/[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 831 - (hide annotations)
Wed May 23 20:03:14 2007 UTC (15 years, 8 months ago) by dpavlin
File size: 30451 byte(s)
 r1220@llin:  dpavlin | 2007-05-20 19:29:36 +0200
 can be called for fields without subfields now...

1 dpavlin 10 package WebPAC::Normalize;
2 dpavlin 536 use Exporter 'import';
3     @EXPORT = qw/
4 dpavlin 538 _set_rec _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 536 tag search display
11 dpavlin 547 marc marc_indicators marc_repeatable_subfield
12 dpavlin 815 marc_compose marc_leader marc_fixed
13 dpavlin 813 marc_duplicate marc_remove marc_count
14 dpavlin 604 marc_original_order
15 dpavlin 540
16 dpavlin 536 rec1 rec2 rec
17     regex prefix suffix surround
18     first lookup join_with
19 dpavlin 707 save_into_lookup
20 dpavlin 562
21     split_rec_on
22 dpavlin 785
23     get set
24 dpavlin 791 count
25 dpavlin 536 /;
26 dpavlin 10
27     use warnings;
28     use strict;
29 dpavlin 536
30     #use base qw/WebPAC::Common/;
31 dpavlin 550 use Data::Dump qw/dump/;
32 dpavlin 574 use Storable qw/dclone/;
33 dpavlin 725 use Carp qw/confess/;
34 dpavlin 10
35 dpavlin 550 # debugging warn(s)
36     my $debug = 0;
37    
38    
39 dpavlin 10 =head1 NAME
40    
41 dpavlin 536 WebPAC::Normalize - describe normalisaton rules using sets
42 dpavlin 10
43     =head1 VERSION
44    
45 dpavlin 819 Version 0.29
46 dpavlin 10
47     =cut
48    
49 dpavlin 819 our $VERSION = '0.29';
50 dpavlin 10
51     =head1 SYNOPSIS
52    
53 dpavlin 536 This module uses C<conf/normalize/*.pl> files to perform normalisation
54     from input records using perl functions which are specialized for set
55     processing.
56 dpavlin 10
57 dpavlin 536 Sets are implemented as arrays, and normalisation file is valid perl, which
58     means that you check it's validity before running WebPAC using
59     C<perl -c normalize.pl>.
60 dpavlin 15
61 dpavlin 536 Normalisation can generate multiple output normalized data. For now, supported output
62 dpavlin 540 types (on the left side of definition) are: C<tag>, C<display>, C<search> and
63 dpavlin 547 C<marc>.
64 dpavlin 15
65 dpavlin 10 =head1 FUNCTIONS
66    
67 dpavlin 538 Functions which start with C<_> are private and used by WebPAC internally.
68     All other functions are available for use within normalisation rules.
69    
70 dpavlin 536 =head2 data_structure
71 dpavlin 10
72 dpavlin 536 Return data structure
73 dpavlin 13
74 dpavlin 538 my $ds = WebPAC::Normalize::data_structure(
75 dpavlin 725 lookup => $lookup_hash,
76 dpavlin 536 row => $row,
77     rules => $normalize_pl_config,
78 dpavlin 541 marc_encoding => 'utf-8',
79 dpavlin 595 config => $config,
80 dpavlin 736 load_row_coderef => sub {
81 dpavlin 725 my ($database,$input,$mfn) = shift;
82 dpavlin 736 $store->load_row( database => $database, input => $input, id => $mfn );
83 dpavlin 725 },
84 dpavlin 13 );
85    
86 dpavlin 707 Options C<row>, C<rules> and C<log> are mandatory while all
87 dpavlin 540 other are optional.
88    
89 dpavlin 736 C<load_row_coderef> is closure only used when executing lookups, so they will
90 dpavlin 725 die if it's not defined.
91    
92 dpavlin 536 This function will B<die> if normalizastion can't be evaled.
93 dpavlin 15
94 dpavlin 538 Since this function isn't exported you have to call it with
95     C<WebPAC::Normalize::data_structure>.
96    
97 dpavlin 10 =cut
98    
99 dpavlin 736 my $load_row_coderef;
100 dpavlin 725
101 dpavlin 536 sub data_structure {
102     my $arg = {@_};
103 dpavlin 13
104 dpavlin 536 die "need row argument" unless ($arg->{row});
105     die "need normalisation argument" unless ($arg->{rules});
106 dpavlin 31
107 dpavlin 536 no strict 'subs';
108 dpavlin 730 _set_lookup( $arg->{lookup} ) if defined($arg->{lookup});
109 dpavlin 538 _set_rec( $arg->{row} );
110 dpavlin 730 _set_config( $arg->{config} ) if defined($arg->{config});
111 dpavlin 541 _clean_ds( %{ $arg } );
112 dpavlin 736 $load_row_coderef = $arg->{load_row_coderef};
113 dpavlin 725
114 dpavlin 536 eval "$arg->{rules}";
115     die "error evaling $arg->{rules}: $@\n" if ($@);
116 dpavlin 540
117 dpavlin 538 return _get_ds();
118 dpavlin 10 }
119    
120 dpavlin 538 =head2 _set_rec
121 dpavlin 13
122 dpavlin 536 Set current record hash
123 dpavlin 433
124 dpavlin 538 _set_rec( $rec );
125 dpavlin 433
126     =cut
127    
128 dpavlin 536 my $rec;
129 dpavlin 433
130 dpavlin 538 sub _set_rec {
131 dpavlin 536 $rec = shift or die "no record hash";
132 dpavlin 433 }
133    
134 dpavlin 595 =head2 _set_config
135    
136     Set current config hash
137    
138     _set_config( $config );
139    
140     Magic keys are:
141    
142     =over 4
143    
144     =item _
145    
146     Code of current database
147    
148     =item _mfn
149    
150     Current MFN
151    
152     =back
153    
154     =cut
155    
156     my $config;
157    
158     sub _set_config {
159     $config = shift;
160     }
161    
162 dpavlin 538 =head2 _get_ds
163    
164     Return hash formatted as data structure
165    
166     my $ds = _get_ds();
167    
168     =cut
169    
170 dpavlin 812 my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader);
171 dpavlin 574 my ($marc_record_offset, $marc_fetch_offset) = (0, 0);
172 dpavlin 538
173     sub _get_ds {
174     return $out;
175     }
176    
177     =head2 _clean_ds
178    
179     Clean data structure hash for next record
180    
181     _clean_ds();
182    
183     =cut
184    
185     sub _clean_ds {
186 dpavlin 541 my $a = {@_};
187 dpavlin 812 ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = ();
188 dpavlin 574 ($marc_record_offset, $marc_fetch_offset) = (0,0);
189 dpavlin 541 $marc_encoding = $a->{marc_encoding};
190 dpavlin 538 }
191    
192     =head2 _set_lookup
193    
194     Set current lookup hash
195    
196     _set_lookup( $lookup );
197    
198     =cut
199    
200     my $lookup;
201    
202     sub _set_lookup {
203     $lookup = shift;
204     }
205    
206 dpavlin 707 =head2 _get_lookup
207    
208     Get current lookup hash
209    
210     my $lookup = _get_lookup();
211    
212     =cut
213    
214     sub _get_lookup {
215     return $lookup;
216     }
217    
218 dpavlin 736 =head2 _set_load_row
219 dpavlin 725
220     Setup code reference which will return L<data_structure> from
221     L<WebPAC::Store>
222    
223 dpavlin 736 _set_load_row(sub {
224 dpavlin 725 my ($database,$input,$mfn) = @_;
225 dpavlin 736 $store->load_row( database => $database, input => $input, id => $mfn );
226 dpavlin 725 });
227    
228     =cut
229    
230 dpavlin 736 sub _set_load_row {
231 dpavlin 725 my $coderef = shift;
232     confess "argument isn't CODE" unless ref($coderef) eq 'CODE';
233    
234 dpavlin 736 $load_row_coderef = $coderef;
235 dpavlin 725 }
236    
237 dpavlin 547 =head2 _get_marc_fields
238 dpavlin 540
239 dpavlin 547 Get all fields defined by calls to C<marc>
240 dpavlin 540
241 dpavlin 547 $marc->add_fields( WebPAC::Normalize:_get_marc_fields() );
242 dpavlin 540
243 dpavlin 543 We are using I<magic> which detect repeatable fields only from
244     sequence of field/subfield data generated by normalization.
245    
246 dpavlin 554 Repeatable field is created when there is second occurence of same subfield or
247     if any of indicators are different.
248 dpavlin 543
249 dpavlin 554 This is sane for most cases. Something like:
250 dpavlin 543
251 dpavlin 554 900a-1 900b-1 900c-1
252     900a-2 900b-2
253     900a-3
254    
255     will be created from any combination of:
256    
257     900a-1 900a-2 900a-3 900b-1 900b-2 900c-1
258    
259     and following rules:
260    
261     marc('900','a', rec('200','a') );
262     marc('900','b', rec('200','b') );
263     marc('900','c', rec('200','c') );
264    
265     which might not be what you have in mind. If you need repeatable subfield,
266     define it using C<marc_repeatable_subfield> like this:
267    
268 dpavlin 574 marc_repeatable_subfield('900','a');
269     marc('900','a', rec('200','a') );
270     marc('900','b', rec('200','b') );
271     marc('900','c', rec('200','c') );
272 dpavlin 554
273 dpavlin 574 will create:
274    
275     900a-1 900a-2 900a-3 900b-1 900c-1
276     900b-2
277    
278     There is also support for returning next or specific using:
279    
280     while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) {
281     # do something with $mf
282     }
283    
284     will always return fields from next MARC record or
285    
286     my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 );
287    
288     will return 42th copy record (if it exists).
289    
290 dpavlin 540 =cut
291    
292 dpavlin 813 my $fetch_pos;
293    
294 dpavlin 547 sub _get_marc_fields {
295 dpavlin 550
296 dpavlin 574 my $arg = {@_};
297     warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2);
298 dpavlin 813 $fetch_pos = $marc_fetch_offset;
299 dpavlin 574 if ($arg->{offset}) {
300 dpavlin 813 $fetch_pos = $arg->{offset};
301 dpavlin 574 } elsif($arg->{fetch_next}) {
302     $marc_fetch_offset++;
303     }
304 dpavlin 550
305 dpavlin 574 return if (! $marc_record || ref($marc_record) ne 'ARRAY');
306    
307     warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2);
308    
309 dpavlin 813 my $marc_rec = $marc_record->[ $fetch_pos ];
310 dpavlin 574
311 dpavlin 813 warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1);
312 dpavlin 574
313     return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0);
314    
315 dpavlin 550 # first, sort all existing fields
316     # XXX might not be needed, but modern perl might randomize elements in hash
317     my @sorted_marc_record = sort {
318 dpavlin 572 $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '')
319 dpavlin 574 } @{ $marc_rec };
320 dpavlin 550
321 dpavlin 574 @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting
322 dpavlin 562
323 dpavlin 550 # output marc fields
324 dpavlin 542 my @m;
325 dpavlin 550
326     # count unique field-subfields (used for offset when walking to next subfield)
327     my $u;
328 dpavlin 572 map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record;
329 dpavlin 550
330     if ($debug) {
331 dpavlin 574 warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield );
332 dpavlin 813 warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/;
333 dpavlin 574 warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/;
334     warn "## subfield count = ", dump( $u ), $/;
335 dpavlin 550 }
336    
337     my $len = $#sorted_marc_record;
338     my $visited;
339     my $i = 0;
340     my $field;
341    
342     foreach ( 0 .. $len ) {
343    
344     # find next element which isn't visited
345     while ($visited->{$i}) {
346     $i = ($i + 1) % ($len + 1);
347 dpavlin 542 }
348    
349 dpavlin 550 # mark it visited
350     $visited->{$i}++;
351    
352 dpavlin 574 my $row = dclone( $sorted_marc_record[$i] );
353 dpavlin 550
354     # field and subfield which is key for
355     # marc_repeatable_subfield and u
356 dpavlin 572 my $fsf = $row->[0] . ( $row->[3] || '' );
357 dpavlin 550
358     if ($debug > 1) {
359    
360     print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/;
361     print "### this [$i]: ", dump( $row ),$/;
362     print "### sf: ", $row->[3], " vs ", $field->[3],
363     $marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/,
364     if ($#$field >= 0);
365    
366     }
367    
368     # if field exists
369     if ( $#$field >= 0 ) {
370     if (
371     $row->[0] ne $field->[0] || # field
372     $row->[1] ne $field->[1] || # i1
373     $row->[2] ne $field->[2] # i2
374     ) {
375     push @m, $field;
376     warn "## saved/1 ", dump( $field ),$/ if ($debug);
377     $field = $row;
378    
379     } elsif (
380     ( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c)
381     ||
382     ( $row->[3] eq $field->[-2] && # same subfield, but not repeatable
383     ! $marc_repeatable_subfield->{ $fsf }
384     )
385     ) {
386     push @m, $field;
387     warn "## saved/2 ", dump( $field ),$/ if ($debug);
388     $field = $row;
389    
390     } else {
391     # append new subfields to existing field
392     push @$field, ( $row->[3], $row->[4] );
393     }
394     } else {
395     # insert first field
396     $field = $row;
397     }
398    
399     if (! $marc_repeatable_subfield->{ $fsf }) {
400     # make step to next subfield
401     $i = ($i + $u->{ $fsf } ) % ($len + 1);
402     }
403 dpavlin 542 }
404    
405 dpavlin 550 if ($#$field >= 0) {
406     push @m, $field;
407     warn "## saved/3 ", dump( $field ),$/ if ($debug);
408     }
409 dpavlin 542
410 dpavlin 579 return \@m;
411 dpavlin 540 }
412    
413 dpavlin 813 =head2 _get_marc_leader
414    
415     Return leader from currently fetched record by L</_get_marc_fields>
416    
417     print WebPAC::Normalize::_get_marc_leader();
418    
419     =cut
420    
421     sub _get_marc_leader {
422     die "no fetch_pos, did you called _get_marc_fields first?" unless ( defined( $fetch_pos ) );
423     return $marc_leader->[ $fetch_pos ];
424     }
425    
426 dpavlin 554 =head2 _debug
427    
428     Change level of debug warnings
429    
430     _debug( 2 );
431    
432     =cut
433    
434     sub _debug {
435     my $l = shift;
436     return $debug unless defined($l);
437 dpavlin 568 warn "debug level $l",$/ if ($l > 0);
438 dpavlin 554 $debug = $l;
439     }
440    
441 dpavlin 540 =head1 Functions to create C<data_structure>
442    
443     Those functions generally have to first in your normalization file.
444    
445 dpavlin 536 =head2 tag
446 dpavlin 433
447 dpavlin 536 Define new tag for I<search> and I<display>.
448 dpavlin 433
449 dpavlin 536 tag('Title', rec('200','a') );
450 dpavlin 13
451    
452     =cut
453    
454 dpavlin 536 sub tag {
455     my $name = shift or die "tag needs name as first argument";
456     my @o = grep { defined($_) && $_ ne '' } @_;
457     return unless (@o);
458     $out->{$name}->{tag} = $name;
459     $out->{$name}->{search} = \@o;
460     $out->{$name}->{display} = \@o;
461     }
462 dpavlin 13
463 dpavlin 536 =head2 display
464 dpavlin 13
465 dpavlin 536 Define tag just for I<display>
466 dpavlin 125
467 dpavlin 536 @v = display('Title', rec('200','a') );
468 dpavlin 125
469 dpavlin 536 =cut
470 dpavlin 125
471 dpavlin 536 sub display {
472     my $name = shift or die "display needs name as first argument";
473     my @o = grep { defined($_) && $_ ne '' } @_;
474     return unless (@o);
475     $out->{$name}->{tag} = $name;
476     $out->{$name}->{display} = \@o;
477     }
478 dpavlin 13
479 dpavlin 536 =head2 search
480 dpavlin 13
481 dpavlin 536 Prepare values just for I<search>
482 dpavlin 13
483 dpavlin 536 @v = search('Title', rec('200','a') );
484 dpavlin 433
485 dpavlin 536 =cut
486 dpavlin 13
487 dpavlin 536 sub search {
488     my $name = shift or die "search needs name as first argument";
489     my @o = grep { defined($_) && $_ ne '' } @_;
490     return unless (@o);
491     $out->{$name}->{tag} = $name;
492     $out->{$name}->{search} = \@o;
493 dpavlin 13 }
494    
495 dpavlin 564 =head2 marc_leader
496    
497     Setup fields within MARC leader or get leader
498    
499     marc_leader('05','c');
500     my $leader = marc_leader();
501    
502     =cut
503    
504     sub marc_leader {
505     my ($offset,$value) = @_;
506    
507     if ($offset) {
508 dpavlin 813 $marc_leader->[ $marc_record_offset ]->{ $offset } = $value;
509 dpavlin 564 } else {
510 dpavlin 813
511     if (defined($marc_leader)) {
512     die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY');
513     return $marc_leader->[ $marc_record_offset ];
514     } else {
515     return;
516     }
517 dpavlin 564 }
518     }
519    
520 dpavlin 815 =head2 marc_fixed
521    
522     Create control/indentifier fields with values in fixed positions
523    
524     marc_fixed('008', 00, '070402');
525     marc_fixed('008', 39, '|');
526    
527     Positions not specified will be filled with spaces (C<0x20>).
528    
529     There will be no effort to extend last specified value to full length of
530     field in standard.
531    
532     =cut
533    
534     sub marc_fixed {
535     my ($f, $pos, $val) = @_;
536     die "need marc(field, position, value)" unless defined($f) && defined($pos);
537    
538     my $update = 0;
539    
540     map {
541     if ($_->[0] eq $f) {
542     my $old = $_->[1];
543     if (length($old) < $pos) {
544     $_->[1] .= ' ' x ( $pos - length($old) ) . $val;
545 dpavlin 817 warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1);
546 dpavlin 815 } else {
547     $_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val));
548 dpavlin 817 warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1);
549 dpavlin 815 }
550     $update++;
551     }
552     } @{ $marc_record->[ $marc_record_offset ] };
553    
554     if (! $update) {
555     my $v = ' ' x $pos . $val;
556     push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
557 dpavlin 817 warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1);
558 dpavlin 815 }
559     }
560    
561 dpavlin 547 =head2 marc
562 dpavlin 540
563     Save value for MARC field
564    
565 dpavlin 547 marc('900','a', rec('200','a') );
566 dpavlin 571 marc('001', rec('000') );
567 dpavlin 540
568     =cut
569    
570 dpavlin 547 sub marc {
571     my $f = shift or die "marc needs field";
572     die "marc field must be numer" unless ($f =~ /^\d+$/);
573 dpavlin 540
574 dpavlin 571 my $sf;
575     if ($f >= 10) {
576     $sf = shift or die "marc needs subfield";
577     }
578 dpavlin 540
579 dpavlin 541 foreach (@_) {
580     my $v = $_; # make var read-write for Encode
581 dpavlin 543 next unless (defined($v) && $v !~ /^\s*$/);
582 dpavlin 548 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
583 dpavlin 571 if (defined $sf) {
584 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ];
585 dpavlin 571 } else {
586 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
587 dpavlin 571 }
588 dpavlin 540 }
589     }
590    
591 dpavlin 547 =head2 marc_repeatable_subfield
592    
593     Save values for MARC repetable subfield
594    
595     marc_repeatable_subfield('910', 'z', rec('909') );
596    
597     =cut
598    
599     sub marc_repeatable_subfield {
600 dpavlin 550 my ($f,$sf) = @_;
601     die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf);
602     $marc_repeatable_subfield->{ $f . $sf }++;
603 dpavlin 547 marc(@_);
604     }
605    
606     =head2 marc_indicators
607    
608     Set both indicators for MARC field
609    
610     marc_indicators('900', ' ', 1);
611    
612     Any indicator value other than C<0-9> will be treated as undefined.
613    
614     =cut
615    
616     sub marc_indicators {
617     my $f = shift || die "marc_indicators need field!\n";
618     my ($i1,$i2) = @_;
619     die "marc_indicators($f, ...) need i1!\n" unless(defined($i1));
620     die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2));
621    
622     $i1 = ' ' if ($i1 !~ /^\d$/);
623     $i2 = ' ' if ($i2 !~ /^\d$/);
624 dpavlin 548 @{ $marc_indicators->{$f} } = ($i1,$i2);
625 dpavlin 547 }
626    
627 dpavlin 562 =head2 marc_compose
628 dpavlin 547
629 dpavlin 562 Save values for each MARC subfield explicitly
630    
631     marc_compose('900',
632     'a', rec('200','a')
633     'b', rec('201','a')
634     'a', rec('200','b')
635     'c', rec('200','c')
636     );
637    
638 dpavlin 603 If you specify C<+> for subfield, value will be appended
639     to previous defined subfield.
640    
641 dpavlin 562 =cut
642    
643     sub marc_compose {
644     my $f = shift or die "marc_compose needs field";
645     die "marc_compose field must be numer" unless ($f =~ /^\d+$/);
646    
647     my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
648     my $m = [ $f, $i1, $i2 ];
649    
650 dpavlin 583 warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2);
651    
652 dpavlin 619 if ($#_ % 2 != 1) {
653     die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n";
654     }
655    
656 dpavlin 562 while (@_) {
657 dpavlin 619 my $sf = shift;
658 dpavlin 565 my $v = shift;
659 dpavlin 562
660     next unless (defined($v) && $v !~ /^\s*$/);
661 dpavlin 568 warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1);
662 dpavlin 603 if ($sf ne '+') {
663     push @$m, ( $sf, $v );
664     } else {
665     $m->[ $#$m ] .= $v;
666     }
667 dpavlin 562 }
668    
669 dpavlin 586 warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1);
670 dpavlin 562
671 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2);
672 dpavlin 562 }
673    
674 dpavlin 574 =head2 marc_duplicate
675 dpavlin 562
676 dpavlin 574 Generate copy of current MARC record and continue working on copy
677    
678     marc_duplicate();
679    
680     Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or
681     C<< _get_marc_fields( offset => 42 ) >>.
682    
683     =cut
684    
685     sub marc_duplicate {
686     my $m = $marc_record->[ -1 ];
687     die "can't duplicate record which isn't defined" unless ($m);
688     push @{ $marc_record }, dclone( $m );
689 dpavlin 813 push @{ $marc_leader }, dclone( marc_leader() );
690     warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1);
691 dpavlin 574 $marc_record_offset = $#{ $marc_record };
692     warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1);
693 dpavlin 813
694 dpavlin 574 }
695    
696     =head2 marc_remove
697    
698     Remove some field or subfield from MARC record.
699    
700     marc_remove('200');
701     marc_remove('200','a');
702    
703     This will erase field C<200> or C<200^a> from current MARC record.
704    
705 dpavlin 786 marc_remove('*');
706    
707     Will remove all fields in current MARC record.
708    
709 dpavlin 574 This is useful after calling C<marc_duplicate> or on it's own (but, you
710     should probably just remove that subfield definition if you are not
711     using C<marc_duplicate>).
712    
713     FIXME: support fields < 10.
714    
715     =cut
716    
717     sub marc_remove {
718     my ($f, $sf) = @_;
719    
720     die "marc_remove needs record number" unless defined($f);
721    
722     my $marc = $marc_record->[ $marc_record_offset ];
723    
724     warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2);
725    
726 dpavlin 786 if ($f eq '*') {
727    
728     delete( $marc_record->[ $marc_record_offset ] );
729 dpavlin 788 warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
730 dpavlin 786
731     } else {
732    
733     my $i = 0;
734     foreach ( 0 .. $#{ $marc } ) {
735     last unless (defined $marc->[$i]);
736     warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3);
737     if ($marc->[$i]->[0] eq $f) {
738     if (! defined $sf) {
739     # remove whole field
740     splice @$marc, $i, 1;
741     warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3);
742     $i--;
743     } else {
744     foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) {
745     my $o = ($j * 2) + 3;
746     if ($marc->[$i]->[$o] eq $sf) {
747     # remove subfield
748     splice @{$marc->[$i]}, $o, 2;
749     warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3);
750     # is record now empty?
751     if ($#{ $marc->[$i] } == 2) {
752     splice @$marc, $i, 1;
753     warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3);
754     $i--;
755     };
756     }
757 dpavlin 574 }
758     }
759     }
760 dpavlin 786 $i++;
761 dpavlin 574 }
762 dpavlin 786
763     warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2);
764    
765     $marc_record->[ $marc_record_offset ] = $marc;
766 dpavlin 574 }
767    
768     warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
769     }
770    
771 dpavlin 604 =head2 marc_original_order
772    
773     Copy all subfields preserving original order to marc field.
774    
775 dpavlin 616 marc_original_order( marc_field_number, original_input_field_number );
776 dpavlin 604
777 dpavlin 616 Please note that field numbers are consistent with other commands (marc
778     field number first), but somewhat counter-intuitive (destination and then
779     source).
780    
781 dpavlin 604 You might want to use this command if you are just renaming subfields or
782     using pre-processing modify_record in C<config.yml> and don't need any
783     post-processing or want to preserve order of original subfields.
784    
785 dpavlin 616
786 dpavlin 604 =cut
787    
788     sub marc_original_order {
789    
790 dpavlin 616 my ($to, $from) = @_;
791 dpavlin 604 die "marc_original_order needs from and to fields\n" unless ($from && $to);
792    
793 dpavlin 616 return unless defined($rec->{$from});
794    
795     my $r = $rec->{$from};
796 dpavlin 604 die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY');
797    
798     my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' ');
799 dpavlin 616 warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1);
800 dpavlin 604
801     foreach my $d (@$r) {
802    
803 dpavlin 605 if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') {
804 dpavlin 616 warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n";
805 dpavlin 605 next;
806     }
807    
808 dpavlin 604 my @sfs = @{ $d->{subfields} };
809    
810     die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1);
811    
812 dpavlin 618 warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2);
813 dpavlin 604
814     my $m = [ $to, $i1, $i2 ];
815    
816     while (my $sf = shift @sfs) {
817 dpavlin 618
818     warn "#--> sf: ",dump($sf), $/ if ($debug > 2);
819 dpavlin 604 my $offset = shift @sfs;
820     die "corrupted sufields specification for field $from\n" unless defined($offset);
821    
822     my $v;
823     if (ref($d->{$sf}) eq 'ARRAY') {
824     $v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset]));
825     } elsif ($offset == 0) {
826     $v = $d->{$sf};
827     } else {
828     die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf});
829     }
830     push @$m, ( $sf, $v ) if (defined($v));
831     }
832    
833     if ($#{$m} > 2) {
834     push @{ $marc_record->[ $marc_record_offset ] }, $m;
835     }
836     }
837    
838     warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1);
839     }
840    
841 dpavlin 813 =head2 marc_count
842 dpavlin 604
843 dpavlin 813 Return number of MARC records created using L</marc_duplicate>.
844    
845     print "created ", marc_count(), " records";
846    
847     =cut
848    
849     sub marc_count {
850     return $#{ $marc_record };
851     }
852    
853    
854 dpavlin 540 =head1 Functions to extract data from input
855    
856     This function should be used inside functions to create C<data_structure> described
857     above.
858    
859 dpavlin 641 =head2 _pack_subfields_hash
860    
861 dpavlin 669 @subfields = _pack_subfields_hash( $h );
862     $subfields = _pack_subfields_hash( $h, 1 );
863 dpavlin 641
864 dpavlin 669 Return each subfield value in array or pack them all together and return scalar
865     with subfields (denoted by C<^>) and values.
866    
867 dpavlin 641 =cut
868    
869     sub _pack_subfields_hash {
870    
871 dpavlin 642 warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1);
872 dpavlin 641
873     my ($h,$include_subfields) = @_;
874    
875 dpavlin 831 # sanity and ease of use
876     return $h if (ref($h) ne 'HASH');
877    
878 dpavlin 641 if ( defined($h->{subfields}) ) {
879     my $sfs = delete $h->{subfields} || die "no subfields?";
880     my @out;
881     while (@$sfs) {
882     my $sf = shift @$sfs;
883     push @out, '^' . $sf if ($include_subfields);
884     my $o = shift @$sfs;
885     if ($o == 0 && ref( $h->{$sf} ) ne 'ARRAY' ) {
886     # single element subfields are not arrays
887 dpavlin 669 #warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n";
888 dpavlin 667
889 dpavlin 641 push @out, $h->{$sf};
890     } else {
891 dpavlin 669 #warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n";
892 dpavlin 641 push @out, $h->{$sf}->[$o];
893     }
894     }
895 dpavlin 667 if ($include_subfields) {
896     return join('', @out);
897     } else {
898     return @out;
899     }
900 dpavlin 641 } else {
901 dpavlin 667 if ($include_subfields) {
902     my $out = '';
903 dpavlin 668 foreach my $sf (sort keys %$h) {
904 dpavlin 667 if (ref($h->{$sf}) eq 'ARRAY') {
905     $out .= '^' . $sf . join('^' . $sf, @{ $h->{$sf} });
906     } else {
907     $out .= '^' . $sf . $h->{$sf};
908     }
909     }
910     return $out;
911     } else {
912     # FIXME this should probably be in alphabetical order instead of hash order
913     values %{$h};
914     }
915 dpavlin 641 }
916     }
917    
918 dpavlin 536 =head2 rec1
919 dpavlin 371
920 dpavlin 536 Return all values in some field
921 dpavlin 371
922 dpavlin 536 @v = rec1('200')
923 dpavlin 15
924 dpavlin 536 TODO: order of values is probably same as in source data, need to investigate that
925 dpavlin 15
926 dpavlin 536 =cut
927 dpavlin 15
928 dpavlin 536 sub rec1 {
929     my $f = shift;
930 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
931 dpavlin 536 return unless (defined($rec) && defined($rec->{$f}));
932 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
933 dpavlin 536 if (ref($rec->{$f}) eq 'ARRAY') {
934 dpavlin 641 my @out;
935     foreach my $h ( @{ $rec->{$f} } ) {
936     if (ref($h) eq 'HASH') {
937     push @out, ( _pack_subfields_hash( $h ) );
938 dpavlin 31 } else {
939 dpavlin 641 push @out, $h;
940 dpavlin 31 }
941 dpavlin 641 }
942     return @out;
943 dpavlin 536 } elsif( defined($rec->{$f}) ) {
944     return $rec->{$f};
945 dpavlin 15 }
946     }
947    
948 dpavlin 536 =head2 rec2
949 dpavlin 15
950 dpavlin 536 Return all values in specific field and subfield
951 dpavlin 13
952 dpavlin 536 @v = rec2('200','a')
953 dpavlin 13
954     =cut
955    
956 dpavlin 536 sub rec2 {
957     my $f = shift;
958     return unless (defined($rec && $rec->{$f}));
959     my $sf = shift;
960 dpavlin 601 warn "rec2($f,$sf) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
961 dpavlin 589 return map {
962     if (ref($_->{$sf}) eq 'ARRAY') {
963     @{ $_->{$sf} };
964     } else {
965     $_->{$sf};
966     }
967     } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} };
968 dpavlin 536 }
969 dpavlin 13
970 dpavlin 536 =head2 rec
971 dpavlin 13
972 dpavlin 536 syntaxtic sugar for
973 dpavlin 13
974 dpavlin 536 @v = rec('200')
975     @v = rec('200','a')
976 dpavlin 13
977 dpavlin 750 If rec() returns just single value, it will
978     return scalar, not array.
979    
980 dpavlin 536 =cut
981 dpavlin 373
982 dpavlin 536 sub rec {
983 dpavlin 583 my @out;
984 dpavlin 536 if ($#_ == 0) {
985 dpavlin 583 @out = rec1(@_);
986 dpavlin 536 } elsif ($#_ == 1) {
987 dpavlin 583 @out = rec2(@_);
988 dpavlin 13 }
989 dpavlin 750 if ($#out == 0 && ! wantarray) {
990     return $out[0];
991     } elsif (@out) {
992 dpavlin 583 return @out;
993     } else {
994     return '';
995     }
996 dpavlin 13 }
997    
998 dpavlin 536 =head2 regex
999 dpavlin 15
1000 dpavlin 536 Apply regex to some or all values
1001 dpavlin 15
1002 dpavlin 536 @v = regex( 's/foo/bar/g', @v );
1003 dpavlin 15
1004     =cut
1005    
1006 dpavlin 536 sub regex {
1007     my $r = shift;
1008     my @out;
1009 dpavlin 550 #warn "r: $r\n", dump(\@_);
1010 dpavlin 536 foreach my $t (@_) {
1011     next unless ($t);
1012     eval "\$t =~ $r";
1013     push @out, $t if ($t && $t ne '');
1014 dpavlin 368 }
1015 dpavlin 536 return @out;
1016 dpavlin 15 }
1017    
1018 dpavlin 536 =head2 prefix
1019 dpavlin 15
1020 dpavlin 536 Prefix all values with a string
1021 dpavlin 15
1022 dpavlin 536 @v = prefix( 'my_', @v );
1023 dpavlin 15
1024     =cut
1025    
1026 dpavlin 536 sub prefix {
1027 dpavlin 819 my $p = shift;
1028     return @_ unless defined( $p );
1029 dpavlin 536 return map { $p . $_ } grep { defined($_) } @_;
1030     }
1031 dpavlin 15
1032 dpavlin 536 =head2 suffix
1033 dpavlin 15
1034 dpavlin 536 suffix all values with a string
1035 dpavlin 15
1036 dpavlin 536 @v = suffix( '_my', @v );
1037 dpavlin 15
1038 dpavlin 536 =cut
1039 dpavlin 15
1040 dpavlin 536 sub suffix {
1041 dpavlin 819 my $s = shift;
1042     return @_ unless defined( $s );
1043 dpavlin 536 return map { $_ . $s } grep { defined($_) } @_;
1044 dpavlin 15 }
1045    
1046 dpavlin 536 =head2 surround
1047 dpavlin 13
1048 dpavlin 536 surround all values with a two strings
1049 dpavlin 13
1050 dpavlin 536 @v = surround( 'prefix_', '_suffix', @v );
1051 dpavlin 13
1052     =cut
1053    
1054 dpavlin 536 sub surround {
1055 dpavlin 819 my $p = shift;
1056     my $s = shift;
1057     $p = '' unless defined( $p );
1058     $s = '' unless defined( $s );
1059 dpavlin 536 return map { $p . $_ . $s } grep { defined($_) } @_;
1060 dpavlin 13 }
1061    
1062 dpavlin 536 =head2 first
1063 dpavlin 13
1064 dpavlin 536 Return first element
1065 dpavlin 15
1066 dpavlin 536 $v = first( @v );
1067 dpavlin 13
1068     =cut
1069    
1070 dpavlin 536 sub first {
1071     my $r = shift;
1072     return $r;
1073 dpavlin 13 }
1074    
1075 dpavlin 536 =head2 lookup
1076 dpavlin 13
1077 dpavlin 536 Consult lookup hashes for some value
1078 dpavlin 13
1079 dpavlin 725 @v = lookup(
1080     sub {
1081     'ffkk/peri/mfn'.rec('000')
1082     },
1083     'ffkk','peri','200-a-200-e',
1084     sub {
1085     first(rec(200,'a')).' '.first(rec('200','e'))
1086     }
1087     );
1088 dpavlin 13
1089 dpavlin 725 Code like above will be B<automatically generated> using L<WebPAC::Parse> from
1090     normal lookup definition in C<conf/lookup/something.pl> which looks like:
1091 dpavlin 707
1092 dpavlin 725 lookup(
1093     # which results to return from record recorded in lookup
1094     sub { 'ffkk/peri/mfn' . rec('000') },
1095     # from which database and input
1096     'ffkk','peri',
1097     # such that following values match
1098     sub { first(rec(200,'a')) . ' ' . first(rec('200','e')) },
1099     # if this part is missing, we will try to match same fields
1100     # from lookup record and current one, or you can override
1101     # which records to use from current record using
1102     sub { rec('900','x') . ' ' . rec('900','y') },
1103     )
1104    
1105     You can think about this lookup as SQL (if that helps):
1106    
1107     select
1108     sub { what }
1109     from
1110     database, input
1111     where
1112     sub { filter from lookuped record }
1113     having
1114     sub { optional filter on current record }
1115    
1116     Easy as pie, right?
1117    
1118 dpavlin 13 =cut
1119    
1120 dpavlin 536 sub lookup {
1121 dpavlin 725 my ($what, $database, $input, $key, $having) = @_;
1122    
1123 dpavlin 766 confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4);
1124 dpavlin 725
1125 dpavlin 729 warn "## lookup ($database, $input, $key)", $/ if ($debug > 1);
1126 dpavlin 725 return unless (defined($lookup->{$database}->{$input}->{$key}));
1127    
1128 dpavlin 736 confess "lookup really need load_row_coderef added to data_structure\n" unless ($load_row_coderef);
1129 dpavlin 725
1130     my $mfns;
1131     my @having = $having->();
1132    
1133 dpavlin 729 warn "## having = ", dump( @having ) if ($debug > 2);
1134 dpavlin 725
1135     foreach my $h ( @having ) {
1136     if (defined($lookup->{$database}->{$input}->{$key}->{$h})) {
1137 dpavlin 752 warn "lookup for $database/$input/$key/$h return ",dump($lookup->{$database}->{$input}->{$key}->{$h}),"\n" if ($debug);
1138 dpavlin 725 $mfns->{$_}++ foreach keys %{ $lookup->{$database}->{$input}->{$key}->{$h} };
1139     }
1140 dpavlin 536 }
1141 dpavlin 725
1142     return unless ($mfns);
1143    
1144     my @mfns = sort keys %$mfns;
1145    
1146 dpavlin 750 warn "# lookup loading $database/$input/$key mfn ", join(",",@mfns)," having ",dump(@having),"\n" if ($debug);
1147 dpavlin 725
1148     my $old_rec = $rec;
1149     my @out;
1150    
1151     foreach my $mfn (@mfns) {
1152 dpavlin 736 $rec = $load_row_coderef->( $database, $input, $mfn );
1153 dpavlin 725
1154 dpavlin 752 warn "got $database/$input/$mfn = ", dump($rec), $/ if ($debug);
1155 dpavlin 725
1156     my @vals = $what->();
1157    
1158     push @out, ( @vals );
1159    
1160 dpavlin 752 warn "lookup for mfn $mfn returned ", dump(@vals), $/ if ($debug);
1161 dpavlin 725 }
1162    
1163     # if (ref($lookup->{$k}) eq 'ARRAY') {
1164     # return @{ $lookup->{$k} };
1165     # } else {
1166     # return $lookup->{$k};
1167     # }
1168    
1169     $rec = $old_rec;
1170    
1171 dpavlin 750 warn "## lookup returns = ", dump(@out), $/ if ($debug);
1172 dpavlin 725
1173 dpavlin 740 if ($#out == 0) {
1174     return $out[0];
1175     } else {
1176     return @out;
1177     }
1178 dpavlin 13 }
1179    
1180 dpavlin 707 =head2 save_into_lookup
1181    
1182 dpavlin 725 Save value into lookup. It associates current database, input
1183     and specific keys with one or more values which will be
1184     associated over MFN.
1185 dpavlin 707
1186 dpavlin 725 MFN will be extracted from first occurence current of field 000
1187     in current record, or if it doesn't exist from L<_set_config> C<_mfn>.
1188    
1189     my $nr = save_into_lookup($database,$input,$key,sub {
1190 dpavlin 707 # code which produce one or more values
1191     });
1192    
1193 dpavlin 725 It returns number of items saved.
1194 dpavlin 707
1195 dpavlin 725 This function shouldn't be called directly, it's called from code created by
1196     L<WebPAC::Parser>.
1197    
1198 dpavlin 707 =cut
1199    
1200     sub save_into_lookup {
1201 dpavlin 712 my ($database,$input,$key,$coderef) = @_;
1202     die "save_into_lookup needs database" unless defined($database);
1203     die "save_into_lookup needs input" unless defined($input);
1204     die "save_into_lookup needs key" unless defined($key);
1205 dpavlin 707 die "save_into_lookup needs CODE" unless ( defined($coderef) && ref($coderef) eq 'CODE' );
1206 dpavlin 725
1207 dpavlin 729 warn "## save_into_lookup rec = ", dump($rec), " config = ", dump($config), $/ if ($debug > 2);
1208 dpavlin 725
1209     my $mfn =
1210     defined($rec->{'000'}->[0]) ? $rec->{'000'}->[0] :
1211     defined($config->{_mfn}) ? $config->{_mfn} :
1212     die "mfn not defined or zero";
1213    
1214     my $nr = 0;
1215    
1216 dpavlin 707 foreach my $v ( $coderef->() ) {
1217 dpavlin 712 $lookup->{$database}->{$input}->{$key}->{$v}->{$mfn}++;
1218 dpavlin 721 warn "# saved lookup $database/$input/$key [$v] $mfn\n" if ($debug > 1);
1219 dpavlin 725 $nr++;
1220 dpavlin 707 }
1221 dpavlin 725
1222     return $nr;
1223 dpavlin 707 }
1224    
1225 dpavlin 595 =head2 config
1226    
1227     Consult config values stored in C<config.yml>
1228    
1229     # return database code (key under databases in yaml)
1230     $database_code = config(); # use _ from hash
1231     $database_name = config('name');
1232     $database_input_name = config('input name');
1233     $tag = config('input normalize tag');
1234    
1235     Up to three levels are supported.
1236    
1237     =cut
1238    
1239     sub config {
1240     return unless ($config);
1241    
1242     my $p = shift;
1243    
1244     $p ||= '';
1245    
1246     my $v;
1247    
1248     warn "### getting config($p)\n" if ($debug > 1);
1249    
1250     my @p = split(/\s+/,$p);
1251     if ($#p < 0) {
1252     $v = $config->{ '_' }; # special, database code
1253     } else {
1254    
1255     my $c = dclone( $config );
1256    
1257     foreach my $k (@p) {
1258     warn "### k: $k c = ",dump($c),$/ if ($debug > 1);
1259     if (ref($c) eq 'ARRAY') {
1260     $c = shift @$c;
1261     warn "config($p) taking first occurence of '$k', probably not what you wanted!\n";
1262     last;
1263     }
1264    
1265     if (! defined($c->{$k}) ) {
1266     $c = undef;
1267     last;
1268     } else {
1269     $c = $c->{$k};
1270     }
1271     }
1272     $v = $c if ($c);
1273    
1274     }
1275    
1276     warn "## config( '$p' ) = ",dump( $v ),$/ if ($v && $debug);
1277     warn "config( '$p' ) is empty\n" if (! $v);
1278    
1279     return $v;
1280     }
1281    
1282     =head2 id
1283    
1284     Returns unique id of this record
1285    
1286     $id = id();
1287    
1288     Returns C<42/2> for 2nd occurence of MFN 42.
1289    
1290     =cut
1291    
1292     sub id {
1293     my $mfn = $config->{_mfn} || die "no _mfn in config data";
1294     return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : '';
1295     }
1296    
1297 dpavlin 536 =head2 join_with
1298 dpavlin 13
1299 dpavlin 536 Joins walues with some delimiter
1300 dpavlin 10
1301 dpavlin 536 $v = join_with(", ", @v);
1302 dpavlin 10
1303 dpavlin 536 =cut
1304 dpavlin 10
1305 dpavlin 536 sub join_with {
1306     my $d = shift;
1307 dpavlin 586 warn "### join_with('$d',",dump(@_),")\n" if ($debug > 2);
1308 dpavlin 583 my $v = join($d, grep { defined($_) && $_ ne '' } @_);
1309     return '' unless defined($v);
1310     return $v;
1311 dpavlin 536 }
1312 dpavlin 10
1313 dpavlin 562 =head2 split_rec_on
1314    
1315     Split record subfield on some regex and take one of parts out
1316    
1317     $a_before_semi_column =
1318     split_rec_on('200','a', /\s*;\s*/, $part);
1319    
1320     C<$part> is optional number of element. First element is
1321     B<1>, not 0!
1322    
1323     If there is no C<$part> parameter or C<$part> is 0, this function will
1324     return all values produced by splitting.
1325    
1326     =cut
1327    
1328     sub split_rec_on {
1329     die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2);
1330    
1331     my ($fld, $sf, $regex, $part) = @_;
1332 dpavlin 568 warn "### regex ", ref($regex), $regex, $/ if ($debug > 2);
1333 dpavlin 562
1334     my @r = rec( $fld, $sf );
1335     my $v = shift @r;
1336 dpavlin 568 warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2);
1337 dpavlin 562
1338 dpavlin 604 return '' if ( ! defined($v) || $v =~ /^\s*$/);
1339 dpavlin 566
1340 dpavlin 562 my @s = split( $regex, $v );
1341 dpavlin 568 warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1);
1342 dpavlin 566 if ($part && $part > 0) {
1343 dpavlin 562 return $s[ $part - 1 ];
1344     } else {
1345 dpavlin 571 return @s;
1346 dpavlin 562 }
1347     }
1348    
1349 dpavlin 785 my $hash;
1350    
1351     =head2 set
1352    
1353     set( key => 'value' );
1354    
1355     =cut
1356    
1357     sub set {
1358     my ($k,$v) = @_;
1359 dpavlin 810 warn "## set ( $k => ", dump($v), " )", $/ if ( $debug );
1360 dpavlin 785 $hash->{$k} = $v;
1361     };
1362    
1363     =head2 get
1364    
1365     get( 'key' );
1366    
1367     =cut
1368    
1369     sub get {
1370     my $k = shift || return;
1371     my $v = $hash->{$k};
1372 dpavlin 810 warn "## get $k = ", dump( $v ), $/ if ( $debug );
1373 dpavlin 785 return $v;
1374     }
1375    
1376 dpavlin 791 =head2 count
1377 dpavlin 785
1378 dpavlin 791 if ( count( @result ) == 1 ) {
1379     # do something if only 1 result is there
1380     }
1381    
1382     =cut
1383    
1384     sub count {
1385 dpavlin 810 warn "## count ",dump(@_),$/ if ( $debug );
1386 dpavlin 791 return @_ . '';
1387     }
1388    
1389 dpavlin 536 # END
1390     1;

  ViewVC Help
Powered by ViewVC 1.1.26