM |
1417 | The last element is green |
1418 | The 2nd last element is yellow | |
f6b705ef | 1419 | |
6ca2e664 | 1420 | =head2 Extra RECNO Methods |
f6b705ef | 1421 | |
045291aa | 1422 | If you are using a version of Perl earlier than 5.004_57, the tied |
6ca2e664 PM |
1423 | array interface is quite limited. In the example script above |
1424 | C<push>, C<pop>, C<shift>, C<unshift> | |
1425 | or determining the array length will not work with a tied array. | |
045291aa PM |
1426 | |
1427 | To make the interface more useful for older versions of Perl, a number | |
1428 | of methods are supplied with B<DB_File> to simulate the missing array | |
1429 | operations. All these methods are accessed via the object returned from | |
1430 | the tie call. | |
f6b705ef | 1431 | |
1432 | Here are the methods: | |
1433 | ||
1434 | =over 5 | |
3b35bae3 | 1435 | |
f6b705ef | 1436 | =item B<$X-E<gt>push(list) ;> |
1437 | ||
1438 | Pushes the elements of C<list> to the end of the array. | |
1439 | ||
1440 | =item B<$value = $X-E<gt>pop ;> | |
1441 | ||
1442 | Removes and returns the last element of the array. | |
1443 | ||
1444 | =item B<$X-E<gt>shift> | |
1445 | ||
1446 | Removes and returns the first element of the array. | |
1447 | ||
1448 | =item B<$X-E<gt>unshift(list) ;> | |
1449 | ||
1450 | Pushes the elements of C<list> to the start of the array. | |
1451 | ||
1452 | =item B<$X-E<gt>length> | |
1453 | ||
1454 | Returns the number of elements in the array. | |
1455 | ||
c5da4faf PM |
1456 | =item B<$X-E<gt>splice(offset, length, elements);> |
1457 | ||
a6d05634 | 1458 | Returns a splice of the array. |
c5da4faf | 1459 | |
f6b705ef | 1460 | =back |
1461 | ||
1462 | =head2 Another Example | |
1463 | ||
1464 | Here is a more complete example that makes use of some of the methods | |
1465 | described above. It also makes use of the API interface directly (see | |
1466 | L<THE API INTERFACE>). | |
1467 | ||
3245f058 | 1468 | use warnings ; |
f6b705ef | 1469 | use strict ; |
962cee9f | 1470 | my (@h, $H, $file, $i) ; |
f6b705ef | 1471 | use DB_File ; |
1472 | use Fcntl ; | |
bbc7dcd2 | 1473 | |
f6b705ef | 1474 | $file = "text" ; |
1475 | ||
1476 | unlink $file ; | |
1477 | ||
45a340cb | 1478 | $H = tie @h, "DB_File", $file, O_RDWR|O_CREAT, 0666, $DB_RECNO |
f6b705ef | 1479 | or die "Cannot open file $file: $!\n" ; |
bbc7dcd2 | 1480 | |
f6b705ef | 1481 | # first create a text file to play with |
1482 | $h[0] = "zero" ; | |
1483 | $h[1] = "one" ; | |
1484 | $h[2] = "two" ; | |
1485 | $h[3] = "three" ; | |
1486 | $h[4] = "four" ; | |
1487 | ||
bbc7dcd2 | 1488 | |
f6b705ef | 1489 | # Print the records in order. |
1490 | # | |
1491 | # The length method is needed here because evaluating a tied | |
1492 | # array in a scalar context does not return the number of | |
1493 | # elements in the array. | |
1494 | ||
1495 | print "\nORIGINAL\n" ; | |
1496 | foreach $i (0 .. $H->length - 1) { | |
1497 | print "$i: $h[$i]\n" ; | |
1498 | } | |
1499 | ||
1500 | # use the push & pop methods | |
1501 | $a = $H->pop ; | |
1502 | $H->push("last") ; | |
1503 | print "\nThe last record was [$a]\n" ; | |
1504 | ||
1505 | # and the shift & unshift methods | |
1506 | $a = $H->shift ; | |
1507 | $H->unshift("first") ; | |
1508 | print "The first record was [$a]\n" ; | |
1509 | ||
1510 | # Use the API to add a new record after record 2. | |
1511 | $i = 2 ; | |
1512 | $H->put($i, "Newbie", R_IAFTER) ; | |
1513 | ||
1514 | # and a new record before record 1. | |
1515 | $i = 1 ; | |
1516 | $H->put($i, "New One", R_IBEFORE) ; | |
1517 | ||
1518 | # delete record 3 | |
1519 | $H->del(3) ; | |
1520 | ||
1521 | # now print the records in reverse order | |
1522 | print "\nREVERSE\n" ; | |
1523 | for ($i = $H->length - 1 ; $i >= 0 ; -- $i) | |
1524 | { print "$i: $h[$i]\n" } | |
1525 | ||
1526 | # same again, but use the API functions instead | |
1527 | print "\nREVERSE again\n" ; | |
610ab055 | 1528 | my ($s, $k, $v) = (0, 0, 0) ; |
f6b705ef | 1529 | for ($s = $H->seq($k, $v, R_LAST) ; |
1530 | $s == 0 ; | |
1531 | $s = $H->seq($k, $v, R_PREV)) | |
1532 | { print "$k: $v\n" } | |
1533 | ||
1534 | undef $H ; | |
1535 | untie @h ; | |
1536 | ||
1537 | and this is what it outputs: | |
1538 | ||
1539 | ORIGINAL | |
1540 | 0: zero | |
1541 | 1: one | |
1542 | 2: two | |
1543 | 3: three | |
1544 | 4: four | |
1545 | ||
1546 | The last record was [four] | |
1547 | The first record was [zero] | |
1548 | ||
1549 | REVERSE | |
1550 | 5: last | |
1551 | 4: three | |
1552 | 3: Newbie | |
1553 | 2: one | |
1554 | 1: New One | |
1555 | 0: first | |
1556 | ||
1557 | REVERSE again | |
1558 | 5: last | |
1559 | 4: three | |
1560 | 3: Newbie | |
1561 | 2: one | |
1562 | 1: New One | |
1563 | 0: first | |
1564 | ||
1565 | Notes: | |
1566 | ||
1567 | =over 5 | |
1568 | ||
1569 | =item 1. | |
1570 | ||
1571 | Rather than iterating through the array, C<@h> like this: | |
1572 | ||
1573 | foreach $i (@h) | |
1574 | ||
1575 | it is necessary to use either this: | |
1576 | ||
1577 | foreach $i (0 .. $H->length - 1) | |
1578 | ||
1579 | or this: | |
1580 | ||
1581 | for ($a = $H->get($k, $v, R_FIRST) ; | |
1582 | $a == 0 ; | |
1583 | $a = $H->get($k, $v, R_NEXT) ) | |
1584 | ||
1585 | =item 2. | |
1586 | ||
1587 | Notice that both times the C<put> method was used the record index was | |
1588 | specified using a variable, C<$i>, rather than the literal value | |
1589 | itself. This is because C<put> will return the record number of the | |
1590 | inserted line via that parameter. | |
1591 | ||
1592 | =back | |
1593 | ||
1594 | =head1 THE API INTERFACE | |
3b35bae3 AD |
1595 | |
1596 | As well as accessing Berkeley DB using a tied hash or array, it is also | |
88108326 | 1597 | possible to make direct use of most of the API functions defined in the |
8e07c86e | 1598 | Berkeley DB documentation. |
3b35bae3 | 1599 | |
88108326 | 1600 | To do this you need to store a copy of the object returned from the tie. |
3b35bae3 | 1601 | |
88108326 | 1602 | $db = tie %hash, "DB_File", "filename" ; |
3b35bae3 | 1603 | |
8e07c86e | 1604 | Once you have done that, you can access the Berkeley DB API functions |
88108326 | 1605 | as B<DB_File> methods directly like this: |
3b35bae3 AD |
1606 | |
1607 | $db->put($key, $value, R_NOOVERWRITE) ; | |
1608 | ||
88108326 | 1609 | B<Important:> If you have saved a copy of the object returned from |
1610 | C<tie>, the underlying database file will I<not> be closed until both | |
1611 | the tied variable is untied and all copies of the saved object are | |
610ab055 | 1612 | destroyed. |
88108326 | 1613 | |
1614 | use DB_File ; | |
1615 | $db = tie %hash, "DB_File", "filename" | |
1616 | or die "Cannot tie filename: $!" ; | |
1617 | ... | |
1618 | undef $db ; | |
1619 | untie %hash ; | |
1620 | ||
9a2c4ce3 | 1621 | See L<The untie() Gotcha> for more details. |
778183f3 | 1622 | |
88108326 | 1623 | All the functions defined in L<dbopen> are available except for |
1624 | close() and dbopen() itself. The B<DB_File> method interface to the | |
1625 | supported functions have been implemented to mirror the way Berkeley DB | |
1626 | works whenever possible. In particular note that: | |
1627 | ||
1628 | =over 5 | |
1629 | ||
1630 | =item * | |
1631 | ||
1632 | The methods return a status value. All return 0 on success. | |
1633 | All return -1 to signify an error and set C<$!> to the exact | |
1634 | error code. The return code 1 generally (but not always) means that the | |
1635 | key specified did not exist in the database. | |
1636 | ||
1637 | Other return codes are defined. See below and in the Berkeley DB | |
1638 | documentation for details. The Berkeley DB documentation should be used | |
1639 | as the definitive source. | |
1640 | ||
1641 | =item * | |
3b35bae3 | 1642 | |
88108326 | 1643 | Whenever a Berkeley DB function returns data via one of its parameters, |
1644 | the equivalent B<DB_File> method does exactly the same. | |
3b35bae3 | 1645 | |
88108326 | 1646 | =item * |
1647 | ||
1648 | If you are careful, it is possible to mix API calls with the tied | |
1649 | hash/array interface in the same piece of code. Although only a few of | |
1650 | the methods used to implement the tied interface currently make use of | |
1651 | the cursor, you should always assume that the cursor has been changed | |
1652 | any time the tied hash/array interface is used. As an example, this | |
1653 | code will probably not do what you expect: | |
1654 | ||
1655 | $X = tie %x, 'DB_File', $filename, O_RDWR|O_CREAT, 0777, $DB_BTREE | |
1656 | or die "Cannot tie $filename: $!" ; | |
1657 | ||
1658 | # Get the first key/value pair and set the cursor | |
1659 | $X->seq($key, $value, R_FIRST) ; | |
1660 | ||
1661 | # this line will modify the cursor | |
1662 | $count = scalar keys %x ; | |
1663 | ||
1664 | # Get the second key/value pair. | |
1665 | # oops, it didn't, it got the last key/value pair! | |
1666 | $X->seq($key, $value, R_NEXT) ; | |
1667 | ||
1668 | The code above can be rearranged to get around the problem, like this: | |
1669 | ||
1670 | $X = tie %x, 'DB_File', $filename, O_RDWR|O_CREAT, 0777, $DB_BTREE | |
1671 | or die "Cannot tie $filename: $!" ; | |
1672 | ||
1673 | # this line will modify the cursor | |
1674 | $count = scalar keys %x ; | |
1675 | ||
1676 | # Get the first key/value pair and set the cursor | |
1677 | $X->seq($key, $value, R_FIRST) ; | |
1678 | ||
1679 | # Get the second key/value pair. | |
1680 | # worked this time. | |
1681 | $X->seq($key, $value, R_NEXT) ; | |
1682 | ||
1683 | =back | |
1684 | ||
1685 | All the constants defined in L<dbopen> for use in the flags parameters | |
1686 | in the methods defined below are also available. Refer to the Berkeley | |
1687 | DB documentation for the precise meaning of the flags values. | |
1688 | ||
1689 | Below is a list of the methods available. | |
3b35bae3 AD |
1690 | |
1691 | =over 5 | |
1692 | ||
f6b705ef | 1693 | =item B<$status = $X-E<gt>get($key, $value [, $flags]) ;> |
88108326 | 1694 | |
1695 | Given a key (C<$key>) this method reads the value associated with it | |
1696 | from the database. The value read from the database is returned in the | |
1697 | C<$value> parameter. | |
3b35bae3 | 1698 | |
88108326 | 1699 | If the key does not exist the method returns 1. |
3b35bae3 | 1700 | |
88108326 | 1701 | No flags are currently defined for this method. |
3b35bae3 | 1702 | |
f6b705ef | 1703 | =item B<$status = $X-E<gt>put($key, $value [, $flags]) ;> |
3b35bae3 | 1704 | |
88108326 | 1705 | Stores the key/value pair in the database. |
1706 | ||
1707 | If you use either the R_IAFTER or R_IBEFORE flags, the C<$key> parameter | |
8e07c86e | 1708 | will have the record number of the inserted key/value pair set. |
3b35bae3 | 1709 | |
88108326 | 1710 | Valid flags are R_CURSOR, R_IAFTER, R_IBEFORE, R_NOOVERWRITE and |
1711 | R_SETCURSOR. | |
1712 | ||
f6b705ef | 1713 | =item B<$status = $X-E<gt>del($key [, $flags]) ;> |
3b35bae3 | 1714 | |
88108326 | 1715 | Removes all key/value pairs with key C<$key> from the database. |
3b35bae3 | 1716 | |
88108326 | 1717 | A return code of 1 means that the requested key was not in the |
1718 | database. | |
3b35bae3 | 1719 | |
88108326 | 1720 | R_CURSOR is the only valid flag at present. |
3b35bae3 | 1721 | |
f6b705ef | 1722 | =item B<$status = $X-E<gt>fd ;> |
3b35bae3 | 1723 | |
88108326 | 1724 | Returns the file descriptor for the underlying database. |
3b35bae3 | 1725 | |
b90e71be GS |
1726 | See L<Locking: The Trouble with fd> for an explanation for why you should |
1727 | not use C<fd> to lock your database. | |
3b35bae3 | 1728 | |
f6b705ef | 1729 | =item B<$status = $X-E<gt>seq($key, $value, $flags) ;> |
3b35bae3 | 1730 | |
88108326 | 1731 | This interface allows sequential retrieval from the database. See |
1732 | L<dbopen> for full details. | |
1733 | ||
1734 | Both the C<$key> and C<$value> parameters will be set to the key/value | |
1735 | pair read from the database. | |
1736 | ||
1737 | The flags parameter is mandatory. The valid flag values are R_CURSOR, | |
1738 | R_FIRST, R_LAST, R_NEXT and R_PREV. | |
1739 | ||
f6b705ef | 1740 | =item B<$status = $X-E<gt>sync([$flags]) ;> |
88108326 | 1741 | |
1742 | Flushes any cached buffers to disk. | |
1743 | ||
1744 | R_RECNOSYNC is the only valid flag at present. | |
3b35bae3 AD |
1745 | |
1746 | =back | |
1747 | ||
cad2e5aa JH |
1748 | =head1 DBM FILTERS |
1749 | ||
1750 | A DBM Filter is a piece of code that is be used when you I<always> | |
1751 | want to make the same transformation to all keys and/or values in a | |
1752 | DBM database. | |
1753 | ||
1754 | There are four methods associated with DBM Filters. All work identically, | |
1755 | and each is used to install (or uninstall) a single DBM Filter. Each | |
1756 | expects a single parameter, namely a reference to a sub. The only | |
1757 | difference between them is the place that the filter is installed. | |
1758 | ||
1759 | To summarise: | |
1760 | ||
1761 | =over 5 | |
1762 | ||
1763 | =item B<filter_store_key> | |
1764 | ||
1765 | If a filter has been installed with this method, it will be invoked | |
1766 | every time you write a key to a DBM database. | |
1767 | ||
1768 | =item B<filter_store_value> | |
1769 | ||
1770 | If a filter has been installed with this method, it will be invoked | |
1771 | every time you write a value to a DBM database. | |
1772 | ||
1773 | ||
1774 | =item B<filter_fetch_key> | |
1775 | ||
1776 | If a filter has been installed with this method, it will be invoked | |
1777 | every time you read a key from a DBM database. | |
1778 | ||
1779 | =item B<filter_fetch_value> | |
1780 | ||
1781 | If a filter has been installed with this method, it will be invoked | |
1782 | every time you read a value from a DBM database. | |
1783 | ||
1784 | =back | |
1785 | ||
1786 | You can use any combination of the methods, from none, to all four. | |
1787 | ||
1788 | All filter methods return the existing filter, if present, or C<undef> | |
1789 | in not. | |
1790 | ||
1791 | To delete a filter pass C<undef> to it. | |
1792 | ||
1793 | =head2 The Filter | |
1794 | ||
1795 | When each filter is called by Perl, a local copy of C<$_> will contain | |
1796 | the key or value to be filtered. Filtering is achieved by modifying | |
1797 | the contents of C<$_>. The return code from the filter is ignored. | |
1798 | ||
1799 | =head2 An Example -- the NULL termination problem. | |
1800 | ||
1801 | Consider the following scenario. You have a DBM database | |
1802 | that you need to share with a third-party C application. The C application | |
1803 | assumes that I<all> keys and values are NULL terminated. Unfortunately | |
1804 | when Perl writes to DBM databases it doesn't use NULL termination, so | |
1805 | your Perl application will have to manage NULL termination itself. When | |
1806 | you write to the database you will have to use something like this: | |
1807 | ||
1808 | $hash{"$key\0"} = "$value\0" ; | |
1809 | ||
1810 | Similarly the NULL needs to be taken into account when you are considering | |
1811 | the length of existing keys/values. | |
1812 | ||
1813 | It would be much better if you could ignore the NULL terminations issue | |
1814 | in the main application code and have a mechanism that automatically | |
1815 | added the terminating NULL to all keys and values whenever you write to | |
1816 | the database and have them removed when you read from the database. As I'm | |
1817 | sure you have already guessed, this is a problem that DBM Filters can | |
1818 | fix very easily. | |
1819 | ||
3245f058 | 1820 | use warnings ; |
cad2e5aa JH |
1821 | use strict ; |
1822 | use DB_File ; | |
1823 | ||
1824 | my %hash ; | |
2359510d | 1825 | my $filename = "filt" ; |
cad2e5aa JH |
1826 | unlink $filename ; |
1827 | ||
1828 | my $db = tie %hash, 'DB_File', $filename, O_CREAT|O_RDWR, 0666, $DB_HASH | |
1829 | or die "Cannot open $filename: $!\n" ; | |
1830 | ||
1831 | # Install DBM Filters | |
1832 | $db->filter_fetch_key ( sub { s/\0$// } ) ; | |
1833 | $db->filter_store_key ( sub { $_ .= "\0" } ) ; | |
1834 | $db->filter_fetch_value( sub { s/\0$// } ) ; | |
1835 | $db->filter_store_value( sub { $_ .= "\0" } ) ; | |
1836 | ||
1837 | $hash{"abc"} = "def" ; | |
1838 | my $a = $hash{"ABC"} ; | |
1839 | # ... | |
1840 | undef $db ; | |
1841 | untie %hash ; | |
1842 | ||
1843 | Hopefully the contents of each of the filters should be | |
1844 | self-explanatory. Both "fetch" filters remove the terminating NULL, | |
1845 | and both "store" filters add a terminating NULL. | |
1846 | ||
1847 | ||
1848 | =head2 Another Example -- Key is a C int. | |
1849 | ||
1850 | Here is another real-life example. By default, whenever Perl writes to | |
1851 | a DBM database it always writes the key and value as strings. So when | |
1852 | you use this: | |
1853 | ||
1854 | $hash{12345} = "soemthing" ; | |
1855 | ||
1856 | the key 12345 will get stored in the DBM database as the 5 byte string | |
1857 | "12345". If you actually want the key to be stored in the DBM database | |
1858 | as a C int, you will have to use C<pack> when writing, and C<unpack> | |
1859 | when reading. | |
1860 | ||
1861 | Here is a DBM Filter that does it: | |
1862 | ||
3245f058 | 1863 | use warnings ; |
cad2e5aa JH |
1864 | use strict ; |
1865 | use DB_File ; | |
1866 | my %hash ; | |
2359510d | 1867 | my $filename = "filt" ; |
cad2e5aa JH |
1868 | unlink $filename ; |
1869 | ||
1870 | ||
1871 | my $db = tie %hash, 'DB_File', $filename, O_CREAT|O_RDWR, 0666, $DB_HASH | |
1872 | or die "Cannot open $filename: $!\n" ; | |
1873 | ||
1874 | $db->filter_fetch_key ( sub { $_ = unpack("i", $_) } ) ; | |
1875 | $db->filter_store_key ( sub { $_ = pack ("i", $_) } ) ; | |
1876 | $hash{123} = "def" ; | |
1877 | # ... | |
1878 | undef $db ; | |
1879 | untie %hash ; | |
1880 | ||
1881 | This time only two filters have been used -- we only need to manipulate | |
1882 | the contents of the key, so it wasn't necessary to install any value | |
1883 | filters. | |
1884 | ||
f6b705ef | 1885 | =head1 HINTS AND TIPS |
3b35bae3 | 1886 | |
3b35bae3 | 1887 | |
b90e71be | 1888 | =head2 Locking: The Trouble with fd |
3b35bae3 | 1889 | |
b90e71be GS |
1890 | Until version 1.72 of this module, the recommended technique for locking |
1891 | B<DB_File> databases was to flock the filehandle returned from the "fd" | |
1892 | function. Unfortunately this technique has been shown to be fundamentally | |
1893 | flawed (Kudos to David Harris for tracking this down). Use it at your own | |
1894 | peril! | |
3b35bae3 | 1895 | |
b90e71be | 1896 | The locking technique went like this. |
cb1a09d0 | 1897 | |
2359510d SD |
1898 | $db = tie(%db, 'DB_File', 'foo.db', O_CREAT|O_RDWR, 0644) |
1899 | || die "dbcreat foo.db $!"; | |
b90e71be GS |
1900 | $fd = $db->fd; |
1901 | open(DB_FH, "+<&=$fd") || die "dup $!"; | |
1902 | flock (DB_FH, LOCK_EX) || die "flock: $!"; | |
1903 | ... | |
1904 | $db{"Tom"} = "Jerry" ; | |
1905 | ... | |
1906 | flock(DB_FH, LOCK_UN); | |
1907 | undef $db; | |
1908 | untie %db; | |
1909 | close(DB_FH); | |
cb1a09d0 | 1910 | |
b90e71be | 1911 | In simple terms, this is what happens: |
cb1a09d0 | 1912 | |
b90e71be | 1913 | =over 5 |
cb1a09d0 | 1914 | |
b90e71be | 1915 | =item 1. |
cb1a09d0 | 1916 | |
b90e71be | 1917 | Use "tie" to open the database. |
cb1a09d0 | 1918 | |
b90e71be | 1919 | =item 2. |
cb1a09d0 | 1920 | |
b90e71be | 1921 | Lock the database with fd & flock. |
cb1a09d0 | 1922 | |
b90e71be | 1923 | =item 3. |
cb1a09d0 | 1924 | |
b90e71be | 1925 | Read & Write to the database. |
cb1a09d0 | 1926 | |
b90e71be | 1927 | =item 4. |
cb1a09d0 | 1928 | |
b90e71be | 1929 | Unlock and close the database. |
cb1a09d0 | 1930 | |
b90e71be GS |
1931 | =back |
1932 | ||
1933 | Here is the crux of the problem. A side-effect of opening the B<DB_File> | |
1934 | database in step 2 is that an initial block from the database will get | |
1935 | read from disk and cached in memory. | |
1936 | ||
1937 | To see why this is a problem, consider what can happen when two processes, | |
1938 | say "A" and "B", both want to update the same B<DB_File> database | |
1939 | using the locking steps outlined above. Assume process "A" has already | |
1940 | opened the database and has a write lock, but it hasn't actually updated | |
1941 | the database yet (it has finished step 2, but not started step 3 yet). Now | |
1942 | process "B" tries to open the same database - step 1 will succeed, | |
1943 | but it will block on step 2 until process "A" releases the lock. The | |
1944 | important thing to notice here is that at this point in time both | |
1945 | processes will have cached identical initial blocks from the database. | |
1946 | ||
1947 | Now process "A" updates the database and happens to change some of the | |
1948 | data held in the initial buffer. Process "A" terminates, flushing | |
1949 | all cached data to disk and releasing the database lock. At this point | |
1950 | the database on disk will correctly reflect the changes made by process | |
1951 | "A". | |
1952 | ||
1953 | With the lock released, process "B" can now continue. It also updates the | |
1954 | database and unfortunately it too modifies the data that was in its | |
1955 | initial buffer. Once that data gets flushed to disk it will overwrite | |
1956 | some/all of the changes process "A" made to the database. | |
1957 | ||
1958 | The result of this scenario is at best a database that doesn't contain | |
1959 | what you expect. At worst the database will corrupt. | |
1960 | ||
1961 | The above won't happen every time competing process update the same | |
1962 | B<DB_File> database, but it does illustrate why the technique should | |
1963 | not be used. | |
1964 | ||
1965 | =head2 Safe ways to lock a database | |
1966 | ||
1967 | Starting with version 2.x, Berkeley DB has internal support for locking. | |
1968 | The companion module to this one, B<BerkeleyDB>, provides an interface | |
1969 | to this locking functionality. If you are serious about locking | |
1970 | Berkeley DB databases, I strongly recommend using B<BerkeleyDB>. | |
1971 | ||
1972 | If using B<BerkeleyDB> isn't an option, there are a number of modules | |
1973 | available on CPAN that can be used to implement locking. Each one | |
1974 | implements locking differently and has different goals in mind. It is | |
1975 | therefore worth knowing the difference, so that you can pick the right | |
1976 | one for your application. Here are the three locking wrappers: | |
1977 | ||
1978 | =over 5 | |
1979 | ||
1980 | =item B<Tie::DB_Lock> | |
1981 | ||
1982 | A B<DB_File> wrapper which creates copies of the database file for | |
1983 | read access, so that you have a kind of a multiversioning concurrent read | |
1984 | system. However, updates are still serial. Use for databases where reads | |
1985 | may be lengthy and consistency problems may occur. | |
1986 | ||
1987 | =item B<Tie::DB_LockFile> | |
1988 | ||
1989 | A B<DB_File> wrapper that has the ability to lock and unlock the database | |
1990 | while it is being used. Avoids the tie-before-flock problem by simply | |
1991 | re-tie-ing the database when you get or drop a lock. Because of the | |
1992 | flexibility in dropping and re-acquiring the lock in the middle of a | |
1993 | session, this can be massaged into a system that will work with long | |
1994 | updates and/or reads if the application follows the hints in the POD | |
1995 | documentation. | |
1996 | ||
1997 | =item B<DB_File::Lock> | |
1998 | ||
1999 | An extremely lightweight B<DB_File> wrapper that simply flocks a lockfile | |
2000 | before tie-ing the database and drops the lock after the untie. Allows | |
2001 | one to use the same lockfile for multiple databases to avoid deadlock | |
2002 | problems, if desired. Use for databases where updates are reads are | |
2003 | quick and simple flock locking semantics are enough. | |
2004 | ||
2005 | =back | |
cb1a09d0 | 2006 | |
68dc0745 | 2007 | =head2 Sharing Databases With C Applications |
f6b705ef | 2008 | |
2009 | There is no technical reason why a Berkeley DB database cannot be | |
2010 | shared by both a Perl and a C application. | |
2011 | ||
2012 | The vast majority of problems that are reported in this area boil down | |
2013 | to the fact that C strings are NULL terminated, whilst Perl strings are | |
cad2e5aa | 2014 | not. See L<DBM FILTERS> for a generic way to work around this problem. |
f6b705ef | 2015 | |
2016 | Here is a real example. Netscape 2.0 keeps a record of the locations you | |
2017 | visit along with the time you last visited them in a DB_HASH database. | |
2018 | This is usually stored in the file F<~/.netscape/history.db>. The key | |
2019 | field in the database is the location string and the value field is the | |
2020 | time the location was last visited stored as a 4 byte binary value. | |
2021 | ||
2022 | If you haven't already guessed, the location string is stored with a | |
2023 | terminating NULL. This means you need to be careful when accessing the | |
2024 | database. | |
2025 | ||
2026 | Here is a snippet of code that is loosely based on Tom Christiansen's | |
2027 | I<ggh> script (available from your nearest CPAN archive in | |
2028 | F<authors/id/TOMC/scripts/nshist.gz>). | |
2029 | ||
3245f058 | 2030 | use warnings ; |
610ab055 | 2031 | use strict ; |
f6b705ef | 2032 | use DB_File ; |
2033 | use Fcntl ; | |
f6b705ef | 2034 | |
962cee9f | 2035 | my ($dotdir, $HISTORY, %hist_db, $href, $binary_time, $date) ; |
f6b705ef | 2036 | $dotdir = $ENV{HOME} || $ENV{LOGNAME}; |
2037 | ||
2038 | $HISTORY = "$dotdir/.netscape/history.db"; | |
2039 | ||
2040 | tie %hist_db, 'DB_File', $HISTORY | |
2041 | or die "Cannot open $HISTORY: $!\n" ;; | |
2042 | ||
2043 | # Dump the complete database | |
2044 | while ( ($href, $binary_time) = each %hist_db ) { | |
2045 | ||
2046 | # remove the terminating NULL | |
2047 | $href =~ s/\x00$// ; | |
2048 | ||
2049 | # convert the binary time into a user friendly string | |
2050 | $date = localtime unpack("V", $binary_time); | |
2051 | print "$date $href\n" ; | |
2052 | } | |
2053 | ||
2054 | # check for the existence of a specific key | |
2055 | # remember to add the NULL | |
2056 | if ( $binary_time = $hist_db{"http://mox.perl.com/\x00"} ) { | |
2057 | $date = localtime unpack("V", $binary_time) ; | |
2058 | print "Last visited mox.perl.com on $date\n" ; | |
2059 | } | |
2060 | else { | |
2061 | print "Never visited mox.perl.com\n" | |
2062 | } | |
2063 | ||
2064 | untie %hist_db ; | |
2065 | ||
68dc0745 | 2066 | =head2 The untie() Gotcha |
778183f3 | 2067 | |
7a2e2cd6 | 2068 | If you make use of the Berkeley DB API, it is I<very> strongly |
68dc0745 | 2069 | recommended that you read L<perltie/The untie Gotcha>. |
778183f3 PM |
2070 | |
2071 | Even if you don't currently make use of the API interface, it is still | |
2072 | worth reading it. | |
2073 | ||
2074 | Here is an example which illustrates the problem from a B<DB_File> | |
2075 | perspective: | |
2076 | ||
2077 | use DB_File ; | |
2078 | use Fcntl ; | |
2079 | ||
2080 | my %x ; | |
2081 | my $X ; | |
2082 | ||
2083 | $X = tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_TRUNC | |
2084 | or die "Cannot tie first time: $!" ; | |
2085 | ||
2086 | $x{123} = 456 ; | |
2087 | ||
2088 | untie %x ; | |
2089 | ||
2090 | tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_CREAT | |
2091 | or die "Cannot tie second time: $!" ; | |
2092 | ||
2093 | untie %x ; | |
2094 | ||
2095 | When run, the script will produce this error message: | |
2096 | ||
2097 | Cannot tie second time: Invalid argument at bad.file line 14. | |
2098 | ||
2099 | Although the error message above refers to the second tie() statement | |
2100 | in the script, the source of the problem is really with the untie() | |
2101 | statement that precedes it. | |
2102 | ||
2103 | Having read L<perltie> you will probably have already guessed that the | |
2104 | error is caused by the extra copy of the tied object stored in C<$X>. | |
2105 | If you haven't, then the problem boils down to the fact that the | |
2106 | B<DB_File> destructor, DESTROY, will not be called until I<all> | |
2107 | references to the tied object are destroyed. Both the tied variable, | |
2108 | C<%x>, and C<$X> above hold a reference to the object. The call to | |
2109 | untie() will destroy the first, but C<$X> still holds a valid | |
2110 | reference, so the destructor will not get called and the database file | |
2111 | F<tst.fil> will remain open. The fact that Berkeley DB then reports the | |
b90e71be | 2112 | attempt to open a database that is already open via the catch-all |
778183f3 PM |
2113 | "Invalid argument" doesn't help. |
2114 | ||
2115 | If you run the script with the C<-w> flag the error message becomes: | |
2116 | ||
2117 | untie attempted while 1 inner references still exist at bad.file line 12. | |
2118 | Cannot tie second time: Invalid argument at bad.file line 14. | |
2119 | ||
2120 | which pinpoints the real problem. Finally the script can now be | |
2121 | modified to fix the original problem by destroying the API object | |
2122 | before the untie: | |
2123 | ||
2124 | ... | |
2125 | $x{123} = 456 ; | |
2126 | ||
2127 | undef $X ; | |
2128 | untie %x ; | |
2129 | ||
2130 | $X = tie %x, 'DB_File', 'tst.fil' , O_RDWR|O_CREAT | |
2131 | ... | |
2132 | ||
f6b705ef | 2133 | |
2134 | =head1 COMMON QUESTIONS | |
2135 | ||
2136 | =head2 Why is there Perl source in my database? | |
2137 | ||
2138 | If you look at the contents of a database file created by DB_File, | |
2139 | there can sometimes be part of a Perl script included in it. | |
2140 | ||
2141 | This happens because Berkeley DB uses dynamic memory to allocate | |
2142 | buffers which will subsequently be written to the database file. Being | |
2143 | dynamic, the memory could have been used for anything before DB | |
2144 | malloced it. As Berkeley DB doesn't clear the memory once it has been | |
2145 | allocated, the unused portions will contain random junk. In the case | |
2146 | where a Perl script gets written to the database, the random junk will | |
2147 | correspond to an area of dynamic memory that happened to be used during | |
2148 | the compilation of the script. | |
2149 | ||
2150 | Unless you don't like the possibility of there being part of your Perl | |
2151 | scripts embedded in a database file, this is nothing to worry about. | |
2152 | ||
2153 | =head2 How do I store complex data structures with DB_File? | |
2154 | ||
2155 | Although B<DB_File> cannot do this directly, there is a module which | |
2156 | can layer transparently over B<DB_File> to accomplish this feat. | |
2157 | ||
2158 | Check out the MLDBM module, available on CPAN in the directory | |
2159 | F<modules/by-module/MLDBM>. | |
2160 | ||
2161 | =head2 What does "Invalid Argument" mean? | |
2162 | ||
2163 | You will get this error message when one of the parameters in the | |
2164 | C<tie> call is wrong. Unfortunately there are quite a few parameters to | |
2165 | get wrong, so it can be difficult to figure out which one it is. | |
2166 | ||
2167 | Here are a couple of possibilities: | |
2168 | ||
2169 | =over 5 | |
2170 | ||
2171 | =item 1. | |
2172 | ||
610ab055 | 2173 | Attempting to reopen a database without closing it. |
f6b705ef | 2174 | |
2175 | =item 2. | |
2176 | ||
2177 | Using the O_WRONLY flag. | |
2178 | ||
2179 | =back | |
2180 | ||
2181 | =head2 What does "Bareword 'DB_File' not allowed" mean? | |
2182 | ||
2183 | You will encounter this particular error message when you have the | |
2184 | C<strict 'subs'> pragma (or the full strict pragma) in your script. | |
2185 | Consider this script: | |
2186 | ||
3245f058 | 2187 | use warnings ; |
f6b705ef | 2188 | use strict ; |
2189 | use DB_File ; | |
07200f1b | 2190 | my %x ; |
f6b705ef | 2191 | tie %x, DB_File, "filename" ; |
2192 | ||
2193 | Running it produces the error in question: | |
2194 | ||
2195 | Bareword "DB_File" not allowed while "strict subs" in use | |
2196 | ||
2197 | To get around the error, place the word C<DB_File> in either single or | |
2198 | double quotes, like this: | |
2199 | ||
2200 | tie %x, "DB_File", "filename" ; | |
2201 | ||
2202 | Although it might seem like a real pain, it is really worth the effort | |
2203 | of having a C<use strict> in all your scripts. | |
2204 | ||
cad2e5aa JH |
2205 | =head1 REFERENCES |
2206 | ||
2207 | Articles that are either about B<DB_File> or make use of it. | |
2208 | ||
2209 | =over 5 | |
2210 | ||
2211 | =item 1. | |
2212 | ||
2213 | I<Full-Text Searching in Perl>, Tim Kientzle (tkientzle@ddj.com), | |
2214 | Dr. Dobb's Journal, Issue 295, January 1999, pp 34-41 | |
2215 | ||
2216 | =back | |
2217 | ||
cb1a09d0 AD |
2218 | =head1 HISTORY |
2219 | ||
1f70e1ea | 2220 | Moved to the Changes file. |
610ab055 | 2221 | |
1f70e1ea | 2222 | =head1 BUGS |
05475680 | 2223 | |
1f70e1ea PM |
2224 | Some older versions of Berkeley DB had problems with fixed length |
2225 | records using the RECNO file format. This problem has been fixed since | |
2226 | version 1.85 of Berkeley DB. | |
e858de61 | 2227 | |
1f70e1ea PM |
2228 | I am sure there are bugs in the code. If you do find any, or can |
2229 | suggest any enhancements, I would welcome your comments. | |
a6ed719b | 2230 | |
1f70e1ea | 2231 | =head1 AVAILABILITY |
a6ed719b | 2232 | |
1f70e1ea PM |
2233 | B<DB_File> comes with the standard Perl source distribution. Look in |
2234 | the directory F<ext/DB_File>. Given the amount of time between releases | |
2235 | of Perl the version that ships with Perl is quite likely to be out of | |
2236 | date, so the most recent version can always be found on CPAN (see | |
5bbd4290 | 2237 | L<perlmodlib/CPAN> for details), in the directory |
1f70e1ea | 2238 | F<modules/by-module/DB_File>. |
a6ed719b | 2239 | |
039d031f PM |
2240 | This version of B<DB_File> will work with either version 1.x, 2.x or |
2241 | 3.x of Berkeley DB, but is limited to the functionality provided by | |
2242 | version 1. | |
a6ed719b | 2243 | |
cad2e5aa | 2244 | The official web site for Berkeley DB is F<http://www.sleepycat.com>. |
039d031f | 2245 | All versions of Berkeley DB are available there. |
93af7a87 | 2246 | |
1f70e1ea PM |
2247 | Alternatively, Berkeley DB version 1 is available at your nearest CPAN |
2248 | archive in F<src/misc/db.1.85.tar.gz>. | |
e858de61 | 2249 | |
1f70e1ea PM |
2250 | If you are running IRIX, then get Berkeley DB version 1 from |
2251 | F<http://reality.sgi.com/ariel>. It has the patches necessary to | |
2252 | compile properly on IRIX 5.3. | |
610ab055 | 2253 | |
1f70e1ea | 2254 | =head1 COPYRIGHT |
3b35bae3 | 2255 | |
dcdb9d1f | 2256 | Copyright (c) 1995-2005 Paul Marquess. All rights reserved. This program |
a9fd575d PM |
2257 | is free software; you can redistribute it and/or modify it under the |
2258 | same terms as Perl itself. | |
3b35bae3 | 2259 | |
1f70e1ea PM |
2260 | Although B<DB_File> is covered by the Perl license, the library it |
2261 | makes use of, namely Berkeley DB, is not. Berkeley DB has its own | |
2262 | copyright and its own license. Please take the time to read it. | |
3b35bae3 | 2263 | |
a9fd575d | 2264 | Here are are few words taken from the Berkeley DB FAQ (at |
b90e71be | 2265 | F<http://www.sleepycat.com>) regarding the license: |
68dc0745 | 2266 | |
a9fd575d | 2267 | Do I have to license DB to use it in Perl scripts? |
3b35bae3 | 2268 | |
a9fd575d PM |
2269 | No. The Berkeley DB license requires that software that uses |
2270 | Berkeley DB be freely redistributable. In the case of Perl, that | |
2271 | software is Perl, and not your scripts. Any Perl scripts that you | |
2272 | write are your property, including scripts that make use of | |
2273 | Berkeley DB. Neither the Perl license nor the Berkeley DB license | |
2274 | place any restriction on what you may do with them. | |
88108326 | 2275 | |
1f70e1ea PM |
2276 | If you are in any doubt about the license situation, contact either the |
2277 | Berkeley DB authors or the author of DB_File. See L<"AUTHOR"> for details. | |
a0b8c8c1 PM |
2278 | |
2279 | ||
3b35bae3 AD |
2280 | =head1 SEE ALSO |
2281 | ||
5bbd4290 PM |
2282 | L<perl>, L<dbopen(3)>, L<hash(3)>, L<recno(3)>, L<btree(3)>, |
2283 | L<perldbmfilter> | |
3b35bae3 | 2284 | |
3b35bae3 AD |
2285 | =head1 AUTHOR |
2286 | ||
8e07c86e | 2287 | The DB_File interface was written by Paul Marquess |
5bbd4290 | 2288 | E<lt>pmqs@cpan.orgE<gt>. |
d3ef3b8a | 2289 | Questions about the DB system itself may be addressed to |
5bbd4290 | 2290 | E<lt>db@sleepycat.comE<gt>. |
3b35bae3 AD |
2291 | |
2292 | =cut |