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_ds _set_lookup |
5 |
_set_load_row |
_set_load_row |
6 |
_get_ds _clean_ds |
_get_ds _clean_ds |
7 |
_debug |
_debug |
8 |
_pack_subfields_hash |
_pack_subfields_hash |
9 |
|
|
10 |
tag search display |
to |
11 |
marc marc_indicators marc_repeatable_subfield |
search_display search display sorted |
|
marc_compose marc_leader marc_fixed |
|
|
marc_duplicate marc_remove marc_count |
|
|
marc_original_order |
|
12 |
|
|
13 |
rec1 rec2 rec |
rec1 rec2 rec |
14 |
|
frec frec_eq frec_ne |
15 |
regex prefix suffix surround |
regex prefix suffix surround |
16 |
first lookup join_with |
first lookup join_with |
17 |
save_into_lookup |
save_into_lookup |
20 |
|
|
21 |
get set |
get set |
22 |
count |
count |
23 |
|
|
24 |
/; |
/; |
25 |
|
|
26 |
use warnings; |
use warnings; |
28 |
|
|
29 |
#use base qw/WebPAC::Common/; |
#use base qw/WebPAC::Common/; |
30 |
use Data::Dump qw/dump/; |
use Data::Dump qw/dump/; |
|
use Storable qw/dclone/; |
|
31 |
use Carp qw/confess/; |
use Carp qw/confess/; |
32 |
|
|
33 |
# debugging warn(s) |
# debugging warn(s) |
34 |
my $debug = 0; |
my $debug = 0; |
35 |
|
_debug( $debug ); |
36 |
|
|
37 |
|
# FIXME |
38 |
|
use WebPAC::Normalize::ISBN; |
39 |
|
push @EXPORT, ( 'isbn_10', 'isbn_13' ); |
40 |
|
|
41 |
=head1 NAME |
use WebPAC::Normalize::MARC; |
42 |
|
push @EXPORT, ( qw/ |
43 |
|
marc marc_indicators marc_repeatable_subfield |
44 |
|
marc_compose marc_leader marc_fixed |
45 |
|
marc_duplicate marc_remove marc_count |
46 |
|
marc_original_order |
47 |
|
marc_template |
48 |
|
/); |
49 |
|
|
50 |
WebPAC::Normalize - describe normalisaton rules using sets |
use Storable qw/dclone/; |
51 |
|
|
52 |
=head1 VERSION |
=head1 NAME |
53 |
|
|
54 |
Version 0.29 |
WebPAC::Normalize - describe normalisaton rules using sets |
55 |
|
|
56 |
=cut |
=cut |
57 |
|
|
58 |
our $VERSION = '0.29'; |
our $VERSION = '0.36'; |
59 |
|
|
60 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
61 |
|
|
68 |
C<perl -c normalize.pl>. |
C<perl -c normalize.pl>. |
69 |
|
|
70 |
Normalisation can generate multiple output normalized data. For now, supported output |
Normalisation can generate multiple output normalized data. For now, supported output |
71 |
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 |
72 |
C<marc>. |
C<marc>. |
73 |
|
|
74 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
87 |
marc_encoding => 'utf-8', |
marc_encoding => 'utf-8', |
88 |
config => $config, |
config => $config, |
89 |
load_row_coderef => sub { |
load_row_coderef => sub { |
90 |
my ($database,$input,$mfn) = shift; |
my ($database,$input,$mfn) = @_; |
91 |
$store->load_row( database => $database, input => $input, id => $mfn ); |
$store->load_row( database => $database, input => $input, id => $mfn ); |
92 |
}, |
}, |
93 |
); |
); |
113 |
die "need row argument" unless ($arg->{row}); |
die "need row argument" unless ($arg->{row}); |
114 |
die "need normalisation argument" unless ($arg->{rules}); |
die "need normalisation argument" unless ($arg->{rules}); |
115 |
|
|
|
no strict 'subs'; |
|
116 |
_set_lookup( $arg->{lookup} ) if defined($arg->{lookup}); |
_set_lookup( $arg->{lookup} ) if defined($arg->{lookup}); |
117 |
_set_rec( $arg->{row} ); |
_set_ds( $arg->{row} ); |
118 |
_set_config( $arg->{config} ) if defined($arg->{config}); |
_set_config( $arg->{config} ) if defined($arg->{config}); |
119 |
_clean_ds( %{ $arg } ); |
_clean_ds( %{ $arg } ); |
120 |
$load_row_coderef = $arg->{load_row_coderef}; |
$load_row_coderef = $arg->{load_row_coderef}; |
121 |
|
|
122 |
eval "$arg->{rules}"; |
no strict 'subs'; |
123 |
|
no warnings 'redefine'; |
124 |
|
eval "$arg->{rules};"; |
125 |
die "error evaling $arg->{rules}: $@\n" if ($@); |
die "error evaling $arg->{rules}: $@\n" if ($@); |
126 |
|
|
127 |
return _get_ds(); |
return _get_ds(); |
128 |
} |
} |
129 |
|
|
130 |
=head2 _set_rec |
=head2 _set_ds |
131 |
|
|
132 |
Set current record hash |
Set current record hash |
133 |
|
|
134 |
_set_rec( $rec ); |
_set_ds( $rec ); |
135 |
|
|
136 |
=cut |
=cut |
137 |
|
|
138 |
my $rec; |
my $rec; |
139 |
|
|
140 |
sub _set_rec { |
sub _set_ds { |
141 |
$rec = shift or die "no record hash"; |
$rec = shift or die "no record hash"; |
142 |
|
$WebPAC::Normalize::MARC::rec = $rec; |
143 |
} |
} |
144 |
|
|
145 |
|
=head2 |
146 |
|
|
147 |
|
my $rec = _get_rec(); |
148 |
|
|
149 |
|
=cut |
150 |
|
|
151 |
|
sub _get_rec { $rec }; |
152 |
|
|
153 |
=head2 _set_config |
=head2 _set_config |
154 |
|
|
155 |
Set current config hash |
Set current config hash |
186 |
|
|
187 |
=cut |
=cut |
188 |
|
|
189 |
my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader); |
my $out; |
|
my ($marc_record_offset, $marc_fetch_offset) = (0, 0); |
|
190 |
|
|
191 |
sub _get_ds { |
sub _get_ds { |
192 |
|
#warn "## out = ",dump($out); |
193 |
return $out; |
return $out; |
194 |
} |
} |
195 |
|
|
203 |
|
|
204 |
sub _clean_ds { |
sub _clean_ds { |
205 |
my $a = {@_}; |
my $a = {@_}; |
206 |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = (); |
$out = undef; |
207 |
($marc_record_offset, $marc_fetch_offset) = (0,0); |
WebPAC::Normalize::MARC::_clean(); |
|
$marc_encoding = $a->{marc_encoding}; |
|
208 |
} |
} |
209 |
|
|
210 |
=head2 _set_lookup |
=head2 _set_lookup |
252 |
$load_row_coderef = $coderef; |
$load_row_coderef = $coderef; |
253 |
} |
} |
254 |
|
|
|
=head2 _get_marc_fields |
|
|
|
|
|
Get all fields defined by calls to C<marc> |
|
|
|
|
|
$marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); |
|
|
|
|
|
We are using I<magic> which detect repeatable fields only from |
|
|
sequence of field/subfield data generated by normalization. |
|
|
|
|
|
Repeatable field is created when there is second occurence of same subfield or |
|
|
if any of indicators are different. |
|
|
|
|
|
This is sane for most cases. Something like: |
|
|
|
|
|
900a-1 900b-1 900c-1 |
|
|
900a-2 900b-2 |
|
|
900a-3 |
|
|
|
|
|
will be created from any combination of: |
|
|
|
|
|
900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 |
|
|
|
|
|
and following rules: |
|
|
|
|
|
marc('900','a', rec('200','a') ); |
|
|
marc('900','b', rec('200','b') ); |
|
|
marc('900','c', rec('200','c') ); |
|
|
|
|
|
which might not be what you have in mind. If you need repeatable subfield, |
|
|
define it using C<marc_repeatable_subfield> like this: |
|
|
|
|
|
marc_repeatable_subfield('900','a'); |
|
|
marc('900','a', rec('200','a') ); |
|
|
marc('900','b', rec('200','b') ); |
|
|
marc('900','c', rec('200','c') ); |
|
|
|
|
|
will create: |
|
|
|
|
|
900a-1 900a-2 900a-3 900b-1 900c-1 |
|
|
900b-2 |
|
|
|
|
|
There is also support for returning next or specific using: |
|
|
|
|
|
while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) { |
|
|
# do something with $mf |
|
|
} |
|
|
|
|
|
will always return fields from next MARC record or |
|
|
|
|
|
my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 ); |
|
|
|
|
|
will return 42th copy record (if it exists). |
|
|
|
|
|
=cut |
|
|
|
|
|
my $fetch_pos; |
|
|
|
|
|
sub _get_marc_fields { |
|
|
|
|
|
my $arg = {@_}; |
|
|
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); |
|
|
$fetch_pos = $marc_fetch_offset; |
|
|
if ($arg->{offset}) { |
|
|
$fetch_pos = $arg->{offset}; |
|
|
} elsif($arg->{fetch_next}) { |
|
|
$marc_fetch_offset++; |
|
|
} |
|
|
|
|
|
return if (! $marc_record || ref($marc_record) ne 'ARRAY'); |
|
|
|
|
|
warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); |
|
|
|
|
|
my $marc_rec = $marc_record->[ $fetch_pos ]; |
|
|
|
|
|
warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1); |
|
|
|
|
|
return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0); |
|
|
|
|
|
# first, sort all existing fields |
|
|
# XXX might not be needed, but modern perl might randomize elements in hash |
|
|
my @sorted_marc_record = sort { |
|
|
$a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '') |
|
|
} @{ $marc_rec }; |
|
|
|
|
|
@sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting |
|
|
|
|
|
# output marc fields |
|
|
my @m; |
|
|
|
|
|
# count unique field-subfields (used for offset when walking to next subfield) |
|
|
my $u; |
|
|
map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record; |
|
|
|
|
|
if ($debug) { |
|
|
warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); |
|
|
warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/; |
|
|
warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; |
|
|
warn "## subfield count = ", dump( $u ), $/; |
|
|
} |
|
|
|
|
|
my $len = $#sorted_marc_record; |
|
|
my $visited; |
|
|
my $i = 0; |
|
|
my $field; |
|
|
|
|
|
foreach ( 0 .. $len ) { |
|
|
|
|
|
# find next element which isn't visited |
|
|
while ($visited->{$i}) { |
|
|
$i = ($i + 1) % ($len + 1); |
|
|
} |
|
|
|
|
|
# mark it visited |
|
|
$visited->{$i}++; |
|
|
|
|
|
my $row = dclone( $sorted_marc_record[$i] ); |
|
|
|
|
|
# field and subfield which is key for |
|
|
# marc_repeatable_subfield and u |
|
|
my $fsf = $row->[0] . ( $row->[3] || '' ); |
|
|
|
|
|
if ($debug > 1) { |
|
|
|
|
|
print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; |
|
|
print "### this [$i]: ", dump( $row ),$/; |
|
|
print "### sf: ", $row->[3], " vs ", $field->[3], |
|
|
$marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, |
|
|
if ($#$field >= 0); |
|
|
|
|
|
} |
|
|
|
|
|
# if field exists |
|
|
if ( $#$field >= 0 ) { |
|
|
if ( |
|
|
$row->[0] ne $field->[0] || # field |
|
|
$row->[1] ne $field->[1] || # i1 |
|
|
$row->[2] ne $field->[2] # i2 |
|
|
) { |
|
|
push @m, $field; |
|
|
warn "## saved/1 ", dump( $field ),$/ if ($debug); |
|
|
$field = $row; |
|
|
|
|
|
} elsif ( |
|
|
( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) |
|
|
|| |
|
|
( $row->[3] eq $field->[-2] && # same subfield, but not repeatable |
|
|
! $marc_repeatable_subfield->{ $fsf } |
|
|
) |
|
|
) { |
|
|
push @m, $field; |
|
|
warn "## saved/2 ", dump( $field ),$/ if ($debug); |
|
|
$field = $row; |
|
|
|
|
|
} else { |
|
|
# append new subfields to existing field |
|
|
push @$field, ( $row->[3], $row->[4] ); |
|
|
} |
|
|
} else { |
|
|
# insert first field |
|
|
$field = $row; |
|
|
} |
|
|
|
|
|
if (! $marc_repeatable_subfield->{ $fsf }) { |
|
|
# make step to next subfield |
|
|
$i = ($i + $u->{ $fsf } ) % ($len + 1); |
|
|
} |
|
|
} |
|
|
|
|
|
if ($#$field >= 0) { |
|
|
push @m, $field; |
|
|
warn "## saved/3 ", dump( $field ),$/ if ($debug); |
|
|
} |
|
|
|
|
|
return \@m; |
|
|
} |
|
|
|
|
|
=head2 _get_marc_leader |
|
|
|
|
|
Return leader from currently fetched record by L</_get_marc_fields> |
|
|
|
|
|
print WebPAC::Normalize::_get_marc_leader(); |
|
|
|
|
|
=cut |
|
|
|
|
|
sub _get_marc_leader { |
|
|
die "no fetch_pos, did you called _get_marc_fields first?" unless ( defined( $fetch_pos ) ); |
|
|
return $marc_leader->[ $fetch_pos ]; |
|
|
} |
|
|
|
|
255 |
=head2 _debug |
=head2 _debug |
256 |
|
|
257 |
Change level of debug warnings |
Change level of debug warnings |
265 |
return $debug unless defined($l); |
return $debug unless defined($l); |
266 |
warn "debug level $l",$/ if ($l > 0); |
warn "debug level $l",$/ if ($l > 0); |
267 |
$debug = $l; |
$debug = $l; |
268 |
|
$WebPAC::Normalize::MARC::debug = $debug; |
269 |
} |
} |
270 |
|
|
271 |
=head1 Functions to create C<data_structure> |
=head1 Functions to create C<data_structure> |
272 |
|
|
273 |
Those functions generally have to first in your normalization file. |
Those functions generally have to first in your normalization file. |
274 |
|
|
275 |
=head2 tag |
=head2 to |
|
|
|
|
Define new tag for I<search> and I<display>. |
|
|
|
|
|
tag('Title', rec('200','a') ); |
|
|
|
|
|
|
|
|
=cut |
|
|
|
|
|
sub tag { |
|
|
my $name = shift or die "tag needs name as first argument"; |
|
|
my @o = grep { defined($_) && $_ ne '' } @_; |
|
|
return unless (@o); |
|
|
$out->{$name}->{tag} = $name; |
|
|
$out->{$name}->{search} = \@o; |
|
|
$out->{$name}->{display} = \@o; |
|
|
} |
|
276 |
|
|
277 |
=head2 display |
Generic way to set values for some name |
278 |
|
|
279 |
Define tag just for I<display> |
to('field-name', 'name-value' => rec('200','a') ); |
280 |
|
|
281 |
@v = display('Title', rec('200','a') ); |
There are many helpers defined below which might be easier to use. |
282 |
|
|
283 |
=cut |
=cut |
284 |
|
|
285 |
sub display { |
sub to { |
286 |
my $name = shift or die "display needs name as first argument"; |
my $type = shift or confess "need type -- BUG?"; |
287 |
|
my $name = shift or confess "needs name as first argument"; |
288 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
289 |
return unless (@o); |
return unless (@o); |
290 |
$out->{$name}->{tag} = $name; |
$out->{$name}->{$type} = \@o; |
|
$out->{$name}->{display} = \@o; |
|
291 |
} |
} |
292 |
|
|
293 |
=head2 search |
=head2 search_display |
294 |
|
|
295 |
Prepare values just for I<search> |
Define output for L<search> and L<display> at the same time |
296 |
|
|
297 |
@v = search('Title', rec('200','a') ); |
search_display('Title', rec('200','a') ); |
298 |
|
|
299 |
=cut |
=cut |
300 |
|
|
301 |
sub search { |
sub search_display { |
302 |
my $name = shift or die "search needs name as first argument"; |
my $name = shift or die "search_display needs name as first argument"; |
303 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
304 |
return unless (@o); |
return unless (@o); |
|
$out->{$name}->{tag} = $name; |
|
305 |
$out->{$name}->{search} = \@o; |
$out->{$name}->{search} = \@o; |
306 |
|
$out->{$name}->{display} = \@o; |
307 |
} |
} |
308 |
|
|
309 |
=head2 marc_leader |
=head2 tag |
|
|
|
|
Setup fields within MARC leader or get leader |
|
|
|
|
|
marc_leader('05','c'); |
|
|
my $leader = marc_leader(); |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_leader { |
|
|
my ($offset,$value) = @_; |
|
|
|
|
|
if ($offset) { |
|
|
$marc_leader->[ $marc_record_offset ]->{ $offset } = $value; |
|
|
} else { |
|
|
|
|
|
if (defined($marc_leader)) { |
|
|
die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY'); |
|
|
return $marc_leader->[ $marc_record_offset ]; |
|
|
} else { |
|
|
return; |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
=head2 marc_fixed |
|
|
|
|
|
Create control/indentifier fields with values in fixed positions |
|
|
|
|
|
marc_fixed('008', 00, '070402'); |
|
|
marc_fixed('008', 39, '|'); |
|
|
|
|
|
Positions not specified will be filled with spaces (C<0x20>). |
|
|
|
|
|
There will be no effort to extend last specified value to full length of |
|
|
field in standard. |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_fixed { |
|
|
my ($f, $pos, $val) = @_; |
|
|
die "need marc(field, position, value)" unless defined($f) && defined($pos); |
|
|
|
|
|
my $update = 0; |
|
|
|
|
|
map { |
|
|
if ($_->[0] eq $f) { |
|
|
my $old = $_->[1]; |
|
|
if (length($old) < $pos) { |
|
|
$_->[1] .= ' ' x ( $pos - length($old) ) . $val; |
|
|
warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1); |
|
|
} else { |
|
|
$_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val)); |
|
|
warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1); |
|
|
} |
|
|
$update++; |
|
|
} |
|
|
} @{ $marc_record->[ $marc_record_offset ] }; |
|
|
|
|
|
if (! $update) { |
|
|
my $v = ' ' x $pos . $val; |
|
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; |
|
|
warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1); |
|
|
} |
|
|
} |
|
|
|
|
|
=head2 marc |
|
|
|
|
|
Save value for MARC field |
|
|
|
|
|
marc('900','a', rec('200','a') ); |
|
|
marc('001', rec('000') ); |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc { |
|
|
my $f = shift or die "marc needs field"; |
|
|
die "marc field must be numer" unless ($f =~ /^\d+$/); |
|
|
|
|
|
my $sf; |
|
|
if ($f >= 10) { |
|
|
$sf = shift or die "marc needs subfield"; |
|
|
} |
|
|
|
|
|
foreach (@_) { |
|
|
my $v = $_; # make var read-write for Encode |
|
|
next unless (defined($v) && $v !~ /^\s*$/); |
|
|
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
|
|
if (defined $sf) { |
|
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; |
|
|
} else { |
|
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; |
|
|
} |
|
|
} |
|
|
} |
|
|
|
|
|
=head2 marc_repeatable_subfield |
|
|
|
|
|
Save values for MARC repetable subfield |
|
|
|
|
|
marc_repeatable_subfield('910', 'z', rec('909') ); |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_repeatable_subfield { |
|
|
my ($f,$sf) = @_; |
|
|
die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); |
|
|
$marc_repeatable_subfield->{ $f . $sf }++; |
|
|
marc(@_); |
|
|
} |
|
|
|
|
|
=head2 marc_indicators |
|
|
|
|
|
Set both indicators for MARC field |
|
|
|
|
|
marc_indicators('900', ' ', 1); |
|
|
|
|
|
Any indicator value other than C<0-9> will be treated as undefined. |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_indicators { |
|
|
my $f = shift || die "marc_indicators need field!\n"; |
|
|
my ($i1,$i2) = @_; |
|
|
die "marc_indicators($f, ...) need i1!\n" unless(defined($i1)); |
|
|
die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2)); |
|
|
|
|
|
$i1 = ' ' if ($i1 !~ /^\d$/); |
|
|
$i2 = ' ' if ($i2 !~ /^\d$/); |
|
|
@{ $marc_indicators->{$f} } = ($i1,$i2); |
|
|
} |
|
|
|
|
|
=head2 marc_compose |
|
|
|
|
|
Save values for each MARC subfield explicitly |
|
|
|
|
|
marc_compose('900', |
|
|
'a', rec('200','a') |
|
|
'b', rec('201','a') |
|
|
'a', rec('200','b') |
|
|
'c', rec('200','c') |
|
|
); |
|
|
|
|
|
If you specify C<+> for subfield, value will be appended |
|
|
to previous defined subfield. |
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_compose { |
|
|
my $f = shift or die "marc_compose needs field"; |
|
|
die "marc_compose field must be numer" unless ($f =~ /^\d+$/); |
|
|
|
|
|
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
|
|
my $m = [ $f, $i1, $i2 ]; |
|
|
|
|
|
warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); |
|
|
|
|
|
if ($#_ % 2 != 1) { |
|
|
die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n"; |
|
|
} |
|
|
|
|
|
while (@_) { |
|
|
my $sf = shift; |
|
|
my $v = shift; |
|
|
|
|
|
next unless (defined($v) && $v !~ /^\s*$/); |
|
|
warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); |
|
|
if ($sf ne '+') { |
|
|
push @$m, ( $sf, $v ); |
|
|
} else { |
|
|
$m->[ $#$m ] .= $v; |
|
|
} |
|
|
} |
|
|
|
|
|
warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); |
|
|
|
|
|
push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2); |
|
|
} |
|
|
|
|
|
=head2 marc_duplicate |
|
|
|
|
|
Generate copy of current MARC record and continue working on copy |
|
|
|
|
|
marc_duplicate(); |
|
310 |
|
|
311 |
Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or |
Old name for L<search_display>, it will probably be removed at one point. |
|
C<< _get_marc_fields( offset => 42 ) >>. |
|
312 |
|
|
313 |
=cut |
=cut |
314 |
|
|
315 |
sub marc_duplicate { |
sub tag { |
316 |
my $m = $marc_record->[ -1 ]; |
search_display( @_ ); |
|
die "can't duplicate record which isn't defined" unless ($m); |
|
|
push @{ $marc_record }, dclone( $m ); |
|
|
push @{ $marc_leader }, dclone( marc_leader() ); |
|
|
warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1); |
|
|
$marc_record_offset = $#{ $marc_record }; |
|
|
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1); |
|
|
|
|
317 |
} |
} |
318 |
|
|
319 |
=head2 marc_remove |
=head2 display |
|
|
|
|
Remove some field or subfield from MARC record. |
|
|
|
|
|
marc_remove('200'); |
|
|
marc_remove('200','a'); |
|
|
|
|
|
This will erase field C<200> or C<200^a> from current MARC record. |
|
|
|
|
|
marc_remove('*'); |
|
|
|
|
|
Will remove all fields in current MARC record. |
|
320 |
|
|
321 |
This is useful after calling C<marc_duplicate> or on it's own (but, you |
Define output just for I<display> |
|
should probably just remove that subfield definition if you are not |
|
|
using C<marc_duplicate>). |
|
322 |
|
|
323 |
FIXME: support fields < 10. |
@v = display('Title', rec('200','a') ); |
324 |
|
|
325 |
=cut |
=cut |
326 |
|
|
327 |
sub marc_remove { |
sub display { to( 'display', @_ ) } |
|
my ($f, $sf) = @_; |
|
|
|
|
|
die "marc_remove needs record number" unless defined($f); |
|
|
|
|
|
my $marc = $marc_record->[ $marc_record_offset ]; |
|
328 |
|
|
329 |
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); |
=head2 search |
|
|
|
|
if ($f eq '*') { |
|
|
|
|
|
delete( $marc_record->[ $marc_record_offset ] ); |
|
|
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
|
|
|
|
|
} else { |
|
|
|
|
|
my $i = 0; |
|
|
foreach ( 0 .. $#{ $marc } ) { |
|
|
last unless (defined $marc->[$i]); |
|
|
warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3); |
|
|
if ($marc->[$i]->[0] eq $f) { |
|
|
if (! defined $sf) { |
|
|
# remove whole field |
|
|
splice @$marc, $i, 1; |
|
|
warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3); |
|
|
$i--; |
|
|
} else { |
|
|
foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) { |
|
|
my $o = ($j * 2) + 3; |
|
|
if ($marc->[$i]->[$o] eq $sf) { |
|
|
# remove subfield |
|
|
splice @{$marc->[$i]}, $o, 2; |
|
|
warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3); |
|
|
# is record now empty? |
|
|
if ($#{ $marc->[$i] } == 2) { |
|
|
splice @$marc, $i, 1; |
|
|
warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3); |
|
|
$i--; |
|
|
}; |
|
|
} |
|
|
} |
|
|
} |
|
|
} |
|
|
$i++; |
|
|
} |
|
|
|
|
|
warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); |
|
|
|
|
|
$marc_record->[ $marc_record_offset ] = $marc; |
|
|
} |
|
|
|
|
|
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
|
|
} |
|
|
|
|
|
=head2 marc_original_order |
|
|
|
|
|
Copy all subfields preserving original order to marc field. |
|
|
|
|
|
marc_original_order( marc_field_number, original_input_field_number ); |
|
|
|
|
|
Please note that field numbers are consistent with other commands (marc |
|
|
field number first), but somewhat counter-intuitive (destination and then |
|
|
source). |
|
330 |
|
|
331 |
You might want to use this command if you are just renaming subfields or |
Prepare values just for I<search> |
|
using pre-processing modify_record in C<config.yml> and don't need any |
|
|
post-processing or want to preserve order of original subfields. |
|
332 |
|
|
333 |
|
@v = search('Title', rec('200','a') ); |
334 |
|
|
335 |
=cut |
=cut |
336 |
|
|
337 |
sub marc_original_order { |
sub search { to( 'search', @_ ) } |
|
|
|
|
my ($to, $from) = @_; |
|
|
die "marc_original_order needs from and to fields\n" unless ($from && $to); |
|
|
|
|
|
return unless defined($rec->{$from}); |
|
|
|
|
|
my $r = $rec->{$from}; |
|
|
die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY'); |
|
|
|
|
|
my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' '); |
|
|
warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1); |
|
|
|
|
|
foreach my $d (@$r) { |
|
|
|
|
|
if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') { |
|
|
warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n"; |
|
|
next; |
|
|
} |
|
|
|
|
|
my @sfs = @{ $d->{subfields} }; |
|
|
|
|
|
die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); |
|
|
|
|
|
warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2); |
|
|
|
|
|
my $m = [ $to, $i1, $i2 ]; |
|
|
|
|
|
while (my $sf = shift @sfs) { |
|
|
|
|
|
warn "#--> sf: ",dump($sf), $/ if ($debug > 2); |
|
|
my $offset = shift @sfs; |
|
|
die "corrupted sufields specification for field $from\n" unless defined($offset); |
|
338 |
|
|
339 |
my $v; |
=head2 sorted |
|
if (ref($d->{$sf}) eq 'ARRAY') { |
|
|
$v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset])); |
|
|
} elsif ($offset == 0) { |
|
|
$v = $d->{$sf}; |
|
|
} else { |
|
|
die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf}); |
|
|
} |
|
|
push @$m, ( $sf, $v ) if (defined($v)); |
|
|
} |
|
|
|
|
|
if ($#{$m} > 2) { |
|
|
push @{ $marc_record->[ $marc_record_offset ] }, $m; |
|
|
} |
|
|
} |
|
|
|
|
|
warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); |
|
|
} |
|
|
|
|
|
=head2 marc_count |
|
340 |
|
|
341 |
Return number of MARC records created using L</marc_duplicate>. |
Insert into lists which will be automatically sorted |
342 |
|
|
343 |
print "created ", marc_count(), " records"; |
sorted('Title', rec('200','a') ); |
344 |
|
|
345 |
=cut |
=cut |
346 |
|
|
347 |
sub marc_count { |
sub sorted { to( 'sorted', @_ ) } |
|
return $#{ $marc_record }; |
|
|
} |
|
348 |
|
|
349 |
|
|
350 |
=head1 Functions to extract data from input |
=head1 Functions to extract data from input |
366 |
|
|
367 |
warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1); |
warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1); |
368 |
|
|
369 |
my ($h,$include_subfields) = @_; |
my ($hash,$include_subfields) = @_; |
370 |
|
|
371 |
# sanity and ease of use |
# sanity and ease of use |
372 |
return $h if (ref($h) ne 'HASH'); |
return $hash if (ref($hash) ne 'HASH'); |
373 |
|
|
374 |
|
my $h = dclone( $hash ); |
375 |
|
|
376 |
if ( defined($h->{subfields}) ) { |
if ( defined($h->{subfields}) ) { |
377 |
my $sfs = delete $h->{subfields} || die "no subfields?"; |
my $sfs = delete $h->{subfields} || die "no subfields?"; |
462 |
} else { |
} else { |
463 |
$_->{$sf}; |
$_->{$sf}; |
464 |
} |
} |
465 |
} grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} }; |
} grep { ref($_) eq 'HASH' && defined $_->{$sf} } @{ $rec->{$f} }; |
466 |
} |
} |
467 |
|
|
468 |
=head2 rec |
=head2 rec |
493 |
} |
} |
494 |
} |
} |
495 |
|
|
496 |
|
=head2 frec |
497 |
|
|
498 |
|
Returns first value from field |
499 |
|
|
500 |
|
$v = frec('200'); |
501 |
|
$v = frec('200','a'); |
502 |
|
|
503 |
|
=cut |
504 |
|
|
505 |
|
sub frec { |
506 |
|
my @out = rec(@_); |
507 |
|
warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0; |
508 |
|
return shift @out; |
509 |
|
} |
510 |
|
|
511 |
|
=head2 frec_eq |
512 |
|
|
513 |
|
=head2 frec_ne |
514 |
|
|
515 |
|
Check if first values from two fields are same or different |
516 |
|
|
517 |
|
if ( frec_eq( 900 => 'a', 910 => 'c' ) ) { |
518 |
|
# values are same |
519 |
|
} else { |
520 |
|
# values are different |
521 |
|
} |
522 |
|
|
523 |
|
Strictly speaking C<frec_eq> and C<frec_ne> wouldn't be needed if you |
524 |
|
could write something like: |
525 |
|
|
526 |
|
if ( frec( '900','a' ) eq frec( '910','c' ) ) { |
527 |
|
# yada tada |
528 |
|
} |
529 |
|
|
530 |
|
but you can't since our parser L<WebPAC::Parser> will remove all whitespaces |
531 |
|
in order to parse text and create invalid function C<eqfrec>. |
532 |
|
|
533 |
|
=cut |
534 |
|
|
535 |
|
sub frec_eq { |
536 |
|
my ( $f1,$sf1, $f2, $sf2 ) = @_; |
537 |
|
return (rec( $f1, $sf1 ))[0] eq (rec( $f2, $sf2 ))[0]; |
538 |
|
} |
539 |
|
|
540 |
|
sub frec_ne { |
541 |
|
return ! frec_eq( @_ ); |
542 |
|
} |
543 |
|
|
544 |
=head2 regex |
=head2 regex |
545 |
|
|
546 |
Apply regex to some or all values |
Apply regex to some or all values |
776 |
$database_code = config(); # use _ from hash |
$database_code = config(); # use _ from hash |
777 |
$database_name = config('name'); |
$database_name = config('name'); |
778 |
$database_input_name = config('input name'); |
$database_input_name = config('input name'); |
|
$tag = config('input normalize tag'); |
|
779 |
|
|
780 |
Up to three levels are supported. |
Up to three levels are supported. |
781 |
|
|
836 |
|
|
837 |
sub id { |
sub id { |
838 |
my $mfn = $config->{_mfn} || die "no _mfn in config data"; |
my $mfn = $config->{_mfn} || die "no _mfn in config data"; |
839 |
return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : ''; |
return $mfn . ( WebPAC::Normalize::MARC::_created_marc_records() || '' ); |
840 |
} |
} |
841 |
|
|
842 |
=head2 join_with |
=head2 join_with |