use MARC::Batch;
use Unicode::Normalize;
use MARC::File::XML ( BinaryEncoding => 'utf-8' );
+use Equinox::Migration::SubfieldMapper;
my $conf = {}; # configuration hashref
my $count = 0; my $scount = 0;
$batch->strict_off();
$batch->warnings_off();
- while ( $record = $batch->next ) {
+ my $record;
+ while ( 1 ) {
+ eval { $record = $batch->next; };
+ if ($@) {
+ import MARC::File::XML;
+ print "bad record\n";
+ next;
+ }
+ last unless $record;
$count++; progress_ticker();
my $marc = undef;
unless ( defined $record )
# populate and normalize marc
$marc = populate_marc($record, $id);
+ # check for manual exclusion
+ next if this_record_is_excluded($record, $marc);
normalize_marc($marc);
unless (marc_isvalid($marc))
{ dump_exception($marc); next; }
}
unless ($marc{date1} and $marc{date1} =~ /\d{4}/) {
my $my_260 = $record->field('260');
- my $date1 = $my_260->subfield('c') if $my_260;
- if (defined $date1 and $date1 =~ /\d{4}/) {
- $marc{date1} = $date1;
- $marc{fudgedate} = 1;
- print XF ">> using 260c as date1 at rec $count\n";
+ if ($my_260 and $my_260->subfield('c')) {
+ my $date1 = $my_260->subfield('c');
+ $date1 =~ s/\D//g;
+ if (defined $date1 and $date1 =~ /\d{4}/) {
+ $marc{date1} = $date1;
+ $marc{fudgedate} = 1;
+ print XF ">> using 260c as date1 at rec $count\n";
+ }
}
}
if (defined $oclc and $oclc =~ /\(OCoLC\)/ and $oclc =~/([0-9]+)/);
}
+ if ($record->field('999')) {
+ my $koha_bib_id = $record->field('999')->subfield('c');
+ $marc{koha_bib_id} = $koha_bib_id if defined $koha_bib_id and $koha_bib_id =~ /^\d+$/;
+ }
+
# "Accompanying material" and check for "copy" (300)
if ($record->field('300')) {
$marc{accomp} = $record->field('300')->subfield('e');
#----------------------------------
# static criteria scoring
#----------------------------------
- $marc->{misc_score} = 999999999999;
- # subtract record id if we want older records to win
- $marc->{misc_score} -= $marc->{id} unless ($conf->{newwins});
- # handle arbitrary adjustments
- if ($conf->{'arbitrarily-decrease-score-above'}) {
- $marc->{misc_score} -= $conf->{'arbitrarily-decrease-score-by'}
- if ($marc->{id} >= $conf->{'arbitrarily-decrease-score-above'});
- }
- if ($conf->{'arbitrarily-decrease-score-below'}) {
- $marc->{misc_score} -= $conf->{'arbitrarily-decrease-score-by'}
- if ($marc->{id} <= $conf->{'arbitrarily-decrease-score-below'});
- }
+ $marc->{misc_score} = 999;
+ $marc->{age_score} = 999999999999;
+
# -1 if 008 has been padded, -2 if it doesn't exist
if ($marc->{tag008})
{ $marc->{misc_score}-- if ($marc->{tag008} =~ /\|$/) }
$marc->{misc_score}--
if (defined $marc->{tag300a} and $marc->{tag300a} =~ /copy/i);
+ # subtract record id if we want older records to win
+ #$marc->{age_score} -= $marc->{id} unless ($conf->{newwins});
+ # handle arbitrary adjustments
+ $marc->{age_score} = 1;
+ if ($conf->{'arbitrarily-lose-above'}) {
+ $marc->{age_score} = 0
+ if ($marc->{id} >= $conf->{'arbitrarily-lose-above'});
+ }
+ if ($conf->{'arbitrarily-lose-below'}) {
+ $marc->{age_score} = 0
+ if ($marc->{id} <= $conf->{'arbitrarily-lose-below'});
+ }
+
#----------------------------------
# dynamic calculated scoring
#----------------------------------
}
$json .= 'misc:' . $marc->{misc_score} . '}';
- my $compact = join('', $marc->{misc_score}, @score);
+ my $compact = join('', $marc->{age_score}, $marc->{misc_score}, @score);
$marc->{score} = "$compact\t$json";
}
}
}
+ if ($conf->{fingerprints}{koha_bib_id} and exists $marc->{koha_bib_id}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "z_koha_bib_id",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type},
+ $marc->{bib_lvl}, $marc->{title},
+ $marc->{koha_bib_id}), "\n";
+ }
+
if ($conf->{fingerprints}{isbn}) {
if ((scalar @{ $marc->{isbns} } > 0) and $marc->{pages}) {
foreach my $isbn ( @{ $marc->{isbns}} ) {
=cut
sub dump_exception {
- my ($marc) = @_;
+ my ($marc, $msg) = @_;
unless (defined $marc) {
print XF "Undefined record at line $count; likely bad XML\n";
return;
}
+
print XF "Record ", $marc->{id}, " excluded: ";
+ if (defined $msg) {
+ print XF "$msg\n";
+ return
+ }
+
print XF "missing item_form; " unless ($marc->{item_form});
unless (defined $marc->{date1})
{ print XF "missing date1; " }
}
+=head2 this_record_is_excluded
+
+Returns 1 if the record B<is> and 0 if the record B<is not> excluded,
+according to the subfield mapping (generated via the C<--excludelist>
+option).
+
+=cut
+
+sub this_record_is_excluded {
+ my ($rec, $marc) = @_;
+ return 0 unless defined $conf->{excludelist};
+
+ for my $tag (keys %{ $conf->{excludelist}->{tags} }) {
+ for my $sub (keys %{$conf->{excludelist}->{tags}{$tag}}) {
+ my $f = $conf->{excludelist}->field($tag, $sub);
+
+ # if this record doesn't have the right tag/sub, it can't be
+ return 0 unless ($rec->field($tag) and $rec->field($tag)->subfield($sub));
+ # but it does, so if there are no filters to check...
+ unless ($conf->{excludelist}->filters($f))
+ { dump_exception($marc, "exclusion $tag$sub"); return 1 }
+
+ my $sub_contents = $rec->field($tag)->subfield($sub);
+ for my $filter (@{ $conf->{excludelist}->filters($f)}) {
+ if ($sub_contents =~ /$filter/i) {
+ # filter matches. no fp.
+ dump_exception($marc, "exclusion $tag$sub '$filter'");
+ return 1;
+ }
+ # no match, no exclude
+ return 0;
+ }
+ }
+ }
+}
+
=head2 initialize
Performs boring script initialization. Handles argument parsing,
'tag|t=s',
'fingerprints=s',
'scores=s',
- 'arbitrarily-decrease-score-above=i',
- 'arbitrarily-decrease-score-below=i',
- 'arbitrarily-decrease-score-by=i',
+ 'arbitrarily-lose-above=i',
+ 'arbitrarily-lose-below=i',
'newwins',
+ 'excludelist=s',
'quiet|q',
'help|h',
);
my %valid_fps = ( oclc => 1, isbn => 1, issn => 1, lccn => 1,
edition => 1, accomp => 1, authpub => 1,
baseline => 1, crap => 1,
+ koha_bib_id => 1,
);
for (split /,/, $c->{fingerprints}) {
die "Invalid fingerprint '$_'\n" unless $valid_fps{$_};
$c->{tag} = 903 unless defined $c->{tag};
$c->{subfield} = 'a' unless defined $c->{subfield};
$c->{marctype} = 'XML' unless defined $c->{marctype};
- $c->{'arbitrarily-decrease-score-by'} = 0
- unless defined $c->{'arbitrarily-decrease-score-by'};
if ($c->{prefix}) {
$c->{output} = join('.',$c->{prefix},'fp');
$c->{exception} = join('.',$c->{prefix},'fp','ex');
}
+ # get SFM object if excludelist was specified
+ if ($c->{excludelist}) {
+ $c->{excludelist} =
+ Equinox::Migration::SubfieldMapper->new( file => $c->{excludelist} );
+ }
+
my @keys = keys %{$c};
show_help() unless (@ARGV and @keys);
for my $key ('tag', 'subfield', 'output', 'exception')
--fingerprints=LIST Fingerprints to generate, comma separated
Default: oclc,isbn,edition,issn,lccn,accomp,authpub
- Others: baseline
+ Others: baseline,koha_bib_id
+ --excludelist=FILE Name of fingerprints exclusions file
--scores=LIST Scores to calculate, comma separated
Default: oclc,dlc,num_650,num_tags,enc_level
--newwins New record IDs score higher (default is old wins)
+ --arbitrarily-lose-above
+ --arbitrarily-lose-below
+ --arbitrarily-decrease-score-by
+ Modify fingerprint scoring of records whose EG id is above or below a
+ given value, inclusive (so 5 is <= 5 or >= 5) such that they lose.
--marctype=TYPE Defaults to 'XML'
HELP
projects / migration-tools.git / blobdiff