/[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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 618 - (show annotations)
Fri Aug 25 12:31:01 2006 UTC (16 years, 5 months ago) by dpavlin
File size: 21564 byte(s)
 r866@llin:  dpavlin | 2006-08-25 12:32:28 +0200
 finish marc_original_order and remove warning

1 package WebPAC::Normalize;
2 use Exporter 'import';
3 @EXPORT = qw/
4 _set_rec _set_lookup
5 _get_ds _clean_ds
6 _debug
7
8 tag search display
9 marc marc_indicators marc_repeatable_subfield
10 marc_compose marc_leader
11 marc_duplicate marc_remove
12 marc_original_order
13
14 rec1 rec2 rec
15 regex prefix suffix surround
16 first lookup join_with
17
18 split_rec_on
19 /;
20
21 use warnings;
22 use strict;
23
24 #use base qw/WebPAC::Common/;
25 use Data::Dump qw/dump/;
26 use Encode qw/from_to/;
27 use Storable qw/dclone/;
28
29 # debugging warn(s)
30 my $debug = 0;
31
32
33 =head1 NAME
34
35 WebPAC::Normalize - describe normalisaton rules using sets
36
37 =head1 VERSION
38
39 Version 0.16
40
41 =cut
42
43 our $VERSION = '0.16';
44
45 =head1 SYNOPSIS
46
47 This module uses C<conf/normalize/*.pl> files to perform normalisation
48 from input records using perl functions which are specialized for set
49 processing.
50
51 Sets are implemented as arrays, and normalisation file is valid perl, which
52 means that you check it's validity before running WebPAC using
53 C<perl -c normalize.pl>.
54
55 Normalisation can generate multiple output normalized data. For now, supported output
56 types (on the left side of definition) are: C<tag>, C<display>, C<search> and
57 C<marc>.
58
59 =head1 FUNCTIONS
60
61 Functions which start with C<_> are private and used by WebPAC internally.
62 All other functions are available for use within normalisation rules.
63
64 =head2 data_structure
65
66 Return data structure
67
68 my $ds = WebPAC::Normalize::data_structure(
69 lookup => $lookup->lookup_hash,
70 row => $row,
71 rules => $normalize_pl_config,
72 marc_encoding => 'utf-8',
73 config => $config,
74 );
75
76 Options C<lookup>, C<row>, C<rules> and C<log> are mandatory while all
77 other are optional.
78
79 This function will B<die> if normalizastion can't be evaled.
80
81 Since this function isn't exported you have to call it with
82 C<WebPAC::Normalize::data_structure>.
83
84 =cut
85
86 sub data_structure {
87 my $arg = {@_};
88
89 die "need row argument" unless ($arg->{row});
90 die "need normalisation argument" unless ($arg->{rules});
91
92 no strict 'subs';
93 _set_lookup( $arg->{lookup} );
94 _set_rec( $arg->{row} );
95 _set_config( $arg->{config} );
96 _clean_ds( %{ $arg } );
97 eval "$arg->{rules}";
98 die "error evaling $arg->{rules}: $@\n" if ($@);
99
100 return _get_ds();
101 }
102
103 =head2 _set_rec
104
105 Set current record hash
106
107 _set_rec( $rec );
108
109 =cut
110
111 my $rec;
112
113 sub _set_rec {
114 $rec = shift or die "no record hash";
115 }
116
117 =head2 _set_config
118
119 Set current config hash
120
121 _set_config( $config );
122
123 Magic keys are:
124
125 =over 4
126
127 =item _
128
129 Code of current database
130
131 =item _mfn
132
133 Current MFN
134
135 =back
136
137 =cut
138
139 my $config;
140
141 sub _set_config {
142 $config = shift;
143 }
144
145 =head2 _get_ds
146
147 Return hash formatted as data structure
148
149 my $ds = _get_ds();
150
151 =cut
152
153 my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators);
154 my ($marc_record_offset, $marc_fetch_offset) = (0, 0);
155
156 sub _get_ds {
157 return $out;
158 }
159
160 =head2 _clean_ds
161
162 Clean data structure hash for next record
163
164 _clean_ds();
165
166 =cut
167
168 sub _clean_ds {
169 my $a = {@_};
170 ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = ();
171 ($marc_record_offset, $marc_fetch_offset) = (0,0);
172 $marc_encoding = $a->{marc_encoding};
173 }
174
175 =head2 _set_lookup
176
177 Set current lookup hash
178
179 _set_lookup( $lookup );
180
181 =cut
182
183 my $lookup;
184
185 sub _set_lookup {
186 $lookup = shift;
187 }
188
189 =head2 _get_marc_fields
190
191 Get all fields defined by calls to C<marc>
192
193 $marc->add_fields( WebPAC::Normalize:_get_marc_fields() );
194
195 We are using I<magic> which detect repeatable fields only from
196 sequence of field/subfield data generated by normalization.
197
198 Repeatable field is created when there is second occurence of same subfield or
199 if any of indicators are different.
200
201 This is sane for most cases. Something like:
202
203 900a-1 900b-1 900c-1
204 900a-2 900b-2
205 900a-3
206
207 will be created from any combination of:
208
209 900a-1 900a-2 900a-3 900b-1 900b-2 900c-1
210
211 and following rules:
212
213 marc('900','a', rec('200','a') );
214 marc('900','b', rec('200','b') );
215 marc('900','c', rec('200','c') );
216
217 which might not be what you have in mind. If you need repeatable subfield,
218 define it using C<marc_repeatable_subfield> like this:
219
220 marc_repeatable_subfield('900','a');
221 marc('900','a', rec('200','a') );
222 marc('900','b', rec('200','b') );
223 marc('900','c', rec('200','c') );
224
225 will create:
226
227 900a-1 900a-2 900a-3 900b-1 900c-1
228 900b-2
229
230 There is also support for returning next or specific using:
231
232 while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) {
233 # do something with $mf
234 }
235
236 will always return fields from next MARC record or
237
238 my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 );
239
240 will return 42th copy record (if it exists).
241
242 =cut
243
244 sub _get_marc_fields {
245
246 my $arg = {@_};
247 warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2);
248 my $offset = $marc_fetch_offset;
249 if ($arg->{offset}) {
250 $offset = $arg->{offset};
251 } elsif($arg->{fetch_next}) {
252 $marc_fetch_offset++;
253 }
254
255 return if (! $marc_record || ref($marc_record) ne 'ARRAY');
256
257 warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2);
258
259 my $marc_rec = $marc_record->[ $offset ];
260
261 warn "## _get_marc_fields (at offset: $offset) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1);
262
263 return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0);
264
265 # first, sort all existing fields
266 # XXX might not be needed, but modern perl might randomize elements in hash
267 my @sorted_marc_record = sort {
268 $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '')
269 } @{ $marc_rec };
270
271 @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting
272
273 # output marc fields
274 my @m;
275
276 # count unique field-subfields (used for offset when walking to next subfield)
277 my $u;
278 map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record;
279
280 if ($debug) {
281 warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield );
282 warn "## marc_record[$offset] = ", dump( $marc_rec ), $/;
283 warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/;
284 warn "## subfield count = ", dump( $u ), $/;
285 }
286
287 my $len = $#sorted_marc_record;
288 my $visited;
289 my $i = 0;
290 my $field;
291
292 foreach ( 0 .. $len ) {
293
294 # find next element which isn't visited
295 while ($visited->{$i}) {
296 $i = ($i + 1) % ($len + 1);
297 }
298
299 # mark it visited
300 $visited->{$i}++;
301
302 my $row = dclone( $sorted_marc_record[$i] );
303
304 # field and subfield which is key for
305 # marc_repeatable_subfield and u
306 my $fsf = $row->[0] . ( $row->[3] || '' );
307
308 if ($debug > 1) {
309
310 print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/;
311 print "### this [$i]: ", dump( $row ),$/;
312 print "### sf: ", $row->[3], " vs ", $field->[3],
313 $marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/,
314 if ($#$field >= 0);
315
316 }
317
318 # if field exists
319 if ( $#$field >= 0 ) {
320 if (
321 $row->[0] ne $field->[0] || # field
322 $row->[1] ne $field->[1] || # i1
323 $row->[2] ne $field->[2] # i2
324 ) {
325 push @m, $field;
326 warn "## saved/1 ", dump( $field ),$/ if ($debug);
327 $field = $row;
328
329 } elsif (
330 ( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c)
331 ||
332 ( $row->[3] eq $field->[-2] && # same subfield, but not repeatable
333 ! $marc_repeatable_subfield->{ $fsf }
334 )
335 ) {
336 push @m, $field;
337 warn "## saved/2 ", dump( $field ),$/ if ($debug);
338 $field = $row;
339
340 } else {
341 # append new subfields to existing field
342 push @$field, ( $row->[3], $row->[4] );
343 }
344 } else {
345 # insert first field
346 $field = $row;
347 }
348
349 if (! $marc_repeatable_subfield->{ $fsf }) {
350 # make step to next subfield
351 $i = ($i + $u->{ $fsf } ) % ($len + 1);
352 }
353 }
354
355 if ($#$field >= 0) {
356 push @m, $field;
357 warn "## saved/3 ", dump( $field ),$/ if ($debug);
358 }
359
360 return \@m;
361 }
362
363 =head2 _debug
364
365 Change level of debug warnings
366
367 _debug( 2 );
368
369 =cut
370
371 sub _debug {
372 my $l = shift;
373 return $debug unless defined($l);
374 warn "debug level $l",$/ if ($l > 0);
375 $debug = $l;
376 }
377
378 =head1 Functions to create C<data_structure>
379
380 Those functions generally have to first in your normalization file.
381
382 =head2 tag
383
384 Define new tag for I<search> and I<display>.
385
386 tag('Title', rec('200','a') );
387
388
389 =cut
390
391 sub tag {
392 my $name = shift or die "tag needs name as first argument";
393 my @o = grep { defined($_) && $_ ne '' } @_;
394 return unless (@o);
395 $out->{$name}->{tag} = $name;
396 $out->{$name}->{search} = \@o;
397 $out->{$name}->{display} = \@o;
398 }
399
400 =head2 display
401
402 Define tag just for I<display>
403
404 @v = display('Title', rec('200','a') );
405
406 =cut
407
408 sub display {
409 my $name = shift or die "display needs name as first argument";
410 my @o = grep { defined($_) && $_ ne '' } @_;
411 return unless (@o);
412 $out->{$name}->{tag} = $name;
413 $out->{$name}->{display} = \@o;
414 }
415
416 =head2 search
417
418 Prepare values just for I<search>
419
420 @v = search('Title', rec('200','a') );
421
422 =cut
423
424 sub search {
425 my $name = shift or die "search needs name as first argument";
426 my @o = grep { defined($_) && $_ ne '' } @_;
427 return unless (@o);
428 $out->{$name}->{tag} = $name;
429 $out->{$name}->{search} = \@o;
430 }
431
432 =head2 marc_leader
433
434 Setup fields within MARC leader or get leader
435
436 marc_leader('05','c');
437 my $leader = marc_leader();
438
439 =cut
440
441 sub marc_leader {
442 my ($offset,$value) = @_;
443
444 if ($offset) {
445 $out->{' leader'}->{ $offset } = $value;
446 } else {
447 return $out->{' leader'};
448 }
449 }
450
451 =head2 marc
452
453 Save value for MARC field
454
455 marc('900','a', rec('200','a') );
456 marc('001', rec('000') );
457
458 =cut
459
460 sub marc {
461 my $f = shift or die "marc needs field";
462 die "marc field must be numer" unless ($f =~ /^\d+$/);
463
464 my $sf;
465 if ($f >= 10) {
466 $sf = shift or die "marc needs subfield";
467 }
468
469 foreach (@_) {
470 my $v = $_; # make var read-write for Encode
471 next unless (defined($v) && $v !~ /^\s*$/);
472 from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding);
473 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
474 if (defined $sf) {
475 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ];
476 } else {
477 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
478 }
479 }
480 }
481
482 =head2 marc_repeatable_subfield
483
484 Save values for MARC repetable subfield
485
486 marc_repeatable_subfield('910', 'z', rec('909') );
487
488 =cut
489
490 sub marc_repeatable_subfield {
491 my ($f,$sf) = @_;
492 die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf);
493 $marc_repeatable_subfield->{ $f . $sf }++;
494 marc(@_);
495 }
496
497 =head2 marc_indicators
498
499 Set both indicators for MARC field
500
501 marc_indicators('900', ' ', 1);
502
503 Any indicator value other than C<0-9> will be treated as undefined.
504
505 =cut
506
507 sub marc_indicators {
508 my $f = shift || die "marc_indicators need field!\n";
509 my ($i1,$i2) = @_;
510 die "marc_indicators($f, ...) need i1!\n" unless(defined($i1));
511 die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2));
512
513 $i1 = ' ' if ($i1 !~ /^\d$/);
514 $i2 = ' ' if ($i2 !~ /^\d$/);
515 @{ $marc_indicators->{$f} } = ($i1,$i2);
516 }
517
518 =head2 marc_compose
519
520 Save values for each MARC subfield explicitly
521
522 marc_compose('900',
523 'a', rec('200','a')
524 'b', rec('201','a')
525 'a', rec('200','b')
526 'c', rec('200','c')
527 );
528
529 If you specify C<+> for subfield, value will be appended
530 to previous defined subfield.
531
532 =cut
533
534 sub marc_compose {
535 my $f = shift or die "marc_compose needs field";
536 die "marc_compose field must be numer" unless ($f =~ /^\d+$/);
537
538 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
539 my $m = [ $f, $i1, $i2 ];
540
541 warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2);
542
543 while (@_) {
544 my $sf = shift or die "marc_compose $f needs subfield";
545 my $v = shift;
546
547 next unless (defined($v) && $v !~ /^\s*$/);
548 from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding);
549 warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1);
550 if ($sf ne '+') {
551 push @$m, ( $sf, $v );
552 } else {
553 $m->[ $#$m ] .= $v;
554 }
555 }
556
557 warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1);
558
559 push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2);
560 }
561
562 =head2 marc_duplicate
563
564 Generate copy of current MARC record and continue working on copy
565
566 marc_duplicate();
567
568 Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or
569 C<< _get_marc_fields( offset => 42 ) >>.
570
571 =cut
572
573 sub marc_duplicate {
574 my $m = $marc_record->[ -1 ];
575 die "can't duplicate record which isn't defined" unless ($m);
576 push @{ $marc_record }, dclone( $m );
577 warn "## marc_duplicate = ", dump(@$marc_record), $/ if ($debug > 1);
578 $marc_record_offset = $#{ $marc_record };
579 warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1);
580 }
581
582 =head2 marc_remove
583
584 Remove some field or subfield from MARC record.
585
586 marc_remove('200');
587 marc_remove('200','a');
588
589 This will erase field C<200> or C<200^a> from current MARC record.
590
591 This is useful after calling C<marc_duplicate> or on it's own (but, you
592 should probably just remove that subfield definition if you are not
593 using C<marc_duplicate>).
594
595 FIXME: support fields < 10.
596
597 =cut
598
599 sub marc_remove {
600 my ($f, $sf) = @_;
601
602 die "marc_remove needs record number" unless defined($f);
603
604 my $marc = $marc_record->[ $marc_record_offset ];
605
606 warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2);
607
608 my $i = 0;
609 foreach ( 0 .. $#{ $marc } ) {
610 last unless (defined $marc->[$i]);
611 warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3);
612 if ($marc->[$i]->[0] eq $f) {
613 if (! defined $sf) {
614 # remove whole field
615 splice @$marc, $i, 1;
616 warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3);
617 $i--;
618 } else {
619 foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) {
620 my $o = ($j * 2) + 3;
621 if ($marc->[$i]->[$o] eq $sf) {
622 # remove subfield
623 splice @{$marc->[$i]}, $o, 2;
624 warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3);
625 # is record now empty?
626 if ($#{ $marc->[$i] } == 2) {
627 splice @$marc, $i, 1;
628 warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3);
629 $i--;
630 };
631 }
632 }
633 }
634 }
635 $i++;
636 }
637
638 warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2);
639
640 $marc_record->[ $marc_record_offset ] = $marc;
641
642 warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
643 }
644
645 =head2 marc_original_order
646
647 Copy all subfields preserving original order to marc field.
648
649 marc_original_order( marc_field_number, original_input_field_number );
650
651 Please note that field numbers are consistent with other commands (marc
652 field number first), but somewhat counter-intuitive (destination and then
653 source).
654
655 You might want to use this command if you are just renaming subfields or
656 using pre-processing modify_record in C<config.yml> and don't need any
657 post-processing or want to preserve order of original subfields.
658
659
660 =cut
661
662 sub marc_original_order {
663
664 my ($to, $from) = @_;
665 die "marc_original_order needs from and to fields\n" unless ($from && $to);
666
667 return unless defined($rec->{$from});
668
669 my $r = $rec->{$from};
670 die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY');
671
672 my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' ');
673 warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1);
674
675 foreach my $d (@$r) {
676
677 if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') {
678 warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n";
679 next;
680 }
681
682 my @sfs = @{ $d->{subfields} };
683
684 die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1);
685
686 warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2);
687
688 my $m = [ $to, $i1, $i2 ];
689
690 while (my $sf = shift @sfs) {
691
692 warn "#--> sf: ",dump($sf), $/ if ($debug > 2);
693 my $offset = shift @sfs;
694 die "corrupted sufields specification for field $from\n" unless defined($offset);
695
696 my $v;
697 if (ref($d->{$sf}) eq 'ARRAY') {
698 $v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset]));
699 } elsif ($offset == 0) {
700 $v = $d->{$sf};
701 } else {
702 die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf});
703 }
704 push @$m, ( $sf, $v ) if (defined($v));
705 }
706
707 if ($#{$m} > 2) {
708 push @{ $marc_record->[ $marc_record_offset ] }, $m;
709 }
710 }
711
712 warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1);
713 }
714
715
716 =head1 Functions to extract data from input
717
718 This function should be used inside functions to create C<data_structure> described
719 above.
720
721 =head2 rec1
722
723 Return all values in some field
724
725 @v = rec1('200')
726
727 TODO: order of values is probably same as in source data, need to investigate that
728
729 =cut
730
731 sub rec1 {
732 my $f = shift;
733 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
734 return unless (defined($rec) && defined($rec->{$f}));
735 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
736 if (ref($rec->{$f}) eq 'ARRAY') {
737 return map {
738 if (ref($_) eq 'HASH') {
739 values %{$_};
740 } else {
741 $_;
742 }
743 } @{ $rec->{$f} };
744 } elsif( defined($rec->{$f}) ) {
745 return $rec->{$f};
746 }
747 }
748
749 =head2 rec2
750
751 Return all values in specific field and subfield
752
753 @v = rec2('200','a')
754
755 =cut
756
757 sub rec2 {
758 my $f = shift;
759 return unless (defined($rec && $rec->{$f}));
760 my $sf = shift;
761 warn "rec2($f,$sf) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
762 return map {
763 if (ref($_->{$sf}) eq 'ARRAY') {
764 @{ $_->{$sf} };
765 } else {
766 $_->{$sf};
767 }
768 } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} };
769 }
770
771 =head2 rec
772
773 syntaxtic sugar for
774
775 @v = rec('200')
776 @v = rec('200','a')
777
778 =cut
779
780 sub rec {
781 my @out;
782 if ($#_ == 0) {
783 @out = rec1(@_);
784 } elsif ($#_ == 1) {
785 @out = rec2(@_);
786 }
787 if (@out) {
788 return @out;
789 } else {
790 return '';
791 }
792 }
793
794 =head2 regex
795
796 Apply regex to some or all values
797
798 @v = regex( 's/foo/bar/g', @v );
799
800 =cut
801
802 sub regex {
803 my $r = shift;
804 my @out;
805 #warn "r: $r\n", dump(\@_);
806 foreach my $t (@_) {
807 next unless ($t);
808 eval "\$t =~ $r";
809 push @out, $t if ($t && $t ne '');
810 }
811 return @out;
812 }
813
814 =head2 prefix
815
816 Prefix all values with a string
817
818 @v = prefix( 'my_', @v );
819
820 =cut
821
822 sub prefix {
823 my $p = shift or return;
824 return map { $p . $_ } grep { defined($_) } @_;
825 }
826
827 =head2 suffix
828
829 suffix all values with a string
830
831 @v = suffix( '_my', @v );
832
833 =cut
834
835 sub suffix {
836 my $s = shift or die "suffix needs string as first argument";
837 return map { $_ . $s } grep { defined($_) } @_;
838 }
839
840 =head2 surround
841
842 surround all values with a two strings
843
844 @v = surround( 'prefix_', '_suffix', @v );
845
846 =cut
847
848 sub surround {
849 my $p = shift or die "surround need prefix as first argument";
850 my $s = shift or die "surround needs suffix as second argument";
851 return map { $p . $_ . $s } grep { defined($_) } @_;
852 }
853
854 =head2 first
855
856 Return first element
857
858 $v = first( @v );
859
860 =cut
861
862 sub first {
863 my $r = shift;
864 return $r;
865 }
866
867 =head2 lookup
868
869 Consult lookup hashes for some value
870
871 @v = lookup( $v );
872 @v = lookup( @v );
873
874 =cut
875
876 sub lookup {
877 my $k = shift or return;
878 return unless (defined($lookup->{$k}));
879 if (ref($lookup->{$k}) eq 'ARRAY') {
880 return @{ $lookup->{$k} };
881 } else {
882 return $lookup->{$k};
883 }
884 }
885
886 =head2 config
887
888 Consult config values stored in C<config.yml>
889
890 # return database code (key under databases in yaml)
891 $database_code = config(); # use _ from hash
892 $database_name = config('name');
893 $database_input_name = config('input name');
894 $tag = config('input normalize tag');
895
896 Up to three levels are supported.
897
898 =cut
899
900 sub config {
901 return unless ($config);
902
903 my $p = shift;
904
905 $p ||= '';
906
907 my $v;
908
909 warn "### getting config($p)\n" if ($debug > 1);
910
911 my @p = split(/\s+/,$p);
912 if ($#p < 0) {
913 $v = $config->{ '_' }; # special, database code
914 } else {
915
916 my $c = dclone( $config );
917
918 foreach my $k (@p) {
919 warn "### k: $k c = ",dump($c),$/ if ($debug > 1);
920 if (ref($c) eq 'ARRAY') {
921 $c = shift @$c;
922 warn "config($p) taking first occurence of '$k', probably not what you wanted!\n";
923 last;
924 }
925
926 if (! defined($c->{$k}) ) {
927 $c = undef;
928 last;
929 } else {
930 $c = $c->{$k};
931 }
932 }
933 $v = $c if ($c);
934
935 }
936
937 warn "## config( '$p' ) = ",dump( $v ),$/ if ($v && $debug);
938 warn "config( '$p' ) is empty\n" if (! $v);
939
940 return $v;
941 }
942
943 =head2 id
944
945 Returns unique id of this record
946
947 $id = id();
948
949 Returns C<42/2> for 2nd occurence of MFN 42.
950
951 =cut
952
953 sub id {
954 my $mfn = $config->{_mfn} || die "no _mfn in config data";
955 return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : '';
956 }
957
958 =head2 join_with
959
960 Joins walues with some delimiter
961
962 $v = join_with(", ", @v);
963
964 =cut
965
966 sub join_with {
967 my $d = shift;
968 warn "### join_with('$d',",dump(@_),")\n" if ($debug > 2);
969 my $v = join($d, grep { defined($_) && $_ ne '' } @_);
970 return '' unless defined($v);
971 return $v;
972 }
973
974 =head2 split_rec_on
975
976 Split record subfield on some regex and take one of parts out
977
978 $a_before_semi_column =
979 split_rec_on('200','a', /\s*;\s*/, $part);
980
981 C<$part> is optional number of element. First element is
982 B<1>, not 0!
983
984 If there is no C<$part> parameter or C<$part> is 0, this function will
985 return all values produced by splitting.
986
987 =cut
988
989 sub split_rec_on {
990 die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2);
991
992 my ($fld, $sf, $regex, $part) = @_;
993 warn "### regex ", ref($regex), $regex, $/ if ($debug > 2);
994
995 my @r = rec( $fld, $sf );
996 my $v = shift @r;
997 warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2);
998
999 return '' if ( ! defined($v) || $v =~ /^\s*$/);
1000
1001 my @s = split( $regex, $v );
1002 warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1);
1003 if ($part && $part > 0) {
1004 return $s[ $part - 1 ];
1005 } else {
1006 return @s;
1007 }
1008 }
1009
1010 # END
1011 1;

  ViewVC Help
Powered by ViewVC 1.1.26