/[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 604 - (hide annotations)
Sun Jul 30 14:19:54 2006 UTC (16 years, 6 months ago) by dpavlin
File size: 21244 byte(s)
 r844@llin:  dpavlin | 2006-07-30 14:24:04 +0200
 first cut at marc_original_order

1 dpavlin 10 package WebPAC::Normalize;
2 dpavlin 536 use Exporter 'import';
3     @EXPORT = qw/
4 dpavlin 538 _set_rec _set_lookup
5     _get_ds _clean_ds
6 dpavlin 554 _debug
7 dpavlin 538
8 dpavlin 536 tag search display
9 dpavlin 547 marc marc_indicators marc_repeatable_subfield
10 dpavlin 564 marc_compose marc_leader
11 dpavlin 574 marc_duplicate marc_remove
12 dpavlin 604 marc_original_order
13 dpavlin 540
14 dpavlin 536 rec1 rec2 rec
15     regex prefix suffix surround
16     first lookup join_with
17 dpavlin 562
18     split_rec_on
19 dpavlin 536 /;
20 dpavlin 10
21     use warnings;
22     use strict;
23 dpavlin 536
24     #use base qw/WebPAC::Common/;
25 dpavlin 550 use Data::Dump qw/dump/;
26 dpavlin 541 use Encode qw/from_to/;
27 dpavlin 574 use Storable qw/dclone/;
28 dpavlin 10
29 dpavlin 550 # debugging warn(s)
30     my $debug = 0;
31    
32    
33 dpavlin 10 =head1 NAME
34    
35 dpavlin 536 WebPAC::Normalize - describe normalisaton rules using sets
36 dpavlin 10
37     =head1 VERSION
38    
39 dpavlin 603 Version 0.15
40 dpavlin 10
41     =cut
42    
43 dpavlin 603 our $VERSION = '0.15';
44 dpavlin 10
45     =head1 SYNOPSIS
46    
47 dpavlin 536 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 dpavlin 10
51 dpavlin 536 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 dpavlin 15
55 dpavlin 536 Normalisation can generate multiple output normalized data. For now, supported output
56 dpavlin 540 types (on the left side of definition) are: C<tag>, C<display>, C<search> and
57 dpavlin 547 C<marc>.
58 dpavlin 15
59 dpavlin 10 =head1 FUNCTIONS
60    
61 dpavlin 538 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 dpavlin 536 =head2 data_structure
65 dpavlin 10
66 dpavlin 536 Return data structure
67 dpavlin 13
68 dpavlin 538 my $ds = WebPAC::Normalize::data_structure(
69 dpavlin 536 lookup => $lookup->lookup_hash,
70     row => $row,
71     rules => $normalize_pl_config,
72 dpavlin 541 marc_encoding => 'utf-8',
73 dpavlin 595 config => $config,
74 dpavlin 13 );
75    
76 dpavlin 540 Options C<lookup>, C<row>, C<rules> and C<log> are mandatory while all
77     other are optional.
78    
79 dpavlin 536 This function will B<die> if normalizastion can't be evaled.
80 dpavlin 15
81 dpavlin 538 Since this function isn't exported you have to call it with
82     C<WebPAC::Normalize::data_structure>.
83    
84 dpavlin 10 =cut
85    
86 dpavlin 536 sub data_structure {
87     my $arg = {@_};
88 dpavlin 13
89 dpavlin 536 die "need row argument" unless ($arg->{row});
90     die "need normalisation argument" unless ($arg->{rules});
91 dpavlin 31
92 dpavlin 536 no strict 'subs';
93 dpavlin 538 _set_lookup( $arg->{lookup} );
94     _set_rec( $arg->{row} );
95 dpavlin 595 _set_config( $arg->{config} );
96 dpavlin 541 _clean_ds( %{ $arg } );
97 dpavlin 536 eval "$arg->{rules}";
98     die "error evaling $arg->{rules}: $@\n" if ($@);
99 dpavlin 540
100 dpavlin 538 return _get_ds();
101 dpavlin 10 }
102    
103 dpavlin 538 =head2 _set_rec
104 dpavlin 13
105 dpavlin 536 Set current record hash
106 dpavlin 433
107 dpavlin 538 _set_rec( $rec );
108 dpavlin 433
109     =cut
110    
111 dpavlin 536 my $rec;
112 dpavlin 433
113 dpavlin 538 sub _set_rec {
114 dpavlin 536 $rec = shift or die "no record hash";
115 dpavlin 433 }
116    
117 dpavlin 595 =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 dpavlin 538 =head2 _get_ds
146    
147     Return hash formatted as data structure
148    
149     my $ds = _get_ds();
150    
151     =cut
152    
153 dpavlin 574 my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators);
154     my ($marc_record_offset, $marc_fetch_offset) = (0, 0);
155 dpavlin 538
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 dpavlin 541 my $a = {@_};
170 dpavlin 550 ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = ();
171 dpavlin 574 ($marc_record_offset, $marc_fetch_offset) = (0,0);
172 dpavlin 541 $marc_encoding = $a->{marc_encoding};
173 dpavlin 538 }
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 dpavlin 547 =head2 _get_marc_fields
190 dpavlin 540
191 dpavlin 547 Get all fields defined by calls to C<marc>
192 dpavlin 540
193 dpavlin 547 $marc->add_fields( WebPAC::Normalize:_get_marc_fields() );
194 dpavlin 540
195 dpavlin 543 We are using I<magic> which detect repeatable fields only from
196     sequence of field/subfield data generated by normalization.
197    
198 dpavlin 554 Repeatable field is created when there is second occurence of same subfield or
199     if any of indicators are different.
200 dpavlin 543
201 dpavlin 554 This is sane for most cases. Something like:
202 dpavlin 543
203 dpavlin 554 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 dpavlin 574 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 dpavlin 554
225 dpavlin 574 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 dpavlin 540 =cut
243    
244 dpavlin 547 sub _get_marc_fields {
245 dpavlin 550
246 dpavlin 574 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 dpavlin 550
255 dpavlin 574 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 dpavlin 550 # 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 dpavlin 572 $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '')
269 dpavlin 574 } @{ $marc_rec };
270 dpavlin 550
271 dpavlin 574 @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting
272 dpavlin 562
273 dpavlin 550 # output marc fields
274 dpavlin 542 my @m;
275 dpavlin 550
276     # count unique field-subfields (used for offset when walking to next subfield)
277     my $u;
278 dpavlin 572 map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record;
279 dpavlin 550
280     if ($debug) {
281 dpavlin 574 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 dpavlin 550 }
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 dpavlin 542 }
298    
299 dpavlin 550 # mark it visited
300     $visited->{$i}++;
301    
302 dpavlin 574 my $row = dclone( $sorted_marc_record[$i] );
303 dpavlin 550
304     # field and subfield which is key for
305     # marc_repeatable_subfield and u
306 dpavlin 572 my $fsf = $row->[0] . ( $row->[3] || '' );
307 dpavlin 550
308     if ($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 dpavlin 542 }
354    
355 dpavlin 550 if ($#$field >= 0) {
356     push @m, $field;
357     warn "## saved/3 ", dump( $field ),$/ if ($debug);
358     }
359 dpavlin 542
360 dpavlin 579 return \@m;
361 dpavlin 540 }
362    
363 dpavlin 554 =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 dpavlin 568 warn "debug level $l",$/ if ($l > 0);
375 dpavlin 554 $debug = $l;
376     }
377    
378 dpavlin 540 =head1 Functions to create C<data_structure>
379    
380     Those functions generally have to first in your normalization file.
381    
382 dpavlin 536 =head2 tag
383 dpavlin 433
384 dpavlin 536 Define new tag for I<search> and I<display>.
385 dpavlin 433
386 dpavlin 536 tag('Title', rec('200','a') );
387 dpavlin 13
388    
389     =cut
390    
391 dpavlin 536 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 dpavlin 13
400 dpavlin 536 =head2 display
401 dpavlin 13
402 dpavlin 536 Define tag just for I<display>
403 dpavlin 125
404 dpavlin 536 @v = display('Title', rec('200','a') );
405 dpavlin 125
406 dpavlin 536 =cut
407 dpavlin 125
408 dpavlin 536 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 dpavlin 13
416 dpavlin 536 =head2 search
417 dpavlin 13
418 dpavlin 536 Prepare values just for I<search>
419 dpavlin 13
420 dpavlin 536 @v = search('Title', rec('200','a') );
421 dpavlin 433
422 dpavlin 536 =cut
423 dpavlin 13
424 dpavlin 536 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 dpavlin 13 }
431    
432 dpavlin 564 =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 dpavlin 547 =head2 marc
452 dpavlin 540
453     Save value for MARC field
454    
455 dpavlin 547 marc('900','a', rec('200','a') );
456 dpavlin 571 marc('001', rec('000') );
457 dpavlin 540
458     =cut
459    
460 dpavlin 547 sub marc {
461     my $f = shift or die "marc needs field";
462     die "marc field must be numer" unless ($f =~ /^\d+$/);
463 dpavlin 540
464 dpavlin 571 my $sf;
465     if ($f >= 10) {
466     $sf = shift or die "marc needs subfield";
467     }
468 dpavlin 540
469 dpavlin 541 foreach (@_) {
470     my $v = $_; # make var read-write for Encode
471 dpavlin 543 next unless (defined($v) && $v !~ /^\s*$/);
472 dpavlin 541 from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding);
473 dpavlin 548 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
474 dpavlin 571 if (defined $sf) {
475 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ];
476 dpavlin 571 } else {
477 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
478 dpavlin 571 }
479 dpavlin 540 }
480     }
481    
482 dpavlin 547 =head2 marc_repeatable_subfield
483    
484     Save values for MARC repetable subfield
485    
486     marc_repeatable_subfield('910', 'z', rec('909') );
487    
488     =cut
489    
490     sub marc_repeatable_subfield {
491 dpavlin 550 my ($f,$sf) = @_;
492     die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf);
493     $marc_repeatable_subfield->{ $f . $sf }++;
494 dpavlin 547 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 dpavlin 548 @{ $marc_indicators->{$f} } = ($i1,$i2);
516 dpavlin 547 }
517    
518 dpavlin 562 =head2 marc_compose
519 dpavlin 547
520 dpavlin 562 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 dpavlin 603 If you specify C<+> for subfield, value will be appended
530     to previous defined subfield.
531    
532 dpavlin 562 =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, $i2 ];
540    
541 dpavlin 583 warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2);
542    
543 dpavlin 562 while (@_) {
544     my $sf = shift or die "marc_compose $f needs subfield";
545 dpavlin 565 my $v = shift;
546 dpavlin 562
547     next unless (defined($v) && $v !~ /^\s*$/);
548     from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding);
549 dpavlin 568 warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1);
550 dpavlin 603 if ($sf ne '+') {
551     push @$m, ( $sf, $v );
552     } else {
553     $m->[ $#$m ] .= $v;
554     }
555 dpavlin 562 }
556    
557 dpavlin 586 warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1);
558 dpavlin 562
559 dpavlin 574 push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2);
560 dpavlin 562 }
561    
562 dpavlin 574 =head2 marc_duplicate
563 dpavlin 562
564 dpavlin 574 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 dpavlin 579 my $i = 0;
609     foreach ( 0 .. $#{ $marc } ) {
610 dpavlin 574 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 dpavlin 579 $i++;
636 dpavlin 574 }
637    
638     warn "### marc_remove($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 dpavlin 604 =head2 marc_original_order
646    
647     Copy all subfields preserving original order to marc field.
648    
649     marc_original_order(210, 260);
650    
651     You might want to use this command if you are just renaming subfields or
652     using pre-processing modify_record in C<config.yml> and don't need any
653     post-processing or want to preserve order of original subfields.
654    
655     =cut
656    
657     sub marc_original_order {
658    
659     my ($from, $to) = @_;
660     die "marc_original_order needs from and to fields\n" unless ($from && $to);
661    
662     my $r = $rec->{$from} || return;
663     die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY');
664    
665     my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' ');
666     warn "## marc_original_order($from,$to) source = ", dump( $r ),$/ if ($debug > 1);
667    
668     foreach my $d (@$r) {
669    
670     my @sfs = @{ $d->{subfields} };
671    
672     die "field $from doesn't have subfields specification\n" unless(@sfs);
673     die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1);
674    
675     warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/;
676    
677     my $m = [ $to, $i1, $i2 ];
678    
679     while (my $sf = shift @sfs) {
680     warn "#--> sf: ",dump($sf), $/;
681     my $offset = shift @sfs;
682     die "corrupted sufields specification for field $from\n" unless defined($offset);
683    
684     my $v;
685     if (ref($d->{$sf}) eq 'ARRAY') {
686     $v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset]));
687     } elsif ($offset == 0) {
688     $v = $d->{$sf};
689     } else {
690     die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf});
691     }
692     push @$m, ( $sf, $v ) if (defined($v));
693     }
694    
695     if ($#{$m} > 2) {
696     push @{ $marc_record->[ $marc_record_offset ] }, $m;
697     }
698     }
699    
700     warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1);
701    
702     warn "# marc_original_order is partly implemented";
703     }
704    
705    
706 dpavlin 540 =head1 Functions to extract data from input
707    
708     This function should be used inside functions to create C<data_structure> described
709     above.
710    
711 dpavlin 536 =head2 rec1
712 dpavlin 371
713 dpavlin 536 Return all values in some field
714 dpavlin 371
715 dpavlin 536 @v = rec1('200')
716 dpavlin 15
717 dpavlin 536 TODO: order of values is probably same as in source data, need to investigate that
718 dpavlin 15
719 dpavlin 536 =cut
720 dpavlin 15
721 dpavlin 536 sub rec1 {
722     my $f = shift;
723 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
724 dpavlin 536 return unless (defined($rec) && defined($rec->{$f}));
725 dpavlin 571 warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
726 dpavlin 536 if (ref($rec->{$f}) eq 'ARRAY') {
727     return map {
728     if (ref($_) eq 'HASH') {
729     values %{$_};
730 dpavlin 31 } else {
731 dpavlin 536 $_;
732 dpavlin 31 }
733 dpavlin 536 } @{ $rec->{$f} };
734     } elsif( defined($rec->{$f}) ) {
735     return $rec->{$f};
736 dpavlin 15 }
737     }
738    
739 dpavlin 536 =head2 rec2
740 dpavlin 15
741 dpavlin 536 Return all values in specific field and subfield
742 dpavlin 13
743 dpavlin 536 @v = rec2('200','a')
744 dpavlin 13
745     =cut
746    
747 dpavlin 536 sub rec2 {
748     my $f = shift;
749     return unless (defined($rec && $rec->{$f}));
750     my $sf = shift;
751 dpavlin 601 warn "rec2($f,$sf) = ", dump( $rec->{$f} ), $/ if ($debug > 1);
752 dpavlin 589 return map {
753     if (ref($_->{$sf}) eq 'ARRAY') {
754     @{ $_->{$sf} };
755     } else {
756     $_->{$sf};
757     }
758     } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} };
759 dpavlin 536 }
760 dpavlin 13
761 dpavlin 536 =head2 rec
762 dpavlin 13
763 dpavlin 536 syntaxtic sugar for
764 dpavlin 13
765 dpavlin 536 @v = rec('200')
766     @v = rec('200','a')
767 dpavlin 13
768 dpavlin 536 =cut
769 dpavlin 373
770 dpavlin 536 sub rec {
771 dpavlin 583 my @out;
772 dpavlin 536 if ($#_ == 0) {
773 dpavlin 583 @out = rec1(@_);
774 dpavlin 536 } elsif ($#_ == 1) {
775 dpavlin 583 @out = rec2(@_);
776 dpavlin 13 }
777 dpavlin 583 if (@out) {
778     return @out;
779     } else {
780     return '';
781     }
782 dpavlin 13 }
783    
784 dpavlin 536 =head2 regex
785 dpavlin 15
786 dpavlin 536 Apply regex to some or all values
787 dpavlin 15
788 dpavlin 536 @v = regex( 's/foo/bar/g', @v );
789 dpavlin 15
790     =cut
791    
792 dpavlin 536 sub regex {
793     my $r = shift;
794     my @out;
795 dpavlin 550 #warn "r: $r\n", dump(\@_);
796 dpavlin 536 foreach my $t (@_) {
797     next unless ($t);
798     eval "\$t =~ $r";
799     push @out, $t if ($t && $t ne '');
800 dpavlin 368 }
801 dpavlin 536 return @out;
802 dpavlin 15 }
803    
804 dpavlin 536 =head2 prefix
805 dpavlin 15
806 dpavlin 536 Prefix all values with a string
807 dpavlin 15
808 dpavlin 536 @v = prefix( 'my_', @v );
809 dpavlin 15
810     =cut
811    
812 dpavlin 536 sub prefix {
813 dpavlin 592 my $p = shift or return;
814 dpavlin 536 return map { $p . $_ } grep { defined($_) } @_;
815     }
816 dpavlin 15
817 dpavlin 536 =head2 suffix
818 dpavlin 15
819 dpavlin 536 suffix all values with a string
820 dpavlin 15
821 dpavlin 536 @v = suffix( '_my', @v );
822 dpavlin 15
823 dpavlin 536 =cut
824 dpavlin 15
825 dpavlin 536 sub suffix {
826     my $s = shift or die "suffix needs string as first argument";
827     return map { $_ . $s } grep { defined($_) } @_;
828 dpavlin 15 }
829    
830 dpavlin 536 =head2 surround
831 dpavlin 13
832 dpavlin 536 surround all values with a two strings
833 dpavlin 13
834 dpavlin 536 @v = surround( 'prefix_', '_suffix', @v );
835 dpavlin 13
836     =cut
837    
838 dpavlin 536 sub surround {
839     my $p = shift or die "surround need prefix as first argument";
840     my $s = shift or die "surround needs suffix as second argument";
841     return map { $p . $_ . $s } grep { defined($_) } @_;
842 dpavlin 13 }
843    
844 dpavlin 536 =head2 first
845 dpavlin 13
846 dpavlin 536 Return first element
847 dpavlin 15
848 dpavlin 536 $v = first( @v );
849 dpavlin 13
850     =cut
851    
852 dpavlin 536 sub first {
853     my $r = shift;
854     return $r;
855 dpavlin 13 }
856    
857 dpavlin 536 =head2 lookup
858 dpavlin 13
859 dpavlin 536 Consult lookup hashes for some value
860 dpavlin 13
861 dpavlin 536 @v = lookup( $v );
862     @v = lookup( @v );
863 dpavlin 13
864     =cut
865    
866 dpavlin 536 sub lookup {
867     my $k = shift or return;
868     return unless (defined($lookup->{$k}));
869     if (ref($lookup->{$k}) eq 'ARRAY') {
870     return @{ $lookup->{$k} };
871     } else {
872     return $lookup->{$k};
873     }
874 dpavlin 13 }
875    
876 dpavlin 595 =head2 config
877    
878     Consult config values stored in C<config.yml>
879    
880     # return database code (key under databases in yaml)
881     $database_code = config(); # use _ from hash
882     $database_name = config('name');
883     $database_input_name = config('input name');
884     $tag = config('input normalize tag');
885    
886     Up to three levels are supported.
887    
888     =cut
889    
890     sub config {
891     return unless ($config);
892    
893     my $p = shift;
894    
895     $p ||= '';
896    
897     my $v;
898    
899     warn "### getting config($p)\n" if ($debug > 1);
900    
901     my @p = split(/\s+/,$p);
902     if ($#p < 0) {
903     $v = $config->{ '_' }; # special, database code
904     } else {
905    
906     my $c = dclone( $config );
907    
908     foreach my $k (@p) {
909     warn "### k: $k c = ",dump($c),$/ if ($debug > 1);
910     if (ref($c) eq 'ARRAY') {
911     $c = shift @$c;
912     warn "config($p) taking first occurence of '$k', probably not what you wanted!\n";
913     last;
914     }
915    
916     if (! defined($c->{$k}) ) {
917     $c = undef;
918     last;
919     } else {
920     $c = $c->{$k};
921     }
922     }
923     $v = $c if ($c);
924    
925     }
926    
927     warn "## config( '$p' ) = ",dump( $v ),$/ if ($v && $debug);
928     warn "config( '$p' ) is empty\n" if (! $v);
929    
930     return $v;
931     }
932    
933     =head2 id
934    
935     Returns unique id of this record
936    
937     $id = id();
938    
939     Returns C<42/2> for 2nd occurence of MFN 42.
940    
941     =cut
942    
943     sub id {
944     my $mfn = $config->{_mfn} || die "no _mfn in config data";
945     return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : '';
946     }
947    
948 dpavlin 536 =head2 join_with
949 dpavlin 13
950 dpavlin 536 Joins walues with some delimiter
951 dpavlin 10
952 dpavlin 536 $v = join_with(", ", @v);
953 dpavlin 10
954 dpavlin 536 =cut
955 dpavlin 10
956 dpavlin 536 sub join_with {
957     my $d = shift;
958 dpavlin 586 warn "### join_with('$d',",dump(@_),")\n" if ($debug > 2);
959 dpavlin 583 my $v = join($d, grep { defined($_) && $_ ne '' } @_);
960     return '' unless defined($v);
961     return $v;
962 dpavlin 536 }
963 dpavlin 10
964 dpavlin 562 =head2 split_rec_on
965    
966     Split record subfield on some regex and take one of parts out
967    
968     $a_before_semi_column =
969     split_rec_on('200','a', /\s*;\s*/, $part);
970    
971     C<$part> is optional number of element. First element is
972     B<1>, not 0!
973    
974     If there is no C<$part> parameter or C<$part> is 0, this function will
975     return all values produced by splitting.
976    
977     =cut
978    
979     sub split_rec_on {
980     die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2);
981    
982     my ($fld, $sf, $regex, $part) = @_;
983 dpavlin 568 warn "### regex ", ref($regex), $regex, $/ if ($debug > 2);
984 dpavlin 562
985     my @r = rec( $fld, $sf );
986     my $v = shift @r;
987 dpavlin 568 warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2);
988 dpavlin 562
989 dpavlin 604 return '' if ( ! defined($v) || $v =~ /^\s*$/);
990 dpavlin 566
991 dpavlin 562 my @s = split( $regex, $v );
992 dpavlin 568 warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1);
993 dpavlin 566 if ($part && $part > 0) {
994 dpavlin 562 return $s[ $part - 1 ];
995     } else {
996 dpavlin 571 return @s;
997 dpavlin 562 }
998     }
999    
1000 dpavlin 536 # END
1001     1;

  ViewVC Help
Powered by ViewVC 1.1.26