3 |
@EXPORT = qw/ |
@EXPORT = qw/ |
4 |
_set_rec _set_lookup |
_set_rec _set_lookup |
5 |
_get_ds _clean_ds |
_get_ds _clean_ds |
6 |
|
_debug |
7 |
|
|
8 |
tag search display |
tag search display |
9 |
marc21 |
marc marc_indicators marc_repeatable_subfield |
10 |
|
marc_compose marc_leader |
11 |
|
|
12 |
rec1 rec2 rec |
rec1 rec2 rec |
13 |
regex prefix suffix surround |
regex prefix suffix surround |
14 |
first lookup join_with |
first lookup join_with |
15 |
|
|
16 |
|
split_rec_on |
17 |
/; |
/; |
18 |
|
|
19 |
use warnings; |
use warnings; |
20 |
use strict; |
use strict; |
21 |
|
|
22 |
#use base qw/WebPAC::Common/; |
#use base qw/WebPAC::Common/; |
23 |
use Data::Dumper; |
use Data::Dump qw/dump/; |
24 |
use Encode qw/from_to/; |
use Encode qw/from_to/; |
25 |
|
|
26 |
|
# debugging warn(s) |
27 |
|
my $debug = 0; |
28 |
|
|
29 |
|
|
30 |
=head1 NAME |
=head1 NAME |
31 |
|
|
32 |
WebPAC::Normalize - describe normalisaton rules using sets |
WebPAC::Normalize - describe normalisaton rules using sets |
33 |
|
|
34 |
=head1 VERSION |
=head1 VERSION |
35 |
|
|
36 |
Version 0.06 |
Version 0.09 |
37 |
|
|
38 |
=cut |
=cut |
39 |
|
|
40 |
our $VERSION = '0.06'; |
our $VERSION = '0.09'; |
41 |
|
|
42 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
43 |
|
|
51 |
|
|
52 |
Normalisation can generate multiple output normalized data. For now, supported output |
Normalisation can generate multiple output normalized data. For now, supported output |
53 |
types (on the left side of definition) are: C<tag>, C<display>, C<search> and |
types (on the left side of definition) are: C<tag>, C<display>, C<search> and |
54 |
C<marc21>. |
C<marc>. |
55 |
|
|
56 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
57 |
|
|
117 |
|
|
118 |
=cut |
=cut |
119 |
|
|
120 |
my $out; |
my ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators); |
|
my $marc21; |
|
|
my $marc_encoding; |
|
121 |
|
|
122 |
sub _get_ds { |
sub _get_ds { |
123 |
return $out; |
return $out; |
133 |
|
|
134 |
sub _clean_ds { |
sub _clean_ds { |
135 |
my $a = {@_}; |
my $a = {@_}; |
136 |
$out = undef; |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = (); |
|
$marc21 = undef; |
|
137 |
$marc_encoding = $a->{marc_encoding}; |
$marc_encoding = $a->{marc_encoding}; |
138 |
} |
} |
139 |
|
|
151 |
$lookup = shift; |
$lookup = shift; |
152 |
} |
} |
153 |
|
|
154 |
=head2 _get_marc21_fields |
=head2 _get_marc_fields |
|
|
|
|
Get all fields defined by calls to C<marc21> |
|
|
|
|
|
$marc->add_fields( WebPAC::Normalize:_get_marc21_fields() ); |
|
155 |
|
|
156 |
|
Get all fields defined by calls to C<marc> |
157 |
|
|
158 |
|
$marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); |
159 |
|
|
160 |
We are using I<magic> which detect repeatable fields only from |
We are using I<magic> which detect repeatable fields only from |
161 |
sequence of field/subfield data generated by normalization. |
sequence of field/subfield data generated by normalization. |
162 |
|
|
163 |
Repeatable field is created if there is second occurence of same subfield or |
Repeatable field is created when there is second occurence of same subfield or |
164 |
if any of indicators are different. This is sane for most cases except for |
if any of indicators are different. |
165 |
non-repeatable fields with repeatable subfields. |
|
166 |
|
This is sane for most cases. Something like: |
167 |
|
|
168 |
|
900a-1 900b-1 900c-1 |
169 |
|
900a-2 900b-2 |
170 |
|
900a-3 |
171 |
|
|
172 |
|
will be created from any combination of: |
173 |
|
|
174 |
|
900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 |
175 |
|
|
176 |
|
and following rules: |
177 |
|
|
178 |
|
marc('900','a', rec('200','a') ); |
179 |
|
marc('900','b', rec('200','b') ); |
180 |
|
marc('900','c', rec('200','c') ); |
181 |
|
|
182 |
|
which might not be what you have in mind. If you need repeatable subfield, |
183 |
|
define it using C<marc_repeatable_subfield> like this: |
184 |
|
|
185 |
B<TODO>: implement exceptions to magic |
.... |
186 |
|
|
187 |
=cut |
=cut |
188 |
|
|
189 |
sub _get_marc21_fields { |
sub _get_marc_fields { |
190 |
|
|
191 |
|
return if (! $marc_record || ref($marc_record) ne 'ARRAY' || $#{ $marc_record } < 0); |
192 |
|
|
193 |
|
# first, sort all existing fields |
194 |
|
# XXX might not be needed, but modern perl might randomize elements in hash |
195 |
|
my @sorted_marc_record = sort { |
196 |
|
$a->[0] . $a->[3] cmp $b->[0] . $b->[3] |
197 |
|
} @{ $marc_record }; |
198 |
|
|
199 |
|
@sorted_marc_record = @{ $marc_record }; ### FIXME disable sorting |
200 |
|
|
201 |
|
# output marc fields |
202 |
my @m; |
my @m; |
203 |
my $last; |
|
204 |
foreach my $row (@{ $marc21 }) { |
# count unique field-subfields (used for offset when walking to next subfield) |
205 |
if ($last && |
my $u; |
206 |
$last->[0] eq $row->[0] && # check if field is same |
map { $u->{ $_->[0] . $_->[3] }++ } @sorted_marc_record; |
207 |
$last->[1] eq $row->[1] && # check for i1 |
|
208 |
$last->[2] eq $row->[2] && # and for i2 |
if ($debug) { |
209 |
$last->[3] ne $row->[3] # and subfield is different |
warn "## marc_repeatable_subfield ", dump( $marc_repeatable_subfield ), $/; |
210 |
) { |
warn "## marc_record ", dump( $marc_record ), $/; |
211 |
push @$last, ( $row->[3] , $row->[4] ); |
warn "## sorted_marc_record ", dump( \@sorted_marc_record ), $/; |
212 |
warn "## ++ added $row->[0] ^$row->[3] to $last->[0]\n"; |
warn "## subfield count ", dump( $u ), $/; |
213 |
next; |
} |
214 |
} elsif ($last) { |
|
215 |
push @m, $last; |
my $len = $#sorted_marc_record; |
216 |
|
my $visited; |
217 |
|
my $i = 0; |
218 |
|
my $field; |
219 |
|
|
220 |
|
foreach ( 0 .. $len ) { |
221 |
|
|
222 |
|
# find next element which isn't visited |
223 |
|
while ($visited->{$i}) { |
224 |
|
$i = ($i + 1) % ($len + 1); |
225 |
} |
} |
226 |
|
|
227 |
$last = $row; |
# mark it visited |
228 |
|
$visited->{$i}++; |
229 |
|
|
230 |
|
my $row = $sorted_marc_record[$i]; |
231 |
|
|
232 |
|
# field and subfield which is key for |
233 |
|
# marc_repeatable_subfield and u |
234 |
|
my $fsf = $row->[0] . $row->[3]; |
235 |
|
|
236 |
|
if ($debug > 1) { |
237 |
|
|
238 |
|
print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; |
239 |
|
print "### this [$i]: ", dump( $row ),$/; |
240 |
|
print "### sf: ", $row->[3], " vs ", $field->[3], |
241 |
|
$marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, |
242 |
|
if ($#$field >= 0); |
243 |
|
|
244 |
|
} |
245 |
|
|
246 |
|
# if field exists |
247 |
|
if ( $#$field >= 0 ) { |
248 |
|
if ( |
249 |
|
$row->[0] ne $field->[0] || # field |
250 |
|
$row->[1] ne $field->[1] || # i1 |
251 |
|
$row->[2] ne $field->[2] # i2 |
252 |
|
) { |
253 |
|
push @m, $field; |
254 |
|
warn "## saved/1 ", dump( $field ),$/ if ($debug); |
255 |
|
$field = $row; |
256 |
|
|
257 |
|
} elsif ( |
258 |
|
( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) |
259 |
|
|| |
260 |
|
( $row->[3] eq $field->[-2] && # same subfield, but not repeatable |
261 |
|
! $marc_repeatable_subfield->{ $fsf } |
262 |
|
) |
263 |
|
) { |
264 |
|
push @m, $field; |
265 |
|
warn "## saved/2 ", dump( $field ),$/ if ($debug); |
266 |
|
$field = $row; |
267 |
|
|
268 |
|
} else { |
269 |
|
# append new subfields to existing field |
270 |
|
push @$field, ( $row->[3], $row->[4] ); |
271 |
|
} |
272 |
|
} else { |
273 |
|
# insert first field |
274 |
|
$field = $row; |
275 |
|
} |
276 |
|
|
277 |
|
if (! $marc_repeatable_subfield->{ $fsf }) { |
278 |
|
# make step to next subfield |
279 |
|
$i = ($i + $u->{ $fsf } ) % ($len + 1); |
280 |
|
} |
281 |
} |
} |
282 |
|
|
283 |
push @m, $last if ($last); |
if ($#$field >= 0) { |
284 |
|
push @m, $field; |
285 |
|
warn "## saved/3 ", dump( $field ),$/ if ($debug); |
286 |
|
} |
287 |
|
|
288 |
return @m; |
return @m; |
289 |
} |
} |
290 |
|
|
291 |
|
=head2 _debug |
292 |
|
|
293 |
|
Change level of debug warnings |
294 |
|
|
295 |
|
_debug( 2 ); |
296 |
|
|
297 |
|
=cut |
298 |
|
|
299 |
|
sub _debug { |
300 |
|
my $l = shift; |
301 |
|
return $debug unless defined($l); |
302 |
|
warn "debug level $l",$/ if ($l > 0); |
303 |
|
$debug = $l; |
304 |
|
} |
305 |
|
|
306 |
=head1 Functions to create C<data_structure> |
=head1 Functions to create C<data_structure> |
307 |
|
|
308 |
Those functions generally have to first in your normalization file. |
Those functions generally have to first in your normalization file. |
357 |
$out->{$name}->{search} = \@o; |
$out->{$name}->{search} = \@o; |
358 |
} |
} |
359 |
|
|
360 |
=head2 marc21 |
=head2 marc_leader |
361 |
|
|
362 |
|
Setup fields within MARC leader or get leader |
363 |
|
|
364 |
|
marc_leader('05','c'); |
365 |
|
my $leader = marc_leader(); |
366 |
|
|
367 |
|
=cut |
368 |
|
|
369 |
|
sub marc_leader { |
370 |
|
my ($offset,$value) = @_; |
371 |
|
|
372 |
|
if ($offset) { |
373 |
|
$out->{' leader'}->{ $offset } = $value; |
374 |
|
} else { |
375 |
|
return $out->{' leader'}; |
376 |
|
} |
377 |
|
} |
378 |
|
|
379 |
|
=head2 marc |
380 |
|
|
381 |
Save value for MARC field |
Save value for MARC field |
382 |
|
|
383 |
marc21('900','a', rec('200','a') ); |
marc('900','a', rec('200','a') ); |
384 |
|
|
385 |
=cut |
=cut |
386 |
|
|
387 |
sub marc21 { |
sub marc { |
388 |
my $f = shift or die "marc21 needs field"; |
my $f = shift or die "marc needs field"; |
389 |
die "marc21 field must be numer" unless ($f =~ /^\d+$/); |
die "marc field must be numer" unless ($f =~ /^\d+$/); |
390 |
|
|
391 |
my $sf = shift or die "marc21 needs subfield"; |
my $sf = shift or die "marc needs subfield"; |
392 |
|
|
393 |
foreach (@_) { |
foreach (@_) { |
394 |
my $v = $_; # make var read-write for Encode |
my $v = $_; # make var read-write for Encode |
395 |
next unless (defined($v) && $v !~ /^\s*$/); |
next unless (defined($v) && $v !~ /^\s*$/); |
396 |
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
397 |
push @{ $marc21 }, [ $f, ' ', ' ', $sf => $v ]; |
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
398 |
|
push @{ $marc_record }, [ $f, $i1, $i2, $sf => $v ]; |
399 |
} |
} |
400 |
} |
} |
401 |
|
|
402 |
|
=head2 marc_repeatable_subfield |
403 |
|
|
404 |
|
Save values for MARC repetable subfield |
405 |
|
|
406 |
|
marc_repeatable_subfield('910', 'z', rec('909') ); |
407 |
|
|
408 |
|
=cut |
409 |
|
|
410 |
|
sub marc_repeatable_subfield { |
411 |
|
my ($f,$sf) = @_; |
412 |
|
die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); |
413 |
|
$marc_repeatable_subfield->{ $f . $sf }++; |
414 |
|
marc(@_); |
415 |
|
} |
416 |
|
|
417 |
|
=head2 marc_indicators |
418 |
|
|
419 |
|
Set both indicators for MARC field |
420 |
|
|
421 |
|
marc_indicators('900', ' ', 1); |
422 |
|
|
423 |
|
Any indicator value other than C<0-9> will be treated as undefined. |
424 |
|
|
425 |
|
=cut |
426 |
|
|
427 |
|
sub marc_indicators { |
428 |
|
my $f = shift || die "marc_indicators need field!\n"; |
429 |
|
my ($i1,$i2) = @_; |
430 |
|
die "marc_indicators($f, ...) need i1!\n" unless(defined($i1)); |
431 |
|
die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2)); |
432 |
|
|
433 |
|
$i1 = ' ' if ($i1 !~ /^\d$/); |
434 |
|
$i2 = ' ' if ($i2 !~ /^\d$/); |
435 |
|
@{ $marc_indicators->{$f} } = ($i1,$i2); |
436 |
|
} |
437 |
|
|
438 |
|
=head2 marc_compose |
439 |
|
|
440 |
|
Save values for each MARC subfield explicitly |
441 |
|
|
442 |
|
marc_compose('900', |
443 |
|
'a', rec('200','a') |
444 |
|
'b', rec('201','a') |
445 |
|
'a', rec('200','b') |
446 |
|
'c', rec('200','c') |
447 |
|
); |
448 |
|
|
449 |
|
=cut |
450 |
|
|
451 |
|
sub marc_compose { |
452 |
|
my $f = shift or die "marc_compose needs field"; |
453 |
|
die "marc_compose field must be numer" unless ($f =~ /^\d+$/); |
454 |
|
|
455 |
|
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
456 |
|
my $m = [ $f, $i1, $i2 ]; |
457 |
|
|
458 |
|
while (@_) { |
459 |
|
my $sf = shift or die "marc_compose $f needs subfield"; |
460 |
|
my $v = shift; |
461 |
|
|
462 |
|
next unless (defined($v) && $v !~ /^\s*$/); |
463 |
|
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
464 |
|
push @$m, ( $sf, $v ); |
465 |
|
warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); |
466 |
|
} |
467 |
|
|
468 |
|
warn "## marc_compose(d) ", dump( $m ),$/ if ($debug > 1); |
469 |
|
|
470 |
|
push @{ $marc_record }, $m if ($#{$m} > 2); |
471 |
|
} |
472 |
|
|
473 |
|
|
474 |
=head1 Functions to extract data from input |
=head1 Functions to extract data from input |
475 |
|
|
476 |
This function should be used inside functions to create C<data_structure> described |
This function should be used inside functions to create C<data_structure> described |
545 |
sub regex { |
sub regex { |
546 |
my $r = shift; |
my $r = shift; |
547 |
my @out; |
my @out; |
548 |
#warn "r: $r\n",Dumper(\@_); |
#warn "r: $r\n", dump(\@_); |
549 |
foreach my $t (@_) { |
foreach my $t (@_) { |
550 |
next unless ($t); |
next unless ($t); |
551 |
eval "\$t =~ $r"; |
eval "\$t =~ $r"; |
639 |
return join($d, grep { defined($_) && $_ ne '' } @_); |
return join($d, grep { defined($_) && $_ ne '' } @_); |
640 |
} |
} |
641 |
|
|
642 |
|
=head2 split_rec_on |
643 |
|
|
644 |
|
Split record subfield on some regex and take one of parts out |
645 |
|
|
646 |
|
$a_before_semi_column = |
647 |
|
split_rec_on('200','a', /\s*;\s*/, $part); |
648 |
|
|
649 |
|
C<$part> is optional number of element. First element is |
650 |
|
B<1>, not 0! |
651 |
|
|
652 |
|
If there is no C<$part> parameter or C<$part> is 0, this function will |
653 |
|
return all values produced by splitting. |
654 |
|
|
655 |
|
=cut |
656 |
|
|
657 |
|
sub split_rec_on { |
658 |
|
die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2); |
659 |
|
|
660 |
|
my ($fld, $sf, $regex, $part) = @_; |
661 |
|
warn "### regex ", ref($regex), $regex, $/ if ($debug > 2); |
662 |
|
|
663 |
|
my @r = rec( $fld, $sf ); |
664 |
|
my $v = shift @r; |
665 |
|
warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2); |
666 |
|
|
667 |
|
return '' if( ! defined($v) || $v =~ /^\s*$/); |
668 |
|
|
669 |
|
my @s = split( $regex, $v ); |
670 |
|
warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1); |
671 |
|
if ($part && $part > 0) { |
672 |
|
return $s[ $part - 1 ]; |
673 |
|
} else { |
674 |
|
return [ @s ]; |
675 |
|
} |
676 |
|
} |
677 |
|
|
678 |
# END |
# END |
679 |
1; |
1; |