From cbcf1fdc302a7e5fdfac5ad67cd21e6c312457e5 Mon Sep 17 00:00:00 2001 From: Dobrica Pavlinusic Date: Mon, 12 Nov 2007 11:10:48 +0000 Subject: [PATCH] r1610@llin: dpavlin | 2007-11-12 12:10:45 +0100 split MARC handling routines into WebPAC::Normalize::MARC [2.31] git-svn-id: svn+ssh://mjesec/home/dpavlin/svn/webpac2/trunk@1036 07558da8-63fa-0310-ba24-9fe276d99e06 --- MANIFEST | 3 +- TODO | 1 + lib/WebPAC.pm | 6 +- lib/WebPAC/Normalize.pm | 587 +--------------------------------- lib/WebPAC/Normalize/MARC.pm | 601 ++++++++++++++++++++++++++++++++++- t/0-load.t | 5 +- t/3-normalize-marc.t | 82 +++-- t/3-normalize.t | 16 +- 8 files changed, 678 insertions(+), 623 deletions(-) diff --git a/MANIFEST b/MANIFEST index 4e617ef..4690c2e 100644 --- a/MANIFEST +++ b/MANIFEST @@ -9,10 +9,9 @@ lib/WebPAC/Store.pm lib/WebPAC/Input.pm lib/WebPAC/Input/ISIS.pm lib/WebPAC/Normalize.pm +lib/WebPAC/Normalize/MARC.pm lib/WebPAC/Output.pm -lib/WebPAC/Output/CDBI.pm lib/WebPAC/Output/Estraier.pm -lib/WebPAC/Output/OAI.pm lib/WebPAC/Output/TT.pm lib/WebPAC/Manual.pod t/0-load.t diff --git a/TODO b/TODO index 7d46a1a..8b6c073 100644 --- a/TODO +++ b/TODO @@ -44,6 +44,7 @@ + leader from WebPAC::Input::MARC is now available as rec('leader') + show i1 and i2 in statistcs specially (sorted first) + ISBN support in normalize ++ marc_template (WebPAC v1 in a command) [2.31] - rewrite WebPAC::Input to be based on Module::Pluggable - indicator(field,nr) nr=1|2 for MARC input - check usage of config in normalize file (database, input name?) diff --git a/lib/WebPAC.pm b/lib/WebPAC.pm index 4936402..3cb79a5 100644 --- a/lib/WebPAC.pm +++ b/lib/WebPAC.pm @@ -7,13 +7,9 @@ use strict; WebPAC - core module -=head1 VERSION - -Version 2.30 - =cut -our $VERSION = '2.30'; +our $VERSION = '2.31'; =head1 SYNOPSIS diff --git a/lib/WebPAC/Normalize.pm b/lib/WebPAC/Normalize.pm index 1d2eb65..084b3ea 100644 --- a/lib/WebPAC/Normalize.pm +++ b/lib/WebPAC/Normalize.pm @@ -9,12 +9,6 @@ our @EXPORT = qw/ search_display search display sorted - marc marc_indicators marc_repeatable_subfield - marc_compose marc_leader marc_fixed - marc_duplicate marc_remove marc_count - marc_original_order - marc_template - rec1 rec2 rec frec frec_eq frec_ne regex prefix suffix surround @@ -44,7 +38,13 @@ use WebPAC::Normalize::ISBN; push @EXPORT, ( 'isbn_10', 'isbn_13' ); use WebPAC::Normalize::MARC; -push @EXPORT, ( 'marc_template' ); +push @EXPORT, ( qw/ + marc marc_indicators marc_repeatable_subfield + marc_compose marc_leader marc_fixed + marc_duplicate marc_remove marc_count + marc_original_order + marc_template +/); =head1 NAME @@ -136,6 +136,7 @@ my $rec; sub _set_ds { $rec = shift or die "no record hash"; + $WebPAC::Normalize::MARC::rec = $rec; } =head2 @@ -182,8 +183,7 @@ Return hash formatted as data structure =cut -my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader); -my ($marc_record_offset, $marc_fetch_offset) = (0, 0); +my $out; sub _get_ds { #warn "## out = ",dump($out); @@ -200,9 +200,8 @@ Clean data structure hash for next record sub _clean_ds { my $a = {@_}; - ($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}; + $out = undef; + WebPAC::Normalize::MARC::_clean(); } =head2 _set_lookup @@ -250,195 +249,6 @@ sub _set_load_row { $load_row_coderef = $coderef; } -=head2 _get_marc_fields - -Get all fields defined by calls to C - - $marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); - -We are using I which detect repeatable fields only from -sequence of field/subfield data generated by normalization. - -Repeatable field is created when there is second occurence of same subfield or -if any of indicators are different. - -This is sane for most cases. Something like: - - 900a-1 900b-1 900c-1 - 900a-2 900b-2 - 900a-3 - -will be created from any combination of: - - 900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 - -and following rules: - - marc('900','a', rec('200','a') ); - marc('900','b', rec('200','b') ); - marc('900','c', rec('200','c') ); - -which might not be what you have in mind. If you need repeatable subfield, -define it using C like this: - - marc_repeatable_subfield('900','a'); - marc('900','a', rec('200','a') ); - marc('900','b', rec('200','b') ); - marc('900','c', rec('200','c') ); - -will create: - - 900a-1 900a-2 900a-3 900b-1 900c-1 - 900b-2 - -There is also support for returning next or specific using: - - while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) { - # do something with $mf - } - -will always return fields from next MARC record or - - my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 ); - -will return 42th copy record (if it exists). - -=cut - -my $fetch_pos; - -sub _get_marc_fields { - - my $arg = {@_}; - warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); - $fetch_pos = $marc_fetch_offset; - if ($arg->{offset}) { - $fetch_pos = $arg->{offset}; - } elsif($arg->{fetch_next}) { - $marc_fetch_offset++; - } - - return if (! $marc_record || ref($marc_record) ne 'ARRAY'); - - warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); - - my $marc_rec = $marc_record->[ $fetch_pos ]; - - 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); - - # first, sort all existing fields - # XXX might not be needed, but modern perl might randomize elements in hash - my @sorted_marc_record = sort { - $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '') - } @{ $marc_rec }; - - @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting - - # output marc fields - my @m; - - # count unique field-subfields (used for offset when walking to next subfield) - my $u; - map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record; - - if ($debug) { - warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); - warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/; - warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; - warn "## subfield count = ", dump( $u ), $/; - } - - my $len = $#sorted_marc_record; - my $visited; - my $i = 0; - my $field; - - foreach ( 0 .. $len ) { - - # find next element which isn't visited - while ($visited->{$i}) { - $i = ($i + 1) % ($len + 1); - } - - # mark it visited - $visited->{$i}++; - - my $row = dclone( $sorted_marc_record[$i] ); - - # field and subfield which is key for - # marc_repeatable_subfield and u - my $fsf = $row->[0] . ( $row->[3] || '' ); - - if ($debug > 1) { - - print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; - print "### this [$i]: ", dump( $row ),$/; - print "### sf: ", $row->[3], " vs ", $field->[3], - $marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, - if ($#$field >= 0); - - } - - # if field exists - if ( $#$field >= 0 ) { - if ( - $row->[0] ne $field->[0] || # field - $row->[1] ne $field->[1] || # i1 - $row->[2] ne $field->[2] # i2 - ) { - push @m, $field; - warn "## saved/1 ", dump( $field ),$/ if ($debug); - $field = $row; - - } elsif ( - ( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) - || - ( $row->[3] eq $field->[-2] && # same subfield, but not repeatable - ! $marc_repeatable_subfield->{ $fsf } - ) - ) { - push @m, $field; - warn "## saved/2 ", dump( $field ),$/ if ($debug); - $field = $row; - - } else { - # append new subfields to existing field - push @$field, ( $row->[3], $row->[4] ); - } - } else { - # insert first field - $field = $row; - } - - if (! $marc_repeatable_subfield->{ $fsf }) { - # make step to next subfield - $i = ($i + $u->{ $fsf } ) % ($len + 1); - } - } - - if ($#$field >= 0) { - push @m, $field; - warn "## saved/3 ", dump( $field ),$/ if ($debug); - } - - return \@m; -} - -=head2 _get_marc_leader - -Return leader from currently fetched record by L - - 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 @@ -452,6 +262,7 @@ sub _debug { return $debug unless defined($l); warn "debug level $l",$/ if ($l > 0); $debug = $l; + $WebPAC::Normalize::MARC::debug = $debug; } =head1 Functions to create C @@ -524,378 +335,6 @@ Insert into lists which will be automatically sorted sub sorted { _field( 'sorted', @_ ) } -=head2 marc_leader - -Setup fields within MARC leader or get leader - - marc_leader('05','c'); - my $leader = marc_leader(); - -=cut - -sub marc_leader { - my ($offset,$value) = @_; - - if ($offset) { - $marc_leader->[ $marc_record_offset ]->{ $offset } = $value; - } else { - - if (defined($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); - } -} - -=head2 marc - -Save value for MARC field - - marc('900','a', rec('200','a') ); - marc('001', rec('000') ); - -=cut - -sub marc { - my $f = shift or die "marc needs field"; - die "marc field must be numer" unless ($f =~ /^\d+$/); - - my $sf; - if ($f >= 10) { - $sf = shift or die "marc needs subfield"; - } - - foreach (@_) { - my $v = $_; # make var read-write for Encode - next unless (defined($v) && $v !~ /^\s*$/); - my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); - if (defined $sf) { - push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; - } else { - push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; - } - } -} - -=head2 marc_repeatable_subfield - -Save values for MARC repetable subfield - - marc_repeatable_subfield('910', 'z', rec('909') ); - -=cut - -sub marc_repeatable_subfield { - my ($f,$sf) = @_; - die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); - $marc_repeatable_subfield->{ $f . $sf }++; - 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); -} - -=head2 marc_compose - -Save values for each MARC subfield explicitly - - marc_compose('900', - 'a', rec('200','a') - 'b', rec('201','a') - 'a', rec('200','b') - 'c', rec('200','c') - ); - -If you specify C<+> for subfield, value will be appended -to previous defined subfield. - -=cut - -sub marc_compose { - my $f = shift or die "marc_compose needs field"; - die "marc_compose field must be numer" unless ($f =~ /^\d+$/); - - my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); - my $m = [ $f, $i1, $i2 ]; - - warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); - - if ($#_ % 2 != 1) { - die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n"; - } - - while (@_) { - my $sf = shift; - my $v = shift; - - next unless (defined($v) && $v !~ /^\s*$/); - warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); - if ($sf ne '+') { - push @$m, ( $sf, $v ); - } else { - $m->[ $#$m ] .= $v; - } - } - - warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); - - push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2); -} - -=head2 marc_duplicate - -Generate copy of current MARC record and continue working on copy - - marc_duplicate(); - -Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or -C<< _get_marc_fields( offset => 42 ) >>. - -=cut - -sub marc_duplicate { - my $m = $marc_record->[ -1 ]; - die "can't duplicate record which isn't defined" unless ($m); - push @{ $marc_record }, dclone( $m ); - 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 - -Remove some field or subfield from MARC record. - - marc_remove('200'); - marc_remove('200','a'); - -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 or on it's own (but, you -should probably just remove that subfield definition if you are not -using C). - -FIXME: support fields < 10. - -=cut - -sub marc_remove { - my ($f, $sf) = @_; - - die "marc_remove needs record number" unless defined($f); - - my $marc = $marc_record->[ $marc_record_offset ]; - - warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); - - 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--; - }; - } - } - } - } - $i++; - } - - warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); - - $marc_record->[ $marc_record_offset ] = $marc; - } - - warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); -} - -=head2 marc_original_order - -Copy all subfields preserving original order to marc field. - - marc_original_order( marc_field_number, original_input_field_number ); - -Please note that field numbers are consistent with other commands (marc -field number first), but somewhat counter-intuitive (destination and then -source). - -You might want to use this command if you are just renaming subfields or -using pre-processing modify_record in C and don't need any -post-processing or want to preserve order of original subfields. - - -=cut - -sub marc_original_order { - - my ($to, $from) = @_; - die "marc_original_order needs from and to fields\n" unless ($from && $to); - - return unless defined($rec->{$from}); - - my $r = $rec->{$from}; - die "record field $from isn't array ",dump( $rec ) unless (ref($r) eq 'ARRAY'); - - my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' '); - warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1); - - foreach my $d (@$r) { - - if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') { - warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n"; - next; - } - - my @sfs = @{ $d->{subfields} }; - - die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); - - warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2); - - my $m = [ $to, $i1, $i2 ]; - - while (my $sf = shift @sfs) { - - warn "#--> sf: ",dump($sf), $/ if ($debug > 2); - my $offset = shift @sfs; - die "corrupted sufields specification for field $from\n" unless defined($offset); - - my $v; - if (ref($d->{$sf}) eq 'ARRAY') { - $v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset])); - } elsif ($offset == 0) { - $v = $d->{$sf}; - } else { - die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf}); - } - push @$m, ( $sf, $v ) if (defined($v)); - } - - if ($#{$m} > 2) { - push @{ $marc_record->[ $marc_record_offset ] }, $m; - } - } - - warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); -} - - -=head2 marc_count - -Return number of MARC records created using L. - - print "created ", marc_count(), " records"; - -=cut - -sub marc_count { - return $#{ $marc_record }; -} - -=head2 _marc_push - - _marc_push( $marc ); - -=cut - -sub _marc_push { - my $marc = shift || die "no marc?"; - push @{ $marc_record->[ $marc_record_offset ] }, $marc; -} - =head1 Functions to extract data from input @@ -1384,7 +823,7 @@ Returns C<42/2> for 2nd occurence of MFN 42. sub id { my $mfn = $config->{_mfn} || die "no _mfn in config data"; - return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : ''; + return $mfn . ( WebPAC::Normalize::MARC::_created_marc_records() || '' ); } =head2 join_with diff --git a/lib/WebPAC/Normalize/MARC.pm b/lib/WebPAC/Normalize/MARC.pm index 5882a0a..c3d8dab 100644 --- a/lib/WebPAC/Normalize/MARC.pm +++ b/lib/WebPAC/Normalize/MARC.pm @@ -1,18 +1,28 @@ package WebPAC::Normalize::MARC; use Exporter 'import'; -@EXPORT = qw/ +our @EXPORT = qw/ + marc marc_indicators marc_repeatable_subfield + marc_compose marc_leader marc_fixed + marc_duplicate marc_remove marc_count + marc_original_order marc_template /; use strict; use warnings; +use Storable qw/dclone/; use Data::Dump qw/dump/; use Carp qw/confess/; use WebPAC::Normalize; -my $debug = 0; +our $debug = 42; + +my ($marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader); +my ($marc_record_offset, $marc_fetch_offset) = (0, 0); + +our $rec; =head1 NAME @@ -230,7 +240,7 @@ sub marc_template { foreach my $marc ( @marc_out ) { warn "+++ ",dump( $marc ),$/ if $debug; - WebPAC::Normalize::_marc_push( $marc ); + _marc_push( $marc ); $recs++; } @@ -239,4 +249,589 @@ sub marc_template { return $recs; } +=head2 marc_leader + +Setup fields within MARC leader or get leader + + marc_leader('05','c'); + my $leader = marc_leader(); + +=cut + +sub marc_leader { + my ($offset,$value) = @_; + + if ($offset) { + $marc_leader->[ $marc_record_offset ]->{ $offset } = $value; + } else { + + if (defined($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); + } +} + +=head2 marc + +Save value for MARC field + + marc('900','a', rec('200','a') ); + marc('001', rec('000') ); + +=cut + +sub marc { + my $f = shift or die "marc needs field"; + die "marc field must be numer" unless ($f =~ /^\d+$/); + + my $sf; + if ($f >= 10) { + $sf = shift or die "marc needs subfield"; + } + + foreach (@_) { + my $v = $_; # make var read-write for Encode + next unless (defined($v) && $v !~ /^\s*$/); + my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); + if (defined $sf) { + push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; + } else { + push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; + } + } +} + +=head2 marc_repeatable_subfield + +Save values for MARC repetable subfield + + marc_repeatable_subfield('910', 'z', rec('909') ); + +=cut + +sub marc_repeatable_subfield { + my ($f,$sf) = @_; + die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); + $marc_repeatable_subfield->{ $f . $sf }++; + 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); +} + +=head2 marc_compose + +Save values for each MARC subfield explicitly + + marc_compose('900', + 'a', rec('200','a') + 'b', rec('201','a') + 'a', rec('200','b') + 'c', rec('200','c') + ); + +If you specify C<+> for subfield, value will be appended +to previous defined subfield. + +=cut + +sub marc_compose { + my $f = shift or die "marc_compose needs field"; + die "marc_compose field must be numer" unless ($f =~ /^\d+$/); + + my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); + my $m = [ $f, $i1, $i2 ]; + + warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); + + if ($#_ % 2 != 1) { + die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n"; + } + + while (@_) { + my $sf = shift; + my $v = shift; + + next unless (defined($v) && $v !~ /^\s*$/); + warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); + if ($sf ne '+') { + push @$m, ( $sf, $v ); + } else { + $m->[ $#$m ] .= $v; + } + } + + warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); + + push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2); +} + +=head2 marc_duplicate + +Generate copy of current MARC record and continue working on copy + + marc_duplicate(); + +Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or +C<< _get_marc_fields( offset => 42 ) >>. + +=cut + +sub marc_duplicate { + my $m = $marc_record->[ -1 ]; + die "can't duplicate record which isn't defined" unless ($m); + push @{ $marc_record }, dclone( $m ); + 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 + +Remove some field or subfield from MARC record. + + marc_remove('200'); + marc_remove('200','a'); + +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 or on it's own (but, you +should probably just remove that subfield definition if you are not +using C). + +FIXME: support fields < 10. + +=cut + +sub marc_remove { + my ($f, $sf) = @_; + + die "marc_remove needs record number" unless defined($f); + + my $marc = $marc_record->[ $marc_record_offset ]; + + warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); + + 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--; + }; + } + } + } + } + $i++; + } + + warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); + + $marc_record->[ $marc_record_offset ] = $marc; + } + + warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); +} + +=head2 marc_original_order + +Copy all subfields preserving original order to marc field. + + marc_original_order( marc_field_number, original_input_field_number ); + +Please note that field numbers are consistent with other commands (marc +field number first), but somewhat counter-intuitive (destination and then +source). + +You might want to use this command if you are just renaming subfields or +using pre-processing modify_record in C and don't need any +post-processing or want to preserve order of original subfields. + + +=cut + +sub marc_original_order { + + my ($to, $from) = @_; + die "marc_original_order needs from and to fields\n" unless ($from && $to); + + return unless defined($rec->{$from}); + + my $r = $rec->{$from}; + die "record field $from isn't array ",dump( $rec ) unless (ref($r) eq 'ARRAY'); + + my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' '); + warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1); + + foreach my $d (@$r) { + + if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') { + warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n"; + next; + } + + my @sfs = @{ $d->{subfields} }; + + die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); + + warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2); + + my $m = [ $to, $i1, $i2 ]; + + while (my $sf = shift @sfs) { + + warn "#--> sf: ",dump($sf), $/ if ($debug > 2); + my $offset = shift @sfs; + die "corrupted sufields specification for field $from\n" unless defined($offset); + + my $v; + if (ref($d->{$sf}) eq 'ARRAY') { + $v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset])); + } elsif ($offset == 0) { + $v = $d->{$sf}; + } else { + die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf}); + } + push @$m, ( $sf, $v ) if (defined($v)); + } + + if ($#{$m} > 2) { + push @{ $marc_record->[ $marc_record_offset ] }, $m; + } + } + + warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); +} + + +=head2 marc_count + +Return number of MARC records created using L. + + print "created ", marc_count(), " records"; + +=cut + +sub marc_count { + return $#{ $marc_record }; +} + +=head2 _marc_push + + _marc_push( $marc ); + +=cut + +sub _marc_push { + my $marc = shift || die "no marc?"; + push @{ $marc_record->[ $marc_record_offset ] }, $marc; +} + +=head1 PRIVATE FUNCTIONS + +=head2 _clean + +Clean internal structures + +=cut + +sub _clean { + ($marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = (); + ($marc_record_offset, $marc_fetch_offset) = (0,0); +} + + +=head2 _get_marc_fields + +Get all fields defined by calls to C + + $marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); + +We are using I which detect repeatable fields only from +sequence of field/subfield data generated by normalization. + +Repeatable field is created when there is second occurence of same subfield or +if any of indicators are different. + +This is sane for most cases. Something like: + + 900a-1 900b-1 900c-1 + 900a-2 900b-2 + 900a-3 + +will be created from any combination of: + + 900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 + +and following rules: + + marc('900','a', rec('200','a') ); + marc('900','b', rec('200','b') ); + marc('900','c', rec('200','c') ); + +which might not be what you have in mind. If you need repeatable subfield, +define it using C like this: + + marc_repeatable_subfield('900','a'); + marc('900','a', rec('200','a') ); + marc('900','b', rec('200','b') ); + marc('900','c', rec('200','c') ); + +will create: + + 900a-1 900a-2 900a-3 900b-1 900c-1 + 900b-2 + +There is also support for returning next or specific using: + + while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) { + # do something with $mf + } + +will always return fields from next MARC record or + + my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 ); + +will return 42th copy record (if it exists). + +=cut + +my $fetch_pos; + +sub _get_marc_fields { + + my $arg = {@_}; + warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); + $fetch_pos = $marc_fetch_offset; + if ($arg->{offset}) { + $fetch_pos = $arg->{offset}; + } elsif($arg->{fetch_next}) { + $marc_fetch_offset++; + } + + return if (! $marc_record || ref($marc_record) ne 'ARRAY'); + + warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); + + my $marc_rec = $marc_record->[ $fetch_pos ]; + + 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); + + # first, sort all existing fields + # XXX might not be needed, but modern perl might randomize elements in hash + my @sorted_marc_record = sort { + $a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '') + } @{ $marc_rec }; + + @sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting + + # output marc fields + my @m; + + # count unique field-subfields (used for offset when walking to next subfield) + my $u; + map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record; + + if ($debug) { + warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); + warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/; + warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; + warn "## subfield count = ", dump( $u ), $/; + } + + my $len = $#sorted_marc_record; + my $visited; + my $i = 0; + my $field; + + foreach ( 0 .. $len ) { + + # find next element which isn't visited + while ($visited->{$i}) { + $i = ($i + 1) % ($len + 1); + } + + # mark it visited + $visited->{$i}++; + + my $row = dclone( $sorted_marc_record[$i] ); + + # field and subfield which is key for + # marc_repeatable_subfield and u + my $fsf = $row->[0] . ( $row->[3] || '' ); + + if ($debug > 1) { + + print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; + print "### this [$i]: ", dump( $row ),$/; + print "### sf: ", $row->[3], " vs ", $field->[3], + $marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, + if ($#$field >= 0); + + } + + # if field exists + if ( $#$field >= 0 ) { + if ( + $row->[0] ne $field->[0] || # field + $row->[1] ne $field->[1] || # i1 + $row->[2] ne $field->[2] # i2 + ) { + push @m, $field; + warn "## saved/1 ", dump( $field ),$/ if ($debug); + $field = $row; + + } elsif ( + ( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) + || + ( $row->[3] eq $field->[-2] && # same subfield, but not repeatable + ! $marc_repeatable_subfield->{ $fsf } + ) + ) { + push @m, $field; + warn "## saved/2 ", dump( $field ),$/ if ($debug); + $field = $row; + + } else { + # append new subfields to existing field + push @$field, ( $row->[3], $row->[4] ); + } + } else { + # insert first field + $field = $row; + } + + if (! $marc_repeatable_subfield->{ $fsf }) { + # make step to next subfield + $i = ($i + $u->{ $fsf } ) % ($len + 1); + } + } + + if ($#$field >= 0) { + push @m, $field; + warn "## saved/3 ", dump( $field ),$/ if ($debug); + } + + return \@m; +} + +=head2 _get_marc_leader + +Return leader from currently fetched record by L + + print WebPAC::Normalize::MARC::_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 _created_marc_records + + my $nr_records = _created_marc_records; + +=cut + +sub _created_marc_records { + return $#{ $marc_record } + 1 if $marc_record; +} + 1; diff --git a/t/0-load.t b/t/0-load.t index d6d43ee..3b01489 100644 --- a/t/0-load.t +++ b/t/0-load.t @@ -2,7 +2,7 @@ use strict; -use Test::More tests => 16; +use Test::More tests => 15; use blib; use lib 'lib'; @@ -14,11 +14,10 @@ use_ok( 'WebPAC::Input::ISIS' ); use_ok( 'WebPAC::Store' ); use_ok( 'WebPAC::Parser' ); use_ok( 'WebPAC::Normalize' ); +use_ok( 'WebPAC::Normalize::MARC' ); use_ok( 'WebPAC::Output' ); use_ok( 'WebPAC::Output::Estraier' ); use_ok( 'WebPAC::Output::TT' ); -use_ok( 'WebPAC::Output::OAI' ); -use_ok( 'WebPAC::Output::CDBI' ); use_ok( 'WebPAC::Output::MARC' ); use_ok( 'WebPAC::Output::JSON' ); use_ok( 'WebPAC::Output::KinoSearch' ); diff --git a/t/3-normalize-marc.t b/t/3-normalize-marc.t index a636b91..0b7ea6a 100755 --- a/t/3-normalize-marc.t +++ b/t/3-normalize-marc.t @@ -50,36 +50,42 @@ my $rec = { }], }; -ok( _set_ds( $rec ), '_set_ds' ); +sub test_marc_template { + my $rec = shift; -ok( marc_template( - from => 225, to => 440, - subfields_rename => [ - 'a' => 'a', - 'x' => 'x', - 'v' => 'v', - 'h' => 'n', - 'i' => 'p', - 'w' => 'v', - ], - isis_template => [ - 'a ; |v. |i', - 'a. |i ; |w', - ], - marc_template => [ - 'a', - 'a ;|v', - 'a.|p', - 'a, |x ; |v. |n, |p ; |v', - 'a ; |v. |p ; |v', - 'v', - ], -), 'marc_template' ); + ok( _set_ds( $rec ), '_set_ds' ); -ok(my $marc = WebPAC::Normalize::_get_marc_fields(), "_get_marc_fields"); -diag " _get_marc_fields = ",dump( $marc ) if $debug; + ok( marc_template( + from => 225, to => 440, + subfields_rename => [ + 'a' => 'a', + 'x' => 'x', + 'v' => 'v', + 'h' => 'n', + 'i' => 'p', + 'w' => 'v', + ], + isis_template => [ + 'a ; |v. |i', + 'a. |i ; |w', + ], + marc_template => [ + 'a', + 'a ;|v', + 'a.|p', + 'a, |x ; |v. |n, |p ; |v', + 'a ; |v. |p ; |v', + 'v', + ], + ), 'marc_template' ); -is_deeply( $marc, [ + ok(my $marc = WebPAC::Normalize::MARC::_get_marc_fields(), "_get_marc_fields"); + diag " _get_marc_fields = ",dump( $marc ) if $debug; + + return $marc; +} + +my $marc_out = [ [440, " ", " ", "a", "a-1-1, ", "x", "x-1-1 ; ", "v", "v-1-1. ", "n", "h-1-1, ", "p", "i-1-1 ; ", "v", "w-1-1"], [440, " ", " ", "v", "v-9-1"], [440, " ", " ", "a", "a-2-1 ; ", "v", "v-2-1. ", "p", "i-2-1"], @@ -89,4 +95,24 @@ is_deeply( $marc, [ [440, " ", " ", "a", "a-6-1. ", "p", "i-6-1 ; ", "v", "w-6-1"], [440, " ", " ", "a", "a-7-1"], [440, " ", " ", "a", "a-8-1 ;", "v", "v-8-1"], -], 'is_deeply'); +]; + +is_deeply( test_marc_template($rec), $marc_out , 'is_deeply'); + +__END__ + +my $max_occ = $#{ $rec->{225} }; + +foreach my $from ( 0 .. $max_occ - 1 ) { + foreach my $to ( 1 .. $max_occ ) { + my $new_rec = { + 225 => [ splice( @{ $rec->{225} }, $from, $to ) ], + }; + diag "new_rec = ",dump( $new_rec ) if $debug; + + my $expect_marc = [ splice( @{ $marc_out }, $from, $to ) ]; + diag "expect_marc = ",dump( $expect_marc ) if $debug; + + is_deeply( [ test_marc_template($rec) ], $expect_marc, 'is_deeply'); + } +} diff --git a/t/3-normalize.t b/t/3-normalize.t index 1327252..5163f73 100755 --- a/t/3-normalize.t +++ b/t/3-normalize.t @@ -485,7 +485,7 @@ sub test_s { test_s(qq{ marc_indicators('900',1,2) }); test_s(qq{ marc('900','a', rec('200') ) }); my $marc; - ok($marc = WebPAC::Normalize::_get_marc_fields(), "_get_marc_fields"); + ok($marc = WebPAC::Normalize::MARC::_get_marc_fields(), "_get_marc_fields"); diag dump( $marc ) if ($debug); is_deeply( $marc, [ @@ -496,7 +496,7 @@ sub test_s { test_s(qq{ marc_indicators('900',' ',9) }); test_s(qq{ marc_repeatable_subfield('900','a', rec('200') ) }); - ok($marc = WebPAC::Normalize::_get_marc_fields(), "_get_marc_fields"); + ok($marc = WebPAC::Normalize::MARC::_get_marc_fields(), "_get_marc_fields"); diag dump( $marc ) if ($debug); is_deeply( $marc, [ @@ -522,7 +522,7 @@ sub test_s { test_s($r) if ($r); } - ok(my $marc = WebPAC::Normalize::_get_marc_fields(), "_get_marc_fields"); + ok(my $marc = WebPAC::Normalize::MARC::_get_marc_fields(), "_get_marc_fields"); diag dump( $marc ) if $debug; diag "expects:\n", dump($struct) if ($debug > 1); is_deeply( $marc, $struct, $msg ); @@ -732,7 +732,7 @@ sub test_s { ); test_s(qq{ marc_remove('*'); }); - ok(! WebPAC::Normalize::_get_marc_fields(), 'marc_remove(*)'); + ok(! WebPAC::Normalize::MARC::_get_marc_fields(), 'marc_remove(*)'); test_rec_rules( 'marc_duplicate', @@ -766,7 +766,7 @@ sub test_s { my $i = 0; foreach my $v ( qw/bar baz bing bong/ ) { - ok($marc = WebPAC::Normalize::_get_marc_fields( offset => $i ), + ok($marc = WebPAC::Normalize::MARC::_get_marc_fields( offset => $i ), "_get_marc_fields( offset => $i )" ); diag "marc $i = ", dump( $marc ) if ($debug); @@ -774,7 +774,7 @@ sub test_s { [ [ '900', ' ', ' ', 'a', 42, 'b', $v ] ], "MARC copy $i has $v", ); - is_deeply(WebPAC::Normalize::_get_marc_leader(), { '06' => 42, 11 => $i }, "_get_marc_leader copy $i"); + is_deeply(WebPAC::Normalize::MARC::_get_marc_leader(), { '06' => 42, 11 => $i }, "_get_marc_leader copy $i"); $i++; } @@ -841,9 +841,9 @@ sub test_s { marc_fixed('000', 10, 'A'); marc_fixed('000', 0, '0'); }); - ok( my $m = WebPAC::Normalize::_get_marc_fields(), '_get_marc_fields'); + ok( my $m = WebPAC::Normalize::MARC::_get_marc_fields(), '_get_marc_fields'); diag dump( $m ); - is_deeply( WebPAC::Normalize::_get_marc_fields(), + is_deeply( WebPAC::Normalize::MARC::_get_marc_fields(), [ ["008", "abcdef"], # 0....5....10 -- 2.20.1