|
|
|
1 |
package MARC::Fast; |
package MARC::Fast; |
2 |
|
|
3 |
use strict; |
use strict; |
4 |
use Carp; |
use Carp; |
5 |
use Data::Dumper; |
use Data::Dump qw/dump/; |
6 |
|
|
7 |
BEGIN { |
BEGIN { |
8 |
use Exporter (); |
use Exporter (); |
9 |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
10 |
$VERSION = 0.04; |
$VERSION = 0.12; |
11 |
@ISA = qw (Exporter); |
@ISA = qw (Exporter); |
12 |
#Give a hoot don't pollute, do not export more than needed by default |
#Give a hoot don't pollute, do not export more than needed by default |
13 |
@EXPORT = qw (); |
@EXPORT = qw (); |
23 |
|
|
24 |
use MARC::Fast; |
use MARC::Fast; |
25 |
|
|
26 |
|
my $marc = new MARC::Fast( |
27 |
|
marcdb => 'unimarc.iso', |
28 |
|
); |
29 |
|
|
30 |
|
foreach my $mfn ( 1 .. $marc->count ) { |
31 |
|
print $marc->to_ascii( $mfn ); |
32 |
|
} |
33 |
|
|
34 |
|
For longer example with command line options look at L<scripts/dump_fastmarc.pl> |
35 |
|
|
36 |
=head1 DESCRIPTION |
=head1 DESCRIPTION |
37 |
|
|
38 |
This is very fast alternative to C<MARC> and C<MARC::Record> modules. |
This is very fast alternative to C<MARC> and C<MARC::Record> modules. |
39 |
|
|
40 |
It's is also very sutable for random access to MARC records (as opposed to |
It's is also very subtable for random access to MARC records (as opposed to |
41 |
sequential one). |
sequential one). |
42 |
|
|
43 |
=head1 METHODS |
=head1 METHODS |
87 |
my $len = read($self->{fh}, $leader, 24); |
my $len = read($self->{fh}, $leader, 24); |
88 |
|
|
89 |
if ($len < 24) { |
if ($len < 24) { |
90 |
carp "short read of leader, aborting\n"; |
warn "short read of leader, aborting\n"; |
91 |
|
$self->{count}--; |
92 |
last; |
last; |
93 |
} |
} |
94 |
|
|
121 |
print STDERR "REC ",$self->{count},": $leader\n" if ($self->{debug}); |
print STDERR "REC ",$self->{count},": $leader\n" if ($self->{debug}); |
122 |
|
|
123 |
# store leader for later |
# store leader for later |
124 |
push @{$self->{leaders}}, $leader; |
push @{$self->{leader}}, $leader; |
125 |
|
|
126 |
# skip to next record |
# skip to next record |
127 |
my $o = substr($leader,0,5); |
my $o = substr($leader,0,5); |
128 |
|
warn "# in record ", $self->{count}," record length isn't number but: ",dump($o),"\n" unless $o =~ m/^\d+$/; |
129 |
if ($o > 24) { |
if ($o > 24) { |
130 |
seek($self->{fh},$o-24,1) if ($o); |
seek($self->{fh},$o-24,1) if ($o); |
131 |
} else { |
} else { |
156 |
|
|
157 |
my $hash = $marc->fetch(42); |
my $hash = $marc->fetch(42); |
158 |
|
|
159 |
|
First record number is C<1> |
160 |
|
|
161 |
=cut |
=cut |
162 |
|
|
163 |
sub fetch { |
sub fetch { |
164 |
my $self = shift; |
my $self = shift; |
165 |
|
|
166 |
my $rec_nr = shift || return; |
my $rec_nr = shift; |
167 |
|
|
168 |
|
if ( ! $rec_nr ) { |
169 |
|
$self->{last_leader} = undef; |
170 |
|
return; |
171 |
|
} |
172 |
|
|
173 |
my $leader = $self->{leaders}->[$rec_nr - 1]; |
my $leader = $self->{leader}->[$rec_nr - 1]; |
174 |
|
$self->{last_leader} = $leader; |
175 |
unless ($leader) { |
unless ($leader) { |
176 |
carp "can't find record $rec_nr"; |
carp "can't find record $rec_nr"; |
177 |
return; |
return; |
253 |
} |
} |
254 |
|
|
255 |
|
|
256 |
|
=head2 last_leader |
257 |
|
|
258 |
|
Returns leader of last record L<fetch>ed |
259 |
|
|
260 |
|
print $marc->last_leader; |
261 |
|
|
262 |
|
Added in version 0.08 of this module, so if you need it use: |
263 |
|
|
264 |
|
use MARC::Fast 0.08; |
265 |
|
|
266 |
|
to be sure that it's supported. |
267 |
|
|
268 |
|
=cut |
269 |
|
|
270 |
|
sub last_leader { |
271 |
|
my $self = shift; |
272 |
|
return $self->{last_leader}; |
273 |
|
} |
274 |
|
|
275 |
|
|
276 |
=head2 to_hash |
=head2 to_hash |
277 |
|
|
278 |
Read record with specified MFN and convert it to hash |
Read record with specified MFN and convert it to hash |
279 |
|
|
280 |
my $hash = $marc->to_hash($mfn); |
my $hash = $marc->to_hash( $mfn, include_subfields => 1, |
281 |
|
hash_filter => sub { my ($l,$tag) = @_; return $l; } |
282 |
|
); |
283 |
|
|
284 |
It has ability to convert characters (using C<hash_filter>) from MARC |
It has ability to convert characters (using C<hash_filter>) from MARC |
285 |
database before creating structures enabling character re-mapping or quick |
database before creating structures enabling character re-mapping or quick |
286 |
fix-up of data. |
fix-up of data. If you specified C<hash_filter> both in C<new> and C<to_hash> |
287 |
|
only the one from C<to_hash> will be used. |
288 |
|
|
289 |
This function returns hash which is like this: |
This function returns hash which is like this: |
290 |
|
|
307 |
|
|
308 |
my $mfn = shift || confess "need mfn!"; |
my $mfn = shift || confess "need mfn!"; |
309 |
|
|
310 |
|
my $args = {@_}; |
311 |
|
my $filter_coderef = $args->{'hash_filter'} || $self->{'hash_filter'}; |
312 |
|
|
313 |
# init record to include MFN as field 000 |
# init record to include MFN as field 000 |
314 |
my $rec = { '000' => [ $mfn ] }; |
my $rec = { '000' => [ $mfn ] }; |
315 |
|
|
316 |
my $row = $self->fetch($mfn) || return; |
my $row = $self->fetch($mfn) || return; |
317 |
|
|
318 |
foreach my $rec_nr (keys %{$row}) { |
foreach my $tag (keys %{$row}) { |
319 |
foreach my $l (@{$row->{$rec_nr}}) { |
foreach my $l (@{$row->{$tag}}) { |
320 |
|
|
321 |
# remove end marker |
# remove end marker |
322 |
$l =~ s/\x1E$//; |
$l =~ s/\x1E$//; |
323 |
|
|
324 |
# filter output |
# filter output |
325 |
$l = $self->{'hash_filter'}->($l, $rec_nr) if ($self->{'hash_filter'}); |
$l = $filter_coderef->($l, $tag) if $filter_coderef; |
326 |
|
|
327 |
my $val; |
my $val; |
328 |
|
|
329 |
# has identifiers? |
# has identifiers? |
330 |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\x1F/\x1F/); |
($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\x1F/\x1F/); |
331 |
|
|
332 |
|
my $sf_usage; |
333 |
|
my @subfields; |
334 |
|
|
335 |
# has subfields? |
# has subfields? |
336 |
if ($l =~ m/\x1F/) { |
if ($l =~ m/\x1F/) { |
337 |
foreach my $t (split(/\x1F/,$l)) { |
foreach my $t (split(/\x1F/,$l)) { |
338 |
next if (! $t); |
next if (! $t); |
339 |
my $f = substr($t,0,1); |
my $f = substr($t,0,1); |
340 |
# repeatable subfileds. When we hit first one, |
my $v = substr($t,1); |
341 |
# store CURRENT (up to that) in first repetition |
|
342 |
# of this record. Then, new record with same |
push @subfields, ( $f, $sf_usage->{$f}++ || 0 ); |
343 |
# identifiers will be created. |
|
344 |
if ($val->{$f}) { |
# repeatable subfiled -- convert it to array |
345 |
push @{$rec->{$rec_nr}}, $val; |
if ( defined $val->{$f} ) { |
346 |
$val = { |
if ( ref($val->{$f}) ne 'ARRAY' ) { |
347 |
i1 => $val->{i1}, |
$val->{$f} = [ $val->{$f}, $v ]; |
348 |
i2 => $val->{i2}, |
} else { |
349 |
}; |
push @{$val->{$f}}, $v; |
350 |
|
} |
351 |
|
} else { |
352 |
|
$val->{$f} = $v; |
353 |
} |
} |
|
$val->{substr($t,0,1)} = substr($t,1); |
|
354 |
} |
} |
355 |
|
$val->{subfields} = [ @subfields ] if $args->{include_subfields}; |
356 |
} else { |
} else { |
357 |
$val = $l; |
$val = $l; |
358 |
} |
} |
359 |
|
|
360 |
push @{$rec->{$rec_nr}}, $val; |
push @{$rec->{$tag}}, $val; |
361 |
} |
} |
362 |
} |
} |
363 |
|
|
364 |
return $rec; |
return $rec; |
365 |
} |
} |
366 |
|
|
367 |
|
=head2 to_ascii |
368 |
|
|
369 |
|
print $marc->to_ascii( 42 ); |
370 |
|
|
371 |
|
=cut |
372 |
|
|
373 |
|
sub to_ascii { |
374 |
|
my $self = shift; |
375 |
|
|
376 |
|
my $mfn = shift || confess "need mfn"; |
377 |
|
my $row = $self->fetch($mfn) || return; |
378 |
|
|
379 |
|
my $out; |
380 |
|
|
381 |
|
foreach my $f (sort keys %{$row}) { |
382 |
|
my $dump = join('', @{ $row->{$f} }); |
383 |
|
$dump =~ s/\x1e$//; |
384 |
|
$dump =~ s/\x1f/\$/g; |
385 |
|
$out .= "$f\t$dump\n"; |
386 |
|
} |
387 |
|
|
388 |
|
return $out; |
389 |
|
} |
390 |
|
|
391 |
1; |
1; |
392 |
__END__ |
__END__ |
393 |
|
|
394 |
=head1 BUGS |
=head1 UTF-8 ENCODING |
395 |
|
|
396 |
|
This module does nothing with encoding. But, since MARC format is byte |
397 |
|
oriented even when using UTF-8 which has variable number of bytes for each |
398 |
|
character, file is opened in binary mode. |
399 |
|
|
400 |
|
As a result, all scalars recturned to perl don't have utf-8 flag. Solution is |
401 |
|
to use C<hash_filter> and L<Encode> to decode utf-8 encoding like this: |
402 |
|
|
403 |
=head1 SUPPORT |
use Encode; |
404 |
|
|
405 |
|
my $marc = new MARC::Fast( |
406 |
|
marcdb => 'utf8.marc', |
407 |
|
hash_filter => sub { |
408 |
|
Encode::decode( 'utf-8', $_[0] ); |
409 |
|
}, |
410 |
|
); |
411 |
|
|
412 |
|
This will affect C<to_hash>, but C<fetch> will still return binary representation |
413 |
|
since it doesn't support C<hash_filter>. |
414 |
|
|
415 |
=head1 AUTHOR |
=head1 AUTHOR |
416 |
|
|
430 |
|
|
431 |
=head1 SEE ALSO |
=head1 SEE ALSO |
432 |
|
|
433 |
perl(1). |
L<Biblio::Isis>, perl(1). |
434 |
|
|
435 |
=cut |
=cut |