X-Git-Url: http://git.equinoxoli.org/?p=migration-tools.git;a=blobdiff_plain;f=extract_holdings;h=27cf58a419ad361aec61220d80d5d432a53eb4da;hp=69ceeff92f11771cfc489c134c18509e5a28bd6a;hb=7a654b50c2edbb1f746895a2e73e6038e59d6e93;hpb=7120b02cebae06d75d650db57789059113548091 diff --git a/extract_holdings b/extract_holdings index 69ceeff..27cf58a 100755 --- a/extract_holdings +++ b/extract_holdings @@ -2,110 +2,181 @@ use strict; use warnings; -use XML::Twig; -use YAML::Tiny; use Getopt::Long; -use Equinox::Migration::SubfieldMapper; +use Equinox::Migration::MapDrivenMARCXMLProc 1.003; +use Equinox::Migration::MARCXMLSampler; -$| = 1; +my $VERSION = '1.001'; -my $c = initialize(); -my $marcxml = shift; -my $copyid = 0; -my $holdings; - -my $count = 0; -my $total = `grep -c 'new; -my $t = XML::Twig->new( twig_handlers => { record => \&record } ); -$t->parsefile($marcxml); -write_sample_fieds(); - -sub record { - my($t, $r)= @_; - $holdings = {}; - - my @dfields = $r->children('datafield'); - for my $d (@dfields) { - process_datafields($d); - } - write_data_out(); - $r->purge; +=pod - $count++; - $percent = int(($count / $total) * 100); - print "\r$percent% done ($count)";# if ($percent != $prevper); - $prevper = $percent; -} +TODO -sub process_datafields { - my ($d) = @_; - my $tag = $d->{'att'}->{'tag'}; - - if ($tag == 903) { - my $s = $d->first_child('subfield'); - $holdings->{id} = $s->text;; - } elsif ($c->{map}->has($tag)) { - push @{$holdings->{copies}}, { tag => $tag }; - my @subs = $d->children('subfield'); - for my $sub (@subs) - { process_subs($tag,$sub) } - } -} + * Have detail mode report on number of subfields per datafield -sub process_subs { - my ($tag, $sub) = @_; - my $code = $sub->{'att'}->{'code'}; +=cut - unless ($c->{map}->has($tag, $code)) { - # this is a subfield code we don't have mapped. report on it if this is a sample tag - push @{$c->{sample}{$tag}}, $code if defined $c->{sample}{tag}; - return; - } +my $c = initialize(); +$| = 1; - my $copy = $holdings->{copies}[-1]; - my $field = $c->{map}->field($tag, $code); - if ($c->{map}->mod($field) eq 'multi') { - my $name = $tag . $code; - push @{$copy->{multi}{$name}}, $sub->text; - } else { - $copy->{uni}{$code} = $sub->text; +# run samples if we've been asked for them +run_samples($c) if ($c->{sample} or $c->{samplemap} or $c->{samplestr}); +extract_holdings($c) if ($c->{map}); + +#-------------------------- + +sub extract_holdings { + my ($c) = @_; + print "Parsing records for extraction... "; + my $m = Equinox::Migration::MapDrivenMARCXMLProc->new( marcfile => $c->{marcfile}, + mapfile => $c->{map}, + verbose => 1, + ); + print "Writing holdings to output file(s)...\n"; + # open main holdings file + open HOLDINGS, '>', ($c->{prefix} . "-HOLDINGS.pg"); + # create multi files + my $multis = $m->get_multis; + my %MULTIFILE = (); + for my $t ( keys %{$multis} ) { + for my $s ( keys %{$multis->{$t}}) + { open my $fh, ">", ($c->{prefix} . "-HOLDINGS-MULT-$t$s.pg"); $MULTIFILE{"$t$s"} = $fh } } -} + my $i = 0; # record counter + my $j = 0; # holdings counter + + while ( $m->{data}{recs}[$i] ) { + print HOLDINGS "BEGIN;\n\negid, hseq, " unless $j; + my $rec = $m->{data}{recs}[$i]; + my $k = 0; # holding-within-record pointer + # for each holdings tag in the record... + while ( defined $rec->{tmap}{ $c->{holdings} }[$k] ) { + my $holdidx = $rec->{tmap}{ $c->{holdings} }[$k]; + my $tagid = $rec->{tags}[$holdidx]{tag}; + $k++; + + my @out = (); # clear the output buffer + push @out, $rec->{egid}; # slug in the egid first thing + push @out, $j; # holding seq goes next + + # grab the unary mappings and slug 'em in + for my $sub ( sort keys %{$rec->{tags}[$holdidx]{uni}} ) { + push @out, $rec->{tags}[$holdidx]{uni}{$sub}; + print HOLDINGS "l_", $m->name($tagid, $sub),", " unless $j; + } -#------------------------------------------------ + # handle holdings multis + for my $sub ( sort keys %{$multis->{$tagid}} ) { + for my $value ( @{$rec->{tags}[$holdidx]{multi}{$sub}} ) { + my $fh = $MULTIFILE{"$tagid$sub"}; + print $fh join("\t", $rec->{egid}, $j, $value), "\n"; + } + } -sub write_data_out { - my $i = 0; - for my $copy (@{$holdings->{copies}}) { - print HOLDINGS $holdings->{id}, "\t$i\t", $copy->{tag}; - for ( sort keys %{ $c->{map}{fields} } ) { - if (defined $copy->{uni}{$_->{sub}}) { - print HOLDINGS "\t", $copy->{uni}{$_->{sub}}; - } else { - print HOLDINGS "\t"; + # now get everything else in the mapping + for my $othertag ( sort keys %{$rec->{tmap}} ) { + next if $othertag eq $c->{holdings}; # ignoring the holdings, o'course + my $idx = $rec->{tmap}{$othertag}[0]; # get index into tags struct + for my $sub ( sort keys %{$rec->{tags}[$idx]{uni}} ) { + push @out, $rec->{tags}[$idx]{uni}{$sub}; + print HOLDINGS "l_", $m->name($rec->{tags}[$idx]{tag}, $sub), ", " + unless $j; + } } - } - print HOLDINGS "\n"; - for my $m (sort keys %{$copy->{multi}}) { - my $fh = $c->{files}{multi}{$m}; - print $fh join("\t", $holdings->{id}, $i, @{$copy->{multi}{$m}}), "\n"; + # and dump it + print HOLDINGS "\n" unless $j; + print HOLDINGS join("\t", @out); + print HOLDINGS "\n"; + $j++; + + @out = undef; } $i++; + print "\r$i"; } + print "\n"; +} + +#-------------------------- + +sub run_samples { + my ($c) = @_; + my $s; + print "Parsing records for sampling... "; + if ($c->{samplemap}) { + $s = Equinox::Migration::MARCXMLSampler->new( marcfile => $c->{marcfile}, + mapfile => $c->{samplemap}); + } elsif ($c->{samplestr}) { + $s = Equinox::Migration::MARCXMLSampler->new( marcfile => $c->{marcfile}, + mapstring => $c->{samplestr}); + } else { + $s = Equinox::Migration::MARCXMLSampler->new( marcfile => $c->{marcfile} ); + } + + dump_sample_overview($c, $s) if $c->{sample}; + dump_sample_detail($c, $s) if ($c->{samplemap} or $c->{samplestr}); } -sub write_sample_fields { +sub dump_sample_detail { + my ($c, $s) = @_; + my $tags = $s->{data}{samp}; + my $count = $s->{data}{rcnt}; + my $scnt = $s->{data}{scnt}; + + open DETAIL, '>', ($c->{prefix} . "-HOLDINGS-DETAIL.txt"); + select DETAIL; + for my $tag (sort keys %{ $tags }) { + print ">>>>> TAG $tag\n\n"; + for my $subkey (sort keys %{ $tags->{$tag} }) { + my $sub = $tags->{$tag}{$subkey}; + print "|| $subkey | ", $sub->{value}, " | ", + $sub->{count}, "/", $sub->{tcnt}, " | ||\n"; + } + print "\n"; + } + close DETAIL; + open SCOUNT, '>', ($c->{prefix} . "-HOLDINGS-SUBCOUNTS.txt"); + select SCOUNT; + for my $tag (sort keys %{ $scnt }) { + print ">>>>> TAG $tag\n\n"; + for my $len (sort keys %{ $scnt->{$tag} }) + { print "|| $len | ", $scnt->{$tag}{$len}, " ||\n" } + print "\n"; + } + select STDOUT; } -#------------------------------------------------ +sub dump_sample_overview { + my ($c, $s) = @_; + my $tags = $s->{data}{tags}; + my $count = $s->{data}{rcnt}; + + my @tagsbyname = sort keys %{$tags}; + my @tagsbycount = reverse sort { $tags->{$a} <=> $tags->{$b} } keys %{$tags}; + + open SAMPLE, '>', ($c->{prefix} . "-HOLDINGS-OVERVIEW.txt"); + select SAMPLE; + print "SAMPLE REPORT FOR ", $c->{prefix},": $count records\n\n"; + print "FOUND TAGS (BY TAG) FOUND TAGS (BY COUNT)\n"; + print "------------------------ --------------------------\n"; + for my $i (0 .. @tagsbyname - 1) { + print $tagsbyname[$i], (" " x (14 - length $tags->{ $tagsbyname[$i] })), + $tags->{ $tagsbyname[$i] }; + print " (", sprintf("%03d", int($tags->{ $tagsbyname[$i] } / $count * 100)), "%)"; + print " "; + print $tagsbycount[$i], (" " x (16 - length $tags->{ $tagsbycount[$i] })), + $tags->{ $tagsbycount[$i] }; + print " (", sprintf("%03d", int($tags->{ $tagsbycount[$i] } / $count * 100)), "%)\n"; + } + select STDOUT; + print "\n"; + close SAMPLE; +} +#-------------------------- sub initialize { my $c = {}; @@ -115,72 +186,66 @@ sub initialize { binmode(STDIN, ':utf8'); my $rc = GetOptions( $c, - 'sample|s=s', - 'map|m=s', - 'ils=s', + 'sample|s', + 'samplemap|sm=s', + 'samplestr|ss=s', + 'marcfile|m=s', + 'map=s', + 'holdings|h=i', + 'copyid|c=s', 'prefix|p=s', - 'help|h', + 'version|v', + 'help', ); show_help() unless $rc; show_help() if ($c->{help}); + show_help("Nothing to do!") + unless ($c->{map} or $c->{sample} or $c->{samplemap} or $c->{samplestr}); + show_help("map, holdings, and copyid must be specified together!") + if ($c->{map} and !($c->{holdings} and $c->{copyid})); + show_version() if $c->{version}; my @keys = keys %{$c}; - show_help() unless (@ARGV and @keys); - for my $key ('prefix', 'map', 'ils') + for my $key ('prefix', 'marcfile') { push @missing, $key unless $c->{$key} } if (@missing) { print "Required option: ", join(', ', @missing), " missing!\n"; show_help(); } - # generate subfield map - $c->{map} = Equinox::Migration::SubfieldMapper->new( file => $c->{map} ); + return $c; +} - # explode sample tags string - if (defined $c->{sample}) { - my $sample = $c->{sample}; - $c->{sample} = {}; - for (split /,/, $c->{sample}) { - $c->{sample}{$_} = []; - } - } +sub show_help { + my ($msg) = @_; + print "\nERROR - $msg\n" if $msg; + print <', ($c->{prefix} . ".holdings.pg"); - for my $f (keys %{$c->{map}{fields}}) { - if ($c->{map}->mod($f)) { - open my $mfh, '>', join('.', $c->{prefix}, "holdings", - $c->{map}{fields}{$f}{tag}, - $c->{map}{fields}{$f}{sub}, "pg"); - $c->{files}{multi}{ ($c->{map}{fields}{$f}{tag} . $c->{map}{fields}{$f}{sub}) } - = $mfh; - } - } +Usage is: extract_holdings -p PREFIX -m MARCFILE [ARGUMENTS] - # print file headers - print HOLDINGS "BEGIN;\n"; - print HOLDINGS "CREATE TABLE ", $c->{prefix}, ".asset_copy_", $c->{ils}; - print HOLDINGS $c->{library} if (defined $c->{library}); - print HOLDINGS " ("; - for ( sort keys %{ $c->{map}{fields} } ) { - print HOLDINGS "l_", $_, " TEXT, "; - } - print HOLDINGS ") INHERITS FROM (", $c->{prefix}, ".asset_copy);\n"; - print HOLDINGS "COPY ", $c->{prefix}, ".asset_copy_", $c->{ils}; - print HOLDINGS $c->{library} if (defined $c->{library}); - print HOLDINGS " ("; - for ( sort keys %{ $c->{map}{fields} } ) { - print HOLDINGS "l_", $_, ", "; - } - print HOLDINGS ") FROM STDIN;\n"; +REQUIRED ARGUMENTS + --prefix -p Prefix string for output filenames + --marcfile -m MARCXML to use as source data +SAMPLING ARGUMENTS + --sample -s Generate a report of all tags in the MARC data + --samplemap -sm Specify a E::M::STL map file which will be used to generate + subfield breakdown reports about specific tags in the MARC + data + --samplestr -ss As above, but with a one-liner map specified on the command + line as a string (e.g. '-ss "852 999"') - return $c; -} + If --samplemap and --samplestr are both specified, --samplemap wins. -sub show_help { - print <