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;
# 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; }
# date1, date2
my $my_008 = $record->field('008');
- $my_008 = $my_008->as_string() if ($my_008);
- if (defined $my_008) {
- unless (length $my_008 == 40) {
- $my_008 = $my_008 . (' ' x (40 - length($my_008)));
- print XF ">> Short 008 padded to ",length($my_008)," at rec $count\n";
+ $marc{tag008} = $my_008->as_string() if ($my_008);
+ if (defined $marc{tag008}) {
+ unless (length $marc{tag008} == 40) {
+ $marc{tag008} = $marc{tag008} . ('|' x (40 - length($marc{tag008})));
+ print XF ">> Short 008 padded to ",length($marc{tag008})," at rec $count\n";
}
- $marc{date1} = substr($my_008,7,4) if ($my_008);
- $marc{date2} = substr($my_008,11,4) if ($my_008); # UNUSED
+ $marc{date1} = substr($marc{tag008},7,4) if ($marc{tag008});
+ $marc{date2} = substr($marc{tag008},11,4) if ($marc{tag008}); # UNUSED
}
unless ($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";
}
}
# item_form
if ( $marc{record_type} =~ /[gkroef]/ ) { # MAP, VIS
- $marc{item_form} = substr($my_008,29,1) if ($my_008);
+ $marc{item_form} = substr($marc{tag008},29,1) if ($marc{tag008});
} else {
- $marc{item_form} = substr($my_008,23,1) if ($my_008);
+ $marc{item_form} = substr($marc{tag008},23,1) if ($marc{tag008});
}
# isbns
if (defined $oclc and $oclc =~ /\(OCoLC\)/ and $oclc =~/([0-9]+)/);
}
- # "Accompanying material" (300e)
- $marc{accomp} = $record->field('300')->subfield('e')
- if $record->field('300');
+ # "Accompanying material" and check for "copy" (300)
+ if ($record->field('300')) {
+ $marc{accomp} = $record->field('300')->subfield('e');
+ $marc{tag300a} = $record->field('300')->subfield('a');
+ }
# issn, lccn, title, desc, pages, pub, pubyear, edition
$marc{lccn} = $record->field('010')->subfield('a') if $record->field('010');
my @score = ();
my $json = '{';
+ #----------------------------------
+ # static criteria scoring
+ #----------------------------------
+ $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} =~ /\|$/) }
+ else
+ { $marc->{misc_score} -= 2 }
+ # -1 if date has been pulled from 260
+ $marc->{misc_score}-- if $marc->{fudgedate};
+ # -1 if this is a copy record
+ $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
+ #----------------------------------
my %scores_code = (
oclc => sub { return $marc->{oclc}[0] ? 1 : 0 },
dlc => sub {
my %levels = ( ' ' => 9, 1 => 8, 2 => 7, 3 => 6, 4 => 5, 5 => 4,
6 => 3, 7 => 2, 8 => 1, 'u' => 0, 'z' => 0 );
return $levels{$enc} || 0;
- }
- );
+ }
+ );
+ #----------------------------------
+ # assemble and store scores
+ #----------------------------------
for ( @{ $conf->{dyn_scores} } ) {
push @score, $scores_code{$_}->($marc, $record);
$json .= $_ . ':' . $score[-1] . ',';
}
- chop($json); # get rid of the trailing comma
-
- $json .= '}';
+ $json .= 'misc:' . $marc->{misc_score} . '}';
- my $compact = join('', @score);
+ my $compact = join('', $marc->{age_score}, $marc->{misc_score}, @score);
$marc->{score} = "$compact\t$json";
}
=cut
sub dump_exception {
- my ($marc) = @_;
+ my ($marc, $msg) = @_;
unless (defined $marc) {
print XF "Undefined record at line $count; likely bad XML\n";
return;
}
- print XF "Record ", $marc->{id}, " did not make the cut: ";
- print XF "Missing item_form. " unless ($marc->{item_form});
+
+ 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. " }
+ { print XF "missing date1; " }
else
- { print XF "Invalid date1: ", $marc->{date1}, " "
+ { print XF "invalid date1: '", $marc->{date1}, "'; "
unless ($marc->{date1} =~ /\d{4}/); }
- print XF "Missing record_type. " unless ($marc->{record_type});
- print XF "Missing bib_lvl. " unless ($marc->{bib_lvl});
- print XF "Missing title. " unless ($marc->{title});
+ print XF "missing record_type; " unless ($marc->{record_type});
+ print XF "missing bib_lvl; " unless ($marc->{bib_lvl});
+ print XF "missing title " unless ($marc->{title});
print XF "\n";
}
+=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-lose-above=i',
+ 'arbitrarily-lose-below=i',
+ 'newwins',
+ 'excludelist=s',
'quiet|q',
'help|h',
);
$c->{exception} = join('.',$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
+ --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