/[webpac2]/trunk/lib/WebPAC/Normalize.pm
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Annotation of /trunk/lib/WebPAC/Normalize.pm

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1018 - (hide annotations)
Fri Nov 9 23:46:44 2007 UTC (15 years, 3 months ago) by dpavlin
File size: 36044 byte(s)
consider only one-char subfields (to correctly skip subfields key which
contains order of subfields), die on subfields which doesn't have
rename mapping defined

1 dpavlin 10 package WebPAC::Normalize;
2 dpavlin 536 use Exporter 'import';
3 dpavlin 980 our @EXPORT = qw/
4 dpavlin 983 _set_ds _set_lookup
5 dpavlin 736 _set_load_row
6 dpavlin 538 _get_ds _clean_ds
7 dpavlin 554 _debug
8 dpavlin 641 _pack_subfields_hash
9 dpavlin 538
10 dpavlin 923 search_display search display sorted
11 dpavlin 912
12 dpavlin 547 marc marc_indicators marc_repeatable_subfield
13 dpavlin 815 marc_compose marc_leader marc_fixed
14 dpavlin 813 marc_duplicate marc_remove marc_count
15 dpavlin 604 marc_original_order
16 dpavlin 1013 marc_template
17 dpavlin 540
18 dpavlin 536 rec1 rec2 rec
19 dpavlin 1015 frec frec_eq frec_ne
20 dpavlin 536 regex prefix suffix surround
21     first lookup join_with
22 dpavlin 707 save_into_lookup
23 dpavlin 562
24     split_rec_on
25 dpavlin 785
26     get set
27 dpavlin 791 count
28 dpavlin 980
29 dpavlin 536 /;
30 dpavlin 10
31     use warnings;
32     use strict;
33 dpavlin 536
34     #use base qw/WebPAC::Common/;
35 dpavlin 550 use Data::Dump qw/dump/;
36 dpavlin 574 use Storable qw/dclone/;
37 dpavlin 725 use Carp qw/confess/;
38 dpavlin 10
39 dpavlin 550 # debugging warn(s)
40     my $debug = 0;
41    
42 dpavlin 980 use WebPAC::Normalize::ISBN;
43     push @EXPORT, ( 'isbn_10', 'isbn_13' );
44 dpavlin 550
45 dpavlin 10 =head1 NAME
46    
47 dpavlin 536 WebPAC::Normalize - describe normalisaton rules using sets
48 dpavlin 10
49     =cut
50    
51 dpavlin 1015 our $VERSION = '0.34';
52 dpavlin 10
53     =head1 SYNOPSIS
54    
55 dpavlin 536 This module uses C<conf/normalize/*.pl> files to perform normalisation
56     from input records using perl functions which are specialized for set
57     processing.
58 dpavlin 10
59 dpavlin 536 Sets are implemented as arrays, and normalisation file is valid perl, which
60     means that you check it's validity before running WebPAC using
61     C<perl -c normalize.pl>.
62 dpavlin 15
63 dpavlin 536 Normalisation can generate multiple output normalized data. For now, supported output
64 dpavlin 912 types (on the left side of definition) are: C<search_display>, C<display>, C<search> and
65 dpavlin 547 C<marc>.
66 dpavlin 15
67 dpavlin 10 =head1 FUNCTIONS
68    
69 dpavlin 538 Functions which start with C<_> are private and used by WebPAC internally.
70     All other functions are available for use within normalisation rules.
71    
72 dpavlin 536 =head2 data_structure
73 dpavlin 10
74 dpavlin 536 Return data structure
75 dpavlin 13
76 dpavlin 538 my $ds = WebPAC::Normalize::data_structure(
77 dpavlin 725 lookup => $lookup_hash,
78 dpavlin 536 row => $row,
79     rules => $normalize_pl_config,
80 dpavlin 541 marc_encoding => 'utf-8',
81 dpavlin 595 config => $config,
82 dpavlin 736 load_row_coderef => sub {
83 dpavlin 979 my ($database,$input,$mfn) = @_;
84 dpavlin 736 $store->load_row( database => $database, input => $input, id => $mfn );
85 dpavlin 725 },
86 dpavlin 13 );
87    
88 dpavlin 707 Options C<row>, C<rules> and C<log> are mandatory while all
89 dpavlin 540 other are optional.
90    
91 dpavlin 736 C<load_row_coderef> is closure only used when executing lookups, so they will
92 dpavlin 725 die if it's not defined.
93    
94 dpavlin 536 This function will B<die> if normalizastion can't be evaled.
95 dpavlin 15
96 dpavlin 538 Since this function isn't exported you have to call it with
97     C<WebPAC::Normalize::data_structure>.
98    
99 dpavlin 10 =cut
100    
101 dpavlin 736 my $load_row_coderef;
102 dpavlin 725
103 dpavlin 536 sub data_structure {
104     my $arg = {@_};
105 dpavlin 13
106 dpavlin 536 die "need row argument" unless ($arg->{row});
107     die "need normalisation argument" unless ($arg->{rules});
108 dpavlin 31
109 dpavlin 730 _set_lookup( $arg->{lookup} ) if defined($arg->{lookup});
110 dpavlin 983 _set_ds( $arg->{row} );
111 dpavlin 730 _set_config( $arg->{config} ) if defined($arg->{config});
112 dpavlin 541 _clean_ds( %{ $arg } );
113 dpavlin 736 $load_row_coderef = $arg->{load_row_coderef};
114 dpavlin 725
115 dpavlin 1011 no strict 'subs';
116     no warnings 'redefine';
117     eval "$arg->{rules};";
118 dpavlin 536 die "error evaling $arg->{rules}: $@\n" if ($@);
119 dpavlin 540
120 dpavlin 538 return _get_ds();
121 dpavlin 10 }
122    
123 dpavlin 983 =head2 _set_ds
124 dpavlin 13
125 dpavlin 536 Set current record hash
126 dpavlin 433
127 dpavlin 983 _set_ds( $rec );
128 dpavlin 433
129     =cut
130    
131 dpavlin 536 my $rec;
132 dpavlin 433
133 dpavlin 983 sub _set_ds {
134 dpavlin 536 $rec = shift or die "no record hash";
135 dpavlin 433 }
136    
137 dpavlin 595 =head2 _set_config
138    
139     Set current config hash
140    
141     _set_config( $config );
142    
143     Magic keys are:
144    
145     =over 4
146    
147     =item _
148    
149     Code of current database
150    
151     =item _mfn
152    
153     Current MFN
154    
155     =back
156    
157     =cut
158    
159     my $config;
160    
161     sub _set_config {
162     $config = shift;
163     }
164    
165 dpavlin 538 =head2 _get_ds
166    
167     Return hash formatted as data structure
168    
169     my $ds = _get_ds();
170    
171     =cut
172    
173 dpavlin 812 my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader);
174 dpavlin 574 my ($marc_record_offset, $marc_fetch_offset) = (0, 0);
175 dpavlin 538
176     sub _get_ds {
177 dpavlin 982 #warn "## out = ",dump($out);
178 dpavlin 538 return $out;
179     }
180    
181     =head2 _clean_ds
182    
183     Clean data structure hash for next record
184    
185     _clean_ds();
186    
187     =cut
188    
189     sub _clean_ds {
190 dpavlin 541 my $a = {@_};
191 dpavlin 812 ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = ();
192 dpavlin 574 ($marc_record_offset, $marc_fetch_offset) = (0,0);
193 dpavlin 541 $marc_encoding = $a->{marc_encoding};
194 dpavlin 538 }
195    
196     =head2 _set_lookup
197    
198     Set current lookup hash
199    
200     _set_lookup( $lookup );
201    
202     =cut
203    
204     my $lookup;
205    
206     sub _set_lookup {
207     $lookup = shift;
208     }
209    
210 dpavlin 707 =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 $lookup;
220     }
221    
222 dpavlin 736 =head2 _set_load_row
223 dpavlin 725
224     Setup code reference which will return L<data_structure> from
225     L<WebPAC::Store>
226    
227 dpavlin 736 _set_load_row(sub {
228 dpavlin 725 my ($database,$input,$mfn) = @_;
229 dpavlin 736 $store->load_row( database => $database, input => $input, id => $mfn );
230 dpavlin 725 });
231    
232     =cut
233    
234 dpavlin 736 sub _set_load_row {
235 dpavlin 725 my $coderef = shift;
236     confess "argument isn't CODE" unless ref($coderef) eq 'CODE';
237    
238 dpavlin 736 $load_row_coderef = $coderef;
239 dpavlin 725 }
240    
241 dpavlin 547 =head2 _get_marc_fields
242 dpavlin 540
243 dpavlin 547 Get all fields defined by calls to C<marc>
244 dpavlin 540
245 dpavlin 547 $marc->add_fields( WebPAC::Normalize:_get_marc_fields() );
246 dpavlin 540
247 dpavlin 543 We are using I<magic> which detect repeatable fields only from
248     sequence of field/subfield data generated by normalization.
249    
250 dpavlin 554 Repeatable field is created when there is second occurence of same subfield or
251     if any of indicators are different.
252 dpavlin 543
253 dpavlin 554 This is sane for most cases. Something like:
254 dpavlin 543
255 dpavlin 554 900a-1 900b-1 900c-1
256     900a-2 900b-2
257     900a-3
258    
259     will be created from any combination of:
260    
261     900a-1 900a-2 900a-3 900b-1 900b-2 900c-1
262    
263     and following rules:
264    
265     marc('900','a', rec('200','a') );
266     marc('900','b', rec('200','b') );
267     marc('900','c', rec('200','c') );
268    
269     which might not be what you have in mind. If you need repeatable subfield,
270     define it using C<marc_repeatable_subfield> like this:
271    
272 dpavlin 574 marc_repeatable_subfield('900','a');
273     marc('900','a', rec('200','a') );
274     marc('900','b', rec('200','b') );
275     marc('900','c', rec('200','c') );
276 dpavlin 554
277 dpavlin 574 will create:
278    
279     900a-1 900a-2 900a-3 900b-1 900c-1
280     900b-2
281    
282     There is also support for returning next or specific using:
283    
284     while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) {
285     # do something with $mf
286     }
287    
288     will always return fields from next MARC record or
289    
290     my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 );
291    
292     will return 42th copy record (if it exists).
293    
294 dpavlin 540 =cut
295    
296 dpavlin 813 my $fetch_pos;
297    
298 dpavlin 547 sub _get_marc_fields {
299 dpavlin 550
300 dpavlin 574 my $arg = {@_};
301     warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2);
302 dpavlin 813 $fetch_pos = $marc_fetch_offset;
303 dpavlin 574 if ($arg->{offset}) {
304 dpavlin 813 $fetch_pos = $arg->{offset};
305 dpavlin 574 } elsif($arg->{fetch_next}) {
306     $marc_fetch_offset++;
307     }
308 dpavlin 550
309 dpavlin 574 return if (! $marc_record || ref($marc_record) ne 'ARRAY');
310    
311     warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2);
312    
313 dpavlin 813 my $marc_rec = $marc_record->[ $fetch_pos ];
314 dpavlin 574
315 dpavlin 813 warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1);
316 dpavlin 574
317     return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0);
318    
319 dpavlin 550 # first, sort all existing fields
320     # XXX might not be needed, but modern perl might randomize elements in hash
321     my @sorted_marc_record = sort {
322 dpavlin 572 $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '')
323 dpavlin 574 } @{ $marc_rec };
324 dpavlin 550
325 dpavlin 574 @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting
326 dpavlin 562
327 dpavlin 550 # output marc fields
328 dpavlin 542 my @m;
329 dpavlin 550
330     # count unique field-subfields (used for offset when walking to next subfield)
331     my $u;
332 dpavlin 572 map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record;
333 dpavlin 550
334     if ($debug) {
335 dpavlin 574 warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield );
336 dpavlin 813 warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/;
337 dpavlin 574 warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/;
338     warn "## subfield count = ", dump( $u ), $/;
339 dpavlin 550 }
340    
341     my $len = $#sorted_marc_record;
342     my $visited;
343     my $i = 0;
344     my $field;
345    
346     foreach ( 0 .. $len ) {
347    
348     # find next element which isn't visited
349     while ($visited->{$i}) {
350     $i = ($i + 1) % ($len + 1);
351 dpavlin 542 }
352    
353 dpavlin 550 # mark it visited
354     $visited->{$i}++;
355    
356 dpavlin 574 my $row = dclone( $sorted_marc_record[$i] );
357 dpavlin 550
358     # field and subfield which is key for
359     # marc_repeatable_subfield and u
360 dpavlin 572 my $fsf = $row->[0] . ( $row->[3] || '' );
361 dpavlin 550
362     if ($debug > 1) {
363    
364     print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/;
365     print "### this [$i]: ", dump( $row ),$/;
366     print "### sf: ", $row->[3], " vs ", $field->[3],
367     $marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/,
368     if ($#$field >= 0);
369    
370     }
371    
372     # if field exists
373     if ( $#$field >= 0 ) {
374     if (
375     $row->[0] ne $field->[0] || # field
376     $row->[1] ne $field->[1] || # i1
377     $row->[2] ne $field->[2] # i2
378     ) {
379     push @m, $field;
380     warn "## saved/1 ", dump( $field ),$/ if ($debug);
381     $field = $row;
382    
383     } elsif (
384     ( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c)
385     ||
386     ( $row->[3] eq $field->[-2] && # same subfield, but not repeatable
387     ! $marc_repeatable_subfield->{ $fsf }
388     )
389     ) {
390     push @m, $field;
391     warn "## saved/2 ", dump( $field ),$/ if ($debug);
392     $field = $row;
393    
394     } else {
395     # append new subfields to existing field
396     push @$field, ( $row->[3], $row->[4] );
397     }
398     } else {
399     # insert first field
400     $field = $row;
401     }
402    
403     if (! $marc_repeatable_subfield->{ $fsf }) {
404     # make step to next subfield
405     $i = ($i + $u->{ $fsf } ) % ($len + 1);
406     }
407 dpavlin 542 }
408    
409 dpavlin 550 if ($#$field >= 0) {
410     push @m, $field;
411     warn "## saved/3 ", dump( $field ),$/ if ($debug);
412     }
413 dpavlin 542
414 dpavlin 579 return \@m;
415 dpavlin 540 }
416    
417 dpavlin 813 =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 dpavlin 554 =head2 _debug
431    
432     Change level of debug warnings
433    
434     _debug( 2 );
435    
436     =cut
437    
438     sub _debug {
439     my $l = shift;
440     return $debug unless defined($l);
441 dpavlin 568 warn "debug level $l",$/ if ($l > 0);
442 dpavlin 554 $debug = $l;
443     }
444    
445 dpavlin 540 =head1 Functions to create C<data_structure>
446    
447     Those functions generally have to first in your normalization file.
448    
449 dpavlin 912 =head2 search_display
450 dpavlin 433
451 dpavlin 912 Define output for L<search> and L<display> at the same time
452 dpavlin 433
453 dpavlin 912 search_display('Title', rec('200','a') );
454 dpavlin 13
455    
456     =cut
457    
458 dpavlin 912 sub search_display {
459     my $name = shift or die "search_display needs name as first argument";
460 dpavlin 536 my @o = grep { defined($_) && $_ ne '' } @_;
461     return unless (@o);
462     $out->{$name}->{search} = \@o;
463     $out->{$name}->{display} = \@o;
464     }
465 dpavlin 13
466 dpavlin 915 =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 dpavlin 536 =head2 display
477 dpavlin 13
478 dpavlin 912 Define output just for I<display>
479 dpavlin 125
480 dpavlin 536 @v = display('Title', rec('200','a') );
481 dpavlin 125
482 dpavlin 536 =cut
483 dpavlin 125
484 dpavlin 923 sub _field {
485     my $type = shift or confess "need type -- BUG?";
486     my $name = shift or confess "needs name as first argument";
487 dpavlin 536 my @o = grep { defined($_) && $_ ne '' } @_;
488     return unless (@o);
489 dpavlin 923 $out->{$name}->{$type} = \@o;
490 dpavlin 536 }
491 dpavlin 13
492 dpavlin 923 sub display { _field( 'display', @_ ) }
493    
494 dpavlin 536 =head2 search
495 dpavlin 13
496 dpavlin 536 Prepare values just for I<search>
497 dpavlin 13
498 dpavlin 536 @v = search('Title', rec('200','a') );
499 dpavlin 433
500 dpavlin 536 =cut
501 dpavlin 13
502 dpavlin 923 sub search { _field( 'search', @_ ) }
503 dpavlin 13
504 dpavlin 923 =head2 sorted
505    
506     Insert into lists which will be automatically sorted
507    
508     sorted('Title', rec('200','a') );
509    
510     =cut
511    
512     sub sorted { _field( 'sorted', @_ ) }
513    
514    
515 dpavlin 564 =head2 marc_leader
516    
517     Setup fields within MARC leader or get leader
518    
519     marc_leader('05','c');
520     my $leader = marc_leader();
521    
522     =cut
523    
524     sub marc_leader {
525     my ($offset,$value) = @_;
526    
527     if ($offset) {
528 dpavlin 813 $marc_leader->[ $marc_record_offset ]->{ $offset } = $value;
529 dpavlin 564 } else {
530 dpavlin 813
531     if (defined($marc_leader)) {
532     die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY');
533     return $marc_leader->[ $marc_record_offset ];
534     } else {
535     return;
536     }
537 dpavlin 564 }
538     }
539    
540 dpavlin 815 =head2 marc_fixed
541    
542     Create control/indentifier fields with values in fixed positions
543    
544     marc_fixed('008', 00, '070402');
545     marc_fixed('008', 39, '|');
546    
547     Positions not specified will be filled with spaces (C<0x20>).
548    
549     There will be no effort to extend last specified value to full length of
550     field in standard.
551    
552     =cut
553    
554     sub marc_fixed {
555     my ($f, $pos, $val) = @_;
556     die "need marc(field, position, value)" unless defined($f) && defined($pos);
557    
558 dpavlin 889 confess "need val" unless defined $val;
559    
560 dpavlin 815 my $update = 0;
561    
562     map {
563     if ($_->[0] eq $f) {
564     my $old = $_->[1];
565 dpavlin 889 if (length($old) <= $pos) {
566 dpavlin 815 $_->[1] .= ' ' x ( $pos - length($old) ) . $val;
567 dpavlin 817 warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1);
568 dpavlin 815 } else {
569     $_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val));
570 dpavlin 817 warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1);
571 dpavlin 815 }
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, $v ];
579 dpavlin 817 warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1);
580 dpavlin 815 }
581     }
582    
583 dpavlin 547 =head2 marc
584 dpavlin 540
585     Save value for MARC field
586    
587 dpavlin 547 marc('900','a', rec('200','a') );
588 dpavlin 571 marc('001', rec('000') );
589 dpavlin 540
590     =cut
591    
592 dpavlin 547 sub marc {
593     my $f = shift or die "marc needs field";
594     die "marc field must be numer" unless ($f =~ /^\d+$/);
595 dpavlin 540
596 dpavlin 571 my $sf;
597     if ($f >= 10) {
598     $sf = shift or die "marc needs subfield";
599     }
600 dpavlin 540
601 dpavlin 541 foreach (@_) {
602     my $v = $_; # make var read-write for Encode
603 dpavlin 543 next unless (defined($v) && $v !~ /^\s*$/);
604 dpavlin 548 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
605 dpavlin 571 if (defined $sf) {
606 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ];
607 dpavlin 571 } else {
608 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
609 dpavlin 571 }
610 dpavlin 540 }
611     }
612    
613 dpavlin 547 =head2 marc_repeatable_subfield
614    
615     Save values for MARC repetable subfield
616    
617     marc_repeatable_subfield('910', 'z', rec('909') );
618    
619     =cut
620    
621     sub marc_repeatable_subfield {
622 dpavlin 550 my ($f,$sf) = @_;
623     die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf);
624     $marc_repeatable_subfield->{ $f . $sf }++;
625 dpavlin 547 marc(@_);
626     }
627    
628     =head2 marc_indicators
629    
630     Set both indicators for MARC field
631    
632     marc_indicators('900', ' ', 1);
633    
634     Any indicator value other than C<0-9> will be treated as undefined.
635    
636     =cut
637    
638     sub marc_indicators {
639     my $f = shift || die "marc_indicators need field!\n";
640     my ($i1,$i2) = @_;
641     die "marc_indicators($f, ...) need i1!\n" unless(defined($i1));
642     die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2));
643    
644     $i1 = ' ' if ($i1 !~ /^\d$/);
645     $i2 = ' ' if ($i2 !~ /^\d$/);
646 dpavlin 548 @{ $marc_indicators->{$f} } = ($i1,$i2);
647 dpavlin 547 }
648    
649 dpavlin 562 =head2 marc_compose
650 dpavlin 547
651 dpavlin 562 Save values for each MARC subfield explicitly
652    
653     marc_compose('900',
654     'a', rec('200','a')
655     'b', rec('201','a')
656     'a', rec('200','b')
657     'c', rec('200','c')
658     );
659    
660 dpavlin 603 If you specify C<+> for subfield, value will be appended
661     to previous defined subfield.
662    
663 dpavlin 562 =cut
664    
665     sub marc_compose {
666     my $f = shift or die "marc_compose needs field";
667     die "marc_compose field must be numer" unless ($f =~ /^\d+$/);
668    
669     my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
670     my $m = [ $f, $i1, $i2 ];
671    
672 dpavlin 583 warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2);
673    
674 dpavlin 619 if ($#_ % 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 dpavlin 562 while (@_) {
679 dpavlin 619 my $sf = shift;
680 dpavlin 565 my $v = shift;
681 dpavlin 562
682     next unless (defined($v) && $v !~ /^\s*$/);
683 dpavlin 568 warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1);
684 dpavlin 603 if ($sf ne '+') {
685     push @$m, ( $sf, $v );
686     } else {
687     $m->[ $#$m ] .= $v;
688     }
689 dpavlin 562 }
690    
691 dpavlin 586 warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1);
692 dpavlin 562
693 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2);
694 dpavlin 562 }
695    
696 dpavlin 574 =head2 marc_duplicate
697 dpavlin 562
698 dpavlin 574 Generate copy of current MARC record and continue working on copy
699    
700     marc_duplicate();
701    
702     Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or
703     C<< _get_marc_fields( offset => 42 ) >>.
704    
705     =cut
706    
707     sub marc_duplicate {
708     my $m = $marc_record->[ -1 ];
709     die "can't duplicate record which isn't defined" unless ($m);
710     push @{ $marc_record }, dclone( $m );
711 dpavlin 813 push @{ $marc_leader }, dclone( marc_leader() );
712     warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1);
713 dpavlin 574 $marc_record_offset = $#{ $marc_record };
714     warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1);
715 dpavlin 813
716 dpavlin 574 }
717    
718     =head2 marc_remove
719    
720     Remove some field or subfield from MARC record.
721    
722     marc_remove('200');
723     marc_remove('200','a');
724    
725     This will erase field C<200> or C<200^a> from current MARC record.
726    
727 dpavlin 786 marc_remove('*');
728    
729     Will remove all fields in current MARC record.
730    
731 dpavlin 574 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
733     using C<marc_duplicate>).
734    
735     FIXME: support fields < 10.
736    
737     =cut
738    
739     sub marc_remove {
740     my ($f, $sf) = @_;
741    
742     die "marc_remove needs record number" unless defined($f);
743    
744     my $marc = $marc_record->[ $marc_record_offset ];
745    
746     warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2);
747    
748 dpavlin 786 if ($f eq '*') {
749    
750     delete( $marc_record->[ $marc_record_offset ] );
751 dpavlin 788 warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
752 dpavlin 786
753     } else {
754    
755     my $i = 0;
756     foreach ( 0 .. $#{ $marc } ) {
757     last unless (defined $marc->[$i]);
758     warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3);
759     if ($marc->[$i]->[0] eq $f) {
760     if (! defined $sf) {
761     # remove whole field
762     splice @$marc, $i, 1;
763     warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3);
764     $i--;
765     } else {
766     foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) {
767     my $o = ($j * 2) + 3;
768     if ($marc->[$i]->[$o] eq $sf) {
769     # remove subfield
770     splice @{$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 dpavlin 574 }
780     }
781     }
782 dpavlin 786 $i++;
783 dpavlin 574 }
784 dpavlin 786
785     warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2);
786    
787     $marc_record->[ $marc_record_offset ] = $marc;
788 dpavlin 574 }
789    
790     warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
791     }
792    
793 dpavlin 604 =head2 marc_original_order
794    
795     Copy all subfields preserving original order to marc field.
796    
797 dpavlin 616 marc_original_order( marc_field_number, original_input_field_number );
798 dpavlin 604
799 dpavlin 616 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 dpavlin 604 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 dpavlin 616
808 dpavlin 604 =cut
809    
810     sub marc_original_order {
811    
812 dpavlin 616 my ($to, $from) = @_;
813 dpavlin 604 die "marc_original_order needs from and to fields\n" unless ($from && $to);
814    
815 dpavlin 616 return unless defined($rec->{$from});
816    
817     my $r = $rec->{$from};
818 dpavlin 1017 die "record field $from isn't array ",dump( $rec ) unless (ref($r) eq 'ARRAY');
819 dpavlin 604
820     my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' ');
821 dpavlin 616 warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1);
822 dpavlin 604
823     foreach my $d (@$r) {
824    
825 dpavlin 605 if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') {
826 dpavlin 616 warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n";
827 dpavlin 605 next;
828     }
829    
830 dpavlin 604 my @sfs = @{ $d->{subfields} };
831    
832     die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1);
833    
834 dpavlin 618 warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2);
835 dpavlin 604
836     my $m = [ $to, $i1, $i2 ];
837    
838     while (my $sf = shift @sfs) {
839 dpavlin 618
840     warn "#--> sf: ",dump($sf), $/ if ($debug > 2);
841 dpavlin 604 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 dpavlin 1013 =head2 marc_template
864    
865     =cut
866    
867     sub marc_template {
868     my $args = {@_};
869     warn "## marc_template(",dump($args),")";
870    
871     foreach ( qw/subfields_rename marc_template/ ) {
872     # warn "ref($_) = ",ref($args->{$_});
873     die "$_ not ARRAY" if ref($args->{$_}) ne 'ARRAY';
874     }
875    
876 dpavlin 1017 my $r = $rec->{ $args->{from} } || return;
877     die "record field ", $args->{from}, " isn't array ",dump( $rec ) unless (ref($r) eq 'ARRAY');
878 dpavlin 1013
879     my @subfields_rename = @{ $args->{subfields_rename} };
880     # warn "### subfields_rename [$#subfields_rename] = ",dump( @subfields_rename );
881    
882     confess "need mapping in pairs for subfields_rename"
883     if $#subfields_rename % 2 != 1;
884    
885     my ( $subfields_rename, $from_subfields, $to_subfields );
886     while ( my ( $from, $to ) = splice(@subfields_rename, 0, 2) ) {
887     my ( $f, $t ) = (
888     $from_subfields->{ $from }++,
889     $to_subfields->{ $to }++
890     );
891     $subfields_rename->{ $from }->[ $f ] = [ $to => $t ];
892     }
893     warn "### subfields_rename = ",dump( $subfields_rename ),$/;
894     warn "### from_subfields = ", dump( $from_subfields ),$/;
895     warn "### to_subfields = ", dump( $to_subfields ),$/;
896    
897     my $fields_re = join('|', keys %$to_subfields );
898    
899     my $pos_templates;
900     my $count;
901     my @marc_order;
902     my $marc_template_order;
903     my $fill_in;
904     my @marc_out;
905    
906     foreach my $template ( @{ $args->{marc_template} } ) {
907     $count = {};
908     @marc_order = ();
909     sub my_count {
910     my $sf = shift;
911     my $nr = $count->{$sf}++;
912     push @marc_order, [ $sf, $nr ];
913     return $sf . $nr;
914     }
915     my $pos_template = $template;
916     $pos_template =~ s/($fields_re)/my_count($1)/ge;
917     my $count_key = dump( $count );
918     warn "### template: |$template| -> |$pos_template| count = $count_key marc_order = ",dump( @marc_order ),$/;
919     $pos_templates->{ $count_key } = $pos_template;
920     $marc_template_order->{ $pos_template } = [ @marc_order ];
921     }
922     warn "### from ",dump( $args->{marc_template} ), " created ", dump( $pos_templates ), " and ", dump( $marc_template_order );
923    
924     my $m;
925    
926     foreach my $r ( @{ $rec->{ $args->{from} } } ) {
927    
928     my $i1 = $r->{i1} || ' ';
929     my $i2 = $r->{i2} || ' ';
930     $m = [ $args->{to}, $i1, $i2 ];
931    
932     warn "### r = ",dump( $r );
933    
934     my ( $new_r, $from_count, $to_count );
935     foreach my $sf ( keys %{$r} ) {
936 dpavlin 1018 # skip everything which isn't one char subfield (e.g. 'subfields')
937     next unless $sf =~ m/^\w$/;
938 dpavlin 1013 my $nr = $from_count->{$sf}++;
939 dpavlin 1018 my $rename_to = $subfields_rename->{ $sf } ||
940     die "can't find subfield rename for $sf/$nr in ", dump( $subfields_rename );
941 dpavlin 1013 warn "### rename $sf/$nr to ", dump( $rename_to->[$nr] ), $/;
942     my ( $to_sf, $to_nr ) = @{ $rename_to->[$nr] };
943     $new_r->{ $to_sf }->[ $to_nr ] = [ $sf => $nr ];
944    
945     $to_count->{ $to_sf }++;
946     }
947    
948     warn "### new_r = ",dump( $new_r );
949    
950     my $from_count_key = dump( $to_count );
951    
952     warn "### from_count = ",dump( $from_count ), $/;
953     warn "### to_count = ",dump( $to_count ), $/;
954    
955     my $template = $pos_templates->{ $from_count_key } ||
956     die "I don't have template for:\n$from_count_key\n## available templates\n", dump( $pos_templates );
957    
958     warn "### selected template: |$template|\n";
959    
960     $fill_in = {};
961    
962     foreach my $sf ( split(/\|/, $template ) ) {
963     sub fill_in {
964     my ( $r, $sf, $nr ) = @_;
965     my ( $from_sf, $from_nr ) = @{ $new_r->{ $sf }->[ $nr ] };
966     my $v = $r->{ $from_sf }; # || die "no $from_sf/$from_nr";
967     warn "#### fill_in( $sf, $nr ) = $from_sf/$from_nr >>>> ",dump( $v ), $/;
968     if ( ref( $v ) eq 'ARRAY' ) {
969     $fill_in->{$sf}->[$nr] = $v->[$from_nr];
970     return $v->[$from_nr];
971     } elsif ( $from_nr == 0 ) {
972     $fill_in->{$sf}->[$nr] = $v;
973     return $v;
974     } else {
975     die "requested subfield $from_sf/$from_nr but it's ",dump( $v );
976     }
977     }
978     warn "#### $sf <<<< $fields_re\n";
979     $sf =~ s/($fields_re)(\d+)/fill_in($r,$1,$2)/ge;
980     warn "#### >>>> $sf with fill_in = ",dump( $fill_in ),$/;
981     }
982    
983     warn "## template: |$template|\n## marc_template_order = ",dump( $marc_template_order );
984    
985     foreach my $sf ( @{ $marc_template_order->{$template} } ) {
986     my ( $sf, $nr ) = @$sf;
987     my $v = $fill_in->{$sf}->[$nr] || die "can't find fill_in $sf/$nr";
988     warn "++ $sf/$nr |$v|\n";
989 dpavlin 1016 push @$m, ( $sf, $v );
990 dpavlin 1013 }
991    
992     warn "#### >>>> created marc: ", dump( $m );
993    
994     push @marc_out, $m;
995     }
996    
997     warn "### marc_template produced: ",dump( @marc_out );
998 dpavlin 1016
999     foreach my $marc ( @marc_out ) {
1000     warn "+++ ",dump( $marc );
1001     push @{ $marc_record->[ $marc_record_offset ] }, $marc;
1002     }
1003 dpavlin 1013 }
1004    
1005 dpavlin 813 =head2 marc_count
1006 dpavlin 604
1007 dpavlin 813 Return number of MARC records created using L</marc_duplicate>.
1008    
1009     print "created ", marc_count(), " records";
1010    
1011     =cut
1012    
1013     sub marc_count {
1014     return $#{ $marc_record };
1015     }
1016    
1017    
1018 dpavlin 540 =head1 Functions to extract data from input
1019    
1020     This function should be used inside functions to create C<data_structure> described
1021     above.
1022    
1023 dpavlin 641 =head2 _pack_subfields_hash
1024    
1025 dpavlin 669 @subfields = _pack_subfields_hash( $h );
1026     $subfields = _pack_subfields_hash( $h, 1 );
1027 dpavlin 641
1028 dpavlin 669 Return each subfield value in array or pack them all together and return scalar
1029     with subfields (denoted by C<^>) and values.
1030    
1031 dpavlin 641 =cut
1032    
1033     sub _pack_subfields_hash {
1034    
1035 dpavlin 642 warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1);
1036 dpavlin 641
1037     my ($h,$include_subfields) = @_;
1038    
1039 dpavlin 831 # sanity and ease of use
1040     return $h if (ref($h) ne 'HASH');
1041    
1042 dpavlin 641 if ( defined($h->{subfields}) ) {
1043     my $sfs = delete $h->{subfields} || die "no subfields?";
1044     my @out;
1045     while (@$sfs) {
1046     my $sf = shift @$sfs;
1047     push @out, '^' . $sf if ($include_subfields);
1048     my $o = shift @$sfs;
1049     if ($o == 0 && ref( $h->{$sf} ) ne 'ARRAY' ) {
1050     # single element subfields are not arrays
1051 dpavlin 669 #warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n";
1052 dpavlin 667
1053 dpavlin 641 push @out, $h->{$sf};
1054     } else {
1055 dpavlin 669 #warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n";
1056 dpavlin 641 push @out, $h->{$sf}->[$o];
1057     }
1058     }
1059 dpavlin 667 if ($include_subfields) {
1060     return join('', @out);
1061     } else {
1062     return @out;
1063     }
1064 dpavlin 641 } else {
1065 dpavlin 667 if ($include_subfields) {
1066     my $out = '';
1067 dpavlin 668 foreach my $sf (sort keys %$h) {
1068 dpavlin 667 if (ref($h->{$sf}) eq 'ARRAY') {
1069     $out .= '^' . $sf . join('^' . $sf, @{ $h->{$sf} });
1070     } else {
1071     $out .= '^' . $sf . $h->{$sf};
1072     }
1073     }
1074     return $out;
1075     } else {
1076     # FIXME this should probably be in alphabetical order instead of hash order
1077     values %{$h};
1078     }
1079 dpavlin 641 }
1080     }
1081    
1082 dpavlin 536 =head2 rec1
1083 dpavlin 371
1084 dpavlin 536 Return all values in some field
1085 dpavlin 371
1086 dpavlin 536 @v = rec1('200')
1087 dpavlin 15
1088 dpavlin 536 TODO: order of values is probably same as in source data, need to investigate that
1089 dpavlin 15
1090 dpavlin 536 =cut
1091 dpavlin 15
1092 dpavlin 536 sub rec1 {
1093     my $f = shift;
1094 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
1095 dpavlin 536 return unless (defined($rec) && defined($rec->{$f}));
1096 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
1097 dpavlin 536 if (ref($rec->{$f}) eq 'ARRAY') {
1098 dpavlin 641 my @out;
1099     foreach my $h ( @{ $rec->{$f} } ) {
1100     if (ref($h) eq 'HASH') {
1101     push @out, ( _pack_subfields_hash( $h ) );
1102 dpavlin 31 } else {
1103 dpavlin 641 push @out, $h;
1104 dpavlin 31 }
1105 dpavlin 641 }
1106     return @out;
1107 dpavlin 536 } elsif( defined($rec->{$f}) ) {
1108     return $rec->{$f};
1109 dpavlin 15 }
1110     }
1111    
1112 dpavlin 536 =head2 rec2
1113 dpavlin 15
1114 dpavlin 536 Return all values in specific field and subfield
1115 dpavlin 13
1116 dpavlin 536 @v = rec2('200','a')
1117 dpavlin 13
1118     =cut
1119    
1120 dpavlin 536 sub rec2 {
1121     my $f = shift;
1122     return unless (defined($rec && $rec->{$f}));
1123     my $sf = shift;
1124 dpavlin 601 warn "rec2($f,$sf) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
1125 dpavlin 589 return map {
1126     if (ref($_->{$sf}) eq 'ARRAY') {
1127     @{ $_->{$sf} };
1128     } else {
1129     $_->{$sf};
1130     }
1131     } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} };
1132 dpavlin 536 }
1133 dpavlin 13
1134 dpavlin 536 =head2 rec
1135 dpavlin 13
1136 dpavlin 536 syntaxtic sugar for
1137 dpavlin 13
1138 dpavlin 536 @v = rec('200')
1139     @v = rec('200','a')
1140 dpavlin 13
1141 dpavlin 750 If rec() returns just single value, it will
1142     return scalar, not array.
1143    
1144 dpavlin 536 =cut
1145 dpavlin 373
1146 dpavlin 536 sub rec {
1147 dpavlin 583 my @out;
1148 dpavlin 536 if ($#_ == 0) {
1149 dpavlin 583 @out = rec1(@_);
1150 dpavlin 536 } elsif ($#_ == 1) {
1151 dpavlin 583 @out = rec2(@_);
1152 dpavlin 13 }
1153 dpavlin 750 if ($#out == 0 && ! wantarray) {
1154     return $out[0];
1155     } elsif (@out) {
1156 dpavlin 583 return @out;
1157     } else {
1158     return '';
1159     }
1160 dpavlin 13 }
1161    
1162 dpavlin 1012 =head2 frec
1163    
1164     Returns first value from field
1165    
1166     $v = frec('200');
1167     $v = frec('200','a');
1168    
1169     =cut
1170    
1171     sub frec {
1172     my @out = rec(@_);
1173     warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0;
1174     return shift @out;
1175     }
1176    
1177 dpavlin 1015 =head2 frec_eq
1178    
1179     =head2 frec_ne
1180    
1181     Check if first values from two fields are same or different
1182    
1183     if ( frec_eq( 900 => 'a', 910 => 'c' ) ) {
1184     # values are same
1185     } else {
1186     # values are different
1187     }
1188    
1189     Strictly speaking C<frec_eq> and C<frec_ne> wouldn't be needed if you
1190     could write something like:
1191    
1192     if ( frec( '900','a' ) eq frec( '910','c' ) ) {
1193     # yada tada
1194     }
1195    
1196     but you can't since our parser L<WebPAC::Parser> will remove all whitespaces
1197     in order to parse text and create invalid function C<eqfrec>.
1198    
1199     =cut
1200    
1201     sub frec_eq {
1202     my ( $f1,$sf1, $f2, $sf2 ) = @_;
1203     return (rec( $f1, $sf1 ))[0] eq (rec( $f2, $sf2 ))[0];
1204     }
1205    
1206     sub frec_ne {
1207     return ! frec_eq( @_ );
1208     }
1209    
1210 dpavlin 536 =head2 regex
1211 dpavlin 15
1212 dpavlin 536 Apply regex to some or all values
1213 dpavlin 15
1214 dpavlin 536 @v = regex( 's/foo/bar/g', @v );
1215 dpavlin 15
1216     =cut
1217    
1218 dpavlin 536 sub regex {
1219     my $r = shift;
1220     my @out;
1221 dpavlin 550 #warn "r: $r\n", dump(\@_);
1222 dpavlin 536 foreach my $t (@_) {
1223     next unless ($t);
1224     eval "\$t =~ $r";
1225     push @out, $t if ($t && $t ne '');
1226 dpavlin 368 }
1227 dpavlin 536 return @out;
1228 dpavlin 15 }
1229    
1230 dpavlin 536 =head2 prefix
1231 dpavlin 15
1232 dpavlin 536 Prefix all values with a string
1233 dpavlin 15
1234 dpavlin 536 @v = prefix( 'my_', @v );
1235 dpavlin 15
1236     =cut
1237    
1238 dpavlin 536 sub prefix {
1239 dpavlin 819 my $p = shift;
1240     return @_ unless defined( $p );
1241 dpavlin 536 return map { $p . $_ } grep { defined($_) } @_;
1242     }
1243 dpavlin 15
1244 dpavlin 536 =head2 suffix
1245 dpavlin 15
1246 dpavlin 536 suffix all values with a string
1247 dpavlin 15
1248 dpavlin 536 @v = suffix( '_my', @v );
1249 dpavlin 15
1250 dpavlin 536 =cut
1251 dpavlin 15
1252 dpavlin 536 sub suffix {
1253 dpavlin 819 my $s = shift;
1254     return @_ unless defined( $s );
1255 dpavlin 536 return map { $_ . $s } grep { defined($_) } @_;
1256 dpavlin 15 }
1257    
1258 dpavlin 536 =head2 surround
1259 dpavlin 13
1260 dpavlin 536 surround all values with a two strings
1261 dpavlin 13
1262 dpavlin 536 @v = surround( 'prefix_', '_suffix', @v );
1263 dpavlin 13
1264     =cut
1265    
1266 dpavlin 536 sub surround {
1267 dpavlin 819 my $p = shift;
1268     my $s = shift;
1269     $p = '' unless defined( $p );
1270     $s = '' unless defined( $s );
1271 dpavlin 536 return map { $p . $_ . $s } grep { defined($_) } @_;
1272 dpavlin 13 }
1273    
1274 dpavlin 536 =head2 first
1275 dpavlin 13
1276 dpavlin 536 Return first element
1277 dpavlin 15
1278 dpavlin 536 $v = first( @v );
1279 dpavlin 13
1280     =cut
1281    
1282 dpavlin 536 sub first {
1283     my $r = shift;
1284     return $r;
1285 dpavlin 13 }
1286    
1287 dpavlin 536 =head2 lookup
1288 dpavlin 13
1289 dpavlin 536 Consult lookup hashes for some value
1290 dpavlin 13
1291 dpavlin 725 @v = lookup(
1292     sub {
1293     'ffkk/peri/mfn'.rec('000')
1294     },
1295     'ffkk','peri','200-a-200-e',
1296     sub {
1297     first(rec(200,'a')).' '.first(rec('200','e'))
1298     }
1299     );
1300 dpavlin 13
1301 dpavlin 725 Code like above will be B<automatically generated> using L<WebPAC::Parse> from
1302     normal lookup definition in C<conf/lookup/something.pl> which looks like:
1303 dpavlin 707
1304 dpavlin 725 lookup(
1305     # which results to return from record recorded in lookup
1306     sub { 'ffkk/peri/mfn' . rec('000') },
1307     # from which database and input
1308     'ffkk','peri',
1309     # such that following values match
1310     sub { first(rec(200,'a')) . ' ' . first(rec('200','e')) },
1311     # if this part is missing, we will try to match same fields
1312     # from lookup record and current one, or you can override
1313     # which records to use from current record using
1314     sub { rec('900','x') . ' ' . rec('900','y') },
1315     )
1316    
1317     You can think about this lookup as SQL (if that helps):
1318    
1319     select
1320     sub { what }
1321     from
1322     database, input
1323     where
1324     sub { filter from lookuped record }
1325     having
1326     sub { optional filter on current record }
1327    
1328     Easy as pie, right?
1329    
1330 dpavlin 13 =cut
1331    
1332 dpavlin 536 sub lookup {
1333 dpavlin 725 my ($what, $database, $input, $key, $having) = @_;
1334    
1335 dpavlin 766 confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4);
1336 dpavlin 725
1337 dpavlin 729 warn "## lookup ($database, $input, $key)", $/ if ($debug > 1);
1338 dpavlin 725 return unless (defined($lookup->{$database}->{$input}->{$key}));
1339    
1340 dpavlin 736 confess "lookup really need load_row_coderef added to data_structure\n" unless ($load_row_coderef);
1341 dpavlin 725
1342     my $mfns;
1343     my @having = $having->();
1344    
1345 dpavlin 729 warn "## having = ", dump( @having ) if ($debug > 2);
1346 dpavlin 725
1347     foreach my $h ( @having ) {
1348     if (defined($lookup->{$database}->{$input}->{$key}->{$h})) {
1349 dpavlin 752 warn "lookup for $database/$input/$key/$h return ",dump($lookup->{$database}->{$input}->{$key}->{$h}),"\n" if ($debug);
1350 dpavlin 725 $mfns->{$_}++ foreach keys %{ $lookup->{$database}->{$input}->{$key}->{$h} };
1351     }
1352 dpavlin 536 }
1353 dpavlin 725
1354     return unless ($mfns);
1355    
1356     my @mfns = sort keys %$mfns;
1357    
1358 dpavlin 750 warn "# lookup loading $database/$input/$key mfn ", join(",",@mfns)," having ",dump(@having),"\n" if ($debug);
1359 dpavlin 725
1360     my $old_rec = $rec;
1361     my @out;
1362    
1363     foreach my $mfn (@mfns) {
1364 dpavlin 736 $rec = $load_row_coderef->( $database, $input, $mfn );
1365 dpavlin 725
1366 dpavlin 752 warn "got $database/$input/$mfn = ", dump($rec), $/ if ($debug);
1367 dpavlin 725
1368     my @vals = $what->();
1369    
1370     push @out, ( @vals );
1371    
1372 dpavlin 752 warn "lookup for mfn $mfn returned ", dump(@vals), $/ if ($debug);
1373 dpavlin 725 }
1374    
1375     # if (ref($lookup->{$k}) eq 'ARRAY') {
1376     # return @{ $lookup->{$k} };
1377     # } else {
1378     # return $lookup->{$k};
1379     # }
1380    
1381     $rec = $old_rec;
1382    
1383 dpavlin 750 warn "## lookup returns = ", dump(@out), $/ if ($debug);
1384 dpavlin 725
1385 dpavlin 740 if ($#out == 0) {
1386     return $out[0];
1387     } else {
1388     return @out;
1389     }
1390 dpavlin 13 }
1391    
1392 dpavlin 707 =head2 save_into_lookup
1393    
1394 dpavlin 725 Save value into lookup. It associates current database, input
1395     and specific keys with one or more values which will be
1396     associated over MFN.
1397 dpavlin 707
1398 dpavlin 725 MFN will be extracted from first occurence current of field 000
1399     in current record, or if it doesn't exist from L<_set_config> C<_mfn>.
1400    
1401     my $nr = save_into_lookup($database,$input,$key,sub {
1402 dpavlin 707 # code which produce one or more values
1403     });
1404    
1405 dpavlin 725 It returns number of items saved.
1406 dpavlin 707
1407 dpavlin 725 This function shouldn't be called directly, it's called from code created by
1408     L<WebPAC::Parser>.
1409    
1410 dpavlin 707 =cut
1411    
1412     sub save_into_lookup {
1413 dpavlin 712 my ($database,$input,$key,$coderef) = @_;
1414     die "save_into_lookup needs database" unless defined($database);
1415     die "save_into_lookup needs input" unless defined($input);
1416     die "save_into_lookup needs key" unless defined($key);
1417 dpavlin 707 die "save_into_lookup needs CODE" unless ( defined($coderef) && ref($coderef) eq 'CODE' );
1418 dpavlin 725
1419 dpavlin 729 warn "## save_into_lookup rec = ", dump($rec), " config = ", dump($config), $/ if ($debug > 2);
1420 dpavlin 725
1421     my $mfn =
1422     defined($rec->{'000'}->[0]) ? $rec->{'000'}->[0] :
1423     defined($config->{_mfn}) ? $config->{_mfn} :
1424     die "mfn not defined or zero";
1425    
1426     my $nr = 0;
1427    
1428 dpavlin 707 foreach my $v ( $coderef->() ) {
1429 dpavlin 712 $lookup->{$database}->{$input}->{$key}->{$v}->{$mfn}++;
1430 dpavlin 721 warn "# saved lookup $database/$input/$key [$v] $mfn\n" if ($debug > 1);
1431 dpavlin 725 $nr++;
1432 dpavlin 707 }
1433 dpavlin 725
1434     return $nr;
1435 dpavlin 707 }
1436    
1437 dpavlin 595 =head2 config
1438    
1439     Consult config values stored in C<config.yml>
1440    
1441     # return database code (key under databases in yaml)
1442     $database_code = config(); # use _ from hash
1443     $database_name = config('name');
1444     $database_input_name = config('input name');
1445    
1446     Up to three levels are supported.
1447    
1448     =cut
1449    
1450     sub config {
1451     return unless ($config);
1452    
1453     my $p = shift;
1454    
1455     $p ||= '';
1456    
1457     my $v;
1458    
1459     warn "### getting config($p)\n" if ($debug > 1);
1460    
1461     my @p = split(/\s+/,$p);
1462     if ($#p < 0) {
1463     $v = $config->{ '_' }; # special, database code
1464     } else {
1465    
1466     my $c = dclone( $config );
1467    
1468     foreach my $k (@p) {
1469     warn "### k: $k c = ",dump($c),$/ if ($debug > 1);
1470     if (ref($c) eq 'ARRAY') {
1471     $c = shift @$c;
1472     warn "config($p) taking first occurence of '$k', probably not what you wanted!\n";
1473     last;
1474     }
1475    
1476     if (! defined($c->{$k}) ) {
1477     $c = undef;
1478     last;
1479     } else {
1480     $c = $c->{$k};
1481     }
1482     }
1483     $v = $c if ($c);
1484    
1485     }
1486    
1487     warn "## config( '$p' ) = ",dump( $v ),$/ if ($v && $debug);
1488     warn "config( '$p' ) is empty\n" if (! $v);
1489    
1490     return $v;
1491     }
1492    
1493     =head2 id
1494    
1495     Returns unique id of this record
1496    
1497     $id = id();
1498    
1499     Returns C<42/2> for 2nd occurence of MFN 42.
1500    
1501     =cut
1502    
1503     sub id {
1504     my $mfn = $config->{_mfn} || die "no _mfn in config data";
1505     return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : '';
1506     }
1507    
1508 dpavlin 536 =head2 join_with
1509 dpavlin 13
1510 dpavlin 536 Joins walues with some delimiter
1511 dpavlin 10
1512 dpavlin 536 $v = join_with(", ", @v);
1513 dpavlin 10
1514 dpavlin 536 =cut
1515 dpavlin 10
1516 dpavlin 536 sub join_with {
1517     my $d = shift;
1518 dpavlin 586 warn "### join_with('$d',",dump(@_),")\n" if ($debug > 2);
1519 dpavlin 583 my $v = join($d, grep { defined($_) && $_ ne '' } @_);
1520     return '' unless defined($v);
1521     return $v;
1522 dpavlin 536 }
1523 dpavlin 10
1524 dpavlin 562 =head2 split_rec_on
1525    
1526     Split record subfield on some regex and take one of parts out
1527    
1528     $a_before_semi_column =
1529     split_rec_on('200','a', /\s*;\s*/, $part);
1530    
1531     C<$part> is optional number of element. First element is
1532     B<1>, not 0!
1533    
1534     If there is no C<$part> parameter or C<$part> is 0, this function will
1535     return all values produced by splitting.
1536    
1537     =cut
1538    
1539     sub split_rec_on {
1540     die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2);
1541    
1542     my ($fld, $sf, $regex, $part) = @_;
1543 dpavlin 568 warn "### regex ", ref($regex), $regex, $/ if ($debug > 2);
1544 dpavlin 562
1545     my @r = rec( $fld, $sf );
1546     my $v = shift @r;
1547 dpavlin 568 warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2);
1548 dpavlin 562
1549 dpavlin 604 return '' if ( ! defined($v) || $v =~ /^\s*$/);
1550 dpavlin 566
1551 dpavlin 562 my @s = split( $regex, $v );
1552 dpavlin 568 warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1);
1553 dpavlin 566 if ($part && $part > 0) {
1554 dpavlin 562 return $s[ $part - 1 ];
1555     } else {
1556 dpavlin 571 return @s;
1557 dpavlin 562 }
1558     }
1559    
1560 dpavlin 785 my $hash;
1561    
1562     =head2 set
1563    
1564     set( key => 'value' );
1565    
1566     =cut
1567    
1568     sub set {
1569     my ($k,$v) = @_;
1570 dpavlin 810 warn "## set ( $k => ", dump($v), " )", $/ if ( $debug );
1571 dpavlin 785 $hash->{$k} = $v;
1572     };
1573    
1574     =head2 get
1575    
1576     get( 'key' );
1577    
1578     =cut
1579    
1580     sub get {
1581     my $k = shift || return;
1582     my $v = $hash->{$k};
1583 dpavlin 810 warn "## get $k = ", dump( $v ), $/ if ( $debug );
1584 dpavlin 785 return $v;
1585     }
1586    
1587 dpavlin 791 =head2 count
1588 dpavlin 785
1589 dpavlin 791 if ( count( @result ) == 1 ) {
1590     # do something if only 1 result is there
1591     }
1592    
1593     =cut
1594    
1595     sub count {
1596 dpavlin 810 warn "## count ",dump(@_),$/ if ( $debug );
1597 dpavlin 791 return @_ . '';
1598     }
1599    
1600 dpavlin 536 # END
1601     1;

  ViewVC Help
Powered by ViewVC 1.1.26