package WebPAC::Normalize;
use Exporter 'import';
-@EXPORT = qw/
+our @EXPORT = qw/
_set_rec _set_lookup
_set_load_row
_get_ds _clean_ds
_debug
_pack_subfields_hash
- tag search display
+ search_display search display sorted
+
marc marc_indicators marc_repeatable_subfield
- marc_compose marc_leader
- marc_duplicate marc_remove
+ marc_compose marc_leader marc_fixed
+ marc_duplicate marc_remove marc_count
marc_original_order
rec1 rec2 rec
+ frec
regex prefix suffix surround
first lookup join_with
save_into_lookup
split_rec_on
+
+ get set
+ count
+
/;
use warnings;
# debugging warn(s)
my $debug = 0;
+use WebPAC::Normalize::ISBN;
+push @EXPORT, ( 'isbn_10', 'isbn_13' );
=head1 NAME
WebPAC::Normalize - describe normalisaton rules using sets
-=head1 VERSION
-
-Version 0.23
-
=cut
-our $VERSION = '0.23';
+our $VERSION = '0.31';
=head1 SYNOPSIS
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
+types (on the left side of definition) are: C<search_display>, C<display>, C<search> and
C<marc>.
=head1 FUNCTIONS
marc_encoding => 'utf-8',
config => $config,
load_row_coderef => sub {
- my ($database,$input,$mfn) = shift;
+ my ($database,$input,$mfn) = @_;
$store->load_row( database => $database, input => $input, id => $mfn );
},
);
=cut
-my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $leader);
+my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader);
my ($marc_record_offset, $marc_fetch_offset) = (0, 0);
sub _get_ds {
+#warn "## out = ",dump($out);
return $out;
}
sub _clean_ds {
my $a = {@_};
- ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $leader) = ();
+ ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = ();
($marc_record_offset, $marc_fetch_offset) = (0,0);
$marc_encoding = $a->{marc_encoding};
}
=cut
+my $fetch_pos;
+
sub _get_marc_fields {
my $arg = {@_};
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2);
- my $offset = $marc_fetch_offset;
+ $fetch_pos = $marc_fetch_offset;
if ($arg->{offset}) {
- $offset = $arg->{offset};
+ $fetch_pos = $arg->{offset};
} elsif($arg->{fetch_next}) {
$marc_fetch_offset++;
}
warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2);
- my $marc_rec = $marc_record->[ $offset ];
+ my $marc_rec = $marc_record->[ $fetch_pos ];
- warn "## _get_marc_fields (at offset: $offset) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1);
+ warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1);
return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0);
if ($debug) {
warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield );
- warn "## marc_record[$offset] = ", dump( $marc_rec ), $/;
+ warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/;
warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/;
warn "## subfield count = ", dump( $u ), $/;
}
return \@m;
}
+=head2 _get_marc_leader
+
+Return leader from currently fetched record by L</_get_marc_fields>
+
+ print WebPAC::Normalize::_get_marc_leader();
+
+=cut
+
+sub _get_marc_leader {
+ die "no fetch_pos, did you called _get_marc_fields first?" unless ( defined( $fetch_pos ) );
+ return $marc_leader->[ $fetch_pos ];
+}
+
=head2 _debug
Change level of debug warnings
Those functions generally have to first in your normalization file.
-=head2 tag
+=head2 search_display
-Define new tag for I<search> and I<display>.
+Define output for L<search> and L<display> at the same time
- tag('Title', rec('200','a') );
+ search_display('Title', rec('200','a') );
=cut
-sub tag {
- my $name = shift or die "tag needs name as first argument";
+sub search_display {
+ my $name = shift or die "search_display needs name as first argument";
my @o = grep { defined($_) && $_ ne '' } @_;
return unless (@o);
- $out->{$name}->{tag} = $name;
$out->{$name}->{search} = \@o;
$out->{$name}->{display} = \@o;
}
+=head2 tag
+
+Old name for L<search_display>, but supported
+
+=cut
+
+sub tag {
+ search_display( @_ );
+}
+
=head2 display
-Define tag just for I<display>
+Define output just for I<display>
@v = display('Title', rec('200','a') );
=cut
-sub display {
- my $name = shift or die "display needs name as first argument";
+sub _field {
+ my $type = shift or confess "need type -- BUG?";
+ my $name = shift or confess "needs name as first argument";
my @o = grep { defined($_) && $_ ne '' } @_;
return unless (@o);
- $out->{$name}->{tag} = $name;
- $out->{$name}->{display} = \@o;
+ $out->{$name}->{$type} = \@o;
}
+sub display { _field( 'display', @_ ) }
+
=head2 search
Prepare values just for I<search>
=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;
-}
+sub search { _field( 'search', @_ ) }
+
+=head2 sorted
+
+Insert into lists which will be automatically sorted
+
+ sorted('Title', rec('200','a') );
+
+=cut
+
+sub sorted { _field( 'sorted', @_ ) }
+
=head2 marc_leader
my ($offset,$value) = @_;
if ($offset) {
- $leader->{ $offset } = $value;
+ $marc_leader->[ $marc_record_offset ]->{ $offset } = $value;
} else {
- return $leader;
+
+ if (defined($marc_leader)) {
+ die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY');
+ return $marc_leader->[ $marc_record_offset ];
+ } else {
+ return;
+ }
+ }
+}
+
+=head2 marc_fixed
+
+Create control/indentifier fields with values in fixed positions
+
+ marc_fixed('008', 00, '070402');
+ marc_fixed('008', 39, '|');
+
+Positions not specified will be filled with spaces (C<0x20>).
+
+There will be no effort to extend last specified value to full length of
+field in standard.
+
+=cut
+
+sub marc_fixed {
+ my ($f, $pos, $val) = @_;
+ die "need marc(field, position, value)" unless defined($f) && defined($pos);
+
+ confess "need val" unless defined $val;
+
+ my $update = 0;
+
+ map {
+ if ($_->[0] eq $f) {
+ my $old = $_->[1];
+ if (length($old) <= $pos) {
+ $_->[1] .= ' ' x ( $pos - length($old) ) . $val;
+ warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1);
+ } else {
+ $_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val));
+ warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1);
+ }
+ $update++;
+ }
+ } @{ $marc_record->[ $marc_record_offset ] };
+
+ if (! $update) {
+ my $v = ' ' x $pos . $val;
+ push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ];
+ warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1);
}
}
my $m = $marc_record->[ -1 ];
die "can't duplicate record which isn't defined" unless ($m);
push @{ $marc_record }, dclone( $m );
- warn "## marc_duplicate = ", dump(@$marc_record), $/ if ($debug > 1);
+ push @{ $marc_leader }, dclone( marc_leader() );
+ warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1);
$marc_record_offset = $#{ $marc_record };
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1);
+
}
=head2 marc_remove
This will erase field C<200> or C<200^a> from current MARC record.
+ marc_remove('*');
+
+Will remove all fields in current MARC record.
+
This is useful after calling C<marc_duplicate> or on it's own (but, you
should probably just remove that subfield definition if you are not
using C<marc_duplicate>).
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2);
- my $i = 0;
- foreach ( 0 .. $#{ $marc } ) {
- last unless (defined $marc->[$i]);
- warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3);
- if ($marc->[$i]->[0] eq $f) {
- if (! defined $sf) {
- # remove whole field
- splice @$marc, $i, 1;
- warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3);
- $i--;
- } else {
- foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) {
- my $o = ($j * 2) + 3;
- if ($marc->[$i]->[$o] eq $sf) {
- # remove subfield
- splice @{$marc->[$i]}, $o, 2;
- warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3);
- # is record now empty?
- if ($#{ $marc->[$i] } == 2) {
- splice @$marc, $i, 1;
- warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3);
- $i--;
- };
+ if ($f eq '*') {
+
+ delete( $marc_record->[ $marc_record_offset ] );
+ warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
+
+ } else {
+
+ my $i = 0;
+ foreach ( 0 .. $#{ $marc } ) {
+ last unless (defined $marc->[$i]);
+ warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3);
+ if ($marc->[$i]->[0] eq $f) {
+ if (! defined $sf) {
+ # remove whole field
+ splice @$marc, $i, 1;
+ warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3);
+ $i--;
+ } else {
+ foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) {
+ my $o = ($j * 2) + 3;
+ if ($marc->[$i]->[$o] eq $sf) {
+ # remove subfield
+ splice @{$marc->[$i]}, $o, 2;
+ warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3);
+ # is record now empty?
+ if ($#{ $marc->[$i] } == 2) {
+ splice @$marc, $i, 1;
+ warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3);
+ $i--;
+ };
+ }
}
}
}
+ $i++;
}
- $i++;
- }
- warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2);
+ warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2);
- $marc_record->[ $marc_record_offset ] = $marc;
+ $marc_record->[ $marc_record_offset ] = $marc;
+ }
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1);
}
warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1);
}
+=head2 marc_count
+
+Return number of MARC records created using L</marc_duplicate>.
+
+ print "created ", marc_count(), " records";
+
+=cut
+
+sub marc_count {
+ return $#{ $marc_record };
+}
+
=head1 Functions to extract data from input
my ($h,$include_subfields) = @_;
+ # sanity and ease of use
+ return $h if (ref($h) ne 'HASH');
+
if ( defined($h->{subfields}) ) {
my $sfs = delete $h->{subfields} || die "no subfields?";
my @out;
=cut
+sub frec {
+ my @out = rec(@_);
+ warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0;
+ return shift @out;
+}
+
sub rec {
my @out;
if ($#_ == 0) {
=cut
sub prefix {
- my $p = shift or return;
+ my $p = shift;
+ return @_ unless defined( $p );
return map { $p . $_ } grep { defined($_) } @_;
}
=cut
sub suffix {
- my $s = shift or die "suffix needs string as first argument";
+ my $s = shift;
+ return @_ unless defined( $s );
return map { $_ . $s } grep { defined($_) } @_;
}
=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";
+ my $p = shift;
+ my $s = shift;
+ $p = '' unless defined( $p );
+ $s = '' unless defined( $s );
return map { $p . $_ . $s } grep { defined($_) } @_;
}
sub lookup {
my ($what, $database, $input, $key, $having) = @_;
- confess "lookup needs 5 arguments: what, database, input, key, having" unless ($#_ == 4);
+ confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4);
warn "## lookup ($database, $input, $key)", $/ if ($debug > 1);
return unless (defined($lookup->{$database}->{$input}->{$key}));
$database_code = config(); # use _ from hash
$database_name = config('name');
$database_input_name = config('input name');
- $tag = config('input normalize tag');
Up to three levels are supported.
}
}
+my $hash;
+
+=head2 set
+
+ set( key => 'value' );
+
+=cut
+
+sub set {
+ my ($k,$v) = @_;
+ warn "## set ( $k => ", dump($v), " )", $/ if ( $debug );
+ $hash->{$k} = $v;
+};
+
+=head2 get
+
+ get( 'key' );
+
+=cut
+
+sub get {
+ my $k = shift || return;
+ my $v = $hash->{$k};
+ warn "## get $k = ", dump( $v ), $/ if ( $debug );
+ return $v;
+}
+
+=head2 count
+
+ if ( count( @result ) == 1 ) {
+ # do something if only 1 result is there
+ }
+
+=cut
+
+sub count {
+ warn "## count ",dump(@_),$/ if ( $debug );
+ return @_ . '';
+}
+
# END
1;
projects / webpac2 / blobdiff