X-Git-Url: http://git.equinoxoli.org/?p=migration-tools.git;a=blobdiff_plain;f=fingerprinter;h=e228c6edeb6751ac791c2bbb9d367bc0222d7694;hp=12e9723b5a58d323a130c434c9aa7c1a6e590474;hb=209c24bd96a2534843af84d1c0176e6dbe80b82d;hpb=04db2af74e905a717c624f703d4ab98e0ffc32a1 diff --git a/fingerprinter b/fingerprinter index 12e9723..e228c6e 100755 --- a/fingerprinter +++ b/fingerprinter @@ -74,29 +74,30 @@ sub populate_marc { # 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 @@ -126,9 +127,11 @@ sub populate_marc { 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'); @@ -208,8 +211,9 @@ sub marc_isvalid { 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(); +Score is constructed by pushing elements onto a list, via a dispatch +table. This allows order of fingerprints in the output file to be +varied. =cut @@ -218,6 +222,39 @@ sub score_marc { 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 { @@ -239,18 +276,19 @@ sub score_marc { 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"; } @@ -341,16 +379,16 @@ sub dump_exception { 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 ", $marc->{id}, " excluded: "; + print XF "missing item_form; " unless ($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"; } @@ -370,16 +408,17 @@ sub initialize { binmode(STDIN, ':utf8'); my $rc = GetOptions( $c, - 'incoming', - 'incumbent', 'exception|x=s', - 'marctype|m=s', 'output|o=s', - 'runtype|r=s', + 'prefix|p=s', + 'marctype|m=s', 'subfield|s=s', 'tag|t=s', 'fingerprints=s', 'scores=s', + 'arbitrarily-lose-above=i', + 'arbitrarily-lose-below=i', + 'newwins', 'quiet|q', 'help|h', ); @@ -421,19 +460,13 @@ sub initialize { $c->{dyn_scores} = [ qw/oclc dlc num_650 num_tags enc_lvl/ ]; } - # set defaults if told to do so - if ($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 $c->{exception}; - } elsif ($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}; + # set defaults + $c->{tag} = 903 unless defined $c->{tag}; + $c->{subfield} = 'a' unless defined $c->{subfield}; + $c->{marctype} = 'XML' unless defined $c->{marctype}; + if ($c->{prefix}) { + $c->{output} = join('.',$c->{prefix},'fp'); + $c->{exception} = join('.',$c->{prefix},'fp','ex'); } my @keys = keys %{$c}; @@ -468,20 +501,30 @@ sub show_help { print < Req'd Arguments - --tag=N -t Which tag to use - --subfield=X -s Which subfield to use - --output= -o Output filename - --exceptions= -x Exception report filename + --output= -o Output filename + --exceptions= -x Exception report filename + or + --prefix=> -p Shared prefix for output/exception files. Will + produce PREFIX.fp and PREFIX.fp.ex Options - --incoming '-t 903 -s a -o incoming.fp -x incoming.ex' - --incumbent '-t 901 -s c -o incumbent.fp -x incumbent.ex' + --tag=N -t Which tag to use (default 903) + --subfield=X -s Which subfield to use (default 'a') + --quiet -q Don't write status messages to STDOUT --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 + --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 exit 1; }