my $term = new Term::ReadLine 'yaz-cleanup';
my $OUT = $term->OUT || \*STDOUT;
-my $conf = {};
-
-my $count = 0;
-my $reccount = 0;
-my $oreccount = 0;
-my $line = '';
-my %trash = (); # hash for tags to be dumped
-
# initialization and setup
+my $conf = {};
initialize($conf);
-
-# read in trash tags file if it exists
populate_trash() if ($conf->{trash});
-my @record = (); # current record storage
+my @record = (); # current record storage
my %recmeta = (); # metadata about current record
-my @context= (); # last 5 lines of file
+my $recptr = 0; # record index pointer
my $input = shift || 'incoming.marc.xml';
print $NUMARC '<collection xmlns="http://www.loc.gov/MARC21/slim">',"\n";
open my $EXMARC, '>:utf8', $conf->{exception};
-print $EXMARC '<?xml version="1.0" encoding="UTF-8"?>',"\n";
-print $EXMARC '<collection xmlns="http://www.loc.gov/MARC21/slim">',"\n";
-open MARC2, '<', $input;
-<MARC2>;
# this is the dispatch table which drives command selection in
# edit(), below
my @spinner = qw(- / | \\);
my $sidx = 0;
-while (my $line = getline()) {
- unless ($count % 2000) {
- print "\rWorking... ", $spinner[$sidx];
- $sidx = ($sidx == $#spinner) ? 0 : $sidx + 1;
- }
- update_linecontext();
+while ( buildrecord() ) {
+ print "\rWorking... ", $spinner[$sidx];
+ $sidx = ($sidx == $#spinner) ? 0 : $sidx + 1;
- next if ($line =~ m|</record>|);
+ do_automated_cleanups();
- # catch empty datafield elements
- if ($line =~ m|</datafield>|) {
- if ($record[-2] =~ m/<datafield tag="..." ind1="." ind2=".">/) {
- pop @record; pop @record;
- message("Empty datafield scrubbed");
- next;
- }
- }
+ $recptr = 0;
+ until ($recptr == $#record) {
+ # naked ampersands
+ if ($record[$recptr] =~ /&/ && $record[$recptr] !~ /&\w+?;/)
+ { edit("Naked ampersand"); $recptr= 0; next }
- # pad short leaders
- if ($line =~ m|<leader>(.+?)</leader>|) {
- my $leader = $1;
- if (length $leader < 24) {
- $leader .= ' ' x (20 - length($leader));
- $leader .= "4500";
- $line = "<leader>$leader</leader>\n";
- message("Short leader padded");
+ # tags must be numeric
+ if ($record[$recptr] =~ /<datafield tag="(.+?)"/) {
+ my $match = $1;
+ if ($match =~ /\D/) {
+ edit("Non-numerics in tag");
+ $recptr = 0;
+ next;
+ }
}
- }
-
- # clean misplaced dollarsigns
- if ($line =~ m|<subfield code="\$">c?\d+\.\d{2}|) {
- $line =~ s|"\$">c?(\d+\.\d{2})|"c">\$$1|;
- message("Dollar sign corrected");
- }
- # clean up tags with spaces in them
- $line =~ s/tag=" /tag="00/g;
- $line =~ s/tag=" /tag="0/g;
- $line =~ s/tag="-/tag="0/g;
- $line =~ s/tag="(\d\d) /tag="0$1/g;
-
- # stow tag data if we're looking at it
- if ($line =~ m/<datafield tag="(.{3})" ind1="(.)" ind2="(.)">/) {
- $recmeta{tag} = $1;
- $recmeta{ind1} = $2;
- $recmeta{ind2} = $3;
+ # subfields can't be non-alphanumeric
+ if ($record[$recptr] =~ /<subfield code="(.*?)"/) {
+ my $match = $1;
+ if ($match =~ /\P{IsAlnum}/ or $match eq '') {
+ edit("Junk in subfield code/Null subfield code");
+ $recptr = 0;
+ next;
+ }
+ }
+ $recptr++;
}
- # automatable subfield maladies
- $line =~ s/code=" ">c/code="c">/;
- $line =~ s/code=" ">\$/code="c"$>/;
+ write_record($NUMARC);
+}
+print $NUMARC "</collection>\n";
+print $OUT "\nDone. \n";
- # and stow line back in record
- $record[-1] = $line;
- # naked ampersands
- if ($line =~ /&/ && $line !~ /&\w+?;/)
- { edit("Naked ampersand", $line); next }
+#-----------------------------------------------------------------------------------
+# cleanup routines
+#-----------------------------------------------------------------------------------
- # tags must be numeric
- if ($line =~ /<datafield tag="(.+?)"/) {
- my $match = $1;
- if ($match =~ /\D/) {
- edit("Non-numerics in tag", $line);
+sub do_automated_cleanups {
+ $recptr = 0;
+ until ($recptr == $#record) {
+ # catch empty datafield elements
+ if ($record[$recptr] =~ m|</datafield>|) {
+ if ($record[$recptr + 1] =~ m/<datafield tag="..." ind1="." ind2=".">/) {
+ my @a = @record[0, $recptr - 1];
+ my @b = @record[$recptr + 1, $#record];
+ @record = (@a, @b);
+ message("Empty datafield scrubbed");
+ $recptr = 0;
+ next;
+ }
+ }
+ # and quasi-empty subfields
+ if ($record[$recptr] =~ m|<subfield code="\s*">\s*</sub|) {
+ delete $record[$recptr];
+ message("Empty subfield scrubbed");
+ $recptr = 0;
next;
}
+ $recptr++;
}
- # subfields can't be non-alphanumeric
- if ($line =~ /<subfield code="(.*?)"/) {
- my $match = $1;
- if ($match =~ /\P{IsAlnum}/ or $match eq '') {
- edit("Junk in subfield code/Null subfield code", $line);
- next;
+ # single-line fixes
+ for $recptr (0 .. $#record) {
+ # pad short leaders
+ if ($record[$recptr] =~ m|<leader>(.+?)</leader>|) {
+ my $leader = $1;
+ if (length $leader < 24) {
+ $leader .= ' ' x (20 - length($leader));
+ $leader .= "4500";
+ $record[$recptr] = "<leader>$leader</leader>\n";
+ message("Short leader padded");
+ }
+ }
+
+ # clean misplaced dollarsigns
+ if ($record[$recptr] =~ m|<subfield code="\$">c?\d+\.\d{2}|) {
+ $record[$recptr] =~ s|"\$">c?(\d+\.\d{2})|"c">\$$1|;
+ message("Dollar sign corrected");
}
- }
+ # clean up tags with spaces in them
+ $record[$recptr] =~ s/tag=" /tag="00/g;
+ $record[$recptr] =~ s/tag=" /tag="0/g;
+ $record[$recptr] =~ s/tag="-/tag="0/g;
+ $record[$recptr] =~ s/tag="(\d\d) /tag="0$1/g;
+
+ # stow tag data if we're looking at it
+ if ($record[$recptr] =~ m/<datafield tag="(.{3})" ind1="(.)" ind2="(.)">/) {
+ $recmeta{tag} = $1;
+ $recmeta{ind1} = $2;
+ $recmeta{ind2} = $3;
+ }
+
+ # automatable subfield maladies
+ $record[$recptr] =~ s/code=" ">c/code="c">/;
+ $record[$recptr] =~ s/code=" ">\$/code="c"$>/;
+ }
}
-print $NUMARC "</collection>\n";
-print $EXMARC "</collection>\n";
-print $OUT "\nDone. \n";
+
+#-----------------------------------------------------------------------------------
+# driver routines
+#-----------------------------------------------------------------------------------
+
=head2 edit
=cut
sub edit {
- my ($msg, $line_in) = @_;
- return if $trash{$recmeta{tag}};
+ my ($msg) = @_;
+ my $trash = $conf->{trash};
+
+ return if $trash->{ $recmeta{tag} };
message($msg);
print_context();
{ next unless (defined $commands{$chunks[0]}) }
if (defined $commands{$chunks[0]}) {
- my $term = $commands{$chunks[0]}->($line_in, @chunks[1..$#chunks]);
+ my $term = $commands{$chunks[0]}->(@chunks[1..$#chunks]);
last if $term;
} else {
if ($context[3] eq " [LINE KILLED]\n") {
}
}
-=head2 getline
+=head2 buildrecord
-Reads from the incoming MARC file; returns lines into the driver
-loop. Batches records for output, and maintains the context listing.
+Constructs record arrays from the incoming MARC file and returns them
+to the driver loop.
=cut
-sub getline {
+sub buildrecord {
my $l = <MARC>;
- $count++;
- if (defined $l) {
- if ($l =~ /<record>/) {
- @record = ($l);
- %recmeta = ();
- $reccount++;
- } elsif ($l =~ m|</record>|) {
- write_record($NUMARC) if $reccount;
- } else {
- push @record, $l;
- }
- }
- return $l;
+ return $l unless defined $l;
+
+ $l = <MARC> until ($l =~ /<record>/);
+ @record = ($l);
+ %recmeta = ();
+ $conf->{ricount}++;
+
+ until ($l =~ m|</record>|)
+ { push @record, $l; $l = <MARC>; }
+ push @record, $l;
+ return 1;
}
sub write_record {
my ($FH) = @_;
- $oreccount++ if ($FH eq $NUMARC);
+ my $trash = $conf->{trash};
+
+ $conf->{rocount}++ if ($FH eq $NUMARC);
print $FH '<!-- ', $recmeta{explanation}, " -->\n"
if(defined $recmeta{explanation});
# excise unwanted tags
- if (keys %trash or $conf->{autoscrub}) {
+ if (keys %{$trash} or $conf->{autoscrub}) {
my @trimmed = ();
my $istrash = 0;
for my $line (@record) {
}
if ($line =~ m/<datafield tag="(.{3})"/) {
my $tag = $1;
- if ($trash{$tag} or ($conf->{autoscrub} and $tag =~ /\D/)) {
+ if ($trash->{$tag} or ($conf->{autoscrub} and $tag =~ /\D/)) {
$istrash = 1;
next
}
print $FH "</record>\n";
}
-sub update_linecontext {
- my $line2 = <MARC2>;
- push @context, $line2;
- shift @context if (@context > 5);
-}
-
sub message {
my ($msg) = @_;
- print $OUT "\r$msg at record $reccount/",$oreccount + 1,"\n";
+ print $OUT "\r$msg at ",$conf->{ricount},"/",$conf->{rocount} + 1,"\n";
}
sub dump_record {
my ($line_in, @explanation) = @_;
$recmeta{explanation} = join(' ', 'Tag', $recmeta{tag}, @explanation);
- my $line = <MARC>; $count++;
- update_linecontext();
- until ($line =~ m|</record>|) {
- push @record, $line;
- $line = <MARC>; $count++;
- update_linecontext();
- }
- push @record, $line;
write_record($EXMARC);
return 1;
}
sub trash_add {
my ($tag, $except) = @_;
+ my $trash = $conf->{trash}
+
die "Trash values must be valid tags (000-999)\n"
unless ($tag >= 0 and $tag <= 999);
+
if ($except) {
- delete $trash{$tag};
+ delete $trash->{$tag};
} else {
die "Trash tag '$tag' specified twice (line $.)\n"
- if $trash{$tag};
- $trash{$tag} = 1;
+ if $trash->{$tag};
+ $trash->{$tag} = 1;
}
}