1 |
package WebPAC::Normalize; |
package WebPAC::Normalize; |
2 |
|
use Exporter 'import'; |
3 |
|
our @EXPORT = qw/ |
4 |
|
_set_ds _set_lookup |
5 |
|
_set_load_row |
6 |
|
_get_ds _clean_ds |
7 |
|
_debug |
8 |
|
_pack_subfields_hash |
9 |
|
|
10 |
|
search_display search display sorted |
11 |
|
|
12 |
|
marc marc_indicators marc_repeatable_subfield |
13 |
|
marc_compose marc_leader marc_fixed |
14 |
|
marc_duplicate marc_remove marc_count |
15 |
|
marc_original_order |
16 |
|
|
17 |
|
rec1 rec2 rec |
18 |
|
frec |
19 |
|
regex prefix suffix surround |
20 |
|
first lookup join_with |
21 |
|
save_into_lookup |
22 |
|
|
23 |
|
split_rec_on |
24 |
|
|
25 |
|
get set |
26 |
|
count |
27 |
|
|
28 |
|
/; |
29 |
|
|
30 |
use warnings; |
use warnings; |
31 |
use strict; |
use strict; |
|
use base 'WebPAC::Common'; |
|
|
use Data::Dumper; |
|
32 |
|
|
33 |
=head1 NAME |
#use base qw/WebPAC::Common/; |
34 |
|
use Data::Dump qw/dump/; |
35 |
|
use Storable qw/dclone/; |
36 |
|
use Carp qw/confess/; |
37 |
|
|
38 |
|
# debugging warn(s) |
39 |
|
my $debug = 0; |
40 |
|
|
41 |
WebPAC::Normalize - data mungling for normalisation |
use WebPAC::Normalize::ISBN; |
42 |
|
push @EXPORT, ( 'isbn_10', 'isbn_13' ); |
43 |
|
|
44 |
=head1 VERSION |
=head1 NAME |
45 |
|
|
46 |
Version 0.04 |
WebPAC::Normalize - describe normalisaton rules using sets |
47 |
|
|
48 |
=cut |
=cut |
49 |
|
|
50 |
our $VERSION = '0.04'; |
our $VERSION = '0.32'; |
51 |
|
|
52 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
53 |
|
|
54 |
This package contains code that mungle data to produce normalized format. |
This module uses C<conf/normalize/*.pl> files to perform normalisation |
55 |
|
from input records using perl functions which are specialized for set |
56 |
|
processing. |
57 |
|
|
58 |
|
Sets are implemented as arrays, and normalisation file is valid perl, which |
59 |
|
means that you check it's validity before running WebPAC using |
60 |
|
C<perl -c normalize.pl>. |
61 |
|
|
62 |
|
Normalisation can generate multiple output normalized data. For now, supported output |
63 |
|
types (on the left side of definition) are: C<search_display>, C<display>, C<search> and |
64 |
|
C<marc>. |
65 |
|
|
66 |
|
=head1 FUNCTIONS |
67 |
|
|
68 |
|
Functions which start with C<_> are private and used by WebPAC internally. |
69 |
|
All other functions are available for use within normalisation rules. |
70 |
|
|
71 |
|
=head2 data_structure |
72 |
|
|
73 |
|
Return data structure |
74 |
|
|
75 |
|
my $ds = WebPAC::Normalize::data_structure( |
76 |
|
lookup => $lookup_hash, |
77 |
|
row => $row, |
78 |
|
rules => $normalize_pl_config, |
79 |
|
marc_encoding => 'utf-8', |
80 |
|
config => $config, |
81 |
|
load_row_coderef => sub { |
82 |
|
my ($database,$input,$mfn) = @_; |
83 |
|
$store->load_row( database => $database, input => $input, id => $mfn ); |
84 |
|
}, |
85 |
|
); |
86 |
|
|
87 |
|
Options C<row>, C<rules> and C<log> are mandatory while all |
88 |
|
other are optional. |
89 |
|
|
90 |
|
C<load_row_coderef> is closure only used when executing lookups, so they will |
91 |
|
die if it's not defined. |
92 |
|
|
93 |
|
This function will B<die> if normalizastion can't be evaled. |
94 |
|
|
95 |
|
Since this function isn't exported you have to call it with |
96 |
|
C<WebPAC::Normalize::data_structure>. |
97 |
|
|
98 |
|
=cut |
99 |
|
|
100 |
|
my $load_row_coderef; |
101 |
|
|
102 |
It contains several assumptions: |
sub data_structure { |
103 |
|
my $arg = {@_}; |
104 |
|
|
105 |
|
die "need row argument" unless ($arg->{row}); |
106 |
|
die "need normalisation argument" unless ($arg->{rules}); |
107 |
|
|
108 |
|
_set_lookup( $arg->{lookup} ) if defined($arg->{lookup}); |
109 |
|
_set_ds( $arg->{row} ); |
110 |
|
_set_config( $arg->{config} ) if defined($arg->{config}); |
111 |
|
_clean_ds( %{ $arg } ); |
112 |
|
$load_row_coderef = $arg->{load_row_coderef}; |
113 |
|
|
114 |
|
no strict 'subs'; |
115 |
|
no warnings 'redefine'; |
116 |
|
eval "$arg->{rules};"; |
117 |
|
die "error evaling $arg->{rules}: $@\n" if ($@); |
118 |
|
|
119 |
=over |
return _get_ds(); |
120 |
|
} |
121 |
|
|
122 |
=item * |
=head2 _set_ds |
123 |
|
|
124 |
format of fields is defined using C<v123^a> notation for repeatable fields |
Set current record hash |
|
or C<s123^a> for single (or first) value, where C<123> is field number and |
|
|
C<a> is subfield. |
|
125 |
|
|
126 |
=item * |
_set_ds( $rec ); |
127 |
|
|
128 |
source data records (C<$rec>) have unique identifiers in field C<000> |
=cut |
129 |
|
|
130 |
=item * |
my $rec; |
131 |
|
|
132 |
optional C<eval{length('v123^a') == 3}> tag at B<beginning of format> will be |
sub _set_ds { |
133 |
perl code that is evaluated before producing output (value of field will be |
$rec = shift or die "no record hash"; |
134 |
interpolated before that) |
} |
135 |
|
|
136 |
=item * |
=head2 _set_config |
137 |
|
|
138 |
optional C<filter{filter_name}> at B<begining of format> will apply perl |
Set current config hash |
|
code defined as code ref on format after field substitution to producing |
|
|
output |
|
139 |
|
|
140 |
=item * |
_set_config( $config ); |
141 |
|
|
142 |
optional C<lookup{...}> will be then performed. See C<WebPAC::Lookups>. |
Magic keys are: |
143 |
|
|
144 |
=item * |
=over 4 |
145 |
|
|
146 |
at end, optional C<format>s rules are resolved. Format rules are similar to |
=item _ |
147 |
C<sprintf> and can also contain C<lookup{...}> which is performed after |
|
148 |
values are inserted in format. |
Code of current database |
149 |
|
|
150 |
|
=item _mfn |
151 |
|
|
152 |
|
Current MFN |
153 |
|
|
154 |
=back |
=back |
155 |
|
|
156 |
This also describes order in which transformations are applied (eval, |
=cut |
157 |
filter, lookup, format) which is important to undestand when deciding how to |
|
158 |
solve your data mungling and normalisation process. |
my $config; |
159 |
|
|
160 |
|
sub _set_config { |
161 |
|
$config = shift; |
162 |
|
} |
163 |
|
|
164 |
|
=head2 _get_ds |
165 |
|
|
166 |
|
Return hash formatted as data structure |
167 |
|
|
168 |
=head1 FUNCTIONS |
my $ds = _get_ds(); |
169 |
|
|
170 |
=head2 new |
=cut |
171 |
|
|
172 |
Create new normalisation object |
my ($out, $marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader); |
173 |
|
my ($marc_record_offset, $marc_fetch_offset) = (0, 0); |
174 |
|
|
175 |
my $n = new WebPAC::Normalize::Something( |
sub _get_ds { |
176 |
filter => { |
#warn "## out = ",dump($out); |
177 |
'filter_name_1' => sub { |
return $out; |
178 |
# filter code |
} |
|
return length($_); |
|
|
}, ... |
|
|
}, |
|
|
db => $db_obj, |
|
|
lookup_regex => $lookup->regex, |
|
|
lookup => $lookup_obj, |
|
|
prefix => 'foobar', |
|
|
); |
|
179 |
|
|
180 |
Parametar C<filter> defines user supplied snippets of perl code which can |
=head2 _clean_ds |
|
be use with C<filter{...}> notation. |
|
181 |
|
|
182 |
C<prefix> is used to form filename for database record (to support multiple |
Clean data structure hash for next record |
|
source files which are joined in one database). |
|
183 |
|
|
184 |
Recommended parametar C<lookup_regex> is used to enable parsing of lookups |
_clean_ds(); |
|
in structures. If you pass this parametar, you must also pass C<lookup> |
|
|
which is C<WebPAC::Lookup> object. |
|
185 |
|
|
186 |
=cut |
=cut |
187 |
|
|
188 |
sub new { |
sub _clean_ds { |
189 |
my $class = shift; |
my $a = {@_}; |
190 |
my $self = {@_}; |
($out,$marc_record, $marc_encoding, $marc_repeatable_subfield, $marc_indicators, $marc_leader) = (); |
191 |
bless($self, $class); |
($marc_record_offset, $marc_fetch_offset) = (0,0); |
192 |
|
$marc_encoding = $a->{marc_encoding}; |
193 |
|
} |
194 |
|
|
195 |
my $r = $self->{'lookup_regex'} ? 1 : 0; |
=head2 _set_lookup |
|
my $l = $self->{'lookup'} ? 1 : 0; |
|
196 |
|
|
197 |
my $log = $self->_get_logger(); |
Set current lookup hash |
198 |
|
|
199 |
# those two must be in pair |
_set_lookup( $lookup ); |
|
if ( ($r & $l) != ($r || $l) ) { |
|
|
my $log = $self->_get_logger(); |
|
|
$log->logdie("lookup_regex and lookup must be in pair"); |
|
|
} |
|
200 |
|
|
201 |
$log->logdie("lookup must be WebPAC::Lookup object") if ($self->{'lookup'} && ! $self->{'lookup'}->isa('WebPAC::Lookup')); |
=cut |
202 |
|
|
203 |
$log->warn("no prefix defined. please check that!") unless ($self->{'prefix'}); |
my $lookup; |
204 |
|
|
205 |
$self ? return $self : return undef; |
sub _set_lookup { |
206 |
|
$lookup = shift; |
207 |
} |
} |
208 |
|
|
209 |
|
=head2 _get_lookup |
210 |
|
|
211 |
=head2 data_structure |
Get current lookup hash |
212 |
|
|
213 |
Create in-memory data structure which represents normalized layout from |
my $lookup = _get_lookup(); |
|
C<conf/normalize/*.xml>. |
|
214 |
|
|
215 |
This structures are used to produce output. |
=cut |
216 |
|
|
217 |
|
sub _get_lookup { |
218 |
|
return $lookup; |
219 |
|
} |
220 |
|
|
221 |
my $ds = $webpac->data_structure($rec); |
=head2 _set_load_row |
222 |
|
|
223 |
|
Setup code reference which will return L<data_structure> from |
224 |
|
L<WebPAC::Store> |
225 |
|
|
226 |
|
_set_load_row(sub { |
227 |
|
my ($database,$input,$mfn) = @_; |
228 |
|
$store->load_row( database => $database, input => $input, id => $mfn ); |
229 |
|
}); |
230 |
|
|
231 |
=cut |
=cut |
232 |
|
|
233 |
sub data_structure { |
sub _set_load_row { |
234 |
my $self = shift; |
my $coderef = shift; |
235 |
|
confess "argument isn't CODE" unless ref($coderef) eq 'CODE'; |
236 |
|
|
237 |
|
$load_row_coderef = $coderef; |
238 |
|
} |
239 |
|
|
240 |
|
=head2 _get_marc_fields |
241 |
|
|
242 |
|
Get all fields defined by calls to C<marc> |
243 |
|
|
244 |
|
$marc->add_fields( WebPAC::Normalize:_get_marc_fields() ); |
245 |
|
|
246 |
|
We are using I<magic> which detect repeatable fields only from |
247 |
|
sequence of field/subfield data generated by normalization. |
248 |
|
|
249 |
|
Repeatable field is created when there is second occurence of same subfield or |
250 |
|
if any of indicators are different. |
251 |
|
|
252 |
|
This is sane for most cases. Something like: |
253 |
|
|
254 |
|
900a-1 900b-1 900c-1 |
255 |
|
900a-2 900b-2 |
256 |
|
900a-3 |
257 |
|
|
258 |
|
will be created from any combination of: |
259 |
|
|
260 |
|
900a-1 900a-2 900a-3 900b-1 900b-2 900c-1 |
261 |
|
|
262 |
|
and following rules: |
263 |
|
|
264 |
|
marc('900','a', rec('200','a') ); |
265 |
|
marc('900','b', rec('200','b') ); |
266 |
|
marc('900','c', rec('200','c') ); |
267 |
|
|
268 |
|
which might not be what you have in mind. If you need repeatable subfield, |
269 |
|
define it using C<marc_repeatable_subfield> like this: |
270 |
|
|
271 |
|
marc_repeatable_subfield('900','a'); |
272 |
|
marc('900','a', rec('200','a') ); |
273 |
|
marc('900','b', rec('200','b') ); |
274 |
|
marc('900','c', rec('200','c') ); |
275 |
|
|
276 |
|
will create: |
277 |
|
|
278 |
my $log = $self->_get_logger(); |
900a-1 900a-2 900a-3 900b-1 900c-1 |
279 |
|
900b-2 |
280 |
|
|
281 |
my $rec = shift; |
There is also support for returning next or specific using: |
|
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
|
282 |
|
|
283 |
$log->debug("data_structure rec = ", sub { Dumper($rec) }); |
while (my $mf = WebPAC::Normalize:_get_marc_fields( fetch_next => 1 ) ) { |
284 |
|
# do something with $mf |
285 |
|
} |
286 |
|
|
287 |
$log->logdie("need unique ID (mfn) in field 000 of record ", sub { Dumper($rec) } ) unless (defined($rec->{'000'})); |
will always return fields from next MARC record or |
288 |
|
|
289 |
my $id = $rec->{'000'}->[0] || $log->logdie("field 000 isn't array!"); |
my $mf = WebPAC::Normalize::_get_marc_fields( offset => 42 ); |
290 |
|
|
291 |
my $cache_file; |
will return 42th copy record (if it exists). |
292 |
|
|
293 |
if ($self->{'db'}) { |
=cut |
294 |
my $ds = $self->{'db'}->load_ds( id => $id, prefix => $self->{prefix} ); |
|
295 |
$log->debug("load_ds( rec = ", sub { Dumper($rec) }, ") = ", sub { Dumper($ds) }); |
my $fetch_pos; |
296 |
return $ds if ($ds); |
|
297 |
$log->debug("cache miss, creating"); |
sub _get_marc_fields { |
298 |
|
|
299 |
|
my $arg = {@_}; |
300 |
|
warn "### _get_marc_fields arg: ", dump($arg), $/ if ($debug > 2); |
301 |
|
$fetch_pos = $marc_fetch_offset; |
302 |
|
if ($arg->{offset}) { |
303 |
|
$fetch_pos = $arg->{offset}; |
304 |
|
} elsif($arg->{fetch_next}) { |
305 |
|
$marc_fetch_offset++; |
306 |
} |
} |
307 |
|
|
308 |
undef $self->{'currnet_filename'}; |
return if (! $marc_record || ref($marc_record) ne 'ARRAY'); |
|
undef $self->{'headline'}; |
|
309 |
|
|
310 |
my @sorted_tags; |
warn "### full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 2); |
311 |
if ($self->{tags_by_order}) { |
|
312 |
@sorted_tags = @{$self->{tags_by_order}}; |
my $marc_rec = $marc_record->[ $fetch_pos ]; |
313 |
} else { |
|
314 |
@sorted_tags = sort { $self->_sort_by_order } keys %{$self->{'import_xml'}->{'indexer'}}; |
warn "## _get_marc_fields (at offset: $fetch_pos) -- marc_record = ", dump( @$marc_rec ), $/ if ($debug > 1); |
315 |
$self->{tags_by_order} = \@sorted_tags; |
|
316 |
|
return if (! $marc_rec || ref($marc_rec) ne 'ARRAY' || $#{ $marc_rec } < 0); |
317 |
|
|
318 |
|
# first, sort all existing fields |
319 |
|
# XXX might not be needed, but modern perl might randomize elements in hash |
320 |
|
my @sorted_marc_record = sort { |
321 |
|
$a->[0] . ( $a->[3] || '' ) cmp $b->[0] . ( $b->[3] || '') |
322 |
|
} @{ $marc_rec }; |
323 |
|
|
324 |
|
@sorted_marc_record = @{ $marc_rec }; ### FIXME disable sorting |
325 |
|
|
326 |
|
# output marc fields |
327 |
|
my @m; |
328 |
|
|
329 |
|
# count unique field-subfields (used for offset when walking to next subfield) |
330 |
|
my $u; |
331 |
|
map { $u->{ $_->[0] . ( $_->[3] || '') }++ } @sorted_marc_record; |
332 |
|
|
333 |
|
if ($debug) { |
334 |
|
warn "## marc_repeatable_subfield = ", dump( $marc_repeatable_subfield ), $/ if ( $marc_repeatable_subfield ); |
335 |
|
warn "## marc_record[$fetch_pos] = ", dump( $marc_rec ), $/; |
336 |
|
warn "## sorted_marc_record = ", dump( \@sorted_marc_record ), $/; |
337 |
|
warn "## subfield count = ", dump( $u ), $/; |
338 |
} |
} |
339 |
|
|
340 |
my $ds; |
my $len = $#sorted_marc_record; |
341 |
|
my $visited; |
342 |
|
my $i = 0; |
343 |
|
my $field; |
344 |
|
|
345 |
$log->debug("tags: ",sub { join(", ",@sorted_tags) }); |
foreach ( 0 .. $len ) { |
346 |
|
|
347 |
foreach my $field (@sorted_tags) { |
# find next element which isn't visited |
348 |
|
while ($visited->{$i}) { |
349 |
|
$i = ($i + 1) % ($len + 1); |
350 |
|
} |
351 |
|
|
352 |
my $row; |
# mark it visited |
353 |
|
$visited->{$i}++; |
354 |
|
|
355 |
#print "field $field [",$self->{'tag'},"] = ",Dumper($self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}); |
my $row = dclone( $sorted_marc_record[$i] ); |
356 |
|
|
357 |
foreach my $tag (@{$self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}}) { |
# field and subfield which is key for |
358 |
my $format; |
# marc_repeatable_subfield and u |
359 |
|
my $fsf = $row->[0] . ( $row->[3] || '' ); |
360 |
|
|
361 |
|
if ($debug > 1) { |
362 |
|
|
363 |
|
print "### field so far [", $#$field, "] : ", dump( $field ), " ", $field ? 'T' : 'F', $/; |
364 |
|
print "### this [$i]: ", dump( $row ),$/; |
365 |
|
print "### sf: ", $row->[3], " vs ", $field->[3], |
366 |
|
$marc_repeatable_subfield->{ $row->[0] . $row->[3] } ? ' (repeatable)' : '', $/, |
367 |
|
if ($#$field >= 0); |
368 |
|
|
369 |
$log->logdie("expected tag HASH and got $tag") unless (ref($tag) eq 'HASH'); |
} |
|
$format = $tag->{'value'} || $tag->{'content'}; |
|
370 |
|
|
371 |
$log->debug("format: $format"); |
# if field exists |
372 |
|
if ( $#$field >= 0 ) { |
373 |
|
if ( |
374 |
|
$row->[0] ne $field->[0] || # field |
375 |
|
$row->[1] ne $field->[1] || # i1 |
376 |
|
$row->[2] ne $field->[2] # i2 |
377 |
|
) { |
378 |
|
push @m, $field; |
379 |
|
warn "## saved/1 ", dump( $field ),$/ if ($debug); |
380 |
|
$field = $row; |
381 |
|
|
382 |
|
} elsif ( |
383 |
|
( $row->[3] lt $field->[-2] ) # subfield which is not next (e.g. a after c) |
384 |
|
|| |
385 |
|
( $row->[3] eq $field->[-2] && # same subfield, but not repeatable |
386 |
|
! $marc_repeatable_subfield->{ $fsf } |
387 |
|
) |
388 |
|
) { |
389 |
|
push @m, $field; |
390 |
|
warn "## saved/2 ", dump( $field ),$/ if ($debug); |
391 |
|
$field = $row; |
392 |
|
|
|
my @v; |
|
|
if ($self->{'lookup_regex'} && $format =~ $self->{'lookup_regex'}) { |
|
|
@v = $self->fill_in_to_arr($rec,$format); |
|
393 |
} else { |
} else { |
394 |
@v = $self->parse_to_arr($rec,$format); |
# append new subfields to existing field |
395 |
|
push @$field, ( $row->[3], $row->[4] ); |
396 |
} |
} |
397 |
next if (! @v); |
} else { |
398 |
|
# insert first field |
399 |
|
$field = $row; |
400 |
|
} |
401 |
|
|
402 |
if ($tag->{'sort'}) { |
if (! $marc_repeatable_subfield->{ $fsf }) { |
403 |
@v = $self->sort_arr(@v); |
# make step to next subfield |
404 |
} |
$i = ($i + $u->{ $fsf } ) % ($len + 1); |
405 |
|
} |
406 |
|
} |
407 |
|
|
408 |
# use format? |
if ($#$field >= 0) { |
409 |
if ($tag->{'format_name'}) { |
push @m, $field; |
410 |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
warn "## saved/3 ", dump( $field ),$/ if ($debug); |
411 |
} |
} |
412 |
|
|
413 |
# delimiter will join repeatable fields |
return \@m; |
414 |
if ($tag->{'delimiter'}) { |
} |
|
@v = ( join($tag->{'delimiter'}, @v) ); |
|
|
} |
|
415 |
|
|
416 |
# default types |
=head2 _get_marc_leader |
|
my @types = qw(display search); |
|
|
# override by type attribute |
|
|
@types = ( $tag->{'type'} ) if ($tag->{'type'}); |
|
|
|
|
|
foreach my $type (@types) { |
|
|
# append to previous line? |
|
|
$log->debug("type: $type ",sub { join(" ",@v) }, $row->{'append'} || 'no append'); |
|
|
if ($tag->{'append'}) { |
|
|
|
|
|
# I will delimit appended part with |
|
|
# delimiter (or ,) |
|
|
my $d = $tag->{'delimiter'}; |
|
|
# default delimiter |
|
|
$d ||= " "; |
|
|
|
|
|
my $last = pop @{$row->{$type}}; |
|
|
$d = "" if (! $last); |
|
|
$last .= $d . join($d, @v); |
|
|
push @{$row->{$type}}, $last; |
|
417 |
|
|
418 |
} else { |
Return leader from currently fetched record by L</_get_marc_fields> |
|
push @{$row->{$type}}, @v; |
|
|
} |
|
|
} |
|
419 |
|
|
420 |
|
print WebPAC::Normalize::_get_marc_leader(); |
421 |
|
|
422 |
} |
=cut |
423 |
|
|
424 |
if ($row) { |
sub _get_marc_leader { |
425 |
$row->{'tag'} = $field; |
die "no fetch_pos, did you called _get_marc_fields first?" unless ( defined( $fetch_pos ) ); |
426 |
|
return $marc_leader->[ $fetch_pos ]; |
427 |
|
} |
428 |
|
|
429 |
# TODO: name_sigular, name_plural |
=head2 _debug |
|
my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; |
|
|
my $row_name = $name ? $self->_x($name) : $field; |
|
|
|
|
|
# post-sort all values in field |
|
|
if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { |
|
|
$log->warn("sort at field tag not implemented"); |
|
|
} |
|
430 |
|
|
431 |
$ds->{$row_name} = $row; |
Change level of debug warnings |
432 |
|
|
433 |
$log->debug("row $field: ",sub { Dumper($row) }); |
_debug( 2 ); |
|
} |
|
434 |
|
|
435 |
} |
=cut |
436 |
|
|
437 |
|
sub _debug { |
438 |
|
my $l = shift; |
439 |
|
return $debug unless defined($l); |
440 |
|
warn "debug level $l",$/ if ($l > 0); |
441 |
|
$debug = $l; |
442 |
|
} |
443 |
|
|
444 |
|
=head1 Functions to create C<data_structure> |
445 |
|
|
446 |
|
Those functions generally have to first in your normalization file. |
447 |
|
|
448 |
$self->{'db'}->save_ds( |
=head2 search_display |
|
id => $id, |
|
|
ds => $ds, |
|
|
prefix => $self->{prefix}, |
|
|
) if ($self->{'db'}); |
|
449 |
|
|
450 |
$log->debug("ds: ", sub { Dumper($ds) }); |
Define output for L<search> and L<display> at the same time |
451 |
|
|
452 |
$log->logconfess("data structure returned is not array any more!") if wantarray; |
search_display('Title', rec('200','a') ); |
453 |
|
|
|
return $ds; |
|
454 |
|
|
455 |
|
=cut |
456 |
|
|
457 |
|
sub search_display { |
458 |
|
my $name = shift or die "search_display needs name as first argument"; |
459 |
|
my @o = grep { defined($_) && $_ ne '' } @_; |
460 |
|
return unless (@o); |
461 |
|
$out->{$name}->{search} = \@o; |
462 |
|
$out->{$name}->{display} = \@o; |
463 |
} |
} |
464 |
|
|
465 |
=head2 parse |
=head2 tag |
466 |
|
|
467 |
Perform smart parsing of string, skipping delimiters for fields which aren't |
Old name for L<search_display>, but supported |
|
defined. It can also eval code in format starting with C<eval{...}> and |
|
|
return output or nothing depending on eval code. |
|
468 |
|
|
469 |
my $text = $webpac->parse($rec,'eval{"v901^a" eq "Deskriptor"}descriptor: v250^a', $i); |
=cut |
470 |
|
|
471 |
|
sub tag { |
472 |
|
search_display( @_ ); |
473 |
|
} |
474 |
|
|
475 |
|
=head2 display |
476 |
|
|
477 |
|
Define output just for I<display> |
478 |
|
|
479 |
|
@v = display('Title', rec('200','a') ); |
480 |
|
|
481 |
=cut |
=cut |
482 |
|
|
483 |
sub parse { |
sub _field { |
484 |
my $self = shift; |
my $type = shift or confess "need type -- BUG?"; |
485 |
|
my $name = shift or confess "needs name as first argument"; |
486 |
|
my @o = grep { defined($_) && $_ ne '' } @_; |
487 |
|
return unless (@o); |
488 |
|
$out->{$name}->{$type} = \@o; |
489 |
|
} |
490 |
|
|
491 |
my ($rec, $format_utf8, $i) = @_; |
sub display { _field( 'display', @_ ) } |
492 |
|
|
493 |
return if (! $format_utf8); |
=head2 search |
494 |
|
|
495 |
my $log = $self->_get_logger(); |
Prepare values just for I<search> |
496 |
|
|
497 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
@v = search('Title', rec('200','a') ); |
498 |
|
|
499 |
$i = 0 if (! $i); |
=cut |
500 |
|
|
501 |
my $format = $self->_x($format_utf8) || $log->logconfess("can't convert '$format_utf8' from UTF-8 to ",$self->{'code_page'}); |
sub search { _field( 'search', @_ ) } |
502 |
|
|
503 |
my @out; |
=head2 sorted |
504 |
|
|
505 |
$log->debug("format: $format"); |
Insert into lists which will be automatically sorted |
506 |
|
|
507 |
my $eval_code; |
sorted('Title', rec('200','a') ); |
|
# remove eval{...} from beginning |
|
|
$eval_code = $1 if ($format =~ s/^eval{([^}]+)}//s); |
|
508 |
|
|
509 |
my $filter_name; |
=cut |
510 |
# remove filter{...} from beginning |
|
511 |
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
sub sorted { _field( 'sorted', @_ ) } |
512 |
|
|
|
my $prefix; |
|
|
my $all_found=0; |
|
513 |
|
|
514 |
while ($format =~ s/^(.*?)(v|s)(\d+)(?:\^(\w))?//s) { |
=head2 marc_leader |
515 |
|
|
516 |
my $del = $1 || ''; |
Setup fields within MARC leader or get leader |
|
$prefix ||= $del if ($all_found == 0); |
|
517 |
|
|
518 |
# repeatable index |
marc_leader('05','c'); |
519 |
my $r = $i; |
my $leader = marc_leader(); |
|
$r = 0 if (lc("$2") eq 's'); |
|
520 |
|
|
521 |
my $found = 0; |
=cut |
|
my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found); |
|
522 |
|
|
523 |
if ($found) { |
sub marc_leader { |
524 |
push @out, $del; |
my ($offset,$value) = @_; |
525 |
push @out, $tmp; |
|
526 |
$all_found += $found; |
if ($offset) { |
527 |
|
$marc_leader->[ $marc_record_offset ]->{ $offset } = $value; |
528 |
|
} else { |
529 |
|
|
530 |
|
if (defined($marc_leader)) { |
531 |
|
die "marc_leader not array = ", dump( $marc_leader ) unless (ref($marc_leader) eq 'ARRAY'); |
532 |
|
return $marc_leader->[ $marc_record_offset ]; |
533 |
|
} else { |
534 |
|
return; |
535 |
} |
} |
536 |
} |
} |
537 |
|
} |
538 |
|
|
539 |
|
=head2 marc_fixed |
540 |
|
|
541 |
|
Create control/indentifier fields with values in fixed positions |
542 |
|
|
543 |
|
marc_fixed('008', 00, '070402'); |
544 |
|
marc_fixed('008', 39, '|'); |
545 |
|
|
546 |
return if (! $all_found); |
Positions not specified will be filled with spaces (C<0x20>). |
547 |
|
|
548 |
my $out = join('',@out); |
There will be no effort to extend last specified value to full length of |
549 |
|
field in standard. |
550 |
|
|
551 |
if ($out) { |
=cut |
|
# add rest of format (suffix) |
|
|
$out .= $format; |
|
552 |
|
|
553 |
# add prefix if not there |
sub marc_fixed { |
554 |
$out = $prefix . $out if ($out !~ m/^\Q$prefix\E/); |
my ($f, $pos, $val) = @_; |
555 |
|
die "need marc(field, position, value)" unless defined($f) && defined($pos); |
556 |
|
|
557 |
|
confess "need val" unless defined $val; |
558 |
|
|
559 |
|
my $update = 0; |
560 |
|
|
561 |
|
map { |
562 |
|
if ($_->[0] eq $f) { |
563 |
|
my $old = $_->[1]; |
564 |
|
if (length($old) <= $pos) { |
565 |
|
$_->[1] .= ' ' x ( $pos - length($old) ) . $val; |
566 |
|
warn "## marc_fixed($f,$pos,'$val') append '$old' -> '$_->[1]'\n" if ($debug > 1); |
567 |
|
} else { |
568 |
|
$_->[1] = substr($old, 0, $pos) . $val . substr($old, $pos + length($val)); |
569 |
|
warn "## marc_fixed($f,$pos,'$val') update '$old' -> '$_->[1]'\n" if ($debug > 1); |
570 |
|
} |
571 |
|
$update++; |
572 |
|
} |
573 |
|
} @{ $marc_record->[ $marc_record_offset ] }; |
574 |
|
|
575 |
$log->debug("result: $out"); |
if (! $update) { |
576 |
|
my $v = ' ' x $pos . $val; |
577 |
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; |
578 |
|
warn "## marc_fixed($f,$pos,'val') created '$v'\n" if ($debug > 1); |
579 |
} |
} |
580 |
|
} |
581 |
|
|
582 |
if ($eval_code) { |
=head2 marc |
583 |
my $eval = $self->fill_in($rec,$eval_code,$i) || return; |
|
584 |
$log->debug("about to eval{$eval} format: $out"); |
Save value for MARC field |
585 |
return if (! $self->_eval($eval)); |
|
586 |
|
marc('900','a', rec('200','a') ); |
587 |
|
marc('001', rec('000') ); |
588 |
|
|
589 |
|
=cut |
590 |
|
|
591 |
|
sub marc { |
592 |
|
my $f = shift or die "marc needs field"; |
593 |
|
die "marc field must be numer" unless ($f =~ /^\d+$/); |
594 |
|
|
595 |
|
my $sf; |
596 |
|
if ($f >= 10) { |
597 |
|
$sf = shift or die "marc needs subfield"; |
598 |
} |
} |
599 |
|
|
600 |
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
foreach (@_) { |
601 |
$log->debug("about to filter{$filter_name} format: $out"); |
my $v = $_; # make var read-write for Encode |
602 |
$out = $self->{'filter'}->{$filter_name}->($out); |
next unless (defined($v) && $v !~ /^\s*$/); |
603 |
return unless(defined($out)); |
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
604 |
$log->debug("filter result: $out"); |
if (defined $sf) { |
605 |
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $i1, $i2, $sf => $v ]; |
606 |
|
} else { |
607 |
|
push @{ $marc_record->[ $marc_record_offset ] }, [ $f, $v ]; |
608 |
|
} |
609 |
} |
} |
610 |
|
} |
611 |
|
|
612 |
return $out; |
=head2 marc_repeatable_subfield |
613 |
|
|
614 |
|
Save values for MARC repetable subfield |
615 |
|
|
616 |
|
marc_repeatable_subfield('910', 'z', rec('909') ); |
617 |
|
|
618 |
|
=cut |
619 |
|
|
620 |
|
sub marc_repeatable_subfield { |
621 |
|
my ($f,$sf) = @_; |
622 |
|
die "marc_repeatable_subfield need field and subfield!\n" unless ($f && $sf); |
623 |
|
$marc_repeatable_subfield->{ $f . $sf }++; |
624 |
|
marc(@_); |
625 |
} |
} |
626 |
|
|
627 |
=head2 parse_to_arr |
=head2 marc_indicators |
628 |
|
|
629 |
|
Set both indicators for MARC field |
630 |
|
|
631 |
Similar to C<parse>, but returns array of all repeatable fields |
marc_indicators('900', ' ', 1); |
632 |
|
|
633 |
my @arr = $webpac->parse_to_arr($rec,'v250^a'); |
Any indicator value other than C<0-9> will be treated as undefined. |
634 |
|
|
635 |
=cut |
=cut |
636 |
|
|
637 |
sub parse_to_arr { |
sub marc_indicators { |
638 |
my $self = shift; |
my $f = shift || die "marc_indicators need field!\n"; |
639 |
|
my ($i1,$i2) = @_; |
640 |
|
die "marc_indicators($f, ...) need i1!\n" unless(defined($i1)); |
641 |
|
die "marc_indicators($f, $i1, ...) need i2!\n" unless(defined($i2)); |
642 |
|
|
643 |
|
$i1 = ' ' if ($i1 !~ /^\d$/); |
644 |
|
$i2 = ' ' if ($i2 !~ /^\d$/); |
645 |
|
@{ $marc_indicators->{$f} } = ($i1,$i2); |
646 |
|
} |
647 |
|
|
648 |
my ($rec, $format_utf8) = @_; |
=head2 marc_compose |
649 |
|
|
650 |
my $log = $self->_get_logger(); |
Save values for each MARC subfield explicitly |
651 |
|
|
652 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
marc_compose('900', |
653 |
return if (! $format_utf8); |
'a', rec('200','a') |
654 |
|
'b', rec('201','a') |
655 |
|
'a', rec('200','b') |
656 |
|
'c', rec('200','c') |
657 |
|
); |
658 |
|
|
659 |
my $i = 0; |
If you specify C<+> for subfield, value will be appended |
660 |
my @arr; |
to previous defined subfield. |
661 |
|
|
662 |
|
=cut |
663 |
|
|
664 |
while (my $v = $self->parse($rec,$format_utf8,$i++)) { |
sub marc_compose { |
665 |
push @arr, $v; |
my $f = shift or die "marc_compose needs field"; |
666 |
|
die "marc_compose field must be numer" unless ($f =~ /^\d+$/); |
667 |
|
|
668 |
|
my ($i1,$i2) = defined($marc_indicators->{$f}) ? @{ $marc_indicators->{$f} } : (' ',' '); |
669 |
|
my $m = [ $f, $i1, $i2 ]; |
670 |
|
|
671 |
|
warn "### marc_compose input subfields = ", dump(@_),$/ if ($debug > 2); |
672 |
|
|
673 |
|
if ($#_ % 2 != 1) { |
674 |
|
die "ERROR: marc_compose",dump($f,@_)," not valid (must be even).\nDo you need to add first() or join() around some argument?\n"; |
675 |
} |
} |
676 |
|
|
677 |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
while (@_) { |
678 |
|
my $sf = shift; |
679 |
|
my $v = shift; |
680 |
|
|
681 |
|
next unless (defined($v) && $v !~ /^\s*$/); |
682 |
|
warn "## ++ marc_compose($f,$sf,$v) ", dump( $m ),$/ if ($debug > 1); |
683 |
|
if ($sf ne '+') { |
684 |
|
push @$m, ( $sf, $v ); |
685 |
|
} else { |
686 |
|
$m->[ $#$m ] .= $v; |
687 |
|
} |
688 |
|
} |
689 |
|
|
690 |
return @arr; |
warn "## marc_compose current marc = ", dump( $m ),$/ if ($debug > 1); |
691 |
|
|
692 |
|
push @{ $marc_record->[ $marc_record_offset ] }, $m if ($#{$m} > 2); |
693 |
} |
} |
694 |
|
|
695 |
|
=head2 marc_duplicate |
696 |
|
|
697 |
=head2 fill_in |
Generate copy of current MARC record and continue working on copy |
698 |
|
|
699 |
Workhourse of all: takes record from in-memory structure of database and |
marc_duplicate(); |
|
strings with placeholders and returns string or array of with substituted |
|
|
values from record. |
|
700 |
|
|
701 |
my $text = $webpac->fill_in($rec,'v250^a'); |
Copies can be accessed using C<< _get_marc_fields( fetch_next => 1 ) >> or |
702 |
|
C<< _get_marc_fields( offset => 42 ) >>. |
703 |
|
|
704 |
Optional argument is ordinal number for repeatable fields. By default, |
=cut |
|
it's assume to be first repeatable field (fields are perl array, so first |
|
|
element is 0). |
|
|
Following example will read second value from repeatable field. |
|
705 |
|
|
706 |
my $text = $webpac->fill_in($rec,'Title: v250^a',1); |
sub marc_duplicate { |
707 |
|
my $m = $marc_record->[ -1 ]; |
708 |
|
die "can't duplicate record which isn't defined" unless ($m); |
709 |
|
push @{ $marc_record }, dclone( $m ); |
710 |
|
push @{ $marc_leader }, dclone( marc_leader() ); |
711 |
|
warn "## marc_duplicate = ", dump(@$marc_leader, @$marc_record), $/ if ($debug > 1); |
712 |
|
$marc_record_offset = $#{ $marc_record }; |
713 |
|
warn "## marc_record_offset = $marc_record_offset", $/ if ($debug > 1); |
714 |
|
|
715 |
This function B<does not> perform parsing of format to inteligenty skip |
} |
|
delimiters before fields which aren't used. |
|
716 |
|
|
717 |
This method will automatically decode UTF-8 string to local code page |
=head2 marc_remove |
|
if needed. |
|
718 |
|
|
719 |
=cut |
Remove some field or subfield from MARC record. |
720 |
|
|
721 |
sub fill_in { |
marc_remove('200'); |
722 |
my $self = shift; |
marc_remove('200','a'); |
723 |
|
|
724 |
my $log = $self->_get_logger(); |
This will erase field C<200> or C<200^a> from current MARC record. |
725 |
|
|
726 |
my $rec = shift || $log->logconfess("need data record"); |
marc_remove('*'); |
|
my $format = shift || $log->logconfess("need format to parse"); |
|
|
# iteration (for repeatable fields) |
|
|
my $i = shift || 0; |
|
727 |
|
|
728 |
$log->logdie("infitite loop in format $format") if ($i > ($self->{'max_mfn'} || 9999)); |
Will remove all fields in current MARC record. |
729 |
|
|
730 |
# FIXME remove for speedup? |
This is useful after calling C<marc_duplicate> or on it's own (but, you |
731 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
should probably just remove that subfield definition if you are not |
732 |
|
using C<marc_duplicate>). |
733 |
|
|
734 |
if (utf8::is_utf8($format)) { |
FIXME: support fields < 10. |
|
$format = $self->_x($format); |
|
|
} |
|
735 |
|
|
736 |
my $found = 0; |
=cut |
737 |
|
|
738 |
my $eval_code; |
sub marc_remove { |
739 |
# remove eval{...} from beginning |
my ($f, $sf) = @_; |
|
$eval_code = $1 if ($format =~ s/^eval{([^}]+)}//s); |
|
740 |
|
|
741 |
my $filter_name; |
die "marc_remove needs record number" unless defined($f); |
|
# remove filter{...} from beginning |
|
|
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
|
742 |
|
|
743 |
# do actual replacement of placeholders |
my $marc = $marc_record->[ $marc_record_offset ]; |
|
# repeatable fields |
|
|
$format =~ s/v(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,$i,\$found)/ges; |
|
|
# non-repeatable fields |
|
|
$format =~ s/s(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,0,\$found)/ges; |
|
744 |
|
|
745 |
if ($found) { |
warn "### marc_remove before = ", dump( $marc ), $/ if ($debug > 2); |
746 |
$log->debug("format: $format"); |
|
747 |
if ($eval_code) { |
if ($f eq '*') { |
748 |
my $eval = $self->fill_in($rec,$eval_code,$i); |
|
749 |
return if (! $self->_eval($eval)); |
delete( $marc_record->[ $marc_record_offset ] ); |
750 |
} |
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
751 |
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
|
752 |
$log->debug("filter '$filter_name' for $format"); |
} else { |
753 |
$format = $self->{'filter'}->{$filter_name}->($format); |
|
754 |
return unless(defined($format)); |
my $i = 0; |
755 |
$log->debug("filter result: $format"); |
foreach ( 0 .. $#{ $marc } ) { |
756 |
} |
last unless (defined $marc->[$i]); |
757 |
# do we have lookups? |
warn "#### working on ",dump( @{ $marc->[$i] }), $/ if ($debug > 3); |
758 |
if ($self->{'lookup'}) { |
if ($marc->[$i]->[0] eq $f) { |
759 |
if ($self->{'lookup'}->can('lookup')) { |
if (! defined $sf) { |
760 |
return $self->{'lookup'}->lookup($format); |
# remove whole field |
761 |
} else { |
splice @$marc, $i, 1; |
762 |
$log->warn("Have lookup object but can't invoke lookup method"); |
warn "#### slice \@\$marc, $i, 1 = ",dump( @{ $marc }), $/ if ($debug > 3); |
763 |
|
$i--; |
764 |
|
} else { |
765 |
|
foreach my $j ( 0 .. (( $#{ $marc->[$i] } - 3 ) / 2) ) { |
766 |
|
my $o = ($j * 2) + 3; |
767 |
|
if ($marc->[$i]->[$o] eq $sf) { |
768 |
|
# remove subfield |
769 |
|
splice @{$marc->[$i]}, $o, 2; |
770 |
|
warn "#### slice \@{\$marc->[$i]}, $o, 2 = ", dump( @{ $marc }), $/ if ($debug > 3); |
771 |
|
# is record now empty? |
772 |
|
if ($#{ $marc->[$i] } == 2) { |
773 |
|
splice @$marc, $i, 1; |
774 |
|
warn "#### slice \@\$marc, $i, 1 = ", dump( @{ $marc }), $/ if ($debug > 3); |
775 |
|
$i--; |
776 |
|
}; |
777 |
|
} |
778 |
|
} |
779 |
|
} |
780 |
} |
} |
781 |
} else { |
$i++; |
|
return $format; |
|
782 |
} |
} |
783 |
} else { |
|
784 |
return; |
warn "### marc_remove($f", $sf ? ",$sf" : "", ") after = ", dump( $marc ), $/ if ($debug > 2); |
785 |
|
|
786 |
|
$marc_record->[ $marc_record_offset ] = $marc; |
787 |
} |
} |
788 |
|
|
789 |
|
warn "## full marc_record = ", dump( @{ $marc_record }), $/ if ($debug > 1); |
790 |
} |
} |
791 |
|
|
792 |
|
=head2 marc_original_order |
793 |
|
|
794 |
=head2 fill_in_to_arr |
Copy all subfields preserving original order to marc field. |
795 |
|
|
796 |
Similar to C<fill_in>, but returns array of all repeatable fields. Usable |
marc_original_order( marc_field_number, original_input_field_number ); |
797 |
for fields which have lookups, so they shouldn't be parsed but rather |
|
798 |
C<fill_id>ed. |
Please note that field numbers are consistent with other commands (marc |
799 |
|
field number first), but somewhat counter-intuitive (destination and then |
800 |
|
source). |
801 |
|
|
802 |
|
You might want to use this command if you are just renaming subfields or |
803 |
|
using pre-processing modify_record in C<config.yml> and don't need any |
804 |
|
post-processing or want to preserve order of original subfields. |
805 |
|
|
|
my @arr = $webpac->fill_in_to_arr($rec,'[v900];;[v250^a]'); |
|
806 |
|
|
807 |
=cut |
=cut |
808 |
|
|
809 |
sub fill_in_to_arr { |
sub marc_original_order { |
|
my $self = shift; |
|
810 |
|
|
811 |
my ($rec, $format_utf8) = @_; |
my ($to, $from) = @_; |
812 |
|
die "marc_original_order needs from and to fields\n" unless ($from && $to); |
813 |
|
|
814 |
my $log = $self->_get_logger(); |
return unless defined($rec->{$from}); |
815 |
|
|
816 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
my $r = $rec->{$from}; |
817 |
return if (! $format_utf8); |
die "record field $from isn't array\n" unless (ref($r) eq 'ARRAY'); |
818 |
|
|
819 |
my $i = 0; |
my ($i1,$i2) = defined($marc_indicators->{$to}) ? @{ $marc_indicators->{$to} } : (' ',' '); |
820 |
my @arr; |
warn "## marc_original_order($to,$from) source = ", dump( $r ),$/ if ($debug > 1); |
821 |
|
|
822 |
while (my @v = $self->fill_in($rec,$format_utf8,$i++)) { |
foreach my $d (@$r) { |
823 |
push @arr, @v; |
|
824 |
} |
if (! defined($d->{subfields}) && ref($d->{subfields}) ne 'ARRAY') { |
825 |
|
warn "# marc_original_order($to,$from): field $from doesn't have subfields specification\n"; |
826 |
|
next; |
827 |
|
} |
828 |
|
|
829 |
|
my @sfs = @{ $d->{subfields} }; |
830 |
|
|
831 |
|
die "field $from doesn't have even number of subfields specifications\n" unless($#sfs % 2 == 1); |
832 |
|
|
833 |
|
warn "#--> d: ",dump($d), "\n#--> sfs: ",dump(@sfs),$/ if ($debug > 2); |
834 |
|
|
835 |
|
my $m = [ $to, $i1, $i2 ]; |
836 |
|
|
837 |
|
while (my $sf = shift @sfs) { |
838 |
|
|
839 |
|
warn "#--> sf: ",dump($sf), $/ if ($debug > 2); |
840 |
|
my $offset = shift @sfs; |
841 |
|
die "corrupted sufields specification for field $from\n" unless defined($offset); |
842 |
|
|
843 |
|
my $v; |
844 |
|
if (ref($d->{$sf}) eq 'ARRAY') { |
845 |
|
$v = $d->{$sf}->[$offset] if (defined($d->{$sf}->[$offset])); |
846 |
|
} elsif ($offset == 0) { |
847 |
|
$v = $d->{$sf}; |
848 |
|
} else { |
849 |
|
die "field $from subfield '$sf' need occurence $offset which doesn't exist", dump($d->{$sf}); |
850 |
|
} |
851 |
|
push @$m, ( $sf, $v ) if (defined($v)); |
852 |
|
} |
853 |
|
|
854 |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
if ($#{$m} > 2) { |
855 |
|
push @{ $marc_record->[ $marc_record_offset ] }, $m; |
856 |
|
} |
857 |
|
} |
858 |
|
|
859 |
return @arr; |
warn "## marc_record = ", dump( $marc_record ),$/ if ($debug > 1); |
860 |
} |
} |
861 |
|
|
862 |
|
=head2 marc_count |
863 |
|
|
864 |
|
Return number of MARC records created using L</marc_duplicate>. |
865 |
|
|
866 |
=head2 get_data |
print "created ", marc_count(), " records"; |
867 |
|
|
868 |
Returns value from record. |
=cut |
869 |
|
|
870 |
|
sub marc_count { |
871 |
|
return $#{ $marc_record }; |
872 |
|
} |
873 |
|
|
|
my $text = $self->get_data(\$rec,$f,$sf,$i,\$found); |
|
874 |
|
|
875 |
Arguments are: |
=head1 Functions to extract data from input |
|
record reference C<$rec>, |
|
|
field C<$f>, |
|
|
optional subfiled C<$sf>, |
|
|
index for repeatable values C<$i>. |
|
876 |
|
|
877 |
Optinal variable C<$found> will be incremeted if there |
This function should be used inside functions to create C<data_structure> described |
878 |
is field. |
above. |
879 |
|
|
880 |
Returns value or empty string. |
=head2 _pack_subfields_hash |
881 |
|
|
882 |
|
@subfields = _pack_subfields_hash( $h ); |
883 |
|
$subfields = _pack_subfields_hash( $h, 1 ); |
884 |
|
|
885 |
|
Return each subfield value in array or pack them all together and return scalar |
886 |
|
with subfields (denoted by C<^>) and values. |
887 |
|
|
888 |
=cut |
=cut |
889 |
|
|
890 |
sub get_data { |
sub _pack_subfields_hash { |
891 |
my $self = shift; |
|
892 |
|
warn "## _pack_subfields_hash( ",dump(@_), " )\n" if ($debug > 1); |
893 |
|
|
894 |
|
my ($h,$include_subfields) = @_; |
895 |
|
|
896 |
|
# sanity and ease of use |
897 |
|
return $h if (ref($h) ne 'HASH'); |
898 |
|
|
899 |
my ($rec,$f,$sf,$i,$found) = @_; |
if ( defined($h->{subfields}) ) { |
900 |
|
my $sfs = delete $h->{subfields} || die "no subfields?"; |
901 |
|
my @out; |
902 |
|
while (@$sfs) { |
903 |
|
my $sf = shift @$sfs; |
904 |
|
push @out, '^' . $sf if ($include_subfields); |
905 |
|
my $o = shift @$sfs; |
906 |
|
if ($o == 0 && ref( $h->{$sf} ) ne 'ARRAY' ) { |
907 |
|
# single element subfields are not arrays |
908 |
|
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
909 |
|
|
910 |
if ($$rec->{$f}) { |
push @out, $h->{$sf}; |
911 |
return '' if (! $$rec->{$f}->[$i]); |
} else { |
912 |
no strict 'refs'; |
#warn "====> $sf $o / $#$sfs ", dump( $sfs, $h->{$sf} ), "\n"; |
913 |
if ($sf && $$rec->{$f}->[$i]->{$sf}) { |
push @out, $h->{$sf}->[$o]; |
914 |
$$found++ if (defined($$found)); |
} |
915 |
return $$rec->{$f}->[$i]->{$sf}; |
} |
916 |
} elsif (! $sf && $$rec->{$f}->[$i]) { |
if ($include_subfields) { |
917 |
$$found++ if (defined($$found)); |
return join('', @out); |
918 |
# it still might have subfield, just |
} else { |
919 |
# not specified, so we'll dump all |
return @out; |
920 |
if ($$rec->{$f}->[$i] =~ /HASH/o) { |
} |
921 |
my $out; |
} else { |
922 |
foreach my $k (keys %{$$rec->{$f}->[$i]}) { |
if ($include_subfields) { |
923 |
$out .= $$rec->{$f}->[$i]->{$k}." "; |
my $out = ''; |
924 |
|
foreach my $sf (sort keys %$h) { |
925 |
|
if (ref($h->{$sf}) eq 'ARRAY') { |
926 |
|
$out .= '^' . $sf . join('^' . $sf, @{ $h->{$sf} }); |
927 |
|
} else { |
928 |
|
$out .= '^' . $sf . $h->{$sf}; |
929 |
} |
} |
930 |
return $out; |
} |
931 |
|
return $out; |
932 |
|
} else { |
933 |
|
# FIXME this should probably be in alphabetical order instead of hash order |
934 |
|
values %{$h}; |
935 |
|
} |
936 |
|
} |
937 |
|
} |
938 |
|
|
939 |
|
=head2 rec1 |
940 |
|
|
941 |
|
Return all values in some field |
942 |
|
|
943 |
|
@v = rec1('200') |
944 |
|
|
945 |
|
TODO: order of values is probably same as in source data, need to investigate that |
946 |
|
|
947 |
|
=cut |
948 |
|
|
949 |
|
sub rec1 { |
950 |
|
my $f = shift; |
951 |
|
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
952 |
|
return unless (defined($rec) && defined($rec->{$f})); |
953 |
|
warn "rec1($f) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
954 |
|
if (ref($rec->{$f}) eq 'ARRAY') { |
955 |
|
my @out; |
956 |
|
foreach my $h ( @{ $rec->{$f} } ) { |
957 |
|
if (ref($h) eq 'HASH') { |
958 |
|
push @out, ( _pack_subfields_hash( $h ) ); |
959 |
} else { |
} else { |
960 |
return $$rec->{$f}->[$i]; |
push @out, $h; |
961 |
} |
} |
962 |
|
} |
963 |
|
return @out; |
964 |
|
} elsif( defined($rec->{$f}) ) { |
965 |
|
return $rec->{$f}; |
966 |
|
} |
967 |
|
} |
968 |
|
|
969 |
|
=head2 rec2 |
970 |
|
|
971 |
|
Return all values in specific field and subfield |
972 |
|
|
973 |
|
@v = rec2('200','a') |
974 |
|
|
975 |
|
=cut |
976 |
|
|
977 |
|
sub rec2 { |
978 |
|
my $f = shift; |
979 |
|
return unless (defined($rec && $rec->{$f})); |
980 |
|
my $sf = shift; |
981 |
|
warn "rec2($f,$sf) = ", dump( $rec->{$f} ), $/ if ($debug > 1); |
982 |
|
return map { |
983 |
|
if (ref($_->{$sf}) eq 'ARRAY') { |
984 |
|
@{ $_->{$sf} }; |
985 |
} else { |
} else { |
986 |
return ''; |
$_->{$sf}; |
987 |
} |
} |
988 |
|
} grep { ref($_) eq 'HASH' && $_->{$sf} } @{ $rec->{$f} }; |
989 |
|
} |
990 |
|
|
991 |
|
=head2 rec |
992 |
|
|
993 |
|
syntaxtic sugar for |
994 |
|
|
995 |
|
@v = rec('200') |
996 |
|
@v = rec('200','a') |
997 |
|
|
998 |
|
If rec() returns just single value, it will |
999 |
|
return scalar, not array. |
1000 |
|
|
1001 |
|
=cut |
1002 |
|
|
1003 |
|
sub frec { |
1004 |
|
my @out = rec(@_); |
1005 |
|
warn "rec(",dump(@_),") has more than one return value, ignoring\n" if $#out > 0; |
1006 |
|
return shift @out; |
1007 |
|
} |
1008 |
|
|
1009 |
|
sub rec { |
1010 |
|
my @out; |
1011 |
|
if ($#_ == 0) { |
1012 |
|
@out = rec1(@_); |
1013 |
|
} elsif ($#_ == 1) { |
1014 |
|
@out = rec2(@_); |
1015 |
|
} |
1016 |
|
if ($#out == 0 && ! wantarray) { |
1017 |
|
return $out[0]; |
1018 |
|
} elsif (@out) { |
1019 |
|
return @out; |
1020 |
} else { |
} else { |
1021 |
return ''; |
return ''; |
1022 |
} |
} |
1023 |
} |
} |
1024 |
|
|
1025 |
|
=head2 regex |
1026 |
|
|
1027 |
=head2 apply_format |
Apply regex to some or all values |
1028 |
|
|
1029 |
Apply format specified in tag with C<format_name="name"> and |
@v = regex( 's/foo/bar/g', @v ); |
|
C<format_delimiter=";;">. |
|
1030 |
|
|
1031 |
my $text = $webpac->apply_format($format_name,$format_delimiter,$data); |
=cut |
1032 |
|
|
1033 |
Formats can contain C<lookup{...}> if you need them. |
sub regex { |
1034 |
|
my $r = shift; |
1035 |
|
my @out; |
1036 |
|
#warn "r: $r\n", dump(\@_); |
1037 |
|
foreach my $t (@_) { |
1038 |
|
next unless ($t); |
1039 |
|
eval "\$t =~ $r"; |
1040 |
|
push @out, $t if ($t && $t ne ''); |
1041 |
|
} |
1042 |
|
return @out; |
1043 |
|
} |
1044 |
|
|
1045 |
|
=head2 prefix |
1046 |
|
|
1047 |
|
Prefix all values with a string |
1048 |
|
|
1049 |
|
@v = prefix( 'my_', @v ); |
1050 |
|
|
1051 |
=cut |
=cut |
1052 |
|
|
1053 |
sub apply_format { |
sub prefix { |
1054 |
my $self = shift; |
my $p = shift; |
1055 |
|
return @_ unless defined( $p ); |
1056 |
|
return map { $p . $_ } grep { defined($_) } @_; |
1057 |
|
} |
1058 |
|
|
1059 |
|
=head2 suffix |
1060 |
|
|
1061 |
my ($name,$delimiter,$data) = @_; |
suffix all values with a string |
1062 |
|
|
1063 |
|
@v = suffix( '_my', @v ); |
1064 |
|
|
1065 |
|
=cut |
1066 |
|
|
1067 |
|
sub suffix { |
1068 |
|
my $s = shift; |
1069 |
|
return @_ unless defined( $s ); |
1070 |
|
return map { $_ . $s } grep { defined($_) } @_; |
1071 |
|
} |
1072 |
|
|
1073 |
my $log = $self->_get_logger(); |
=head2 surround |
1074 |
|
|
1075 |
if (! $self->{'import_xml'}->{'format'}->{$name}) { |
surround all values with a two strings |
1076 |
$log->warn("<format name=\"$name\"> is not defined in ",$self->{'import_xml_file'}); |
|
1077 |
return $data; |
@v = surround( 'prefix_', '_suffix', @v ); |
1078 |
|
|
1079 |
|
=cut |
1080 |
|
|
1081 |
|
sub surround { |
1082 |
|
my $p = shift; |
1083 |
|
my $s = shift; |
1084 |
|
$p = '' unless defined( $p ); |
1085 |
|
$s = '' unless defined( $s ); |
1086 |
|
return map { $p . $_ . $s } grep { defined($_) } @_; |
1087 |
|
} |
1088 |
|
|
1089 |
|
=head2 first |
1090 |
|
|
1091 |
|
Return first element |
1092 |
|
|
1093 |
|
$v = first( @v ); |
1094 |
|
|
1095 |
|
=cut |
1096 |
|
|
1097 |
|
sub first { |
1098 |
|
my $r = shift; |
1099 |
|
return $r; |
1100 |
|
} |
1101 |
|
|
1102 |
|
=head2 lookup |
1103 |
|
|
1104 |
|
Consult lookup hashes for some value |
1105 |
|
|
1106 |
|
@v = lookup( |
1107 |
|
sub { |
1108 |
|
'ffkk/peri/mfn'.rec('000') |
1109 |
|
}, |
1110 |
|
'ffkk','peri','200-a-200-e', |
1111 |
|
sub { |
1112 |
|
first(rec(200,'a')).' '.first(rec('200','e')) |
1113 |
} |
} |
1114 |
|
); |
1115 |
|
|
1116 |
|
Code like above will be B<automatically generated> using L<WebPAC::Parse> from |
1117 |
|
normal lookup definition in C<conf/lookup/something.pl> which looks like: |
1118 |
|
|
1119 |
|
lookup( |
1120 |
|
# which results to return from record recorded in lookup |
1121 |
|
sub { 'ffkk/peri/mfn' . rec('000') }, |
1122 |
|
# from which database and input |
1123 |
|
'ffkk','peri', |
1124 |
|
# such that following values match |
1125 |
|
sub { first(rec(200,'a')) . ' ' . first(rec('200','e')) }, |
1126 |
|
# if this part is missing, we will try to match same fields |
1127 |
|
# from lookup record and current one, or you can override |
1128 |
|
# which records to use from current record using |
1129 |
|
sub { rec('900','x') . ' ' . rec('900','y') }, |
1130 |
|
) |
1131 |
|
|
1132 |
|
You can think about this lookup as SQL (if that helps): |
1133 |
|
|
1134 |
|
select |
1135 |
|
sub { what } |
1136 |
|
from |
1137 |
|
database, input |
1138 |
|
where |
1139 |
|
sub { filter from lookuped record } |
1140 |
|
having |
1141 |
|
sub { optional filter on current record } |
1142 |
|
|
1143 |
|
Easy as pie, right? |
1144 |
|
|
1145 |
|
=cut |
1146 |
|
|
1147 |
$log->warn("no delimiter for format $name") if (! $delimiter); |
sub lookup { |
1148 |
|
my ($what, $database, $input, $key, $having) = @_; |
1149 |
|
|
1150 |
my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); |
confess "lookup needs 5 arguments: what, database, input, key, having\n" unless ($#_ == 4); |
1151 |
|
|
1152 |
my @data = split(/\Q$delimiter\E/, $data); |
warn "## lookup ($database, $input, $key)", $/ if ($debug > 1); |
1153 |
|
return unless (defined($lookup->{$database}->{$input}->{$key})); |
1154 |
|
|
1155 |
my $out = sprintf($format, @data); |
confess "lookup really need load_row_coderef added to data_structure\n" unless ($load_row_coderef); |
|
$log->debug("using format $name [$format] on $data to produce: $out"); |
|
1156 |
|
|
1157 |
if ($self->{'lookup_regex'} && $out =~ $self->{'lookup_regex'}) { |
my $mfns; |
1158 |
return $self->{'lookup'}->lookup($out); |
my @having = $having->(); |
1159 |
|
|
1160 |
|
warn "## having = ", dump( @having ) if ($debug > 2); |
1161 |
|
|
1162 |
|
foreach my $h ( @having ) { |
1163 |
|
if (defined($lookup->{$database}->{$input}->{$key}->{$h})) { |
1164 |
|
warn "lookup for $database/$input/$key/$h return ",dump($lookup->{$database}->{$input}->{$key}->{$h}),"\n" if ($debug); |
1165 |
|
$mfns->{$_}++ foreach keys %{ $lookup->{$database}->{$input}->{$key}->{$h} }; |
1166 |
|
} |
1167 |
|
} |
1168 |
|
|
1169 |
|
return unless ($mfns); |
1170 |
|
|
1171 |
|
my @mfns = sort keys %$mfns; |
1172 |
|
|
1173 |
|
warn "# lookup loading $database/$input/$key mfn ", join(",",@mfns)," having ",dump(@having),"\n" if ($debug); |
1174 |
|
|
1175 |
|
my $old_rec = $rec; |
1176 |
|
my @out; |
1177 |
|
|
1178 |
|
foreach my $mfn (@mfns) { |
1179 |
|
$rec = $load_row_coderef->( $database, $input, $mfn ); |
1180 |
|
|
1181 |
|
warn "got $database/$input/$mfn = ", dump($rec), $/ if ($debug); |
1182 |
|
|
1183 |
|
my @vals = $what->(); |
1184 |
|
|
1185 |
|
push @out, ( @vals ); |
1186 |
|
|
1187 |
|
warn "lookup for mfn $mfn returned ", dump(@vals), $/ if ($debug); |
1188 |
|
} |
1189 |
|
|
1190 |
|
# if (ref($lookup->{$k}) eq 'ARRAY') { |
1191 |
|
# return @{ $lookup->{$k} }; |
1192 |
|
# } else { |
1193 |
|
# return $lookup->{$k}; |
1194 |
|
# } |
1195 |
|
|
1196 |
|
$rec = $old_rec; |
1197 |
|
|
1198 |
|
warn "## lookup returns = ", dump(@out), $/ if ($debug); |
1199 |
|
|
1200 |
|
if ($#out == 0) { |
1201 |
|
return $out[0]; |
1202 |
} else { |
} else { |
1203 |
return $out; |
return @out; |
1204 |
|
} |
1205 |
|
} |
1206 |
|
|
1207 |
|
=head2 save_into_lookup |
1208 |
|
|
1209 |
|
Save value into lookup. It associates current database, input |
1210 |
|
and specific keys with one or more values which will be |
1211 |
|
associated over MFN. |
1212 |
|
|
1213 |
|
MFN will be extracted from first occurence current of field 000 |
1214 |
|
in current record, or if it doesn't exist from L<_set_config> C<_mfn>. |
1215 |
|
|
1216 |
|
my $nr = save_into_lookup($database,$input,$key,sub { |
1217 |
|
# code which produce one or more values |
1218 |
|
}); |
1219 |
|
|
1220 |
|
It returns number of items saved. |
1221 |
|
|
1222 |
|
This function shouldn't be called directly, it's called from code created by |
1223 |
|
L<WebPAC::Parser>. |
1224 |
|
|
1225 |
|
=cut |
1226 |
|
|
1227 |
|
sub save_into_lookup { |
1228 |
|
my ($database,$input,$key,$coderef) = @_; |
1229 |
|
die "save_into_lookup needs database" unless defined($database); |
1230 |
|
die "save_into_lookup needs input" unless defined($input); |
1231 |
|
die "save_into_lookup needs key" unless defined($key); |
1232 |
|
die "save_into_lookup needs CODE" unless ( defined($coderef) && ref($coderef) eq 'CODE' ); |
1233 |
|
|
1234 |
|
warn "## save_into_lookup rec = ", dump($rec), " config = ", dump($config), $/ if ($debug > 2); |
1235 |
|
|
1236 |
|
my $mfn = |
1237 |
|
defined($rec->{'000'}->[0]) ? $rec->{'000'}->[0] : |
1238 |
|
defined($config->{_mfn}) ? $config->{_mfn} : |
1239 |
|
die "mfn not defined or zero"; |
1240 |
|
|
1241 |
|
my $nr = 0; |
1242 |
|
|
1243 |
|
foreach my $v ( $coderef->() ) { |
1244 |
|
$lookup->{$database}->{$input}->{$key}->{$v}->{$mfn}++; |
1245 |
|
warn "# saved lookup $database/$input/$key [$v] $mfn\n" if ($debug > 1); |
1246 |
|
$nr++; |
1247 |
} |
} |
1248 |
|
|
1249 |
|
return $nr; |
1250 |
} |
} |
1251 |
|
|
1252 |
=head2 sort_arr |
=head2 config |
1253 |
|
|
1254 |
Sort array ignoring case and html in data |
Consult config values stored in C<config.yml> |
1255 |
|
|
1256 |
my @sorted = $webpac->sort_arr(@unsorted); |
# return database code (key under databases in yaml) |
1257 |
|
$database_code = config(); # use _ from hash |
1258 |
|
$database_name = config('name'); |
1259 |
|
$database_input_name = config('input name'); |
1260 |
|
|
1261 |
|
Up to three levels are supported. |
1262 |
|
|
1263 |
=cut |
=cut |
1264 |
|
|
1265 |
sub sort_arr { |
sub config { |
1266 |
my $self = shift; |
return unless ($config); |
1267 |
|
|
1268 |
|
my $p = shift; |
1269 |
|
|
1270 |
my $log = $self->_get_logger(); |
$p ||= ''; |
1271 |
|
|
1272 |
# FIXME add Schwartzian Transformation? |
my $v; |
1273 |
|
|
1274 |
my @sorted = sort { |
warn "### getting config($p)\n" if ($debug > 1); |
1275 |
$a =~ s#<[^>]+/*>##; |
|
1276 |
$b =~ s#<[^>]+/*>##; |
my @p = split(/\s+/,$p); |
1277 |
lc($b) cmp lc($a) |
if ($#p < 0) { |
1278 |
} @_; |
$v = $config->{ '_' }; # special, database code |
1279 |
$log->debug("sorted values: ",sub { join(", ",@sorted) }); |
} else { |
1280 |
|
|
1281 |
|
my $c = dclone( $config ); |
1282 |
|
|
1283 |
|
foreach my $k (@p) { |
1284 |
|
warn "### k: $k c = ",dump($c),$/ if ($debug > 1); |
1285 |
|
if (ref($c) eq 'ARRAY') { |
1286 |
|
$c = shift @$c; |
1287 |
|
warn "config($p) taking first occurence of '$k', probably not what you wanted!\n"; |
1288 |
|
last; |
1289 |
|
} |
1290 |
|
|
1291 |
|
if (! defined($c->{$k}) ) { |
1292 |
|
$c = undef; |
1293 |
|
last; |
1294 |
|
} else { |
1295 |
|
$c = $c->{$k}; |
1296 |
|
} |
1297 |
|
} |
1298 |
|
$v = $c if ($c); |
1299 |
|
|
1300 |
|
} |
1301 |
|
|
1302 |
return @sorted; |
warn "## config( '$p' ) = ",dump( $v ),$/ if ($v && $debug); |
1303 |
|
warn "config( '$p' ) is empty\n" if (! $v); |
1304 |
|
|
1305 |
|
return $v; |
1306 |
} |
} |
1307 |
|
|
1308 |
|
=head2 id |
1309 |
|
|
1310 |
=head1 INTERNAL METHODS |
Returns unique id of this record |
1311 |
|
|
1312 |
=head2 _sort_by_order |
$id = id(); |
1313 |
|
|
1314 |
Sort xml tags data structure accoding to C<order=""> attribute. |
Returns C<42/2> for 2nd occurence of MFN 42. |
1315 |
|
|
1316 |
=cut |
=cut |
1317 |
|
|
1318 |
sub _sort_by_order { |
sub id { |
1319 |
my $self = shift; |
my $mfn = $config->{_mfn} || die "no _mfn in config data"; |
1320 |
|
return $mfn . $#{$marc_record} ? $#{$marc_record} + 1 : ''; |
1321 |
|
} |
1322 |
|
|
1323 |
|
=head2 join_with |
1324 |
|
|
1325 |
my $va = $self->{'import_xml'}->{'indexer'}->{$a}->{'order'} || |
Joins walues with some delimiter |
1326 |
$self->{'import_xml'}->{'indexer'}->{$a}; |
|
1327 |
my $vb = $self->{'import_xml'}->{'indexer'}->{$b}->{'order'} || |
$v = join_with(", ", @v); |
1328 |
$self->{'import_xml'}->{'indexer'}->{$b}; |
|
1329 |
|
=cut |
1330 |
|
|
1331 |
return $va <=> $vb; |
sub join_with { |
1332 |
|
my $d = shift; |
1333 |
|
warn "### join_with('$d',",dump(@_),")\n" if ($debug > 2); |
1334 |
|
my $v = join($d, grep { defined($_) && $_ ne '' } @_); |
1335 |
|
return '' unless defined($v); |
1336 |
|
return $v; |
1337 |
} |
} |
1338 |
|
|
1339 |
=head2 _x |
=head2 split_rec_on |
1340 |
|
|
1341 |
Convert strings from C<conf/normalize/*.xml> encoding into application |
Split record subfield on some regex and take one of parts out |
|
specific encoding (optinally specified using C<code_page> to C<new> |
|
|
constructor). |
|
1342 |
|
|
1343 |
my $text = $n->_x('normalize text string'); |
$a_before_semi_column = |
1344 |
|
split_rec_on('200','a', /\s*;\s*/, $part); |
1345 |
|
|
1346 |
This is a stub so that other modules doesn't have to implement it. |
C<$part> is optional number of element. First element is |
1347 |
|
B<1>, not 0! |
1348 |
|
|
1349 |
|
If there is no C<$part> parameter or C<$part> is 0, this function will |
1350 |
|
return all values produced by splitting. |
1351 |
|
|
1352 |
=cut |
=cut |
1353 |
|
|
1354 |
sub _x { |
sub split_rec_on { |
1355 |
my $self = shift; |
die "split_rec_on need (fld,sf,regex[,part]" if ($#_ < 2); |
1356 |
return shift; |
|
1357 |
|
my ($fld, $sf, $regex, $part) = @_; |
1358 |
|
warn "### regex ", ref($regex), $regex, $/ if ($debug > 2); |
1359 |
|
|
1360 |
|
my @r = rec( $fld, $sf ); |
1361 |
|
my $v = shift @r; |
1362 |
|
warn "### first rec($fld,$sf) = ",dump($v),$/ if ($debug > 2); |
1363 |
|
|
1364 |
|
return '' if ( ! defined($v) || $v =~ /^\s*$/); |
1365 |
|
|
1366 |
|
my @s = split( $regex, $v ); |
1367 |
|
warn "## split_rec_on($fld,$sf,$regex,$part) = ",dump(@s),$/ if ($debug > 1); |
1368 |
|
if ($part && $part > 0) { |
1369 |
|
return $s[ $part - 1 ]; |
1370 |
|
} else { |
1371 |
|
return @s; |
1372 |
|
} |
1373 |
} |
} |
1374 |
|
|
1375 |
|
my $hash; |
1376 |
|
|
1377 |
=head1 AUTHOR |
=head2 set |
1378 |
|
|
1379 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
set( key => 'value' ); |
1380 |
|
|
1381 |
=head1 COPYRIGHT & LICENSE |
=cut |
1382 |
|
|
1383 |
|
sub set { |
1384 |
|
my ($k,$v) = @_; |
1385 |
|
warn "## set ( $k => ", dump($v), " )", $/ if ( $debug ); |
1386 |
|
$hash->{$k} = $v; |
1387 |
|
}; |
1388 |
|
|
1389 |
|
=head2 get |
1390 |
|
|
1391 |
|
get( 'key' ); |
1392 |
|
|
1393 |
|
=cut |
1394 |
|
|
1395 |
|
sub get { |
1396 |
|
my $k = shift || return; |
1397 |
|
my $v = $hash->{$k}; |
1398 |
|
warn "## get $k = ", dump( $v ), $/ if ( $debug ); |
1399 |
|
return $v; |
1400 |
|
} |
1401 |
|
|
1402 |
Copyright 2005 Dobrica Pavlinusic, All Rights Reserved. |
=head2 count |
1403 |
|
|
1404 |
This program is free software; you can redistribute it and/or modify it |
if ( count( @result ) == 1 ) { |
1405 |
under the same terms as Perl itself. |
# do something if only 1 result is there |
1406 |
|
} |
1407 |
|
|
1408 |
=cut |
=cut |
1409 |
|
|
1410 |
1; # End of WebPAC::Normalize |
sub count { |
1411 |
|
warn "## count ",dump(@_),$/ if ( $debug ); |
1412 |
|
return @_ . ''; |
1413 |
|
} |
1414 |
|
|
1415 |
|
# END |
1416 |
|
1; |