use Getopt::Long;
use MARC::Batch;
-use MARC::File::XML ( BinaryEncoding => 'utf-8' );
-use MARC::Field;
use Unicode::Normalize;
+use MARC::File::XML ( BinaryEncoding => 'utf-8' );
my $conf = {}; # configuration hashref
-my $marc = {}; # MARC record hashref
-my $count = 0;
-initialyze($conf);
+my $count = 0; my $scount = 0;
+my $start = time;
+$| = 1;
-for my $file (@ARGV) {
+initialize($conf);
- print STDERR "Processing $file\n";
+open OF, '>', $conf->{output};
+binmode(OF, ':utf8');
+open XF, '>', $conf->{exception};
+binmode(XF, ':utf8');
- open my $M, '<:utf8', $file;
+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;
- my $batch = MARC::Batch->new('XML',$M);
+ #$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();
- while ( my $record = $batch->next ) {
- $count++;
+ while ( $record = $batch->next ) {
+ $count++; progress_ticker();
+ my $marc = undef;
+ unless ( defined $record )
+ { dump_exception($marc); next; }
my $id = $record->field($conf->{tag});
unless ($id) {
- print STDERR "ERROR: This record is missing a", $conf->{tag},
- "field.\n", $record->as_formatted(), "\n=====\n";
+ print XF "ERROR: Record $count in $file is missing a ",
+ $conf->{tag}, " field.\n", $record->as_formatted(), "\n=====\n";
next;
}
- $id = $id->as_string($conf->{subfield});
-
- my $leader = $record->leader();
- my $record_type = substr($leader,6,1);
- my $bib_lvl = substr($leader,7,1);
-
- my $my_008 = $record->field('008');
- $my_008 = $my_008->as_string() if ($my_008);
- my $date1 = substr($my_008,7,4) if ($my_008);
- my $date2 = substr($my_008,11,4) if ($my_008);
-
- my $item_form;
- if ( $record_type =~ /[gkroef]/ ) { # MAP, VIS
- $item_form = substr($my_008,29,1) if ($my_008);
- } else {
- $item_form = substr($my_008,23,1) if ($my_008);
+
+ $marc = populate_marc($record, $id);
+ $marc = normalize_marc($marc);
+ unless (marc_isvalid($marc))
+ { dump_exception($marc); next; }
+ dump_fingerprints($marc);
+ $scount++; progress_ticker();
+ }
+}
+
+print "\nSuccessfully processed:\t$count\n" unless $conf->{quiet};
+
+=head2 populate_marc
+
+Constructs a hash containing the relevant MARC data for a record and
+returns a reference to it.
+
+=cut
+
+sub populate_marc {
+ my ($record, $id) = @_;
+ my %marc = (); $marc{isbns} = [];
+
+ # id, stringified
+ $marc{id} = $id->as_string($conf->{subfield});
+
+ # record_type, bib_lvl
+ $marc{record_type} = substr($record->leader, 6, 1);
+ $marc{bib_lvl} = substr($record->leader, 7, 1);
+
+ # date1, date2
+ my $my_008 = $record->field('008');
+ $my_008 = $my_008->as_string() if ($my_008);
+ unless (defined $my_008 and length $my_008 == 40)
+ { print XF ">> Bad 008 length in rec ",$marc{id},"\n"; return \%marc }
+ $marc{date1} = substr($my_008,7,4) if ($my_008);
+ $marc{date2} = substr($my_008,11,4) if ($my_008); # UNUSED
+
+ # item_form
+ if ( $marc{record_type} =~ /[gkroef]/ ) { # MAP, VIS
+ $marc{item_form} = substr($my_008,29,1) if ($my_008);
+ } else {
+ $marc{item_form} = substr($my_008,23,1) if ($my_008);
+ }
+
+ # isbns
+ my @isbns = $record->field('020') if $record->field('020');
+ push @isbns, $record->field('024') if $record->field('024');
+ for my $f ( @isbns ) {
+ push @{ $marc{isbns} }, $1 if ( defined $f->subfield('a') and
+ $f->subfield('a')=~/(\S+)/ );
+ }
+
+ # author
+ for my $rec_field (100, 110, 111) {
+ if ($record->field($rec_field)) {
+ $marc{author} = $record->field($rec_field)->subfield('a');
+ last;
}
+ }
- my $title = $record->field('245');
- $title = $title->subfield('a') if $title;
-
- my @isbns = ();
- my @isbns_020;
- if ($record->field('020')) { @isbns_020 = $record->field('020'); }
- foreach my $f ( @isbns_020 ) {
- if ($f->subfield('a')) {
- if ( $f->subfield('a')=~/(\S+)/ ) {
- push @isbns, $1;
- }
- }
+ # "Accompanying material" (300e)
+ $marc{accomp} = $record->field('300')->subfield('e')
+ if $record->field('300');
+
+ # issn, lccn, title, desc, pages, pub, pubyear, edition
+ $marc{lccn} = $record->field('010')->subfield('a') if $record->field('010');
+ $marc{issn} = $record->field('022')->subfield('a') if $record->field('022');
+ $marc{desc} = $record->field('300')->subfield('a') if $record->field('300');
+ $marc{pages} = $1 if (defined $marc{desc} and $marc{desc} =~ /(\d+)/);
+ $marc{title} = $record->field('245')->subfield('a')
+ if $record->field('245');
+ $marc{edition} = $record->field('250')->subfield('a')
+ if $record->field('250');
+ if ($record->field('260')) {
+ $marc{publisher} = $record->field('260')->subfield('b');
+ $marc{pubyear} = $record->field('260')->subfield('c');
+ $marc{pubyear} =
+ (defined $marc{pubyear} and $marc{pubyear} =~ /(\d{4})/) ? $1 : '';
+ }
+ return \%marc;
+}
+
+
+
+=head2 normalize_marc
+
+Gently massages your data.
+
+=cut
+
+sub normalize_marc {
+ my ($marc) = @_;
+
+ $marc->{record_type }= 'a' if ($marc->{record_type} eq ' ');
+ if ($marc->{title}) {
+ $marc->{title} = NFD($marc->{title});
+ $marc->{title} =~ s/[\x{80}-\x{ffff}]//go;
+ $marc->{title} = lc($marc->{title});
+ $marc->{title} =~ s/\W+$//go;
+ }
+ if ($marc->{author}) {
+ $marc->{author} = NFD($marc->{author});
+ $marc->{author} =~ s/[\x{80}-\x{ffff}]//go;
+ $marc->{author} = lc($marc->{author});
+ $marc->{author} =~ s/\W+$//go;
+ if ($marc->{author} =~ /^(\w+)/) {
+ $marc->{author} = $1;
}
- my @isbns_024;
- if ($record->field('024')) { @isbns_024 = $record->field('024'); }
- foreach my $f ( @isbns_024 ) {
- if ($f->subfield('a')) {
- if ( $f->subfield('a')=~/(\S+)/ ) {
- push @isbns, $1;
- }
- }
+ }
+ if ($marc->{publisher}) {
+ $marc->{publisher} = NFD($marc->{publisher});
+ $marc->{publisher} =~ s/[\x{80}-\x{ffff}]//go;
+ $marc->{publisher} = lc($marc->{publisher});
+ $marc->{publisher} =~ s/\W+$//go;
+ if ($marc->{publisher} =~ /^(\w+)/) {
+ $marc->{publisher} = $1;
}
+ }
+ return $marc;
+}
- my $issn = $record->field('022');
- $issn = $issn->subfield('a') if $issn;
- my $lccn = $record->field('010');
- $lccn = $lccn->subfield('a') if $lccn;
- my $author;
- if ($record->field('100'))
- { $author = $record->field('100')->subfield('a'); }
- unless ( $author ) {
- $author = $record->field('110')->subfield('a')
- if ($record->field('110'));
- $author = $record->field('111')->subfield('a')
- if ($record->field('111'));
- }
+=head2 marc_isvalid
+
+Checks MARC record to see if neccessary fingerprinting data is
+available
+
+=cut
+
+sub marc_isvalid {
+ my ($marc) = @_;
+ return 1 if ($marc->{item_form} and ($marc->{date1} =~ /\d{4}/) and
+ $marc->{record_type} and $marc->{bib_lvl} and $marc->{title});
+ return 0;
+}
- my $desc = $record->field('300');
- $desc = $desc->subfield('a') if $desc;
- my $pages;
- $pages = $1 if (defined $desc and $desc =~ /(\d+)/);
+=head2 dump_fingerprints
+
+=cut
- my $my_260 = $record->field('260');
- my $publisher = $my_260->subfield('b') if $my_260;
- my $pubyear = $my_260->subfield('c') if $my_260;
- if ( $pubyear ) {
- if ( $pubyear =~ /(\d\d\d\d)/ )
- { $pubyear = $1; }
- else
- { $pubyear = ''; }
+sub dump_fingerprints {
+ my ($marc) = @_;
+
+ 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) && $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";
+ }
}
- my $edition = $record->field('250');
- $edition = $edition->subfield('a') if $edition;
+ if ($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";
+ }
- # NORMALIZE
- $record_type = 'a' if ($record_type eq ' ');
- if ($title) {
- $title = NFD($title); $title =~ s/[\x{80}-\x{ffff}]//go;
- $title = lc($title);
- $title =~ s/\W+$//go;
+ 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}, $marc->{issn}), "\n";
}
- if ($author) {
- $author = NFD($author); $author =~ s/[\x{80}-\x{ffff}]//go;
- $author = lc($author);
- $author =~ s/\W+$//go;
- if ($author =~ /^(\w+)/) {
- $author = $1;
- }
+
+ if ($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}, $marc->{lccn}) ,"\n";
}
- if ($publisher) {
- $publisher = NFD($publisher); $publisher =~ s/[\x{80}-\x{ffff}]//go;
- $publisher = lc($publisher);
- $publisher =~ s/\W+$//go;
- if ($publisher =~ /^(\w+)/) {
- $publisher = $1;
- }
+
+ if ($marc->{accomp}) { # case e : accomp
+ print OF join("\t", $marc->{id}, "case d",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{accomp}) ,"\n";
}
- # SPIT OUT FINGERPRINTS FROM THE "LOIS ALGORITHM"
- # If we're not getting good matches, we may want to change this.
- # The same thing goes for some other fields.
- if ($item_form && ($date1 =~ /\d\d\d\d/)
- && $record_type && $bib_lvl && $title) {
- if ($conf->{runtype} eq "primary") {
- print STDOUT
- join("\t",$id,$item_form,$date1,$record_type,$bib_lvl,$title)
- ,"\n";
- } else {
- # case a : isbn and pages
- if (scalar(@isbns)>0 && $pages) {
- foreach my $isbn ( @isbns ) {
- print STDOUT
- join("\t", $id, "case a", $item_form, $date1,
- $record_type, $bib_lvl, $title, $isbn, $pages)
- ,"\n";
- }
- }
- # case b : edition
- if ($edition) {
- print STDOUT
- join("\t", $id, "case b", $item_form, $date1,
- $record_type, $bib_lvl, $title,$edition), "\n";
- }
- # case c : issn
- if ($issn) {
- print STDOUT join("\t", $id, "case c", $item_form, $date1,
- $record_type, $bib_lvl, $title, $issn)
- ,"\n";
- }
- # case d : lccn
- if ($lccn) {
- print STDOUT join("\t", $id, "case d", $item_form, $date1,
- $record_type, $bib_lvl, $title, $lccn)
- ,"\n";
- }
- # case e : author, publisher, pubyear, pages
- if ($author && $publisher && $pubyear && $pages) {
- print STDOUT join("\t", $id, "case e", $item_form, $date1,
- $record_type, $bib_lvl, $title, $author,
- $publisher, $pubyear, $pages), "\n";
- }
- }
- } else {
- print STDERR "Record " . $id . " did not make the cut: ";
- print STDERR "Missing item_form. " unless ($item_form);
- print STDERR "Missing valid date1. " unless ($date1 =~ /\d\d\d\d/);
- print STDERR "Missing record_type. " unless ($record_type);
- print STDERR "Missing bib_lvl. " unless ($bib_lvl);
- print STDERR "Missing title. " unless ($title);
- print STDERR "\n";
+ # 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 e",
+ $marc->{item_form}, $marc->{date1},
+ $marc->{record_type}, $marc->{bib_lvl},
+ $marc->{title}, $marc->{author},
+ $marc->{publisher}, $marc->{pubyear},
+ $marc->{pages}), "\n";
}
}
- print STDERR "Processed $count records\n";
}
-=head2 initialyze
+
+
+=head2 dump_exception
+
+Write line of exception report
+
+=cut
+
+sub dump_exception {
+ my ($marc) = @_;
+ 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});
+ unless (defined $marc->{date1})
+ { print XF "Missing date1. " }
+ else
+ { 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 ($marc->{title});
+ print XF "\n";
+}
+
+
+=head2 initialize
Performs boring script initialization. Handles argument parsing,
mostly.
=cut
-sub initialyze {
+sub initialize {
my ($c) = @_;
my @missing = ();
# set mode on existing filehandles
- binmode(STDOUT, ':utf8');
binmode(STDIN, ':utf8');
my $rc = GetOptions( $c,
+ 'incoming',
+ 'incumbent',
+ 'exception|x=s',
+ 'marctype|m=s',
+ 'output|o=s',
'runtype|r=s',
- 'tag|t=s',
'subfield|s=s',
+ 'tag|t=s',
+ 'quiet|q',
+ 'help|h',
);
show_help() unless $rc;
+ show_help() if ($c->{help});
+
+ # set defaults if told to do so
+ if ($c->{incoming}) {
+ $c->{tag} = 903 unless defined $c->{tag};
+ $c->{subfield} = 'a' unless defined $c->{subfield};
+ $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') {
- push @missing, $key unless $c->{$key}
- }
+ for my $key ('runtype', 'tag', 'subfield', 'output', 'exception')
+ { push @missing, $key unless $c->{$key} }
if (@missing) {
print "Required option: ", join(', ', @missing), " missing!\n";
show_help();
}
}
+
+=head2 progress_ticker
+
+=cut
+
+sub progress_ticker {
+ return if $conf->{quiet};
+ printf("\r> %d recs seen; %d processed", $count, $scount);
+ printf(" (%d/s)", ($count / (time - $start + 1)))
+ if ($count % 500 == 0);
+}
+
=head2 show_help
+Display usage message when things go wrong
+
=cut
sub show_help {
print <<HELP;
-Usage is: fingerprinter [REQUIRED ARGS] [OPTIONS] <filelist>
+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
- None yet...
+ --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'
+
+ --quiet -q Don't write status messages to STDOUT
HELP
exit 1;
}