/[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!

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

Parent Directory | Revision Log | View Patch Patch

-revision 10 by dpavlin,
Sat Jul 16 20:35:30 2005 UTC
+revision 548 by dpavlin,
Thu Jun 29 23:29:02 2006 UTC
 Line 1
  package WebPAC::Normalize;
+ use Exporter 'import';
+ @EXPORT = qw/
+         _set_rec _set_lookup
+         _get_ds _clean_ds
+         tag search display
+         marc marc_indicators marc_repeatable_subfield
+         rec1 rec2 rec
+         regex prefix suffix surround
+         first lookup join_with
+ /;
  use warnings;
  use strict;
+ #use base qw/WebPAC::Common/;
+ use Data::Dumper;
+ use Encode qw/from_to/;
  =head1 NAME
- WebPAC::Normalize - normalisation of source file
+ WebPAC::Normalize - describe normalisaton rules using sets
  =head1 VERSION
- Version 0.01
+ Version 0.06
  =cut
- our $VERSION = '0.01';
+ our $VERSION = '0.06';
  =head1 SYNOPSIS
- This package contains code that could be helpful in implementing different
+ This module uses C<conf/normalize/*.pl> files to perform normalisation
- normalisation front-ends.
+ from input records using perl functions which are specialized for set
+ processing.
+ Sets are implemented as arrays, and normalisation file is valid perl, which
+ means that you check it's validity before running WebPAC using
+ C<perl -c normalize.pl>.
+ Normalisation can generate multiple output normalized data. For now, supported output
+ types (on the left side of definition) are: C<tag>, C<display>, C<search> and
+ C<marc>.
  =head1 FUNCTIONS
- =head2 none_yet
+ Functions which start with C<_> are private and used by WebPAC internally.
+ All other functions are available for use within normalisation rules.
+ =head2 data_structure
+ Return data structure
+   my $ds = WebPAC::Normalize::data_structure(
+         lookup => $lookup->lookup_hash,
+         row => $row,
+         rules => $normalize_pl_config,
+         marc_encoding => 'utf-8',
+   );
+ Options C<lookup>, C<row>, C<rules> and C<log> are mandatory while all
+ other are optional.
+ This function will B<die> if normalizastion can't be evaled.
+ Since this function isn't exported you have to call it with
+ C<WebPAC::Normalize::data_structure>.
+ =cut
+ sub data_structure {
+         my $arg = {@_};
+         die "need row argument" unless ($arg->{row});
+         die "need normalisation argument" unless ($arg->{rules});
+         no strict 'subs';
+         _set_lookup( $arg->{lookup} );
+         _set_rec( $arg->{row} );
+         _clean_ds( %{ $arg } );
+         eval "$arg->{rules}";
+         die "error evaling $arg->{rules}: $@\n" if ($@);
+         return _get_ds();
+ }
+ =head2 _set_rec
+ Set current record hash
+   _set_rec( $rec );
  =cut
- sub none_yet {
+ my $rec;
+ sub _set_rec {
+         $rec = shift or die "no record hash";
  }
- =head1 AUTHOR
+ =head2 _get_ds
+ Return hash formatted as data structure
+   my $ds = _get_ds();
+ =cut
+ my ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators);
+ sub _get_ds {
+         return $out;
+ }
+ =head2 _clean_ds
+ Clean data structure hash for next record
+   _clean_ds();
+ =cut
+ sub _clean_ds {
+         my $a = {@_};
+         ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = (undef);
+         $marc_encoding = $a->{marc_encoding};
+ }
+ =head2 _set_lookup
+ Set current lookup hash
+   _set_lookup( $lookup );
+ =cut
+ my $lookup;
+ sub _set_lookup {
+         $lookup = shift;
+ }
+ =head2 _get_marc_fields
+ Get all fields defined by calls to C<marc>
+         $marc->add_fields( WebPAC::Normalize:_get_marc_fields() );
- Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
- =head1 COPYRIGHT & LICENSE
- Copyright 2005 Dobrica Pavlinusic, All Rights Reserved.
+ We are using I<magic> which detect repeatable fields only from
+ sequence of field/subfield data generated by normalization.
+ Repeatable field is created if there is second occurence of same subfield or
+ if any of indicators are different. This is sane for most cases except for
+ non-repeatable fields with repeatable subfields.
+ You can change behaviour of that using C<marc_repeatable_subfield>.
+ =cut
+ sub _get_marc_fields {
+         my @m;
+         my $last;
+         foreach my $row (@{ $marc_record }) {
+                 if ($last &&
+                         $last->[0] eq $row->[0] &&              # check if field is same
+                         $last->[1] eq $row->[1] &&              # check for i1
+                         $last->[2] eq $row->[2] &&              # and for i2
+                                 ( $last->[3] ne $row->[3] ||                            # and subfield is different
+                                 $last->[3] eq $row->[3] &&                                      # or subfield is same,
+                                 $marc_repeatable_subfield->{ $row->[3] }        # but is repeatable
+                         )
+                 ) {
+                         push @$last, ( $row->[3] , $row->[4] );
+                         next;
+                 } elsif ($last) {
+                         push @m, $last;
+                 }
+                 $last = $row;
+         }
+         push @m, $last if ($last);
+         return @m;
+ }
+ =head1 Functions to create C<data_structure>
+ Those functions generally have to first in your normalization file.
+ =head2 tag
+ Define new tag for I<search> and I<display>.
+   tag('Title', rec('200','a') );
+ =cut
+ sub tag {
+         my $name = shift or die "tag needs name as first argument";
+         my @o = grep { defined($_) && $_ ne '' } @_;
+         return unless (@o);
+         $out->{$name}->{tag} = $name;
+         $out->{$name}->{search} = \@o;
+         $out->{$name}->{display} = \@o;
+ }
+ =head2 display
+ Define tag just for I<display>
+   @v = display('Title', rec('200','a') );
+ =cut
+ sub display {
+         my $name = shift or die "display needs name as first argument";
+         my @o = grep { defined($_) && $_ ne '' } @_;
+         return unless (@o);
+         $out->{$name}->{tag} = $name;
+         $out->{$name}->{display} = \@o;
+ }
+ =head2 search
+ Prepare values just for I<search>
+   @v = search('Title', rec('200','a') );
+ =cut
+ sub search {
+         my $name = shift or die "search needs name as first argument";
+         my @o = grep { defined($_) && $_ ne '' } @_;
+         return unless (@o);
+         $out->{$name}->{tag} = $name;
+         $out->{$name}->{search} = \@o;
+ }
+ =head2 marc
+ Save value for MARC field
+   marc('900','a', rec('200','a') );
+ =cut
+ sub marc {
+         my $f = shift or die "marc needs field";
+         die "marc field must be numer" unless ($f =~ /^\d+$/);
+         my $sf = shift or die "marc needs subfield";
+         foreach (@_) {
+                 my $v = $_;             # make var read-write for Encode
+                 next unless (defined($v) && $v !~ /^\s*$/);
+                 from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding);
+                 my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' ');
+                 push @{ $marc_record }, [ $f, $i1, $i2, $sf => $v ];
+         }
+ }
+ =head2 marc_repeatable_subfield
+ Save values for MARC repetable subfield
- This program is free software; you can redistribute it and/or modify it
+   marc_repeatable_subfield('910', 'z', rec('909') );
- under the same terms as Perl itself.
  =cut
-; # End of WebPAC::DB
+ sub marc_repeatable_subfield {
+         die "marc_repeatable_subfield need subfield!\n" unless (defined($_[1]));
+         $marc_repeatable_subfield->{ $_[1] }++;
+         marc(@_);
+ }
+ =head2 marc_indicators
+ Set both indicators for MARC field
+   marc_indicators('900', ' ', 1);
+ Any indicator value other than C<0-9> will be treated as undefined.
+ =cut
+ sub marc_indicators {
+         my $f = shift || die "marc_indicators need field!\n";
+         my ($i1,$i2) = @_;
+         die "marc_indicators($f, ...) need i1!\n" unless(defined($i1));
+         die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2));
+         $i1 = ' ' if ($i1 !~ /^\d$/);
+         $i2 = ' ' if ($i2 !~ /^\d$/);
+         @{ $marc_indicators->{$f} } = ($i1,$i2);
+ }
+ =head1 Functions to extract data from input
+ This function should be used inside functions to create C<data_structure> described
+ above.
+ =head2 rec1
+ Return all values in some field
+   @v = rec1('200')
+ TODO: order of values is probably same as in source data, need to investigate that
+ =cut
+ sub rec1 {
+         my $f = shift;
+         return unless (defined($rec) && defined($rec->{$f}));
+         if (ref($rec->{$f}) eq 'ARRAY') {
+                 return map {
+                         if (ref($_) eq 'HASH') {
+                                 values %{$_};
+                         } else {
+                                 $_;
+                         }
+                 } @{ $rec->{$f} };
+         } elsif( defined($rec->{$f}) ) {
+                 return $rec->{$f};
+         }
+ }
+ =head2 rec2
+ Return all values in specific field and subfield
+   @v = rec2('200','a')
+ =cut
+ sub rec2 {
+         my $f = shift;
+         return unless (defined($rec && $rec->{$f}));
+         my $sf = shift;
+         return map { $_->{$sf} } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} };
+ }
+ =head2 rec
+ syntaxtic sugar for
+   @v = rec('200')
+   @v = rec('200','a')
+ =cut
+ sub rec {
+         if ($#_ == 0) {
+                 return rec1(@_);
+         } elsif ($#_ == 1) {
+                 return rec2(@_);
+         }
+ }
+ =head2 regex
+ Apply regex to some or all values
+   @v = regex( 's/foo/bar/g', @v );
+ =cut
+ sub regex {
+         my $r = shift;
+         my @out;
+         #warn "r: $r\n",Dumper(\@_);
+         foreach my $t (@_) {
+                 next unless ($t);
+                 eval "\$t =~ $r";
+                 push @out, $t if ($t && $t ne '');
+         }
+         return @out;
+ }
+ =head2 prefix
+ Prefix all values with a string
+   @v = prefix( 'my_', @v );
+ =cut
+ sub prefix {
+         my $p = shift or die "prefix needs string as first argument";
+         return map { $p . $_ } grep { defined($_) } @_;
+ }
+ =head2 suffix
+ suffix all values with a string
+   @v = suffix( '_my', @v );
+ =cut
+ sub suffix {
+         my $s = shift or die "suffix needs string as first argument";
+         return map { $_ . $s } grep { defined($_) } @_;
+ }
+ =head2 surround
+ surround all values with a two strings
+   @v = surround( 'prefix_', '_suffix', @v );
+ =cut
+ sub surround {
+         my $p = shift or die "surround need prefix as first argument";
+         my $s = shift or die "surround needs suffix as second argument";
+         return map { $p . $_ . $s } grep { defined($_) } @_;
+ }
+ =head2 first
+ Return first element
+   $v = first( @v );
+ =cut
+ sub first {
+         my $r = shift;
+         return $r;
+ }
+ =head2 lookup
+ Consult lookup hashes for some value
+   @v = lookup( $v );
+   @v = lookup( @v );
+ =cut
+ sub lookup {
+         my $k = shift or return;
+         return unless (defined($lookup->{$k}));
+         if (ref($lookup->{$k}) eq 'ARRAY') {
+                 return @{ $lookup->{$k} };
+         } else {
+                 return $lookup->{$k};
+         }
+ }
+ =head2 join_with
+ Joins walues with some delimiter
+   $v = join_with(", ", @v);
+ =cut
+ sub join_with {
+         my $d = shift;
+         return join($d, grep { defined($_) && $_ ne '' } @_);
+ }
+ # END
+;

 Legend:



Removed from v.10
 


changed lines


 
Added in v.548
 Legend:



Removed from v.10
 


changed lines


 
Added in v.548
-Removed from v.10
+Added in v.548

	ViewVC Help
Powered by ViewVC 1.1.26