9 my $term = new Term::ReadLine 'yaz-cleanup';
10 my $OUT = $term->OUT || \*STDOUT;
12 # initialization and setup
15 populate_trash() if ($conf->{trash});
17 my @record = (); # current record storage
18 my %recmeta = (); # metadata about current record
19 my $ptr = 0; # record index pointer
21 my $input = shift || 'incoming.marc.xml';
23 open MARC, '<:utf8', $input;
24 open my $NUMARC, '>:utf8', $conf->{output};
25 print $NUMARC '<?xml version="1.0" encoding="UTF-8"?>',"\n";
26 print $NUMARC '<collection xmlns="http://www.loc.gov/MARC21/slim">',"\n";
28 open my $EXMARC, '>:utf8', $conf->{exception};
30 # this is the dispatch table which drives command selection in
32 my %commands = ( c => \&print_context,
33 C => \&print_linecontext,
35 O => \&insert_original,
51 my @spinner = qw(- / | \\);
54 while ( buildrecord() ) {
55 unless ($conf->{ricount} % 100) {
56 print "\rWorking... ", $spinner[$sidx];
57 $sidx = ($sidx == $#spinner) ? 0 : $sidx + 1;
60 do_automated_cleanups();
63 until ($ptr == $#record) {
65 if ($record[$ptr] =~ /&/ && $record[$ptr] !~ /&\w+?;/)
66 { edit("Naked ampersand"); $ptr= 0; next }
68 # tags must be numeric
69 if ($record[$ptr] =~ /<datafield tag="(.+?)"/) {
72 edit("Non-numerics in tag");
78 # subfields can't be non-alphanumeric
79 if ($record[$ptr] =~ /<subfield code="(.*?)"/) {
81 if ($match =~ /\P{IsAlnum}/ or $match eq '') {
82 edit("Junk in subfield code/Null subfield code");
89 write_record($NUMARC);
91 print $NUMARC "</collection>\n";
92 print $OUT "\nDone. \n";
95 #-----------------------------------------------------------------------------------
97 #-----------------------------------------------------------------------------------
99 sub do_automated_cleanups {
101 until ($ptr == $#record) {
102 # catch empty datafield elements
103 if ($record[$ptr] =~ m/<datafield tag="..." ind1="." ind2=".">/) {
104 if ($record[$ptr + 1] =~ m|</datafield>|) {
105 my @a = @record[0 .. $ptr - 1];
106 my @b = @record[$ptr + 2 .. $#record];
108 message("Empty datafield scrubbed");
113 # and quasi-empty subfields
114 if ($record[$ptr] =~ m|<subfield code="(.*?)">(.*?)</sub|) {
115 my $code = $1; my $content = $2;
116 if ($code =~ /\W/ and ($content =~ /\s+/ or $content eq '')) {
117 my @a = @record[0 .. $ptr - 1];
118 my @b = @record[$ptr + 1 .. $#record];
120 message("Empty subfield scrubbed");
129 for $ptr (0 .. $#record) {
131 if ($record[$ptr] =~ m|<leader>(.+?)</leader>|) {
133 if (length $leader < 24) {
134 $leader .= ' ' x (20 - length($leader));
136 $record[$ptr] = "<leader>$leader</leader>\n";
137 message("Short leader padded");
140 if ($record[$ptr] =~ m|<controlfield tag="008">(.+?)</control|) {
143 if (length $content < 40) {
144 $content .= ' ' x (40 - length($content));
145 $record[$ptr] = "<controlfield tag=\"008\">$content</controlfield>\n";
146 message("Short 008 padded");
150 # clean misplaced dollarsigns
151 if ($record[$ptr] =~ m|<subfield code="\$">c?\d+\.\d{2}|) {
152 $record[$ptr] =~ s|"\$">c?(\d+\.\d{2})|"c">\$$1|;
153 message("Dollar sign corrected");
156 # clean up tags with spaces in them
157 $record[$ptr] =~ s/tag=" /tag="00/g;
158 $record[$ptr] =~ s/tag=" /tag="0/g;
159 $record[$ptr] =~ s/tag="-/tag="0/g;
160 $record[$ptr] =~ s/tag="(\d\d) /tag="0$1/g;
162 # stow tag data if we're looking at it
163 if ($record[$ptr] =~ m/<datafield tag="(.{3})" ind1="(.)" ind2="(.)">/) {
169 # automatable subfield maladies
170 $record[$ptr] =~ s/code=" ">c/code="c">/;
171 $record[$ptr] =~ s/code=" ">\$/code="c"$>/;
175 #-----------------------------------------------------------------------------------
177 #-----------------------------------------------------------------------------------
182 Handles the Term::ReadLine loop
189 return if $conf->{trash}{ $recmeta{tag} };
193 # stow original problem line
194 $conf->{origline} = $record[$ptr];
197 my $line = $term->readline('marc-cleanup>');
198 my @chunks = split /\s+/, $line;
200 if (length $chunks[0] == 1)
201 { next unless (defined $commands{$chunks[0]}) }
203 if (defined $commands{$chunks[0]}) {
204 my $term = $commands{$chunks[0]}->(@chunks[1..$#chunks]);
207 $record[$ptr] = "$line\n";
215 Constructs record arrays from the incoming MARC file and returns them
222 $l = <MARC> while (defined $l and $l !~ /<record>/);
223 return $l unless defined $l;
228 until ($l =~ m|</record>|)
229 { push @record, $l; $l = <MARC>; }
236 my $trash = $conf->{trash};
238 $conf->{rocount}++ if ($FH eq $NUMARC);
239 print $FH '<!-- ', $recmeta{explanation}, " -->\n"
240 if(defined $recmeta{explanation});
242 # excise unwanted tags
243 if (keys %{$trash} or $conf->{autoscrub}) {
246 for my $line (@record) {
248 $istrash = 0 if $line =~ m|</datafield|;
251 if ($line =~ m/<datafield tag="(.{3})"/) {
253 if ($trash->{$tag} or ($conf->{autoscrub} and $tag =~ /\D/)) {
258 push @trimmed, $line;
264 unless ($conf->{nocollapse}) {
265 s/\n// for (@record);
268 # add 903(?) with new record id
270 if ($conf->{'renumber-from'}) {
271 $renumber = join('', '<datafield tag="', $conf->{'renumber-tag'},
272 '" ind1=" " ind2=" ">',
273 '<subfield code="', $conf->{'renumber-subfield'}, '">',
274 $conf->{'renumber-from'}, '</subfield></datafield>');
275 $renumber .= "\n" if $conf->{nocollapse};
276 push @record, $renumber;
277 $conf->{'renumber-from'}++;
281 print $FH "</record>\n";
285 print " Tag:",$recmeta{tag}, " Ind1:'",
286 $recmeta{ind1},"' Ind2:'", $recmeta{ind2}, "'\n";
291 sub print_linecontext {
292 my $low = ($ptr - 3 < 0) ? 0 : $ptr - 3;
293 print $OUT ' |', $record[$_] for ($low .. $ptr - 1);
294 print $OUT '==> |', $record[$ptr];
295 print $OUT ' |', $record[$ptr + 1], "\n";
300 my ($msg, $new) = @_;
301 print $OUT "\r", ' ' x 72, "\n" if $new;
302 print $OUT "\r$msg at ",$conf->{ricount},"/",$conf->{rocount} + 1,"\n";
306 #-----------------------------------------------------------------------------------
308 #-----------------------------------------------------------------------------------
311 my ($line_in, @chunks) = @_;
313 my $ofrom = shift @chunks;
314 if ($ofrom =~ /^'/ or !@chunks) {
315 until ($ofrom =~ /'$/)
316 { $ofrom .= join(' ','',shift @chunks) }
317 $ofrom =~ s/^'//; $ofrom =~ s/'$//;
319 my $to = shift @chunks;
321 until ($to =~ /'$/ or !@chunks)
322 { $to .= join(' ','',shift @chunks) }
323 $to =~ s/^'//; $to =~ s/'$//;
327 for my $char (split(//,$ofrom)) {
328 $char = "\\" . $char if ($char =~ /\W/);
329 $from = join('', $from, $char);
332 $conf->{prevline} = $record[$ptr];
333 $record[$ptr] =~ s/$from/$to/;
339 $conf->{prevline} = $record[$ptr];
340 # remove <subfield stuff; extract (probably wrong) subfield code
341 $record[$ptr] =~ s/^\s*<subfield code="(.*?)">//;
342 # and move to front of line
343 $record[$ptr] = join(' ', $1 , $record[$ptr]);
344 # tear off trailing subfield tag from preceeding line
345 $record[$ptr - 1] =~ s|</subfield>\n||;
346 # join current line onto preceeding line
347 $record[$ptr - 1] = join('', $record[$ptr - 1], $record[$ptr]);
349 my @a = @record[0 .. $ptr - 1];
350 my @b = @record[$ptr + 1 .. $#record];
352 # move record pointer to previous line
359 my $temp = $record[$ptr];
360 $record[$ptr] = $conf->{prevline};
361 $conf->{prevline} = $temp;
367 $conf->{prevline} = $record[$ptr];
368 my @a = @record[0 .. $ptr - 1];
369 my @b = @record[$ptr + 1 .. $#record];
376 my @a = @record[0 .. $ptr - 1];
377 my @b = @record[$ptr .. $#record];
378 @record = (@a, $conf->{prevline}, @b);
385 print $OUT "\n", $conf->{origline}, "\n";
390 my ($line_in, @explanation) = @_;
391 $recmeta{explanation} = join(' ', 'Tag', $recmeta{tag}, @explanation);
392 write_record($EXMARC);
397 $ptr++ unless ($ptr == $#record);;
403 $ptr-- unless ($ptr == 0);
408 sub commit_edit { return 1 }
413 Type a replacement for the indicated line, or enter a command.
415 Commands: c Show record context again ('C' for brief context)
416 k Kill indicated line (remove from record)
417 m Merge indicated line with previous line
419 s Substitute ARG1 for ARG2 in indicated line
420 t Commit changes and resume stream edit
421 x Write this record to the exception file instead of output
430 #-----------------------------------------------------------------------------------
432 #-----------------------------------------------------------------------------------
433 # defined a domain-specific language for specifying MARC tags to be dropped from
434 # records during processing. it is line oriented, and is specified as follows:
436 # each line may specify any number of tags to be included, either singly (\d{1,3})
437 # or as a range (\d{1,3}\.\.\d{1,3}
439 # if a single number is given, it must be between '000' and '999', inclusive.
441 # ranges obey the previous rule, and also the first number of the range must be less
442 # than the second number
444 # finally, any single range in a line may be followed by the keyword 'except'. every
445 # number or range after 'except' is excluded from the range specified. all these
446 # numbers must actually be within the range.
448 # specifying a tag twice is an error, to help prevent typos
451 print $OUT ">>> TRASHTAGS FILE FOUND. LOADING TAGS TO BE STRIPPED FROM OUTPUT\n";
452 open TRASH, '<', $conf->{trash}
453 or die "Can't open trash tags file!\n";
455 my $lastwasrange = 0;
456 my %lastrange = ( high => 0, low => 0);
459 my @chunks = split /\s+/;
460 while (my $chunk = shift @chunks) {
463 if ($chunk =~ /^\d{1,3}$/) {
464 trash_add($chunk, $except);
470 if ($chunk =~ /^\d{1,3}\.\.\d{1,3}$/) {
471 my ($low, $high) = trash_add_range($chunk, $except, \%lastrange);
473 %lastrange = (low => $low, high => $high)
479 if ($chunk eq 'except') {
480 die "Keyword 'except' can only follow a range (line $.)\n"
481 unless $lastwasrange;
482 die "Keyword 'except' may only occur once per line (line $.)\n"
488 die "Unknown chunk $chunk in .trashtags file (line $.)\n";
492 # remove original id sequence tag from trash hash if we know it
493 trash_add($conf->{'original-tag'}, 1)
494 if ($conf->{'original-tag'} and $conf->{trash}{ $conf->{'original-tag'} });
497 sub trash_add_range {
498 my ($chunk, $except, $range) = @_;
499 my ($low,$high) = split /\.\./, $chunk;
500 die "Ranges must be 'low..high' ($low is greater than $high on line $.)\n"
503 die "Exception ranges must be within last addition range (line $.)\n"
504 if ($low < $range->{low} or $high > $range->{high});
506 for my $tag ($low..$high) {
507 trash_add($tag, $except)
513 my ($tag, $except) = @_;
514 my $trash = $conf->{trash};
516 die "Trash values must be valid tags (000-999)\n"
517 unless ($tag >= 0 and $tag <= 999);
520 delete $trash->{$tag};
522 die "Trash tag '$tag' specified twice (line $.)\n"
528 #-----------------------------------------------------------------------
532 Performs boring script initialization. Handles argument parsing,
541 # set mode on existing filehandles
542 binmode(STDIN, ':utf8');
544 my $rc = GetOptions( $c,
549 'renumber-from|rf=i',
552 'renumber-subfield|rt=i',
556 show_help() unless $rc;
557 show_help() if ($c->{help});
560 $c->{output} = 'incoming.cleaned.marc.xml' unless defined $c->{output};
561 $c->{exception} = 'incoming.exception.marc.xml' unless defined $c->{exception};
562 $c->{'renumber-tag'} = 903 unless defined $c->{'renumber-tag'};
563 $c->{'renumber-subfield'} = 'a' unless defined $c->{'renumber-subfield'};
565 my @keys = keys %{$c};
566 show_help() unless (@ARGV and @keys);
567 #for my $key ('runtype', 'tag', 'subfield', 'output', 'exception')
568 # { push @missing, $key unless $c->{$key} }
570 # print "Required option: ", join(', ', @missing), " missing!\n";
577 Usage is: $0 [OPTIONS] <filelist>
579 --output -o Cleaned MARCXML output filename (default: incoming.cleaned.marc.xml)
580 --exception -x Exception (dumped records) MARCXML filename (incoming.exception.marc.xml)