11 my $term = new Term::ReadLine 'yaz-cleanup';
12 my $OUT = $term->OUT || \*STDOUT;
20 my %trash = (); # hash for tags to be dumped
22 # initialization and setup
25 # read in trash tags file if it exists
26 populate_trash() if (-e '.trashtags');
28 my @record = (); # current record storage
29 my %recmeta = (); # metadata about current record
30 my @context= (); # last 5 lines of file
32 my $input = shift || 'incoming.marc.xml';
34 open MARC, '<', $input;
35 open my $NUMARC, '>', $conf->{output};
36 print $NUMARC '<?xml version="1.0" encoding="UTF-8"?>',"\n";
37 print $NUMARC '<collection xmlns="http://www.loc.gov/MARC21/slim">',"\n";
39 open my $EXMARC, '>', $conf->{exception};
40 print $EXMARC '<?xml version="1.0" encoding="UTF-8"?>',"\n";
41 print $EXMARC '<collection xmlns="http://www.loc.gov/MARC21/slim">',"\n";
42 open MARC2, '<', $input;
45 # this is the dispatch table which drives command selection in
47 my %commands = ( c => \&print_context,
48 C => \&print_linecontext,
61 my @spinner = qw(- / | \\);
64 while (my $line = getline()) {
65 unless ($count % 2000) {
66 print "\rWorking... ", $spinner[$sidx];
67 $sidx = ($sidx == $#spinner) ? 0 : $sidx + 1;
71 next if ($line =~ m|</record>|);
73 # catch empty datafield elements
74 if ($line =~ m|</datafield>|) {
75 if ($record[-2] =~ m/<datafield tag="..." ind1="." ind2=".">/) {
76 pop @record; pop @record;
77 message("Empty datafield scrubbed");
83 if ($line =~ m|<leader>(.+?)</leader>|) {
85 if (length $leader < 24) {
86 $leader .= ' ' x (20 - length($leader));
88 $line = "<leader>$leader</leader>\n";
89 message("Short leader padded");
93 # clean misplaced dollarsigns
94 if ($line =~ m|<subfield code="\$">c?\d+\.\d{2}|) {
95 $line =~ s|"\$">c?(\d+\.\d{2})|"c">\$$1|;
96 message("Dollar sign corrected");
99 # clean up tags with spaces in them
100 $line =~ s/tag=" /tag="00/g;
101 $line =~ s/tag=" /tag="0/g;
102 $line =~ s/tag="-/tag="0/g;
103 $line =~ s/tag="(\d\d) /tag="0$1/g;
105 # stow tag data if we're looking at it
106 if ($line =~ m/<datafield tag="(.{3})" ind1="(.)" ind2="(.)">/) {
112 # and stow line back in record
116 if ($line =~ /&/ && $line !~ /&\w+?;/)
117 { edit("Naked ampersand", $line); next }
119 # tags must be numeric
120 if ($line =~ /<datafield tag="(.+?)"/) {
122 if ($match =~ /\D/) {
123 edit("Non-numerics in tag", $line);
128 # subfields can't be non-alphanumeric
129 if ($line =~ /<subfield code="(.+?)"/) {
131 if ($match =~ /\P{IsAlnum}/) {
132 edit("Junk in subfield code", $line);
138 print $NUMARC "</collection>\n";
139 print $EXMARC "</collection>\n";
140 print $OUT "\nDone. \n";
144 Handles the Term::ReadLine loop
149 my ($msg, $line_in) = @_;
150 return if $trash{$recmeta{tag}};
155 my $line = $term->readline('marc-cleanup>');
156 my @chunks = split /\s+/, $line;
158 if (length $chunks[0] == 1)
159 { next unless (defined $commands{$chunks[0]}) }
161 if (defined $commands{$chunks[0]}) {
162 my $term = $commands{$chunks[0]}->($line_in, @chunks[1..$#chunks]);
165 if ($context[3] eq " [LINE KILLED]\n") {
166 push @record, "$line\n"
168 $record[-1] = "$line\n";
170 $context[3] = "$line\n";
178 Reads from the incoming MARC file; returns lines into the driver
179 loop. Batches records for output, and maintains the context listing.
187 if ($l =~ /<record>/) {
191 } elsif ($l =~ m|</record>|) {
192 write_record($NUMARC) if $reccount;
202 $oreccount++ if ($FH eq $NUMARC);
203 print $FH '<!-- ', $recmeta{explanation}, " -->\n"
204 if(defined $recmeta{explanation});
206 # LOOP OVER %trash TO EXCISE UNWANTED TAGS1
210 for my $line (@record) {
212 $istrash = 0 if $line =~ m|</datafield|;
215 if ($line =~ m/<datafield tag="(.{3})"/) {
217 if ($trash{$tag} or ($conf->{autoscrub} and $tag =~ /\D/)) {
222 push @trimmed, $line;
228 unless ($conf->{nocollapse}) {
229 s/\n// for (@record);
232 # add 903(?) with new record id
233 if ($conf->{'renumber-from'}) {
234 print $FH '<datafield tag="', $conf->{'renumber-tag'}, '"><subfield code="a">',
235 $conf->{'renumber-from'}, '</subfield></datafield>';
236 print $FH "\n" unless $conf->{oneperline};
237 $conf->{'renumber-from'}++;
241 print $FH '</record>\n';
244 sub update_linecontext {
246 push @context, $line2;
247 shift @context if (@context > 5);
252 print $OUT "\r$msg at record $reccount/",$oreccount + 1,"\n";
256 #-----------------------------------------------------------------------------------
258 #-----------------------------------------------------------------------------------
261 my ($line_in, @chunks) = @_;
262 my $ofrom = shift @chunks;
263 if ($ofrom =~ /^'/ or !@chunks) {
264 until ($ofrom =~ /'$/)
265 { $ofrom .= join(' ','',shift @chunks) }
266 $ofrom =~ s/^'//; $ofrom =~ s/'$//;
268 my $to = shift @chunks;
270 until ($to =~ /'$/ or !@chunks)
271 { $to .= join(' ','',shift @chunks) }
272 $to =~ s/^'//; $to =~ s/'$//;
276 for my $char (split(//,$ofrom)) {
277 $char = "\\" . $char if ($char =~ /\W/);
278 $from = join('', $from, $char);
280 $record[-1] =~ s/$from/$to/;
281 $context[3] = $record[-1];
287 my $last = pop @record;
289 $record[-1] =~ s/\n//;
290 $record[-1] = join('', $record[-1], $last);
292 push @temp, @context[0..1];
293 $temp[3] = $record[-1];
294 $temp[4] = $context[4];
302 $context[3] = " [LINE KILLED]\n";
308 my ($line_in, @explanation) = @_;
309 $recmeta{explanation} = join(' ', 'Tag', $recmeta{tag}, @explanation);
310 my $line = <MARC>; $count++;
311 update_linecontext();
312 until ($line =~ m|</record>|) {
314 $line = <MARC>; $count++;
315 update_linecontext();
318 write_record($EXMARC);
322 sub commit_edit { return 1 }
325 print "\n Tag:",$recmeta{tag}, " Ind1:'",
326 $recmeta{ind1},"' Ind2:'", $recmeta{ind2}, "'";
331 sub print_linecontext {
332 print $OUT "\n", join(' |','',@context[0..2]);
333 print $OUT '==> |', $context[3];
334 print $OUT ' |', $context[4],"\n";
340 print $OUT "\n$line_in\n";
347 Type a replacement for the indicated line, or enter a command.
349 Commands: c Show record context ('C' for brief context)
350 k Kill indicated line (remove from record)
351 m Merge indicated line with previous line
353 s Substitute ARG1 for ARG2 in indicated line
354 t Commit changes and resume stream edit
355 x Write this record to the exception file instead of output
364 #-----------------------------------------------------------------------------------
366 #-----------------------------------------------------------------------------------
367 # defined a domain-specific language for specifying MARC tags to be dropped from
368 # records during processing. it is line oriented, and is specified as follows:
370 # each line may specify any number of tags to be included, either singly (\d{1,3})
371 # or as a range (\d{1,3}\.\.\d{1,3}
373 # if a single number is given, it must be between '000' and '999', inclusive.
375 # ranges obey the previous rule, and also the first number of the range must be less
376 # than the second number
378 # finally, any single range in a line may be followed by the keyword 'except'. every
379 # number or range after 'except' is excluded from the range specified. all these
380 # numbers must actually be within the range.
382 # specifying a tag twice is an error, to help prevent typos
385 print $OUT ">>> TRASHTAGS FILE FOUND. LOADING TAGS TO BE STRIPPED FROM OUTPUT...\n";
386 open TRASH, '<', '.trashtags';
388 my $lastwasrange = 0;
389 my %lastrange = ( high => 0, low => 0);
392 my @chunks = split /\s+/;
393 while (my $chunk = shift @chunks) {
396 if ($chunk =~ /^\d{1,3}$/) {
397 trash_add($chunk, $except);
403 if ($chunk =~ /^\d{1,3}\.\.\d{1,3}$/) {
404 my ($low, $high) = trash_add_range($chunk, $except, \%lastrange);
406 %lastrange = (low => $low, high => $high)
412 if ($chunk eq 'except') {
413 die "Keyword 'except' can only follow a range (line $.)\n"
414 unless $lastwasrange;
415 die "Keyword 'except' may only occur once per line (line $.)\n"
421 die "Unknown chunk $chunk in .trashtags file (line $.)\n";
425 # remove original id sequence tag from trash hash if we know it
426 trash_add($conf->{'original-tag'}, 1)
427 if ($conf->{'original-tag'} and $trash{$conf->{'original-tag'}});
430 sub trash_add_range {
431 my ($chunk, $except, $range) = @_;
432 my ($low,$high) = split /\.\./, $chunk;
433 die "Ranges must be 'low..high' ($low is greater than $high on line $.)\n"
436 die "Exception ranges must be within last addition range (line $.)\n"
437 if ($low < $range->{low} or $high > $range->{high});
439 for my $tag ($low..$high) {
440 trash_add($tag, $except)
446 my ($tag, $except) = @_;
447 die "Trash values must be valid tags (000-999)\n"
448 unless ($tag >= 0 and $tag <= 999);
452 die "Trash tag '$tag' specified twice (line $.)\n"
458 #-----------------------------------------------------------------------
462 Performs boring script initialization. Handles argument parsing,
471 # set mode on existing filehandles
472 binmode(STDIN, ':utf8');
474 my $rc = GetOptions( $c,
479 'renumber-from|rf=i',
484 show_help() unless $rc;
485 show_help() if ($c->{help});
488 $c->{output} = 'incoming.cleaned.marc.xml' unless defined $c->{output};
489 $c->{exception} = 'incoming.exception.marc.xml' unless defined $c->{exception};
490 $c->{'renumber-tag'} = 903 unless defined $c->{'renumber-tag'};
492 my @keys = keys %{$c};
493 show_help() unless (@ARGV and @keys);
494 #for my $key ('runtype', 'tag', 'subfield', 'output', 'exception')
495 # { push @missing, $key unless $c->{$key} }
497 # print "Required option: ", join(', ', @missing), " missing!\n";
504 Usage is: $0 [OPTIONS] <filelist>
506 --output -o Cleaned MARCXML output filename (default: incoming.cleaned.marc.xml)
507 --exception -x Exception (dumped records) MARCXML filename (incoming.exception.marc.xml)