for my $file (@ARGV) {
print XF "Processing $file\n";
- #open my $records, '<:utf8', $file; # This hack dosn't play nice with the use MARC::File::XML ( BinaryEncoding => 'utf-8' ); hack
my $batch = undef; my $record = undef;
- #$batch = MARC::Batch->new('XML', $records); # The other part of the hack
$batch = MARC::Batch->new($conf->{marctype}, $file);
$batch->strict_off();
$batch->warnings_off();
next;
}
+ # populate and normalize marc
$marc = populate_marc($record, $id);
- $marc = normalize_marc($marc);
+ normalize_marc($marc);
unless (marc_isvalid($marc))
{ dump_exception($marc); next; }
+
+ # if everything looks good, score it and dump fingerprints
+ score_marc($marc, $record);
dump_fingerprints($marc);
$scount++; progress_ticker();
}
}
-=head2 dump_fingerprints
+=head2 score_marc
+
+Assign a score to the record based on various criteria.
+
+Score is constructed by pushing elements onto a list. At the end of
+the routine, the list is flattened into a string via join();
=cut
-sub dump_fingerprints {
- my ($marc) = @_;
- my $good_fp = 0;
+sub score_marc {
+ my ($marc, $record) = @_;
+ my @score = ();
+ my $chunk;
+
+ # Is this an OCLC record?
+ if ($conf->{scores}{oclc}) {
+ if ($record->field('008')) {
+ $chunk = $record->field('008')->as_string();
+ push @score, ( $chunk =~ /^o/i ? 1 : 0 );
+ } else {
+ push @score, 0;
+ }
+ }
- if ($conf->{runtype} eq "primary") {
- print OF join("\t",$marc->{id}, $marc->{item_form},
- $marc->{date1}, $marc->{record_type},
- $marc->{bib_lvl}, $marc->{title}), "\n";
- } else {
- if ((scalar @{ $marc->{isbns} } > 0) and $marc->{pages}) {
- # case a : isbn and pages
- foreach my $isbn ( @{ $marc->{isbns}} ) {
- print OF join("\t", $marc->{id}, "case a",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type},
- $marc->{bib_lvl}, $marc->{title},
- $isbn, $marc->{pages}), "\n";
- }
- $good_fp = 1;
+ # does 040a contain "dlc"?
+ if ($conf->{scores}{dlc}) {
+ if ($record->field('040') and $record->field('040')->subfield('a')) {
+ $chunk = $record->field('040')->subfield('a');
+ push @score, ( $chunk =~ /dlc/i ? 1 : 0 );
+ } else {
+ push @score, 0;
}
+ }
-=pod
+ # number of 650 datafields
+ # zero-padded to 4 digits with printf
+ if ($conf->{scores}{num_650}) {
+ if ($record->field('650')) {
+ my @tags = $record->field('650');
+ push @score, scalar @tags;
+ } else {
+ push @score, '0000';
+ }
+ }
-Looks like case b isn't a terribly good idea most of the time -- it
-will generate many spurious matches in series books in particular. So
-for now, simply turning it off.
+ # number of tags in total
+ # zero-padded to 4 digits with printf
+ if ($conf->{scores}{num_tags}) {
+ my @tags = $record->fields;
+ push @score, scalar @tags;
+ }
- if ($marc->{edition}) { # case b : edition
- print OF join("\t", $marc->{id}, "case b",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $marc->{edition}), "\n";
- $good_fp = 1;
- }
+ # encoding level
+ if ($conf->{scores}{enc_lvl}) {
+ my $enc = substr($record->leader, 17, 1);
+ my %levels = ( ' ' => 9, 1 => 8, 2 => 7, 3 => 6, 4 => 5, 5 => 4,
+ 6 => 3, 7 => 2, 8 => 1, 'u' => 0, 'z' => 0 );
+ if (defined $enc and $levels{$enc})
+ { push @score, $levels{$enc} }
+ else
+ { push @score, 0 }
+ }
+
+ # put score in marc hash
+ $marc->{score} = join('', '{oclc:', $score[0], ',dlc:', $score[1],
+ ',num_650:', $score[2], ',num_tags:', $score[3],
+ ',enc_lvl:', $score[4], '}');
+}
+
+=head2 dump_fingerprints
=cut
- if ($marc->{issn}) { # case c : issn
- print OF join("\t", $marc->{id}, "case c",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $marc->{issn}), "\n";
- $good_fp = 1;
- }
+sub dump_fingerprints {
+ my ($marc) = @_;
- if ($marc->{lccn}) { # case d : lccn
- print OF join("\t", $marc->{id}, "case d",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $marc->{lccn}) ,"\n";
- $good_fp = 1;
- }
+ if ($conf->{fingerprints}{baseline}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, 'baseline',
+ $marc->{item_form}, $marc->{date1}, $marc->{record_type},
+ $marc->{bib_lvl}, $marc->{title}), "\n";
+ }
- if ($marc->{accomp}) { # case e : accomp
- print OF join("\t", $marc->{id}, "case e",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $marc->{accomp}) ,"\n";
- $good_fp = 1;
+ if ($conf->{fingerprints}{oclc} and scalar @{$marc->{oclc} }) {
+ for (@{$marc->{oclc} }) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "oclc",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $_, "\n");
}
+ }
- # case o: oclc
- if (scalar @{$marc->{oclc} }) {
- for (@{$marc->{oclc} }) {
- print OF join("\t", $marc->{id}, "case o",
+ if ($conf->{fingerprints}{isbn}) {
+ if ((scalar @{ $marc->{isbns} } > 0) and $marc->{pages}) {
+ foreach my $isbn ( @{ $marc->{isbns}} ) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "isbn",
$marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $_, "\n");
+ $marc->{record_type},
+ $marc->{bib_lvl}, $marc->{title},
+ $isbn, $marc->{pages}), "\n";
}
- $good_fp = 1;
}
+ }
- # case z : author, publisher, pubyear, pages
- if ($marc->{author} and $marc->{publisher} and $marc->{pubyear}
- and $marc->{pages}) {
- print OF join("\t", $marc->{id}, "case z",
- $marc->{item_form}, $marc->{date1},
- $marc->{record_type}, $marc->{bib_lvl},
- $marc->{title}, $marc->{author},
- $marc->{publisher}, $marc->{pubyear},
- $marc->{pages}), "\n";
- $good_fp = 1;
- }
+ if ($conf->{fingerprints}{edition} and $marc->{edition}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "edition",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{edition}), "\n";
+ }
- # case poo : nothing good; dump a primary and move on
- unless ($good_fp) {
- print OF join("\t", $marc->{id}, "case poo", $marc->{item_form},
- $marc->{date1}, $marc->{record_type},
- $marc->{bib_lvl}, $marc->{title}), "\n";
- }
+ if ($conf->{fingerprints}{issn} and $marc->{issn}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "issn",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{issn}), "\n";
+ }
+
+ if ($conf->{fingerprints}{lccn} and $marc->{lccn}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "lccn",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{lccn}) ,"\n";
+ }
+
+ if ($conf->{fingerprints}{accomp} and $marc->{accomp}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "accomp",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{accomp}) ,"\n";
+ }
+
+ if ($conf->{fingerprints}{authpub} and $marc->{author} and
+ $marc->{publisher} and $marc->{pubyear} and $marc->{pages}) {
+ print OF join("\t", $marc->{score}, $marc->{id}, "authpub",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{author},
+ $marc->{publisher}, $marc->{pubyear},
+ $marc->{pages}), "\n";
}
}
'runtype|r=s',
'subfield|s=s',
'tag|t=s',
+ 'fingerprints=s',
+ 'scores=s',
'quiet|q',
'help|h',
);
show_help() unless $rc;
show_help() if ($c->{help});
+ # check fingerprints list for validity
+ if ($c->{fingerprints}) {
+ my %fps = ();
+ my %valid_fps = ( oclc => 1, isbn => 1, issn => 1, lccn => 1,
+ edition => 1, accomp => 1, authpub => 1,
+ baseline => 1, crap => 1,
+ );
+ for (split /,/, $c->{fingerprints}) {
+ die "Invalid fingerprint '$_'\n" unless $valid_fps{$_};
+ $fps{$_} = 1;
+ }
+ $c->{fingerprints} = \%fps
+ } else {
+ $c->{fingerprints} = {oclc => 1, isbn => 1, edition => 1, issn => 1,
+ lccn => 1, accomp => 1, authpub => 1};
+ }
+ # check scores list for validity
+ if ($c->{scores}) {
+ my %scores = ();
+ my %valid_scores = ( oclc => 1, dlc => 1, num_650 => 1,
+ num_tags => 1, enc_lvl => 1,
+ );
+ for (split /,/, $c->{scores}) {
+ die "Invalid score mode '$_'\n" unless $valid_scores{$_};
+ $scores{$_} = 1;
+ }
+ $c->{scores} = \%scores;
+ } else {
+ $c->{scores} = {oclc => 1, dlc => 1, num_650 => 1,
+ num_tags => 1, enc_lvl => 1};
+ }
+
# set defaults if told to do so
if ($c->{incoming}) {
$c->{tag} = 903 unless defined $c->{tag};
$c->{marctype} = 'XML' unless defined $c->{marctype};
$c->{output} = 'incoming.fp' unless defined $c->{output};
$c->{exception} = 'incoming.ex' unless defined $c->{exception};
- $c->{runtype} = 'full' unless defined $c->{runtype};
} elsif ($c->{incumbent}) {
$c->{tag} = 901 unless defined $c->{tag};
$c->{subfield} = 'c' unless defined $c->{subfield};
$c->{marctype} = 'XML' unless defined $c->{marctype};
$c->{output} = 'incumbent.fp' unless defined $c->{output};
$c->{exception} = 'incumbent.ex' unless defined $c->{exception};
- $c->{runtype} = 'full' unless defined $c->{runtype};
}
my @keys = keys %{$c};
show_help() unless (@ARGV and @keys);
- for my $key ('runtype', 'tag', 'subfield', 'output', 'exception')
+ for my $key ('tag', 'subfield', 'output', 'exception')
{ push @missing, $key unless $c->{$key} }
if (@missing) {
print "Required option: ", join(', ', @missing), " missing!\n";
print <<HELP;
Usage is: $0 [REQUIRED ARGS] [OPTIONS] <filelist>
Req'd Arguments
- --runtype=(primary|full) -r Do 'primary' or 'full' fingerprinting
--tag=N -t Which tag to use
--subfield=X -s Which subfield to use
--output=<file> -o Output filename
--exceptions=<file> -x Exception report filename
Options
- --incoming Set -r to 'full'; -t, -s, -o, -x to incoming defaults
- --incumbent Set -r to 'full'; -t, -s, -o, -x to incumbent defaults
-
- Example: '$0 --incoming' is equivalent to
- '$0 -r full -t 903 -s a -o incoming.fp -x incoming.ex'
-
+ --incoming '-t 903 -s a -o incoming.fp -x incoming.ex'
+ --incumbent '-t 901 -s c -o incumbent.fp -x incumbent.ex'
+
+ --fingerprints=LIST Fingerprints to generate, comma separated
+ Default: oclc,isbn,edition,issn,lccn,accomp,authpub
+ Others: baseline
+ --scores=LIST Scores to calculate, comma separated
+ Default: oclc,dlc,num_650,num_tags,enc_level
--quiet -q Don't write status messages to STDOUT
HELP
exit 1;