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 |
6 |
_get_ds _clean_ds |
_get_ds _clean_ds |
7 |
_debug |
_debug |
8 |
|
_pack_subfields_hash |
9 |
|
|
10 |
tag search display |
search_display search display sorted |
|
marc marc_indicators marc_repeatable_subfield |
|
|
marc_compose marc_leader |
|
|
marc_duplicate marc_remove |
|
|
marc_original_order |
|
11 |
|
|
12 |
rec1 rec2 rec |
rec1 rec2 rec |
13 |
|
frec frec_eq frec_ne |
14 |
regex prefix suffix surround |
regex prefix suffix surround |
15 |
first lookup join_with |
first lookup join_with |
16 |
|
save_into_lookup |
17 |
|
|
18 |
split_rec_on |
split_rec_on |
19 |
|
|
20 |
|
get set |
21 |
|
count |
22 |
|
|
23 |
/; |
/; |
24 |
|
|
25 |
use warnings; |
use warnings; |
27 |
|
|
28 |
#use base qw/WebPAC::Common/; |
#use base qw/WebPAC::Common/; |
29 |
use Data::Dump qw/dump/; |
use Data::Dump qw/dump/; |
|
use Encode qw/from_to/; |
|
30 |
use Storable qw/dclone/; |
use Storable qw/dclone/; |
31 |
|
use Carp qw/confess/; |
32 |
|
|
33 |
# debugging warn(s) |
# debugging warn(s) |
34 |
my $debug = 0; |
my $debug = 0; |
35 |
|
|
36 |
|
# FIXME |
37 |
|
use WebPAC::Normalize::ISBN; |
38 |
|
push @EXPORT, ( 'isbn_10', 'isbn_13' ); |
39 |
|
|
40 |
|
use WebPAC::Normalize::MARC; |
41 |
|
push @EXPORT, ( qw/ |
42 |
|
marc marc_indicators marc_repeatable_subfield |
43 |
|
marc_compose marc_leader marc_fixed |
44 |
|
marc_duplicate marc_remove marc_count |
45 |
|
marc_original_order |
46 |
|
marc_template |
47 |
|
/); |
48 |
|
|
49 |
=head1 NAME |
=head1 NAME |
50 |
|
|
51 |
WebPAC::Normalize - describe normalisaton rules using sets |
WebPAC::Normalize - describe normalisaton rules using sets |
52 |
|
|
|
=head1 VERSION |
|
|
|
|
|
Version 0.16 |
|
|
|
|
53 |
=cut |
=cut |
54 |
|
|
55 |
our $VERSION = '0.16'; |
our $VERSION = '0.35'; |
56 |
|
|
57 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
58 |
|
|
65 |
C<perl -c normalize.pl>. |
C<perl -c normalize.pl>. |
66 |
|
|
67 |
Normalisation can generate multiple output normalized data. For now, supported output |
Normalisation can generate multiple output normalized data. For now, supported output |
68 |
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 |
69 |
C<marc>. |
C<marc>. |
70 |
|
|
71 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
78 |
Return data structure |
Return data structure |
79 |
|
|
80 |
my $ds = WebPAC::Normalize::data_structure( |
my $ds = WebPAC::Normalize::data_structure( |
81 |
lookup => $lookup->lookup_hash, |
lookup => $lookup_hash, |
82 |
row => $row, |
row => $row, |
83 |
rules => $normalize_pl_config, |
rules => $normalize_pl_config, |
84 |
marc_encoding => 'utf-8', |
marc_encoding => 'utf-8', |
85 |
config => $config, |
config => $config, |
86 |
|
load_row_coderef => sub { |
87 |
|
my ($database,$input,$mfn) = @_; |
88 |
|
$store->load_row( database => $database, input => $input, id => $mfn ); |
89 |
|
}, |
90 |
); |
); |
91 |
|
|
92 |
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 |
93 |
other are optional. |
other are optional. |
94 |
|
|
95 |
|
C<load_row_coderef> is closure only used when executing lookups, so they will |
96 |
|
die if it's not defined. |
97 |
|
|
98 |
This function will B<die> if normalizastion can't be evaled. |
This function will B<die> if normalizastion can't be evaled. |
99 |
|
|
100 |
Since this function isn't exported you have to call it with |
Since this function isn't exported you have to call it with |
102 |
|
|
103 |
=cut |
=cut |
104 |
|
|
105 |
|
my $load_row_coderef; |
106 |
|
|
107 |
sub data_structure { |
sub data_structure { |
108 |
my $arg = {@_}; |
my $arg = {@_}; |
109 |
|
|
110 |
die "need row argument" unless ($arg->{row}); |
die "need row argument" unless ($arg->{row}); |
111 |
die "need normalisation argument" unless ($arg->{rules}); |
die "need normalisation argument" unless ($arg->{rules}); |
112 |
|
|
113 |
no strict 'subs'; |
_set_lookup( $arg->{lookup} ) if defined($arg->{lookup}); |
114 |
_set_lookup( $arg->{lookup} ); |
_set_ds( $arg->{row} ); |
115 |
_set_rec( $arg->{row} ); |
_set_config( $arg->{config} ) if defined($arg->{config}); |
|
_set_config( $arg->{config} ); |
|
116 |
_clean_ds( %{ $arg } ); |
_clean_ds( %{ $arg } ); |
117 |
eval "$arg->{rules}"; |
$load_row_coderef = $arg->{load_row_coderef}; |
118 |
|
|
119 |
|
no strict 'subs'; |
120 |
|
no warnings 'redefine'; |
121 |
|
eval "$arg->{rules};"; |
122 |
die "error evaling $arg->{rules}: $@\n" if ($@); |
die "error evaling $arg->{rules}: $@\n" if ($@); |
123 |
|
|
124 |
return _get_ds(); |
return _get_ds(); |
125 |
} |
} |
126 |
|
|
127 |
=head2 _set_rec |
=head2 _set_ds |
128 |
|
|
129 |
Set current record hash |
Set current record hash |
130 |
|
|
131 |
_set_rec( $rec ); |
_set_ds( $rec ); |
132 |
|
|
133 |
=cut |
=cut |
134 |
|
|
135 |
my $rec; |
my $rec; |
136 |
|
|
137 |
sub _set_rec { |
sub _set_ds { |
138 |
$rec = shift or die "no record hash"; |
$rec = shift or die "no record hash"; |
139 |
|
$WebPAC::Normalize::MARC::rec = $rec; |
140 |
} |
} |
141 |
|
|
142 |
|
=head2 |
143 |
|
|
144 |
|
my $rec = _get_rec(); |
145 |
|
|
146 |
|
=cut |
147 |
|
|
148 |
|
sub _get_rec { $rec }; |
149 |
|
|
150 |
=head2 _set_config |
=head2 _set_config |
151 |
|
|
152 |
Set current config hash |
Set current config hash |
183 |
|
|
184 |
=cut |
=cut |
185 |
|
|
186 |
my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators); |
my $out; |
|
my ($marc_record_offset, $marc_fetch_offset) = (0, 0); |
|
187 |
|
|
188 |
sub _get_ds { |
sub _get_ds { |
189 |
|
#warn "## out = ",dump($out); |
190 |
return $out; |
return $out; |
191 |
} |
} |
192 |
|
|
200 |
|
|
201 |
sub _clean_ds { |
sub _clean_ds { |
202 |
my $a = {@_}; |
my $a = {@_}; |
203 |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = (); |
$out = undef; |
204 |
($marc_record_offset, $marc_fetch_offset) = (0,0); |
WebPAC::Normalize::MARC::_clean(); |
|
$marc_encoding = $a->{marc_encoding}; |
|
205 |
} |
} |
206 |
|
|
207 |
=head2 _set_lookup |
=head2 _set_lookup |
218 |
$lookup = shift; |
$lookup = shift; |
219 |
} |
} |
220 |
|
|
221 |
=head2 _get_marc_fields |
=head2 _get_lookup |
|
|
|
|
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: |
|
222 |
|
|
223 |
while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) { |
Get current lookup hash |
|
# do something with $mf |
|
|
} |
|
|
|
|
|
will always return fields from next MARC record or |
|
|
|
|
|
my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 ); |
|
224 |
|
|
225 |
will return 42th copy record (if it exists). |
my $lookup = _get_lookup(); |
226 |
|
|
227 |
=cut |
=cut |
228 |
|
|
229 |
sub _get_marc_fields { |
sub _get_lookup { |
230 |
|
return $lookup; |
231 |
my $arg = {@_}; |
} |
|
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); |
|
|
my $offset = $marc_fetch_offset; |
|
|
if ($arg->{offset}) { |
|
|
$offset = $arg->{offset}; |
|
|
} elsif($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->[ $offset ]; |
|
|
|
|
|
warn "## _get_marc_fields (at offset: $offset) -- marc_record = ", dump( @$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 ( $marc_repeatable_subfield ); |
|
|
warn "## marc_record[$offset] = ", 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] ); |
|
232 |
|
|
233 |
# field and subfield which is key for |
=head2 _set_load_row |
|
# 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); |
|
234 |
|
|
235 |
} |
Setup code reference which will return L<data_structure> from |
236 |
|
L<WebPAC::Store> |
237 |
|
|
238 |
# if field exists |
_set_load_row(sub { |
239 |
if ( $#$field >= 0 ) { |
my ($database,$input,$mfn) = @_; |
240 |
if ( |
$store->load_row( database => $database, input => $input, id => $mfn ); |
241 |
$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 = $row; |
|
242 |
|
|
243 |
} else { |
=cut |
|
# append new subfields to existing field |
|
|
push @$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); |
|
|
} |
|
|
} |
|
244 |
|
|
245 |
if ($#$field >= 0) { |
sub _set_load_row { |
246 |
push @m, $field; |
my $coderef = shift; |
247 |
warn "## saved/3 ", dump( $field ),$/ if ($debug); |
confess "argument isn't CODE" unless ref($coderef) eq 'CODE'; |
|
} |
|
248 |
|
|
249 |
return \@m; |
$load_row_coderef = $coderef; |
250 |
} |
} |
251 |
|
|
252 |
=head2 _debug |
=head2 _debug |
262 |
return $debug unless defined($l); |
return $debug unless defined($l); |
263 |
warn "debug level $l",$/ if ($l > 0); |
warn "debug level $l",$/ if ($l > 0); |
264 |
$debug = $l; |
$debug = $l; |
265 |
|
$WebPAC::Normalize::MARC::debug = $debug; |
266 |
} |
} |
267 |
|
|
268 |
=head1 Functions to create C<data_structure> |
=head1 Functions to create C<data_structure> |
269 |
|
|
270 |
Those functions generally have to first in your normalization file. |
Those functions generally have to first in your normalization file. |
271 |
|
|
272 |
=head2 tag |
=head2 search_display |
273 |
|
|
274 |
Define new tag for I<search> and I<display>. |
Define output for L<search> and L<display> at the same time |
275 |
|
|
276 |
tag('Title', rec('200','a') ); |
search_display('Title', rec('200','a') ); |
277 |
|
|
278 |
|
|
279 |
=cut |
=cut |
280 |
|
|
281 |
sub tag { |
sub search_display { |
282 |
my $name = shift or die "tag needs name as first argument"; |
my $name = shift or die "search_display needs name as first argument"; |
283 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
284 |
return unless (@o); |
return unless (@o); |
|
$out->{$name}->{tag} = $name; |
|
285 |
$out->{$name}->{search} = \@o; |
$out->{$name}->{search} = \@o; |
286 |
$out->{$name}->{display} = \@o; |
$out->{$name}->{display} = \@o; |
287 |
} |
} |
288 |
|
|
289 |
=head2 display |
=head2 tag |
|
|
|
|
Define tag just for I<display> |
|
290 |
|
|
291 |
@v = display('Title', rec('200','a') ); |
Old name for L<search_display>, but supported |
292 |
|
|
293 |
=cut |
=cut |
294 |
|
|
295 |
sub display { |
sub tag { |
296 |
my $name = shift or die "display needs name as first argument"; |
search_display( @_ ); |
|
my @o = grep { defined($_) && $_ ne '' } @_; |
|
|
return unless (@o); |
|
|
$out->{$name}->{tag} = $name; |
|
|
$out->{$name}->{display} = \@o; |
|
297 |
} |
} |
298 |
|
|
299 |
=head2 search |
=head2 display |
300 |
|
|
301 |
Prepare values just for I<search> |
Define output just for I<display> |
302 |
|
|
303 |
@v = search('Title', rec('200','a') ); |
@v = display('Title', rec('200','a') ); |
304 |
|
|
305 |
=cut |
=cut |
306 |
|
|
307 |
sub search { |
sub _field { |
308 |
my $name = shift or die "search needs name as first argument"; |
my $type = shift or confess "need type -- BUG?"; |
309 |
|
my $name = shift or confess "needs name as first argument"; |
310 |
my @o = grep { defined($_) && $_ ne '' } @_; |
my @o = grep { defined($_) && $_ ne '' } @_; |
311 |
return unless (@o); |
return unless (@o); |
312 |
$out->{$name}->{tag} = $name; |
$out->{$name}->{$type} = \@o; |
|
$out->{$name}->{search} = \@o; |
|
|
} |
|
|
|
|
|
=head2 marc_leader |
|
|
|
|
|
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) { |
|
|
$out->{' leader'}->{ $offset } = $value; |
|
|
} else { |
|
|
return $out->{' leader'}; |
|
|
} |
|
|
} |
|
|
|
|
|
=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) && $v !~ /^\s*$/); |
|
|
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
|
|
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, $v ]; |
|
|
} |
|
|
} |
|
313 |
} |
} |
314 |
|
|
315 |
=head2 marc_repeatable_subfield |
sub display { _field( 'display', @_ ) } |
|
|
|
|
Save values for MARC repetable subfield |
|
316 |
|
|
317 |
marc_repeatable_subfield('910', 'z', rec('909') ); |
=head2 search |
|
|
|
|
=cut |
|
|
|
|
|
sub marc_repeatable_subfield { |
|
|
my ($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 |
|
318 |
|
|
319 |
marc_indicators('900', ' ', 1); |
Prepare values just for I<search> |
320 |
|
|
321 |
Any indicator value other than C<0-9> will be treated as undefined. |
@v = search('Title', rec('200','a') ); |
322 |
|
|
323 |
=cut |
=cut |
324 |
|
|
325 |
sub marc_indicators { |
sub search { _field( 'search', @_ ) } |
|
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 |
|
326 |
|
|
327 |
Save values for each MARC subfield explicitly |
=head2 sorted |
328 |
|
|
329 |
marc_compose('900', |
Insert into lists which will be automatically sorted |
|
'a', rec('200','a') |
|
|
'b', rec('201','a') |
|
|
'a', rec('200','b') |
|
|
'c', rec('200','c') |
|
|
); |
|
330 |
|
|
331 |
If you specify C<+> for subfield, value will be appended |
sorted('Title', rec('200','a') ); |
|
to previous defined subfield. |
|
332 |
|
|
333 |
=cut |
=cut |
334 |
|
|
335 |
sub marc_compose { |
sub sorted { _field( 'sorted', @_ ) } |
|
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, $i2 ]; |
|
336 |
|
|
|
warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); |
|
337 |
|
|
|
while (@_) { |
|
|
my $sf = shift or die "marc_compose $f needs subfield"; |
|
|
my $v = shift; |
|
338 |
|
|
339 |
next unless (defined($v) && $v !~ /^\s*$/); |
=head1 Functions to extract data from input |
|
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
|
|
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); |
|
|
} |
|
340 |
|
|
341 |
=head2 marc_duplicate |
This function should be used inside functions to create C<data_structure> described |
342 |
|
above. |
343 |
|
|
344 |
Generate copy of current MARC record and continue working on copy |
=head2 _pack_subfields_hash |
345 |
|
|
346 |
marc_duplicate(); |
@subfields = _pack_subfields_hash( $h ); |
347 |
|
$subfields = _pack_subfields_hash( $h, 1 ); |
348 |
|
|
349 |
Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or |
Return each subfield value in array or pack them all together and return scalar |
350 |
C<< _get_marc_fields( offset => 42 ) >>. |
with subfields (denoted by C<^>) and values. |
351 |
|
|
352 |
=cut |
=cut |
353 |
|
|
354 |
sub marc_duplicate { |
sub _pack_subfields_hash { |
|
my $m = $marc_record->[ -1 ]; |
|
|
die "can't duplicate record which isn't defined" unless ($m); |
|
|
push @{ $marc_record }, dclone( $m ); |
|
|
warn "## marc_duplicate = ", dump(@$marc_record), $/ if ($debug > 1); |
|
|
$marc_record_offset = $#{ $marc_record }; |
|
|
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1); |
|
|
} |
|
|
|
|
|
=head2 marc_remove |
|
|
|
|
|
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. |
|
|
|
|
|
This is useful after calling C<marc_duplicate> or on it's own (but, you |
|
|
should probably just remove that subfield definition if you are not |
|
|
using C<marc_duplicate>). |
|
355 |
|
|
356 |
FIXME: support fields < 10. |
warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1); |
357 |
|
|
358 |
=cut |
my ($h,$include_subfields) = @_; |
|
|
|
|
sub marc_remove { |
|
|
my ($f, $sf) = @_; |
|
|
|
|
|
die "marc_remove needs record number" unless defined($f); |
|
359 |
|
|
360 |
my $marc = $marc_record->[ $marc_record_offset ]; |
# sanity and ease of use |
361 |
|
return $h if (ref($h) ne 'HASH'); |
362 |
|
|
363 |
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); |
if ( defined($h->{subfields}) ) { |
364 |
|
my $sfs = delete $h->{subfields} || die "no subfields?"; |
365 |
|
my @out; |
366 |
|
while (@$sfs) { |
367 |
|
my $sf = shift @$sfs; |
368 |
|
push @out, '^' . $sf if ($include_subfields); |
369 |
|
my $o = shift @$sfs; |
370 |
|
if ($o == 0 && ref( $h->{$sf} ) ne 'ARRAY' ) { |
371 |
|
# single element subfields are not arrays |
372 |
|
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
373 |
|
|
374 |
my $i = 0; |
push @out, $h->{$sf}; |
|
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--; |
|
375 |
} else { |
} else { |
376 |
foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) { |
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
377 |
my $o = ($j * 2) + 3; |
push @out, $h->{$sf}->[$o]; |
|
if ($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--; |
|
|
}; |
|
|
} |
|
|
} |
|
378 |
} |
} |
379 |
} |
} |
380 |
$i++; |
if ($include_subfields) { |
381 |
} |
return join('', @out); |
382 |
|
} else { |
383 |
warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); |
return @out; |
|
|
|
|
$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). |
|
|
|
|
|
You might want to use this command if you are just renaming subfields or |
|
|
using pre-processing modify_record in C<config.yml> and don't need any |
|
|
post-processing or want to preserve order of original subfields. |
|
|
|
|
|
|
|
|
=cut |
|
|
|
|
|
sub marc_original_order { |
|
|
|
|
|
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; |
|
384 |
} |
} |
385 |
|
} else { |
386 |
my @sfs = @{ $d->{subfields} }; |
if ($include_subfields) { |
387 |
|
my $out = ''; |
388 |
die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); |
foreach my $sf (sort keys %$h) { |
389 |
|
if (ref($h->{$sf}) eq 'ARRAY') { |
390 |
warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/; |
$out .= '^' . $sf . join('^' . $sf, @{ $h->{$sf} }); |
391 |
|
} else { |
392 |
my $m = [ $to, $i1, $i2 ]; |
$out .= '^' . $sf . $h->{$sf}; |
393 |
|
} |
|
while (my $sf = shift @sfs) { |
|
|
warn "#--> sf: ",dump($sf), $/; |
|
|
my $offset = shift @sfs; |
|
|
die "corrupted sufields specification for field $from\n" unless defined($offset); |
|
|
|
|
|
my $v; |
|
|
if (ref($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->{$sf}); |
|
394 |
} |
} |
395 |
push @$m, ( $sf, $v ) if (defined($v)); |
return $out; |
396 |
} |
} else { |
397 |
|
# FIXME this should probably be in alphabetical order instead of hash order |
398 |
if ($#{$m} > 2) { |
values %{$h}; |
|
push @{ $marc_record->[ $marc_record_offset ] }, $m; |
|
399 |
} |
} |
400 |
} |
} |
|
|
|
|
warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); |
|
|
|
|
|
warn "# marc_original_order is partly implemented"; |
|
401 |
} |
} |
402 |
|
|
|
|
|
|
=head1 Functions to extract data from input |
|
|
|
|
|
This function should be used inside functions to create C<data_structure> described |
|
|
above. |
|
|
|
|
403 |
=head2 rec1 |
=head2 rec1 |
404 |
|
|
405 |
Return all values in some field |
Return all values in some field |
416 |
return unless (defined($rec) && defined($rec->{$f})); |
return unless (defined($rec) && defined($rec->{$f})); |
417 |
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
418 |
if (ref($rec->{$f}) eq 'ARRAY') { |
if (ref($rec->{$f}) eq 'ARRAY') { |
419 |
return map { |
my @out; |
420 |
if (ref($_) eq 'HASH') { |
foreach my $h ( @{ $rec->{$f} } ) { |
421 |
values %{$_}; |
if (ref($h) eq 'HASH') { |
422 |
|
push @out, ( _pack_subfields_hash( $h ) ); |
423 |
} else { |
} else { |
424 |
$_; |
push @out, $h; |
425 |
} |
} |
426 |
} @{ $rec->{$f} }; |
} |
427 |
|
return @out; |
428 |
} elsif( defined($rec->{$f}) ) { |
} elsif( defined($rec->{$f}) ) { |
429 |
return $rec->{$f}; |
return $rec->{$f}; |
430 |
} |
} |
459 |
@v = rec('200') |
@v = rec('200') |
460 |
@v = rec('200','a') |
@v = rec('200','a') |
461 |
|
|
462 |
|
If rec() returns just single value, it will |
463 |
|
return scalar, not array. |
464 |
|
|
465 |
=cut |
=cut |
466 |
|
|
467 |
sub rec { |
sub rec { |
471 |
} elsif ($#_ == 1) { |
} elsif ($#_ == 1) { |
472 |
@out = rec2(@_); |
@out = rec2(@_); |
473 |
} |
} |
474 |
if (@out) { |
if ($#out == 0 && ! wantarray) { |
475 |
|
return $out[0]; |
476 |
|
} elsif (@out) { |
477 |
return @out; |
return @out; |
478 |
} else { |
} else { |
479 |
return ''; |
return ''; |
480 |
} |
} |
481 |
} |
} |
482 |
|
|
483 |
|
=head2 frec |
484 |
|
|
485 |
|
Returns first value from field |
486 |
|
|
487 |
|
$v = frec('200'); |
488 |
|
$v = frec('200','a'); |
489 |
|
|
490 |
|
=cut |
491 |
|
|
492 |
|
sub frec { |
493 |
|
my @out = rec(@_); |
494 |
|
warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0; |
495 |
|
return shift @out; |
496 |
|
} |
497 |
|
|
498 |
|
=head2 frec_eq |
499 |
|
|
500 |
|
=head2 frec_ne |
501 |
|
|
502 |
|
Check if first values from two fields are same or different |
503 |
|
|
504 |
|
if ( frec_eq( 900 => 'a', 910 => 'c' ) ) { |
505 |
|
# values are same |
506 |
|
} else { |
507 |
|
# values are different |
508 |
|
} |
509 |
|
|
510 |
|
Strictly speaking C<frec_eq> and C<frec_ne> wouldn't be needed if you |
511 |
|
could write something like: |
512 |
|
|
513 |
|
if ( frec( '900','a' ) eq frec( '910','c' ) ) { |
514 |
|
# yada tada |
515 |
|
} |
516 |
|
|
517 |
|
but you can't since our parser L<WebPAC::Parser> will remove all whitespaces |
518 |
|
in order to parse text and create invalid function C<eqfrec>. |
519 |
|
|
520 |
|
=cut |
521 |
|
|
522 |
|
sub frec_eq { |
523 |
|
my ( $f1,$sf1, $f2, $sf2 ) = @_; |
524 |
|
return (rec( $f1, $sf1 ))[0] eq (rec( $f2, $sf2 ))[0]; |
525 |
|
} |
526 |
|
|
527 |
|
sub frec_ne { |
528 |
|
return ! frec_eq( @_ ); |
529 |
|
} |
530 |
|
|
531 |
=head2 regex |
=head2 regex |
532 |
|
|
533 |
Apply regex to some or all values |
Apply regex to some or all values |
557 |
=cut |
=cut |
558 |
|
|
559 |
sub prefix { |
sub prefix { |
560 |
my $p = shift or return; |
my $p = shift; |
561 |
|
return @_ unless defined( $p ); |
562 |
return map { $p . $_ } grep { defined($_) } @_; |
return map { $p . $_ } grep { defined($_) } @_; |
563 |
} |
} |
564 |
|
|
571 |
=cut |
=cut |
572 |
|
|
573 |
sub suffix { |
sub suffix { |
574 |
my $s = shift or die "suffix needs string as first argument"; |
my $s = shift; |
575 |
|
return @_ unless defined( $s ); |
576 |
return map { $_ . $s } grep { defined($_) } @_; |
return map { $_ . $s } grep { defined($_) } @_; |
577 |
} |
} |
578 |
|
|
585 |
=cut |
=cut |
586 |
|
|
587 |
sub surround { |
sub surround { |
588 |
my $p = shift or die "surround need prefix as first argument"; |
my $p = shift; |
589 |
my $s = shift or die "surround needs suffix as second argument"; |
my $s = shift; |
590 |
|
$p = '' unless defined( $p ); |
591 |
|
$s = '' unless defined( $s ); |
592 |
return map { $p . $_ . $s } grep { defined($_) } @_; |
return map { $p . $_ . $s } grep { defined($_) } @_; |
593 |
} |
} |
594 |
|
|
609 |
|
|
610 |
Consult lookup hashes for some value |
Consult lookup hashes for some value |
611 |
|
|
612 |
@v = lookup( $v ); |
@v = lookup( |
613 |
@v = lookup( @v ); |
sub { |
614 |
|
'ffkk/peri/mfn'.rec('000') |
615 |
|
}, |
616 |
|
'ffkk','peri','200-a-200-e', |
617 |
|
sub { |
618 |
|
first(rec(200,'a')).' '.first(rec('200','e')) |
619 |
|
} |
620 |
|
); |
621 |
|
|
622 |
|
Code like above will be B<automatically generated> using L<WebPAC::Parse> from |
623 |
|
normal lookup definition in C<conf/lookup/something.pl> which looks like: |
624 |
|
|
625 |
|
lookup( |
626 |
|
# which results to return from record recorded in lookup |
627 |
|
sub { 'ffkk/peri/mfn' . rec('000') }, |
628 |
|
# from which database and input |
629 |
|
'ffkk','peri', |
630 |
|
# such that following values match |
631 |
|
sub { first(rec(200,'a')) . ' ' . first(rec('200','e')) }, |
632 |
|
# if this part is missing, we will try to match same fields |
633 |
|
# from lookup record and current one, or you can override |
634 |
|
# which records to use from current record using |
635 |
|
sub { rec('900','x') . ' ' . rec('900','y') }, |
636 |
|
) |
637 |
|
|
638 |
|
You can think about this lookup as SQL (if that helps): |
639 |
|
|
640 |
|
select |
641 |
|
sub { what } |
642 |
|
from |
643 |
|
database, input |
644 |
|
where |
645 |
|
sub { filter from lookuped record } |
646 |
|
having |
647 |
|
sub { optional filter on current record } |
648 |
|
|
649 |
|
Easy as pie, right? |
650 |
|
|
651 |
=cut |
=cut |
652 |
|
|
653 |
sub lookup { |
sub lookup { |
654 |
my $k = shift or return; |
my ($what, $database, $input, $key, $having) = @_; |
655 |
return unless (defined($lookup->{$k})); |
|
656 |
if (ref($lookup->{$k}) eq 'ARRAY') { |
confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4); |
657 |
return @{ $lookup->{$k} }; |
|
658 |
|
warn "## lookup ($database, $input, $key)", $/ if ($debug > 1); |
659 |
|
return unless (defined($lookup->{$database}->{$input}->{$key})); |
660 |
|
|
661 |
|
confess "lookup really need load_row_coderef added to data_structure\n" unless ($load_row_coderef); |
662 |
|
|
663 |
|
my $mfns; |
664 |
|
my @having = $having->(); |
665 |
|
|
666 |
|
warn "## having = ", dump( @having ) if ($debug > 2); |
667 |
|
|
668 |
|
foreach my $h ( @having ) { |
669 |
|
if (defined($lookup->{$database}->{$input}->{$key}->{$h})) { |
670 |
|
warn "lookup for $database/$input/$key/$h return ",dump($lookup->{$database}->{$input}->{$key}->{$h}),"\n" if ($debug); |
671 |
|
$mfns->{$_}++ foreach keys %{ $lookup->{$database}->{$input}->{$key}->{$h} }; |
672 |
|
} |
673 |
|
} |
674 |
|
|
675 |
|
return unless ($mfns); |
676 |
|
|
677 |
|
my @mfns = sort keys %$mfns; |
678 |
|
|
679 |
|
warn "# lookup loading $database/$input/$key mfn ", join(",",@mfns)," having ",dump(@having),"\n" if ($debug); |
680 |
|
|
681 |
|
my $old_rec = $rec; |
682 |
|
my @out; |
683 |
|
|
684 |
|
foreach my $mfn (@mfns) { |
685 |
|
$rec = $load_row_coderef->( $database, $input, $mfn ); |
686 |
|
|
687 |
|
warn "got $database/$input/$mfn = ", dump($rec), $/ if ($debug); |
688 |
|
|
689 |
|
my @vals = $what->(); |
690 |
|
|
691 |
|
push @out, ( @vals ); |
692 |
|
|
693 |
|
warn "lookup for mfn $mfn returned ", dump(@vals), $/ if ($debug); |
694 |
|
} |
695 |
|
|
696 |
|
# if (ref($lookup->{$k}) eq 'ARRAY') { |
697 |
|
# return @{ $lookup->{$k} }; |
698 |
|
# } else { |
699 |
|
# return $lookup->{$k}; |
700 |
|
# } |
701 |
|
|
702 |
|
$rec = $old_rec; |
703 |
|
|
704 |
|
warn "## lookup returns = ", dump(@out), $/ if ($debug); |
705 |
|
|
706 |
|
if ($#out == 0) { |
707 |
|
return $out[0]; |
708 |
} else { |
} else { |
709 |
return $lookup->{$k}; |
return @out; |
710 |
|
} |
711 |
|
} |
712 |
|
|
713 |
|
=head2 save_into_lookup |
714 |
|
|
715 |
|
Save value into lookup. It associates current database, input |
716 |
|
and specific keys with one or more values which will be |
717 |
|
associated over MFN. |
718 |
|
|
719 |
|
MFN will be extracted from first occurence current of field 000 |
720 |
|
in current record, or if it doesn't exist from L<_set_config> C<_mfn>. |
721 |
|
|
722 |
|
my $nr = save_into_lookup($database,$input,$key,sub { |
723 |
|
# code which produce one or more values |
724 |
|
}); |
725 |
|
|
726 |
|
It returns number of items saved. |
727 |
|
|
728 |
|
This function shouldn't be called directly, it's called from code created by |
729 |
|
L<WebPAC::Parser>. |
730 |
|
|
731 |
|
=cut |
732 |
|
|
733 |
|
sub save_into_lookup { |
734 |
|
my ($database,$input,$key,$coderef) = @_; |
735 |
|
die "save_into_lookup needs database" unless defined($database); |
736 |
|
die "save_into_lookup needs input" unless defined($input); |
737 |
|
die "save_into_lookup needs key" unless defined($key); |
738 |
|
die "save_into_lookup needs CODE" unless ( defined($coderef) && ref($coderef) eq 'CODE' ); |
739 |
|
|
740 |
|
warn "## save_into_lookup rec = ", dump($rec), " config = ", dump($config), $/ if ($debug > 2); |
741 |
|
|
742 |
|
my $mfn = |
743 |
|
defined($rec->{'000'}->[0]) ? $rec->{'000'}->[0] : |
744 |
|
defined($config->{_mfn}) ? $config->{_mfn} : |
745 |
|
die "mfn not defined or zero"; |
746 |
|
|
747 |
|
my $nr = 0; |
748 |
|
|
749 |
|
foreach my $v ( $coderef->() ) { |
750 |
|
$lookup->{$database}->{$input}->{$key}->{$v}->{$mfn}++; |
751 |
|
warn "# saved lookup $database/$input/$key [$v] $mfn\n" if ($debug > 1); |
752 |
|
$nr++; |
753 |
} |
} |
754 |
|
|
755 |
|
return $nr; |
756 |
} |
} |
757 |
|
|
758 |
=head2 config |
=head2 config |
763 |
$database_code = config(); # use _ from hash |
$database_code = config(); # use _ from hash |
764 |
$database_name = config('name'); |
$database_name = config('name'); |
765 |
$database_input_name = config('input name'); |
$database_input_name = config('input name'); |
|
$tag = config('input normalize tag'); |
|
766 |
|
|
767 |
Up to three levels are supported. |
Up to three levels are supported. |
768 |
|
|
823 |
|
|
824 |
sub id { |
sub id { |
825 |
my $mfn = $config->{_mfn} || die "no _mfn in config data"; |
my $mfn = $config->{_mfn} || die "no _mfn in config data"; |
826 |
return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : ''; |
return $mfn . ( WebPAC::Normalize::MARC::_created_marc_records() || '' ); |
827 |
} |
} |
828 |
|
|
829 |
=head2 join_with |
=head2 join_with |
878 |
} |
} |
879 |
} |
} |
880 |
|
|
881 |
|
my $hash; |
882 |
|
|
883 |
|
=head2 set |
884 |
|
|
885 |
|
set( key => 'value' ); |
886 |
|
|
887 |
|
=cut |
888 |
|
|
889 |
|
sub set { |
890 |
|
my ($k,$v) = @_; |
891 |
|
warn "## set ( $k => ", dump($v), " )", $/ if ( $debug ); |
892 |
|
$hash->{$k} = $v; |
893 |
|
}; |
894 |
|
|
895 |
|
=head2 get |
896 |
|
|
897 |
|
get( 'key' ); |
898 |
|
|
899 |
|
=cut |
900 |
|
|
901 |
|
sub get { |
902 |
|
my $k = shift || return; |
903 |
|
my $v = $hash->{$k}; |
904 |
|
warn "## get $k = ", dump( $v ), $/ if ( $debug ); |
905 |
|
return $v; |
906 |
|
} |
907 |
|
|
908 |
|
=head2 count |
909 |
|
|
910 |
|
if ( count( @result ) == 1 ) { |
911 |
|
# do something if only 1 result is there |
912 |
|
} |
913 |
|
|
914 |
|
=cut |
915 |
|
|
916 |
|
sub count { |
917 |
|
warn "## count ",dump(@_),$/ if ( $debug ); |
918 |
|
return @_ . ''; |
919 |
|
} |
920 |
|
|
921 |
# END |
# END |
922 |
1; |
1; |