1 |
package WebPAC::Normalize; |
2 |
use Exporter 'import'; |
3 |
@EXPORT = qw/ |
4 |
_set_rec _set_lookup |
5 |
_get_ds _clean_ds |
6 |
_debug |
7 |
|
8 |
tag search display |
9 |
marc marc_indicators marc_repeatable_subfield |
10 |
|
11 |
rec1 rec2 rec |
12 |
regex prefix suffix surround |
13 |
first lookup join_with |
14 |
/; |
15 |
|
16 |
use warnings; |
17 |
use strict; |
18 |
|
19 |
#use base qw/WebPAC::Common/; |
20 |
use Data::Dump qw/dump/; |
21 |
use Encode qw/from_to/; |
22 |
|
23 |
# debugging warn(s) |
24 |
my $debug = 0; |
25 |
|
26 |
|
27 |
=head1 NAME |
28 |
|
29 |
WebPAC::Normalize - describe normalisaton rules using sets |
30 |
|
31 |
=head1 VERSION |
32 |
|
33 |
Version 0.07 |
34 |
|
35 |
=cut |
36 |
|
37 |
our $VERSION = '0.07'; |
38 |
|
39 |
=head1 SYNOPSIS |
40 |
|
41 |
This module uses C<conf/normalize/*.pl> files to perform normalisation |
42 |
from input records using perl functions which are specialized for set |
43 |
processing. |
44 |
|
45 |
Sets are implemented as arrays, and normalisation file is valid perl, which |
46 |
means that you check it's validity before running WebPAC using |
47 |
C<perl -c normalize.pl>. |
48 |
|
49 |
Normalisation can generate multiple output normalized data. For now, supported output |
50 |
types (on the left side of definition) are: C<tag>, C<display>, C<search> and |
51 |
C<marc>. |
52 |
|
53 |
=head1 FUNCTIONS |
54 |
|
55 |
Functions which start with C<_> are private and used by WebPAC internally. |
56 |
All other functions are available for use within normalisation rules. |
57 |
|
58 |
=head2 data_structure |
59 |
|
60 |
Return data structure |
61 |
|
62 |
my $ds = WebPAC::Normalize::data_structure( |
63 |
lookup => $lookup->lookup_hash, |
64 |
row => $row, |
65 |
rules => $normalize_pl_config, |
66 |
marc_encoding => 'utf-8', |
67 |
); |
68 |
|
69 |
Options C<lookup>, C<row>, C<rules> and C<log> are mandatory while all |
70 |
other are optional. |
71 |
|
72 |
This function will B<die> if normalizastion can't be evaled. |
73 |
|
74 |
Since this function isn't exported you have to call it with |
75 |
C<WebPAC::Normalize::data_structure>. |
76 |
|
77 |
=cut |
78 |
|
79 |
sub data_structure { |
80 |
my $arg = {@_}; |
81 |
|
82 |
die "need row argument" unless ($arg->{row}); |
83 |
die "need normalisation argument" unless ($arg->{rules}); |
84 |
|
85 |
no strict 'subs'; |
86 |
_set_lookup( $arg->{lookup} ); |
87 |
_set_rec( $arg->{row} ); |
88 |
_clean_ds( %{ $arg } ); |
89 |
eval "$arg->{rules}"; |
90 |
die "error evaling $arg->{rules}: $@\n" if ($@); |
91 |
|
92 |
return _get_ds(); |
93 |
} |
94 |
|
95 |
=head2 _set_rec |
96 |
|
97 |
Set current record hash |
98 |
|
99 |
_set_rec( $rec ); |
100 |
|
101 |
=cut |
102 |
|
103 |
my $rec; |
104 |
|
105 |
sub _set_rec { |
106 |
$rec = shift or die "no record hash"; |
107 |
} |
108 |
|
109 |
=head2 _get_ds |
110 |
|
111 |
Return hash formatted as data structure |
112 |
|
113 |
my $ds = _get_ds(); |
114 |
|
115 |
=cut |
116 |
|
117 |
my ($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators); |
118 |
|
119 |
sub _get_ds { |
120 |
return $out; |
121 |
} |
122 |
|
123 |
=head2 _clean_ds |
124 |
|
125 |
Clean data structure hash for next record |
126 |
|
127 |
_clean_ds(); |
128 |
|
129 |
=cut |
130 |
|
131 |
sub _clean_ds { |
132 |
my $a = {@_}; |
133 |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators) = (); |
134 |
$marc_encoding = $a->{marc_encoding}; |
135 |
} |
136 |
|
137 |
=head2 _set_lookup |
138 |
|
139 |
Set current lookup hash |
140 |
|
141 |
_set_lookup( $lookup ); |
142 |
|
143 |
=cut |
144 |
|
145 |
my $lookup; |
146 |
|
147 |
sub _set_lookup { |
148 |
$lookup = shift; |
149 |
} |
150 |
|
151 |
=head2 _get_marc_fields |
152 |
|
153 |
Get all fields defined by calls to C<marc> |
154 |
|
155 |
$marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); |
156 |
|
157 |
We are using I<magic> which detect repeatable fields only from |
158 |
sequence of field/subfield data generated by normalization. |
159 |
|
160 |
Repeatable field is created when there is second occurence of same subfield or |
161 |
if any of indicators are different. |
162 |
|
163 |
This is sane for most cases. Something like: |
164 |
|
165 |
900a-1 900b-1 900c-1 |
166 |
900a-2 900b-2 |
167 |
900a-3 |
168 |
|
169 |
will be created from any combination of: |
170 |
|
171 |
900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 |
172 |
|
173 |
and following rules: |
174 |
|
175 |
marc('900','a', rec('200','a') ); |
176 |
marc('900','b', rec('200','b') ); |
177 |
marc('900','c', rec('200','c') ); |
178 |
|
179 |
which might not be what you have in mind. If you need repeatable subfield, |
180 |
define it using C<marc_repeatable_subfield> like this: |
181 |
|
182 |
.... |
183 |
|
184 |
=cut |
185 |
|
186 |
sub _get_marc_fields { |
187 |
|
188 |
return if (! $marc_record || ref($marc_record) ne 'ARRAY' || $#{ $marc_record } < 0); |
189 |
|
190 |
# first, sort all existing fields |
191 |
# XXX might not be needed, but modern perl might randomize elements in hash |
192 |
my @sorted_marc_record = sort { |
193 |
$a->[0] . $a->[3] cmp $b->[0] . $b->[3] |
194 |
} @{ $marc_record }; |
195 |
|
196 |
# output marc fields |
197 |
my @m; |
198 |
|
199 |
# count unique field-subfields (used for offset when walking to next subfield) |
200 |
my $u; |
201 |
map { $u->{ $_->[0] . $_->[3] }++ } @sorted_marc_record; |
202 |
|
203 |
if ($debug) { |
204 |
warn "## marc_repeatable_subfield ", dump( $marc_repeatable_subfield ), $/; |
205 |
warn "## marc_record ", dump( $marc_record ), $/; |
206 |
warn "## sorted_marc_record ", dump( \@sorted_marc_record ), $/; |
207 |
warn "## subfield count ", dump( $u ), $/; |
208 |
} |
209 |
|
210 |
my $len = $#sorted_marc_record; |
211 |
my $visited; |
212 |
my $i = 0; |
213 |
my $field; |
214 |
|
215 |
foreach ( 0 .. $len ) { |
216 |
|
217 |
# find next element which isn't visited |
218 |
while ($visited->{$i}) { |
219 |
$i = ($i + 1) % ($len + 1); |
220 |
} |
221 |
|
222 |
# mark it visited |
223 |
$visited->{$i}++; |
224 |
|
225 |
my $row = $sorted_marc_record[$i]; |
226 |
|
227 |
# field and subfield which is key for |
228 |
# marc_repeatable_subfield and u |
229 |
my $fsf = $row->[0] . $row->[3]; |
230 |
|
231 |
if ($debug > 1) { |
232 |
|
233 |
print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; |
234 |
print "### this [$i]: ", dump( $row ),$/; |
235 |
print "### sf: ", $row->[3], " vs ", $field->[3], |
236 |
$marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, |
237 |
if ($#$field >= 0); |
238 |
|
239 |
} |
240 |
|
241 |
# if field exists |
242 |
if ( $#$field >= 0 ) { |
243 |
if ( |
244 |
$row->[0] ne $field->[0] || # field |
245 |
$row->[1] ne $field->[1] || # i1 |
246 |
$row->[2] ne $field->[2] # i2 |
247 |
) { |
248 |
push @m, $field; |
249 |
warn "## saved/1 ", dump( $field ),$/ if ($debug); |
250 |
$field = $row; |
251 |
|
252 |
} elsif ( |
253 |
( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) |
254 |
|| |
255 |
( $row->[3] eq $field->[-2] && # same subfield, but not repeatable |
256 |
! $marc_repeatable_subfield->{ $fsf } |
257 |
) |
258 |
) { |
259 |
push @m, $field; |
260 |
warn "## saved/2 ", dump( $field ),$/ if ($debug); |
261 |
$field = $row; |
262 |
|
263 |
} else { |
264 |
# append new subfields to existing field |
265 |
push @$field, ( $row->[3], $row->[4] ); |
266 |
} |
267 |
} else { |
268 |
# insert first field |
269 |
$field = $row; |
270 |
} |
271 |
|
272 |
if (! $marc_repeatable_subfield->{ $fsf }) { |
273 |
# make step to next subfield |
274 |
$i = ($i + $u->{ $fsf } ) % ($len + 1); |
275 |
} |
276 |
} |
277 |
|
278 |
if ($#$field >= 0) { |
279 |
push @m, $field; |
280 |
warn "## saved/3 ", dump( $field ),$/ if ($debug); |
281 |
} |
282 |
|
283 |
return @m; |
284 |
} |
285 |
|
286 |
=head2 _debug |
287 |
|
288 |
Change level of debug warnings |
289 |
|
290 |
_debug( 2 ); |
291 |
|
292 |
=cut |
293 |
|
294 |
sub _debug { |
295 |
my $l = shift; |
296 |
return $debug unless defined($l); |
297 |
$debug = $l; |
298 |
} |
299 |
|
300 |
=head1 Functions to create C<data_structure> |
301 |
|
302 |
Those functions generally have to first in your normalization file. |
303 |
|
304 |
=head2 tag |
305 |
|
306 |
Define new tag for I<search> and I<display>. |
307 |
|
308 |
tag('Title', rec('200','a') ); |
309 |
|
310 |
|
311 |
=cut |
312 |
|
313 |
sub tag { |
314 |
my $name = shift or die "tag needs name as first argument"; |
315 |
my @o = grep { defined($_) && $_ ne '' } @_; |
316 |
return unless (@o); |
317 |
$out->{$name}->{tag} = $name; |
318 |
$out->{$name}->{search} = \@o; |
319 |
$out->{$name}->{display} = \@o; |
320 |
} |
321 |
|
322 |
=head2 display |
323 |
|
324 |
Define tag just for I<display> |
325 |
|
326 |
@v = display('Title', rec('200','a') ); |
327 |
|
328 |
=cut |
329 |
|
330 |
sub display { |
331 |
my $name = shift or die "display needs name as first argument"; |
332 |
my @o = grep { defined($_) && $_ ne '' } @_; |
333 |
return unless (@o); |
334 |
$out->{$name}->{tag} = $name; |
335 |
$out->{$name}->{display} = \@o; |
336 |
} |
337 |
|
338 |
=head2 search |
339 |
|
340 |
Prepare values just for I<search> |
341 |
|
342 |
@v = search('Title', rec('200','a') ); |
343 |
|
344 |
=cut |
345 |
|
346 |
sub search { |
347 |
my $name = shift or die "search needs name as first argument"; |
348 |
my @o = grep { defined($_) && $_ ne '' } @_; |
349 |
return unless (@o); |
350 |
$out->{$name}->{tag} = $name; |
351 |
$out->{$name}->{search} = \@o; |
352 |
} |
353 |
|
354 |
=head2 marc |
355 |
|
356 |
Save value for MARC field |
357 |
|
358 |
marc('900','a', rec('200','a') ); |
359 |
|
360 |
=cut |
361 |
|
362 |
sub marc { |
363 |
my $f = shift or die "marc needs field"; |
364 |
die "marc field must be numer" unless ($f =~ /^\d+$/); |
365 |
|
366 |
my $sf = shift or die "marc needs subfield"; |
367 |
|
368 |
foreach (@_) { |
369 |
my $v = $_; # make var read-write for Encode |
370 |
next unless (defined($v) && $v !~ /^\s*$/); |
371 |
from_to($v, 'iso-8859-2', $marc_encoding) if ($marc_encoding); |
372 |
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
373 |
push @{ $marc_record }, [ $f, $i1, $i2, $sf => $v ]; |
374 |
} |
375 |
} |
376 |
|
377 |
=head2 marc_repeatable_subfield |
378 |
|
379 |
Save values for MARC repetable subfield |
380 |
|
381 |
marc_repeatable_subfield('910', 'z', rec('909') ); |
382 |
|
383 |
=cut |
384 |
|
385 |
sub marc_repeatable_subfield { |
386 |
my ($f,$sf) = @_; |
387 |
die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); |
388 |
$marc_repeatable_subfield->{ $f . $sf }++; |
389 |
marc(@_); |
390 |
} |
391 |
|
392 |
=head2 marc_indicators |
393 |
|
394 |
Set both indicators for MARC field |
395 |
|
396 |
marc_indicators('900', ' ', 1); |
397 |
|
398 |
Any indicator value other than C<0-9> will be treated as undefined. |
399 |
|
400 |
=cut |
401 |
|
402 |
sub marc_indicators { |
403 |
my $f = shift || die "marc_indicators need field!\n"; |
404 |
my ($i1,$i2) = @_; |
405 |
die "marc_indicators($f, ...) need i1!\n" unless(defined($i1)); |
406 |
die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2)); |
407 |
|
408 |
$i1 = ' ' if ($i1 !~ /^\d$/); |
409 |
$i2 = ' ' if ($i2 !~ /^\d$/); |
410 |
@{ $marc_indicators->{$f} } = ($i1,$i2); |
411 |
} |
412 |
|
413 |
|
414 |
=head1 Functions to extract data from input |
415 |
|
416 |
This function should be used inside functions to create C<data_structure> described |
417 |
above. |
418 |
|
419 |
=head2 rec1 |
420 |
|
421 |
Return all values in some field |
422 |
|
423 |
@v = rec1('200') |
424 |
|
425 |
TODO: order of values is probably same as in source data, need to investigate that |
426 |
|
427 |
=cut |
428 |
|
429 |
sub rec1 { |
430 |
my $f = shift; |
431 |
return unless (defined($rec) && defined($rec->{$f})); |
432 |
if (ref($rec->{$f}) eq 'ARRAY') { |
433 |
return map { |
434 |
if (ref($_) eq 'HASH') { |
435 |
values %{$_}; |
436 |
} else { |
437 |
$_; |
438 |
} |
439 |
} @{ $rec->{$f} }; |
440 |
} elsif( defined($rec->{$f}) ) { |
441 |
return $rec->{$f}; |
442 |
} |
443 |
} |
444 |
|
445 |
=head2 rec2 |
446 |
|
447 |
Return all values in specific field and subfield |
448 |
|
449 |
@v = rec2('200','a') |
450 |
|
451 |
=cut |
452 |
|
453 |
sub rec2 { |
454 |
my $f = shift; |
455 |
return unless (defined($rec && $rec->{$f})); |
456 |
my $sf = shift; |
457 |
return map { $_->{$sf} } grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} }; |
458 |
} |
459 |
|
460 |
=head2 rec |
461 |
|
462 |
syntaxtic sugar for |
463 |
|
464 |
@v = rec('200') |
465 |
@v = rec('200','a') |
466 |
|
467 |
=cut |
468 |
|
469 |
sub rec { |
470 |
if ($#_ == 0) { |
471 |
return rec1(@_); |
472 |
} elsif ($#_ == 1) { |
473 |
return rec2(@_); |
474 |
} |
475 |
} |
476 |
|
477 |
=head2 regex |
478 |
|
479 |
Apply regex to some or all values |
480 |
|
481 |
@v = regex( 's/foo/bar/g', @v ); |
482 |
|
483 |
=cut |
484 |
|
485 |
sub regex { |
486 |
my $r = shift; |
487 |
my @out; |
488 |
#warn "r: $r\n", dump(\@_); |
489 |
foreach my $t (@_) { |
490 |
next unless ($t); |
491 |
eval "\$t =~ $r"; |
492 |
push @out, $t if ($t && $t ne ''); |
493 |
} |
494 |
return @out; |
495 |
} |
496 |
|
497 |
=head2 prefix |
498 |
|
499 |
Prefix all values with a string |
500 |
|
501 |
@v = prefix( 'my_', @v ); |
502 |
|
503 |
=cut |
504 |
|
505 |
sub prefix { |
506 |
my $p = shift or die "prefix needs string as first argument"; |
507 |
return map { $p . $_ } grep { defined($_) } @_; |
508 |
} |
509 |
|
510 |
=head2 suffix |
511 |
|
512 |
suffix all values with a string |
513 |
|
514 |
@v = suffix( '_my', @v ); |
515 |
|
516 |
=cut |
517 |
|
518 |
sub suffix { |
519 |
my $s = shift or die "suffix needs string as first argument"; |
520 |
return map { $_ . $s } grep { defined($_) } @_; |
521 |
} |
522 |
|
523 |
=head2 surround |
524 |
|
525 |
surround all values with a two strings |
526 |
|
527 |
@v = surround( 'prefix_', '_suffix', @v ); |
528 |
|
529 |
=cut |
530 |
|
531 |
sub surround { |
532 |
my $p = shift or die "surround need prefix as first argument"; |
533 |
my $s = shift or die "surround needs suffix as second argument"; |
534 |
return map { $p . $_ . $s } grep { defined($_) } @_; |
535 |
} |
536 |
|
537 |
=head2 first |
538 |
|
539 |
Return first element |
540 |
|
541 |
$v = first( @v ); |
542 |
|
543 |
=cut |
544 |
|
545 |
sub first { |
546 |
my $r = shift; |
547 |
return $r; |
548 |
} |
549 |
|
550 |
=head2 lookup |
551 |
|
552 |
Consult lookup hashes for some value |
553 |
|
554 |
@v = lookup( $v ); |
555 |
@v = lookup( @v ); |
556 |
|
557 |
=cut |
558 |
|
559 |
sub lookup { |
560 |
my $k = shift or return; |
561 |
return unless (defined($lookup->{$k})); |
562 |
if (ref($lookup->{$k}) eq 'ARRAY') { |
563 |
return @{ $lookup->{$k} }; |
564 |
} else { |
565 |
return $lookup->{$k}; |
566 |
} |
567 |
} |
568 |
|
569 |
=head2 join_with |
570 |
|
571 |
Joins walues with some delimiter |
572 |
|
573 |
$v = join_with(", ", @v); |
574 |
|
575 |
=cut |
576 |
|
577 |
sub join_with { |
578 |
my $d = shift; |
579 |
return join($d, grep { defined($_) && $_ ne '' } @_); |
580 |
} |
581 |
|
582 |
# END |
583 |
1; |