1 |
package WebPAC::Normalize; |
package WebPAC::Normalize; |
2 |
use Exporter 'import'; |
use Exporter 'import'; |
3 |
@EXPORT = qw/ |
our @EXPORT = qw/ |
4 |
_set_rec _set_lookup |
_set_rec _set_lookup |
5 |
|
_set_load_row |
6 |
_get_ds _clean_ds |
_get_ds _clean_ds |
7 |
_debug |
_debug |
8 |
|
_pack_subfields_hash |
9 |
|
|
10 |
|
search_display search display sorted |
11 |
|
|
|
tag search display |
|
12 |
marc marc_indicators marc_repeatable_subfield |
marc marc_indicators marc_repeatable_subfield |
13 |
marc_compose marc_leader |
marc_compose marc_leader marc_fixed |
14 |
marc_duplicate marc_remove |
marc_duplicate marc_remove marc_count |
15 |
|
marc_original_order |
16 |
|
|
17 |
rec1 rec2 rec |
rec1 rec2 rec |
18 |
|
frec |
19 |
regex prefix suffix surround |
regex prefix suffix surround |
20 |
first lookup join_with |
first lookup join_with |
21 |
|
save_into_lookup |
22 |
|
|
23 |
split_rec_on |
split_rec_on |
24 |
|
|
25 |
|
get set |
26 |
|
count |
27 |
|
|
28 |
/; |
/; |
29 |
|
|
30 |
use warnings; |
use warnings; |
32 |
|
|
33 |
#use base qw/WebPAC::Common/; |
#use base qw/WebPAC::Common/; |
34 |
use Data::Dump qw/dump/; |
use Data::Dump qw/dump/; |
|
use Encode qw/from_to/; |
|
35 |
use Storable qw/dclone/; |
use Storable qw/dclone/; |
36 |
|
use Carp qw/confess/; |
37 |
|
|
38 |
# debugging warn(s) |
# debugging warn(s) |
39 |
my $debug = 0; |
my $debug = 0; |
40 |
|
|
41 |
|
use WebPAC::Normalize::ISBN; |
42 |
|
push @EXPORT, ( 'isbn_10', 'isbn_13' ); |
43 |
|
|
44 |
=head1 NAME |
=head1 NAME |
45 |
|
|
46 |
WebPAC::Normalize - describe normalisaton rules using sets |
WebPAC::Normalize - describe normalisaton rules using sets |
47 |
|
|
|
=head1 VERSION |
|
|
|
|
|
Version 0.14 |
|
|
|
|
48 |
=cut |
=cut |
49 |
|
|
50 |
our $VERSION = '0.14'; |
our $VERSION = '0.31'; |
51 |
|
|
52 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
53 |
|
|
60 |
C<perl -c normalize.pl>. |
C<perl -c normalize.pl>. |
61 |
|
|
62 |
Normalisation can generate multiple output normalized data. For now, supported output |
Normalisation can generate multiple output normalized data. For now, supported output |
63 |
types (on the left side of definition) are: C<tag>, C<display>, C<search> and |
types (on the left side of definition) are: C<search_display>, C<display>, C<search> and |
64 |
C<marc>. |
C<marc>. |
65 |
|
|
66 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
73 |
Return data structure |
Return data structure |
74 |
|
|
75 |
my $ds = WebPAC::Normalize::data_structure( |
my $ds = WebPAC::Normalize::data_structure( |
76 |
lookup => $lookup->lookup_hash, |
lookup => $lookup_hash, |
77 |
row => $row, |
row => $row, |
78 |
rules => $normalize_pl_config, |
rules => $normalize_pl_config, |
79 |
marc_encoding => 'utf-8', |
marc_encoding => 'utf-8', |
80 |
config => $config, |
config => $config, |
81 |
|
load_row_coderef => sub { |
82 |
|
my ($database,$input,$mfn) = @_; |
83 |
|
$store->load_row( database => $database, input => $input, id => $mfn ); |
84 |
|
}, |
85 |
); |
); |
86 |
|
|
87 |
Options C<lookup>, C<row>, C<rules> and C<log> are mandatory while all |
Options C<row>, C<rules> and C<log> are mandatory while all |
88 |
other are optional. |
other are optional. |
89 |
|
|
90 |
|
C<load_row_coderef> is closure only used when executing lookups, so they will |
91 |
|
die if it's not defined. |
92 |
|
|
93 |
This function will B<die> if normalizastion can't be evaled. |
This function will B<die> if normalizastion can't be evaled. |
94 |
|
|
95 |
Since this function isn't exported you have to call it with |
Since this function isn't exported you have to call it with |
97 |
|
|
98 |
=cut |
=cut |
99 |
|
|
100 |
|
my $load_row_coderef; |
101 |
|
|
102 |
sub data_structure { |
sub data_structure { |
103 |
my $arg = {@_}; |
my $arg = {@_}; |
104 |
|
|
106 |
die "need normalisation argument" unless ($arg->{rules}); |
die "need normalisation argument" unless ($arg->{rules}); |
107 |
|
|
108 |
no strict 'subs'; |
no strict 'subs'; |
109 |
_set_lookup( $arg->{lookup} ); |
_set_lookup( $arg->{lookup} ) if defined($arg->{lookup}); |
110 |
_set_rec( $arg->{row} ); |
_set_rec( $arg->{row} ); |
111 |
_set_config( $arg->{config} ); |
_set_config( $arg->{config} ) if defined($arg->{config}); |
112 |
_clean_ds( %{ $arg } ); |
_clean_ds( %{ $arg } ); |
113 |
|
$load_row_coderef = $arg->{load_row_coderef}; |
114 |
|
|
115 |
eval "$arg->{rules}"; |
eval "$arg->{rules}"; |
116 |
die "error evaling $arg->{rules}: $@\n" if ($@); |
die "error evaling $arg->{rules}: $@\n" if ($@); |
117 |
|
|
168 |
|
|
169 |
=cut |
=cut |
170 |
|
|
171 |
my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators); |
my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader); |
172 |
my ($marc_record_offset, $marc_fetch_offset) = (0, 0); |
my ($marc_record_offset, $marc_fetch_offset) = (0, 0); |
173 |
|
|
174 |
sub _get_ds { |
sub _get_ds { |
175 |
|
|
176 |
|
warn "## out = ",dump($out); |
177 |
|
|
178 |
return $out; |
return $out; |
179 |
} |
} |
180 |
|
|
188 |
|
|
189 |
sub _clean_ds { |
sub _clean_ds { |
190 |
my $a = {@_}; |
my $a = {@_}; |
191 |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = (); |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = (); |
192 |
($marc_record_offset, $marc_fetch_offset) = (0,0); |
($marc_record_offset, $marc_fetch_offset) = (0,0); |
193 |
$marc_encoding = $a->{marc_encoding}; |
$marc_encoding = $a->{marc_encoding}; |
194 |
} |
} |
207 |
$lookup = shift; |
$lookup = shift; |
208 |
} |
} |
209 |
|
|
210 |
|
=head2 _get_lookup |
211 |
|
|
212 |
|
Get current lookup hash |
213 |
|
|
214 |
|
my $lookup = _get_lookup(); |
215 |
|
|
216 |
|
=cut |
217 |
|
|
218 |
|
sub _get_lookup { |
219 |
|
return $lookup; |
220 |
|
} |
221 |
|
|
222 |
|
=head2 _set_load_row |
223 |
|
|
224 |
|
Setup code reference which will return L<data_structure> from |
225 |
|
L<WebPAC::Store> |
226 |
|
|
227 |
|
_set_load_row(sub { |
228 |
|
my ($database,$input,$mfn) = @_; |
229 |
|
$store->load_row( database => $database, input => $input, id => $mfn ); |
230 |
|
}); |
231 |
|
|
232 |
|
=cut |
233 |
|
|
234 |
|
sub _set_load_row { |
235 |
|
my $coderef = shift; |
236 |
|
confess "argument isn't CODE" unless ref($coderef) eq 'CODE'; |
237 |
|
|
238 |
|
$load_row_coderef = $coderef; |
239 |
|
} |
240 |
|
|
241 |
=head2 _get_marc_fields |
=head2 _get_marc_fields |
242 |
|
|
243 |
Get all fields defined by calls to C<marc> |
Get all fields defined by calls to C<marc> |
293 |
|
|
294 |
=cut |
=cut |
295 |
|
|
296 |
|
my $fetch_pos; |
297 |
|
|
298 |
sub _get_marc_fields { |
sub _get_marc_fields { |
299 |
|
|
300 |
my $arg = {@_}; |
my $arg = {@_}; |
301 |
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); |
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); |
302 |
my $offset = $marc_fetch_offset; |
$fetch_pos = $marc_fetch_offset; |
303 |
if ($arg->{offset}) { |
if ($arg->{offset}) { |
304 |
$offset = $arg->{offset}; |
$fetch_pos = $arg->{offset}; |
305 |
} elsif($arg->{fetch_next}) { |
} elsif($arg->{fetch_next}) { |
306 |
$marc_fetch_offset++; |
$marc_fetch_offset++; |
307 |
} |
} |
310 |
|
|
311 |
warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); |
warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); |
312 |
|
|
313 |
my $marc_rec = $marc_record->[ $offset ]; |
my $marc_rec = $marc_record->[ $fetch_pos ]; |
314 |
|
|
315 |
warn "## _get_marc_fields (at offset: $offset) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1); |
warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1); |
316 |
|
|
317 |
return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0); |
return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0); |
318 |
|
|
333 |
|
|
334 |
if ($debug) { |
if ($debug) { |
335 |
warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); |
warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); |
336 |
warn "## marc_record[$offset] = ", dump( $marc_rec ), $/; |
warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/; |
337 |
warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; |
warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; |
338 |
warn "## subfield count = ", dump( $u ), $/; |
warn "## subfield count = ", dump( $u ), $/; |
339 |
} |
} |
414 |
return \@m; |
return \@m; |
415 |
} |
} |
416 |
|
|
417 |
|
=head2 _get_marc_leader |
418 |
|
|
419 |
|
Return leader from currently fetched record by L</_get_marc_fields> |
420 |
|
|
421 |
|
print WebPAC::Normalize::_get_marc_leader(); |
422 |
|
|
423 |
|
=cut |
424 |
|
|
425 |
|
sub _get_marc_leader { |
426 |
|
die "no fetch_pos, did you called _get_marc_fields first?" unless ( defined( $fetch_pos ) ); |
427 |
|
return $marc_leader->[ $fetch_pos ]; |
428 |
|
} |
429 |
|
|
430 |
=head2 _debug |
=head2 _debug |
431 |
|
|
432 |
Change level of debug warnings |
Change level of debug warnings |
446 |
|
|
447 |
Those functions generally have to first in your normalization file. |
Those functions generally have to first in your normalization file. |
448 |
|
|
449 |
=head2 tag |
=head2 search_display |
450 |
|
|
451 |
Define new tag for I<search> and I<display>. |
Define output for L<search> and L<display> at the same time |
452 |
|
|
453 |
tag('Title', rec('200','a') ); |
search_display('Title', rec('200','a') ); |
454 |
|
|
455 |
|
|
456 |
=cut |
=cut |
457 |
|
|
458 |
sub tag { |
sub search_display { |
459 |
my $name = shift or die "tag needs name as first argument"; |
my $name = shift or die "search_display needs name as first argument"; |
460 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
461 |
return unless (@o); |
return unless (@o); |
|
$out->{$name}->{tag} = $name; |
|
462 |
$out->{$name}->{search} = \@o; |
$out->{$name}->{search} = \@o; |
463 |
$out->{$name}->{display} = \@o; |
$out->{$name}->{display} = \@o; |
464 |
} |
} |
465 |
|
|
466 |
|
=head2 tag |
467 |
|
|
468 |
|
Old name for L<search_display>, but supported |
469 |
|
|
470 |
|
=cut |
471 |
|
|
472 |
|
sub tag { |
473 |
|
search_display( @_ ); |
474 |
|
} |
475 |
|
|
476 |
=head2 display |
=head2 display |
477 |
|
|
478 |
Define tag just for I<display> |
Define output just for I<display> |
479 |
|
|
480 |
@v = display('Title', rec('200','a') ); |
@v = display('Title', rec('200','a') ); |
481 |
|
|
482 |
=cut |
=cut |
483 |
|
|
484 |
sub display { |
sub _field { |
485 |
my $name = shift or die "display needs name as first argument"; |
my $type = shift or confess "need type -- BUG?"; |
486 |
|
my $name = shift or confess "needs name as first argument"; |
487 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
488 |
return unless (@o); |
return unless (@o); |
489 |
$out->{$name}->{tag} = $name; |
$out->{$name}->{$type} = \@o; |
|
$out->{$name}->{display} = \@o; |
|
490 |
} |
} |
491 |
|
|
492 |
|
sub display { _field( 'display', @_ ) } |
493 |
|
|
494 |
=head2 search |
=head2 search |
495 |
|
|
496 |
Prepare values just for I<search> |
Prepare values just for I<search> |
499 |
|
|
500 |
=cut |
=cut |
501 |
|
|
502 |
sub search { |
sub search { _field( 'search', @_ ) } |
503 |
my $name = shift or die "search needs name as first argument"; |
|
504 |
my @o = grep { defined($_) && $_ ne '' } @_; |
=head2 sorted |
505 |
return unless (@o); |
|
506 |
$out->{$name}->{tag} = $name; |
Insert into lists which will be automatically sorted |
507 |
$out->{$name}->{search} = \@o; |
|
508 |
} |
sorted('Title', rec('200','a') ); |
509 |
|
|
510 |
|
=cut |
511 |
|
|
512 |
|
sub sorted { _field( 'sorted', @_ ) } |
513 |
|
|
514 |
|
|
515 |
=head2 marc_leader |
=head2 marc_leader |
516 |
|
|
525 |
my ($offset,$value) = @_; |
my ($offset,$value) = @_; |
526 |
|
|
527 |
if ($offset) { |
if ($offset) { |
528 |
$out->{' leader'}->{ $offset } = $value; |
$marc_leader->[ $marc_record_offset ]->{ $offset } = $value; |
529 |
} else { |
} else { |
530 |
return $out->{' leader'}; |
|
531 |
|
if (defined($marc_leader)) { |
532 |
|
die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY'); |
533 |
|
return $marc_leader->[ $marc_record_offset ]; |
534 |
|
} else { |
535 |
|
return; |
536 |
|
} |
537 |
|
} |
538 |
|
} |
539 |
|
|
540 |
|
=head2 marc_fixed |
541 |
|
|
542 |
|
Create control/indentifier fields with values in fixed positions |
543 |
|
|
544 |
|
marc_fixed('008', 00, '070402'); |
545 |
|
marc_fixed('008', 39, '|'); |
546 |
|
|
547 |
|
Positions not specified will be filled with spaces (C<0x20>). |
548 |
|
|
549 |
|
There will be no effort to extend last specified value to full length of |
550 |
|
field in standard. |
551 |
|
|
552 |
|
=cut |
553 |
|
|
554 |
|
sub marc_fixed { |
555 |
|
my ($f, $pos, $val) = @_; |
556 |
|
die "need marc(field, position, value)" unless defined($f) && defined($pos); |
557 |
|
|
558 |
|
confess "need val" unless defined $val; |
559 |
|
|
560 |
|
my $update = 0; |
561 |
|
|
562 |
|
map { |
563 |
|
if ($_->[0] eq $f) { |
564 |
|
my $old = $_->[1]; |
565 |
|
if (length($old) <= $pos) { |
566 |
|
$_->[1] .= ' ' x ( $pos - length($old) ) . $val; |
567 |
|
warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1); |
568 |
|
} else { |
569 |
|
$_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val)); |
570 |
|
warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1); |
571 |
|
} |
572 |
|
$update++; |
573 |
|
} |
574 |
|
} @{ $marc_record->[ $marc_record_offset ] }; |
575 |
|
|
576 |
|
if (! $update) { |
577 |
|
my $v = ' ' x $pos . $val; |
578 |
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; |
579 |
|
warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1); |
580 |
} |
} |
581 |
} |
} |
582 |
|
|
601 |
foreach (@_) { |
foreach (@_) { |
602 |
my $v = $_; # make var read-write for Encode |
my $v = $_; # make var read-write for Encode |
603 |
next unless (defined($v) && $v !~ /^\s*$/); |
next unless (defined($v) && $v !~ /^\s*$/); |
|
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
|
604 |
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
605 |
if (defined $sf) { |
if (defined $sf) { |
606 |
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; |
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; |
657 |
'c', rec('200','c') |
'c', rec('200','c') |
658 |
); |
); |
659 |
|
|
660 |
|
If you specify C<+> for subfield, value will be appended |
661 |
|
to previous defined subfield. |
662 |
|
|
663 |
=cut |
=cut |
664 |
|
|
665 |
sub marc_compose { |
sub marc_compose { |
671 |
|
|
672 |
warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); |
warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); |
673 |
|
|
674 |
|
if ($#_ % 2 != 1) { |
675 |
|
die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n"; |
676 |
|
} |
677 |
|
|
678 |
while (@_) { |
while (@_) { |
679 |
my $sf = shift or die "marc_compose $f needs subfield"; |
my $sf = shift; |
680 |
my $v = shift; |
my $v = shift; |
681 |
|
|
682 |
next unless (defined($v) && $v !~ /^\s*$/); |
next unless (defined($v) && $v !~ /^\s*$/); |
|
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
|
|
push @$m, ( $sf, $v ); |
|
683 |
warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); |
warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); |
684 |
|
if ($sf ne '+') { |
685 |
|
push @$m, ( $sf, $v ); |
686 |
|
} else { |
687 |
|
$m->[ $#$m ] .= $v; |
688 |
|
} |
689 |
} |
} |
690 |
|
|
691 |
warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); |
warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); |
708 |
my $m = $marc_record->[ -1 ]; |
my $m = $marc_record->[ -1 ]; |
709 |
die "can't duplicate record which isn't defined" unless ($m); |
die "can't duplicate record which isn't defined" unless ($m); |
710 |
push @{ $marc_record }, dclone( $m ); |
push @{ $marc_record }, dclone( $m ); |
711 |
warn "## marc_duplicate = ", dump(@$marc_record), $/ if ($debug > 1); |
push @{ $marc_leader }, dclone( marc_leader() ); |
712 |
|
warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1); |
713 |
$marc_record_offset = $#{ $marc_record }; |
$marc_record_offset = $#{ $marc_record }; |
714 |
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1); |
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1); |
715 |
|
|
716 |
} |
} |
717 |
|
|
718 |
=head2 marc_remove |
=head2 marc_remove |
724 |
|
|
725 |
This will erase field C<200> or C<200^a> from current MARC record. |
This will erase field C<200> or C<200^a> from current MARC record. |
726 |
|
|
727 |
|
marc_remove('*'); |
728 |
|
|
729 |
|
Will remove all fields in current MARC record. |
730 |
|
|
731 |
This is useful after calling C<marc_duplicate> or on it's own (but, you |
This is useful after calling C<marc_duplicate> or on it's own (but, you |
732 |
should probably just remove that subfield definition if you are not |
should probably just remove that subfield definition if you are not |
733 |
using C<marc_duplicate>). |
using C<marc_duplicate>). |
745 |
|
|
746 |
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); |
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); |
747 |
|
|
748 |
my $i = 0; |
if ($f eq '*') { |
749 |
foreach ( 0 .. $#{ $marc } ) { |
|
750 |
last unless (defined $marc->[$i]); |
delete( $marc_record->[ $marc_record_offset ] ); |
751 |
warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3); |
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
752 |
if ($marc->[$i]->[0] eq $f) { |
|
753 |
if (! defined $sf) { |
} else { |
754 |
# remove whole field |
|
755 |
splice @$marc, $i, 1; |
my $i = 0; |
756 |
warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3); |
foreach ( 0 .. $#{ $marc } ) { |
757 |
$i--; |
last unless (defined $marc->[$i]); |
758 |
} else { |
warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3); |
759 |
foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) { |
if ($marc->[$i]->[0] eq $f) { |
760 |
my $o = ($j * 2) + 3; |
if (! defined $sf) { |
761 |
if ($marc->[$i]->[$o] eq $sf) { |
# remove whole field |
762 |
# remove subfield |
splice @$marc, $i, 1; |
763 |
splice @{$marc->[$i]}, $o, 2; |
warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3); |
764 |
warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3); |
$i--; |
765 |
# is record now empty? |
} else { |
766 |
if ($#{ $marc->[$i] } == 2) { |
foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) { |
767 |
splice @$marc, $i, 1; |
my $o = ($j * 2) + 3; |
768 |
warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3); |
if ($marc->[$i]->[$o] eq $sf) { |
769 |
$i--; |
# remove subfield |
770 |
}; |
splice @{$marc->[$i]}, $o, 2; |
771 |
|
warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3); |
772 |
|
# is record now empty? |
773 |
|
if ($#{ $marc->[$i] } == 2) { |
774 |
|
splice @$marc, $i, 1; |
775 |
|
warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3); |
776 |
|
$i--; |
777 |
|
}; |
778 |
|
} |
779 |
} |
} |
780 |
} |
} |
781 |
} |
} |
782 |
|
$i++; |
783 |
} |
} |
|
$i++; |
|
|
} |
|
784 |
|
|
785 |
warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); |
warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); |
786 |
|
|
787 |
$marc_record->[ $marc_record_offset ] = $marc; |
$marc_record->[ $marc_record_offset ] = $marc; |
788 |
|
} |
789 |
|
|
790 |
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
791 |
} |
} |
792 |
|
|
793 |
|
=head2 marc_original_order |
794 |
|
|
795 |
|
Copy all subfields preserving original order to marc field. |
796 |
|
|
797 |
|
marc_original_order( marc_field_number, original_input_field_number ); |
798 |
|
|
799 |
|
Please note that field numbers are consistent with other commands (marc |
800 |
|
field number first), but somewhat counter-intuitive (destination and then |
801 |
|
source). |
802 |
|
|
803 |
|
You might want to use this command if you are just renaming subfields or |
804 |
|
using pre-processing modify_record in C<config.yml> and don't need any |
805 |
|
post-processing or want to preserve order of original subfields. |
806 |
|
|
807 |
|
|
808 |
|
=cut |
809 |
|
|
810 |
|
sub marc_original_order { |
811 |
|
|
812 |
|
my ($to, $from) = @_; |
813 |
|
die "marc_original_order needs from and to fields\n" unless ($from && $to); |
814 |
|
|
815 |
|
return unless defined($rec->{$from}); |
816 |
|
|
817 |
|
my $r = $rec->{$from}; |
818 |
|
die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY'); |
819 |
|
|
820 |
|
my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' '); |
821 |
|
warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1); |
822 |
|
|
823 |
|
foreach my $d (@$r) { |
824 |
|
|
825 |
|
if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') { |
826 |
|
warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n"; |
827 |
|
next; |
828 |
|
} |
829 |
|
|
830 |
|
my @sfs = @{ $d->{subfields} }; |
831 |
|
|
832 |
|
die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); |
833 |
|
|
834 |
|
warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2); |
835 |
|
|
836 |
|
my $m = [ $to, $i1, $i2 ]; |
837 |
|
|
838 |
|
while (my $sf = shift @sfs) { |
839 |
|
|
840 |
|
warn "#--> sf: ",dump($sf), $/ if ($debug > 2); |
841 |
|
my $offset = shift @sfs; |
842 |
|
die "corrupted sufields specification for field $from\n" unless defined($offset); |
843 |
|
|
844 |
|
my $v; |
845 |
|
if (ref($d->{$sf}) eq 'ARRAY') { |
846 |
|
$v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset])); |
847 |
|
} elsif ($offset == 0) { |
848 |
|
$v = $d->{$sf}; |
849 |
|
} else { |
850 |
|
die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf}); |
851 |
|
} |
852 |
|
push @$m, ( $sf, $v ) if (defined($v)); |
853 |
|
} |
854 |
|
|
855 |
|
if ($#{$m} > 2) { |
856 |
|
push @{ $marc_record->[ $marc_record_offset ] }, $m; |
857 |
|
} |
858 |
|
} |
859 |
|
|
860 |
|
warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); |
861 |
|
} |
862 |
|
|
863 |
|
=head2 marc_count |
864 |
|
|
865 |
|
Return number of MARC records created using L</marc_duplicate>. |
866 |
|
|
867 |
|
print "created ", marc_count(), " records"; |
868 |
|
|
869 |
|
=cut |
870 |
|
|
871 |
|
sub marc_count { |
872 |
|
return $#{ $marc_record }; |
873 |
|
} |
874 |
|
|
875 |
|
|
876 |
=head1 Functions to extract data from input |
=head1 Functions to extract data from input |
877 |
|
|
878 |
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 |
879 |
above. |
above. |
880 |
|
|
881 |
|
=head2 _pack_subfields_hash |
882 |
|
|
883 |
|
@subfields = _pack_subfields_hash( $h ); |
884 |
|
$subfields = _pack_subfields_hash( $h, 1 ); |
885 |
|
|
886 |
|
Return each subfield value in array or pack them all together and return scalar |
887 |
|
with subfields (denoted by C<^>) and values. |
888 |
|
|
889 |
|
=cut |
890 |
|
|
891 |
|
sub _pack_subfields_hash { |
892 |
|
|
893 |
|
warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1); |
894 |
|
|
895 |
|
my ($h,$include_subfields) = @_; |
896 |
|
|
897 |
|
# sanity and ease of use |
898 |
|
return $h if (ref($h) ne 'HASH'); |
899 |
|
|
900 |
|
if ( defined($h->{subfields}) ) { |
901 |
|
my $sfs = delete $h->{subfields} || die "no subfields?"; |
902 |
|
my @out; |
903 |
|
while (@$sfs) { |
904 |
|
my $sf = shift @$sfs; |
905 |
|
push @out, '^' . $sf if ($include_subfields); |
906 |
|
my $o = shift @$sfs; |
907 |
|
if ($o == 0 && ref( $h->{$sf} ) ne 'ARRAY' ) { |
908 |
|
# single element subfields are not arrays |
909 |
|
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
910 |
|
|
911 |
|
push @out, $h->{$sf}; |
912 |
|
} else { |
913 |
|
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
914 |
|
push @out, $h->{$sf}->[$o]; |
915 |
|
} |
916 |
|
} |
917 |
|
if ($include_subfields) { |
918 |
|
return join('', @out); |
919 |
|
} else { |
920 |
|
return @out; |
921 |
|
} |
922 |
|
} else { |
923 |
|
if ($include_subfields) { |
924 |
|
my $out = ''; |
925 |
|
foreach my $sf (sort keys %$h) { |
926 |
|
if (ref($h->{$sf}) eq 'ARRAY') { |
927 |
|
$out .= '^' . $sf . join('^' . $sf, @{ $h->{$sf} }); |
928 |
|
} else { |
929 |
|
$out .= '^' . $sf . $h->{$sf}; |
930 |
|
} |
931 |
|
} |
932 |
|
return $out; |
933 |
|
} else { |
934 |
|
# FIXME this should probably be in alphabetical order instead of hash order |
935 |
|
values %{$h}; |
936 |
|
} |
937 |
|
} |
938 |
|
} |
939 |
|
|
940 |
=head2 rec1 |
=head2 rec1 |
941 |
|
|
942 |
Return all values in some field |
Return all values in some field |
953 |
return unless (defined($rec) && defined($rec->{$f})); |
return unless (defined($rec) && defined($rec->{$f})); |
954 |
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
955 |
if (ref($rec->{$f}) eq 'ARRAY') { |
if (ref($rec->{$f}) eq 'ARRAY') { |
956 |
return map { |
my @out; |
957 |
if (ref($_) eq 'HASH') { |
foreach my $h ( @{ $rec->{$f} } ) { |
958 |
values %{$_}; |
if (ref($h) eq 'HASH') { |
959 |
|
push @out, ( _pack_subfields_hash( $h ) ); |
960 |
} else { |
} else { |
961 |
$_; |
push @out, $h; |
962 |
} |
} |
963 |
} @{ $rec->{$f} }; |
} |
964 |
|
return @out; |
965 |
} elsif( defined($rec->{$f}) ) { |
} elsif( defined($rec->{$f}) ) { |
966 |
return $rec->{$f}; |
return $rec->{$f}; |
967 |
} |
} |
996 |
@v = rec('200') |
@v = rec('200') |
997 |
@v = rec('200','a') |
@v = rec('200','a') |
998 |
|
|
999 |
|
If rec() returns just single value, it will |
1000 |
|
return scalar, not array. |
1001 |
|
|
1002 |
=cut |
=cut |
1003 |
|
|
1004 |
|
sub frec { |
1005 |
|
my @out = rec(@_); |
1006 |
|
warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0; |
1007 |
|
return shift @out; |
1008 |
|
} |
1009 |
|
|
1010 |
sub rec { |
sub rec { |
1011 |
my @out; |
my @out; |
1012 |
if ($#_ == 0) { |
if ($#_ == 0) { |
1014 |
} elsif ($#_ == 1) { |
} elsif ($#_ == 1) { |
1015 |
@out = rec2(@_); |
@out = rec2(@_); |
1016 |
} |
} |
1017 |
if (@out) { |
if ($#out == 0 && ! wantarray) { |
1018 |
|
return $out[0]; |
1019 |
|
} elsif (@out) { |
1020 |
return @out; |
return @out; |
1021 |
} else { |
} else { |
1022 |
return ''; |
return ''; |
1052 |
=cut |
=cut |
1053 |
|
|
1054 |
sub prefix { |
sub prefix { |
1055 |
my $p = shift or return; |
my $p = shift; |
1056 |
|
return @_ unless defined( $p ); |
1057 |
return map { $p . $_ } grep { defined($_) } @_; |
return map { $p . $_ } grep { defined($_) } @_; |
1058 |
} |
} |
1059 |
|
|
1066 |
=cut |
=cut |
1067 |
|
|
1068 |
sub suffix { |
sub suffix { |
1069 |
my $s = shift or die "suffix needs string as first argument"; |
my $s = shift; |
1070 |
|
return @_ unless defined( $s ); |
1071 |
return map { $_ . $s } grep { defined($_) } @_; |
return map { $_ . $s } grep { defined($_) } @_; |
1072 |
} |
} |
1073 |
|
|
1080 |
=cut |
=cut |
1081 |
|
|
1082 |
sub surround { |
sub surround { |
1083 |
my $p = shift or die "surround need prefix as first argument"; |
my $p = shift; |
1084 |
my $s = shift or die "surround needs suffix as second argument"; |
my $s = shift; |
1085 |
|
$p = '' unless defined( $p ); |
1086 |
|
$s = '' unless defined( $s ); |
1087 |
return map { $p . $_ . $s } grep { defined($_) } @_; |
return map { $p . $_ . $s } grep { defined($_) } @_; |
1088 |
} |
} |
1089 |
|
|
1104 |
|
|
1105 |
Consult lookup hashes for some value |
Consult lookup hashes for some value |
1106 |
|
|
1107 |
@v = lookup( $v ); |
@v = lookup( |
1108 |
@v = lookup( @v ); |
sub { |
1109 |
|
'ffkk/peri/mfn'.rec('000') |
1110 |
|
}, |
1111 |
|
'ffkk','peri','200-a-200-e', |
1112 |
|
sub { |
1113 |
|
first(rec(200,'a')).' '.first(rec('200','e')) |
1114 |
|
} |
1115 |
|
); |
1116 |
|
|
1117 |
|
Code like above will be B<automatically generated> using L<WebPAC::Parse> from |
1118 |
|
normal lookup definition in C<conf/lookup/something.pl> which looks like: |
1119 |
|
|
1120 |
|
lookup( |
1121 |
|
# which results to return from record recorded in lookup |
1122 |
|
sub { 'ffkk/peri/mfn' . rec('000') }, |
1123 |
|
# from which database and input |
1124 |
|
'ffkk','peri', |
1125 |
|
# such that following values match |
1126 |
|
sub { first(rec(200,'a')) . ' ' . first(rec('200','e')) }, |
1127 |
|
# if this part is missing, we will try to match same fields |
1128 |
|
# from lookup record and current one, or you can override |
1129 |
|
# which records to use from current record using |
1130 |
|
sub { rec('900','x') . ' ' . rec('900','y') }, |
1131 |
|
) |
1132 |
|
|
1133 |
|
You can think about this lookup as SQL (if that helps): |
1134 |
|
|
1135 |
|
select |
1136 |
|
sub { what } |
1137 |
|
from |
1138 |
|
database, input |
1139 |
|
where |
1140 |
|
sub { filter from lookuped record } |
1141 |
|
having |
1142 |
|
sub { optional filter on current record } |
1143 |
|
|
1144 |
|
Easy as pie, right? |
1145 |
|
|
1146 |
=cut |
=cut |
1147 |
|
|
1148 |
sub lookup { |
sub lookup { |
1149 |
my $k = shift or return; |
my ($what, $database, $input, $key, $having) = @_; |
1150 |
return unless (defined($lookup->{$k})); |
|
1151 |
if (ref($lookup->{$k}) eq 'ARRAY') { |
confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4); |
1152 |
return @{ $lookup->{$k} }; |
|
1153 |
|
warn "## lookup ($database, $input, $key)", $/ if ($debug > 1); |
1154 |
|
return unless (defined($lookup->{$database}->{$input}->{$key})); |
1155 |
|
|
1156 |
|
confess "lookup really need load_row_coderef added to data_structure\n" unless ($load_row_coderef); |
1157 |
|
|
1158 |
|
my $mfns; |
1159 |
|
my @having = $having->(); |
1160 |
|
|
1161 |
|
warn "## having = ", dump( @having ) if ($debug > 2); |
1162 |
|
|
1163 |
|
foreach my $h ( @having ) { |
1164 |
|
if (defined($lookup->{$database}->{$input}->{$key}->{$h})) { |
1165 |
|
warn "lookup for $database/$input/$key/$h return ",dump($lookup->{$database}->{$input}->{$key}->{$h}),"\n" if ($debug); |
1166 |
|
$mfns->{$_}++ foreach keys %{ $lookup->{$database}->{$input}->{$key}->{$h} }; |
1167 |
|
} |
1168 |
|
} |
1169 |
|
|
1170 |
|
return unless ($mfns); |
1171 |
|
|
1172 |
|
my @mfns = sort keys %$mfns; |
1173 |
|
|
1174 |
|
warn "# lookup loading $database/$input/$key mfn ", join(",",@mfns)," having ",dump(@having),"\n" if ($debug); |
1175 |
|
|
1176 |
|
my $old_rec = $rec; |
1177 |
|
my @out; |
1178 |
|
|
1179 |
|
foreach my $mfn (@mfns) { |
1180 |
|
$rec = $load_row_coderef->( $database, $input, $mfn ); |
1181 |
|
|
1182 |
|
warn "got $database/$input/$mfn = ", dump($rec), $/ if ($debug); |
1183 |
|
|
1184 |
|
my @vals = $what->(); |
1185 |
|
|
1186 |
|
push @out, ( @vals ); |
1187 |
|
|
1188 |
|
warn "lookup for mfn $mfn returned ", dump(@vals), $/ if ($debug); |
1189 |
|
} |
1190 |
|
|
1191 |
|
# if (ref($lookup->{$k}) eq 'ARRAY') { |
1192 |
|
# return @{ $lookup->{$k} }; |
1193 |
|
# } else { |
1194 |
|
# return $lookup->{$k}; |
1195 |
|
# } |
1196 |
|
|
1197 |
|
$rec = $old_rec; |
1198 |
|
|
1199 |
|
warn "## lookup returns = ", dump(@out), $/ if ($debug); |
1200 |
|
|
1201 |
|
if ($#out == 0) { |
1202 |
|
return $out[0]; |
1203 |
} else { |
} else { |
1204 |
return $lookup->{$k}; |
return @out; |
1205 |
|
} |
1206 |
|
} |
1207 |
|
|
1208 |
|
=head2 save_into_lookup |
1209 |
|
|
1210 |
|
Save value into lookup. It associates current database, input |
1211 |
|
and specific keys with one or more values which will be |
1212 |
|
associated over MFN. |
1213 |
|
|
1214 |
|
MFN will be extracted from first occurence current of field 000 |
1215 |
|
in current record, or if it doesn't exist from L<_set_config> C<_mfn>. |
1216 |
|
|
1217 |
|
my $nr = save_into_lookup($database,$input,$key,sub { |
1218 |
|
# code which produce one or more values |
1219 |
|
}); |
1220 |
|
|
1221 |
|
It returns number of items saved. |
1222 |
|
|
1223 |
|
This function shouldn't be called directly, it's called from code created by |
1224 |
|
L<WebPAC::Parser>. |
1225 |
|
|
1226 |
|
=cut |
1227 |
|
|
1228 |
|
sub save_into_lookup { |
1229 |
|
my ($database,$input,$key,$coderef) = @_; |
1230 |
|
die "save_into_lookup needs database" unless defined($database); |
1231 |
|
die "save_into_lookup needs input" unless defined($input); |
1232 |
|
die "save_into_lookup needs key" unless defined($key); |
1233 |
|
die "save_into_lookup needs CODE" unless ( defined($coderef) && ref($coderef) eq 'CODE' ); |
1234 |
|
|
1235 |
|
warn "## save_into_lookup rec = ", dump($rec), " config = ", dump($config), $/ if ($debug > 2); |
1236 |
|
|
1237 |
|
my $mfn = |
1238 |
|
defined($rec->{'000'}->[0]) ? $rec->{'000'}->[0] : |
1239 |
|
defined($config->{_mfn}) ? $config->{_mfn} : |
1240 |
|
die "mfn not defined or zero"; |
1241 |
|
|
1242 |
|
my $nr = 0; |
1243 |
|
|
1244 |
|
foreach my $v ( $coderef->() ) { |
1245 |
|
$lookup->{$database}->{$input}->{$key}->{$v}->{$mfn}++; |
1246 |
|
warn "# saved lookup $database/$input/$key [$v] $mfn\n" if ($debug > 1); |
1247 |
|
$nr++; |
1248 |
} |
} |
1249 |
|
|
1250 |
|
return $nr; |
1251 |
} |
} |
1252 |
|
|
1253 |
=head2 config |
=head2 config |
1258 |
$database_code = config(); # use _ from hash |
$database_code = config(); # use _ from hash |
1259 |
$database_name = config('name'); |
$database_name = config('name'); |
1260 |
$database_input_name = config('input name'); |
$database_input_name = config('input name'); |
|
$tag = config('input normalize tag'); |
|
1261 |
|
|
1262 |
Up to three levels are supported. |
Up to three levels are supported. |
1263 |
|
|
1362 |
my $v = shift @r; |
my $v = shift @r; |
1363 |
warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2); |
warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2); |
1364 |
|
|
1365 |
return '' if( ! defined($v) || $v =~ /^\s*$/); |
return '' if ( ! defined($v) || $v =~ /^\s*$/); |
1366 |
|
|
1367 |
my @s = split( $regex, $v ); |
my @s = split( $regex, $v ); |
1368 |
warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1); |
warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1); |
1373 |
} |
} |
1374 |
} |
} |
1375 |
|
|
1376 |
|
my $hash; |
1377 |
|
|
1378 |
|
=head2 set |
1379 |
|
|
1380 |
|
set( key => 'value' ); |
1381 |
|
|
1382 |
|
=cut |
1383 |
|
|
1384 |
|
sub set { |
1385 |
|
my ($k,$v) = @_; |
1386 |
|
warn "## set ( $k => ", dump($v), " )", $/ if ( $debug ); |
1387 |
|
$hash->{$k} = $v; |
1388 |
|
}; |
1389 |
|
|
1390 |
|
=head2 get |
1391 |
|
|
1392 |
|
get( 'key' ); |
1393 |
|
|
1394 |
|
=cut |
1395 |
|
|
1396 |
|
sub get { |
1397 |
|
my $k = shift || return; |
1398 |
|
my $v = $hash->{$k}; |
1399 |
|
warn "## get $k = ", dump( $v ), $/ if ( $debug ); |
1400 |
|
return $v; |
1401 |
|
} |
1402 |
|
|
1403 |
|
=head2 count |
1404 |
|
|
1405 |
|
if ( count( @result ) == 1 ) { |
1406 |
|
# do something if only 1 result is there |
1407 |
|
} |
1408 |
|
|
1409 |
|
=cut |
1410 |
|
|
1411 |
|
sub count { |
1412 |
|
warn "## count ",dump(@_),$/ if ( $debug ); |
1413 |
|
return @_ . ''; |
1414 |
|
} |
1415 |
|
|
1416 |
# END |
# END |
1417 |
1; |
1; |