2 |
|
|
3 |
use warnings; |
use warnings; |
4 |
use strict; |
use strict; |
5 |
|
use base 'WebPAC::Common'; |
6 |
use Data::Dumper; |
use Data::Dumper; |
|
use Storable; |
|
7 |
|
|
8 |
=head1 NAME |
=head1 NAME |
9 |
|
|
10 |
WebPAC::Normalize - normalisation of source file |
WebPAC::Normalize - data mungling for normalisation |
11 |
|
|
12 |
=head1 VERSION |
=head1 VERSION |
13 |
|
|
14 |
Version 0.01 |
Version 0.08 |
15 |
|
|
16 |
=cut |
=cut |
17 |
|
|
18 |
our $VERSION = '0.01'; |
our $VERSION = '0.08'; |
19 |
|
|
20 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
21 |
|
|
22 |
This package contains code that could be helpful in implementing different |
This package contains code that mungle data to produce normalized format. |
23 |
normalisation front-ends. |
|
24 |
|
It contains several assumptions: |
25 |
|
|
26 |
|
=over |
27 |
|
|
28 |
|
=item * |
29 |
|
|
30 |
|
format of fields is defined using C<v123^a> notation for repeatable fields |
31 |
|
or C<s123^a> for single (or first) value, where C<123> is field number and |
32 |
|
C<a> is subfield. |
33 |
|
|
34 |
|
=item * |
35 |
|
|
36 |
|
source data records (C<$rec>) have unique identifiers in field C<000> |
37 |
|
|
38 |
|
=item * |
39 |
|
|
40 |
|
optional C<eval{length('v123^a') == 3}> tag at B<beginning of format> will be |
41 |
|
perl code that is evaluated before producing output (value of field will be |
42 |
|
interpolated before that) |
43 |
|
|
44 |
|
=item * |
45 |
|
|
46 |
|
optional C<filter{filter_name}> at B<begining of format> will apply perl |
47 |
|
code defined as code ref on format after field substitution to producing |
48 |
|
output |
49 |
|
|
50 |
|
There is one built-in filter called C<regex> which can be use like this: |
51 |
|
|
52 |
|
filter{regex(s/foo/bar/)} |
53 |
|
|
54 |
|
=item * |
55 |
|
|
56 |
|
optional C<lookup{...}> will be then performed. See C<WebPAC::Lookups>. |
57 |
|
|
58 |
|
=item * |
59 |
|
|
60 |
|
at end, optional C<format>s rules are resolved. Format rules are similar to |
61 |
|
C<sprintf> and can also contain C<lookup{...}> which is performed after |
62 |
|
values are inserted in format. |
63 |
|
|
64 |
|
=back |
65 |
|
|
66 |
|
This also describes order in which transformations are applied (eval, |
67 |
|
filter, lookup, format) which is important to undestand when deciding how to |
68 |
|
solve your data mungling and normalisation process. |
69 |
|
|
70 |
|
|
71 |
|
|
72 |
|
|
73 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
74 |
|
|
77 |
Create new normalisation object |
Create new normalisation object |
78 |
|
|
79 |
my $n = new WebPAC::Normalize::Something( |
my $n = new WebPAC::Normalize::Something( |
80 |
cache_data_structure => './cache/ds/', |
filter => { |
81 |
|
'filter_name_1' => sub { |
82 |
|
# filter code |
83 |
|
return length($_); |
84 |
|
}, ... |
85 |
|
}, |
86 |
|
db => $db_obj, |
87 |
lookup_regex => $lookup->regex, |
lookup_regex => $lookup->regex, |
88 |
|
lookup => $lookup_obj, |
89 |
|
prefix => 'foobar', |
90 |
); |
); |
91 |
|
|
92 |
Optional parameter C<cache_data_structure> defines path to directory |
Parametar C<filter> defines user supplied snippets of perl code which can |
93 |
in which cache file for C<data_structure> call will be created. |
be use with C<filter{...}> notation. |
94 |
|
|
95 |
|
C<prefix> is used to form filename for database record (to support multiple |
96 |
|
source files which are joined in one database). |
97 |
|
|
98 |
Recommended parametar C<lookup_regex> is used to enable parsing of lookups |
Recommended parametar C<lookup_regex> is used to enable parsing of lookups |
99 |
in structures. |
in structures. If you pass this parametar, you must also pass C<lookup> |
100 |
|
which is C<WebPAC::Lookup> object. |
101 |
|
|
102 |
=cut |
=cut |
103 |
|
|
106 |
my $self = {@_}; |
my $self = {@_}; |
107 |
bless($self, $class); |
bless($self, $class); |
108 |
|
|
109 |
$self->setup_cache_dir( $self->{'cache_data_structure'} ); |
my $r = $self->{'lookup_regex'} ? 1 : 0; |
110 |
|
my $l = $self->{'lookup'} ? 1 : 0; |
|
$self ? return $self : return undef; |
|
|
} |
|
|
|
|
|
=head2 setup_cache_dir |
|
111 |
|
|
112 |
Check if specified cache directory exist, and if not, disable caching. |
my $log = $self->_get_logger(); |
|
|
|
|
$setup_cache_dir('./cache/ds/'); |
|
|
|
|
|
If you pass false or zero value to this function, it will disable |
|
|
cacheing. |
|
113 |
|
|
114 |
=cut |
# those two must be in pair |
115 |
|
if ( ($r & $l) != ($r || $l) ) { |
116 |
|
my $log = $self->_get_logger(); |
117 |
|
$log->logdie("lookup_regex and lookup must be in pair"); |
118 |
|
} |
119 |
|
|
120 |
sub setup_cache_dir { |
$log->logdie("lookup must be WebPAC::Lookup object") if ($self->{'lookup'} && ! $self->{'lookup'}->isa('WebPAC::Lookup')); |
|
my $self = shift; |
|
121 |
|
|
122 |
my $dir = shift; |
$log->warn("no prefix defined. please check that!") unless ($self->{'prefix'}); |
123 |
|
|
124 |
my $log = $self->_get_logger(); |
$log->debug("using lookup regex: ", $self->{lookup_regex}) if ($r && $l); |
125 |
|
|
126 |
if ($dir) { |
if (! $self->{filter} || ! $self->{filter}->{regex}) { |
127 |
my $msg; |
$log->debug("adding built-in filter regex"); |
128 |
if (! -e $dir) { |
$self->{filter}->{regex} = sub { |
129 |
$msg = "doesn't exist"; |
my ($val, $regex) = @_; |
130 |
} elsif (! -d $dir) { |
eval "\$val =~ $regex"; |
131 |
$msg = "is not directory"; |
return $val; |
132 |
} elsif (! -w $dir) { |
}; |
|
$msg = "not writable"; |
|
|
} |
|
|
|
|
|
if ($msg) { |
|
|
undef $self->{'cache_data_structure'}; |
|
|
$log->warn("cache_data_structure $dir $msg, disabling..."); |
|
|
} else { |
|
|
$log->debug("using cache dir $dir"); |
|
|
} |
|
|
} else { |
|
|
$log->debug("disabling cache"); |
|
|
undef $self->{'cache_data_structure'}; |
|
133 |
} |
} |
134 |
|
|
135 |
|
$self ? return $self : return undef; |
136 |
} |
} |
137 |
|
|
138 |
|
|
143 |
|
|
144 |
This structures are used to produce output. |
This structures are used to produce output. |
145 |
|
|
146 |
my @ds = $webpac->data_structure($rec); |
my $ds = $webpac->data_structure($rec); |
|
|
|
|
B<Note: historical oddity follows> |
|
|
|
|
|
This method will also set C<< $webpac->{'currnet_filename'} >> if there is |
|
|
C<< <filename> >> tag and C<< $webpac->{'headline'} >> if there is |
|
|
C<< <headline> >> tag. |
|
147 |
|
|
148 |
=cut |
=cut |
149 |
|
|
155 |
my $rec = shift; |
my $rec = shift; |
156 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
157 |
|
|
158 |
|
$log->debug("data_structure rec = ", sub { Dumper($rec) }); |
159 |
|
|
160 |
|
$log->logdie("need unique ID (mfn) in field 000 of record " . Dumper($rec) ) unless (defined($rec->{'000'})); |
161 |
|
|
162 |
|
my $id = $rec->{'000'}->[0] || $log->logdie("field 000 isn't array!"); |
163 |
|
|
164 |
my $cache_file; |
my $cache_file; |
165 |
|
|
166 |
if (my $cache_path = $self->{'cache_data_structure'}) { |
if ($self->{'db'}) { |
167 |
my $id = $rec->{'000'}; |
my $ds = $self->{'db'}->load_ds( id => $id, prefix => $self->{prefix} ); |
168 |
$id = $rec->{'000'}->[0] if ($id =~ m/^ARRAY/o); |
$log->debug("load_ds( rec = ", sub { Dumper($rec) }, ") = ", sub { Dumper($ds) }); |
169 |
unless (defined($id)) { |
return $ds if ($ds); |
170 |
$log->warn("Can't use cache_data_structure on records without unique identifier in field 000"); |
$log->debug("cache miss, creating"); |
|
undef $self->{'cache_data_structure'}; |
|
|
} else { |
|
|
$cache_file = "$cache_path/$id"; |
|
|
if (-r $cache_file) { |
|
|
my $ds_ref = retrieve($cache_file); |
|
|
if ($ds_ref) { |
|
|
$log->debug("cache hit: $cache_file"); |
|
|
my $ok = 1; |
|
|
foreach my $f (qw(current_filename headline)) { |
|
|
if ($ds_ref->{$f}) { |
|
|
$self->{$f} = $ds_ref->{$f}; |
|
|
} else { |
|
|
$ok = 0; |
|
|
} |
|
|
}; |
|
|
if ($ok && $ds_ref->{'ds'}) { |
|
|
return @{ $ds_ref->{'ds'} }; |
|
|
} else { |
|
|
$log->warn("cache_data_structure $cache_path corrupt. Use rm $cache_path/* to re-create it on next run!"); |
|
|
undef $self->{'cache_data_structure'}; |
|
|
} |
|
|
} |
|
|
} |
|
|
} |
|
171 |
} |
} |
172 |
|
|
|
undef $self->{'currnet_filename'}; |
|
|
undef $self->{'headline'}; |
|
|
|
|
173 |
my @sorted_tags; |
my @sorted_tags; |
174 |
if ($self->{tags_by_order}) { |
if ($self->{tags_by_order}) { |
175 |
@sorted_tags = @{$self->{tags_by_order}}; |
@sorted_tags = @{$self->{tags_by_order}}; |
178 |
$self->{tags_by_order} = \@sorted_tags; |
$self->{tags_by_order} = \@sorted_tags; |
179 |
} |
} |
180 |
|
|
181 |
my @ds; |
my $ds; |
182 |
|
|
183 |
$log->debug("tags: ",sub { join(", ",@sorted_tags) }); |
$log->debug("tags: ",sub { join(", ",@sorted_tags) }); |
184 |
|
|
189 |
#print "field $field [",$self->{'tag'},"] = ",Dumper($self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}); |
#print "field $field [",$self->{'tag'},"] = ",Dumper($self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}); |
190 |
|
|
191 |
foreach my $tag (@{$self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}}) { |
foreach my $tag (@{$self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}}) { |
192 |
my $format = $tag->{'value'} || $tag->{'content'}; |
my $format; |
193 |
|
|
194 |
$log->debug("format: $format"); |
$log->logdie("expected tag HASH and got $tag") unless (ref($tag) eq 'HASH'); |
195 |
|
$format = $tag->{'value'} || $tag->{'content'}; |
196 |
|
|
197 |
my @v; |
my @v; |
198 |
if ($self->{'lookup_regex'} && $format =~ $self->{'lookup_regex'}) { |
if ($self->{'lookup_regex'} && $format =~ $self->{'lookup_regex'}) { |
200 |
} else { |
} else { |
201 |
@v = $self->parse_to_arr($rec,$format); |
@v = $self->parse_to_arr($rec,$format); |
202 |
} |
} |
203 |
next if (! @v); |
if (! @v) { |
204 |
|
$log->debug("$field <",$self->{tag},"> format: $format no values"); |
205 |
|
# next; |
206 |
|
} else { |
207 |
|
$log->debug("$field <",$self->{tag},"> format: $format values: ", join(",", @v)); |
208 |
|
} |
209 |
|
|
210 |
if ($tag->{'sort'}) { |
if ($tag->{'sort'}) { |
211 |
@v = $self->sort_arr(@v); |
@v = $self->sort_arr(@v); |
216 |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
217 |
} |
} |
218 |
|
|
|
if ($field eq 'filename') { |
|
|
$self->{'current_filename'} = join('',@v); |
|
|
$log->debug("filename: ",$self->{'current_filename'}); |
|
|
} elsif ($field eq 'headline') { |
|
|
$self->{'headline'} .= join('',@v); |
|
|
$log->debug("headline: ",$self->{'headline'}); |
|
|
next; # don't return headline in data_structure! |
|
|
} |
|
|
|
|
219 |
# delimiter will join repeatable fields |
# delimiter will join repeatable fields |
220 |
if ($tag->{'delimiter'}) { |
if ($tag->{'delimiter'}) { |
221 |
@v = ( join($tag->{'delimiter'}, @v) ); |
@v = ( join($tag->{'delimiter'}, @v) ); |
222 |
} |
} |
223 |
|
|
224 |
# default types |
# default types |
225 |
my @types = qw(display swish); |
my @types = qw(display search); |
226 |
# override by type attribute |
# override by type attribute |
227 |
@types = ( $tag->{'type'} ) if ($tag->{'type'}); |
@types = ( $tag->{'type'} ) if ($tag->{'type'}); |
228 |
|
|
229 |
foreach my $type (@types) { |
foreach my $type (@types) { |
230 |
# append to previous line? |
# append to previous line? |
231 |
$log->debug("type: $type ",sub { join(" ",@v) }, $row->{'append'} || 'no append'); |
$log->debug("tag $field / $type [",sub { join(",",@v) }, "] ", $row->{'append'} || 'no append'); |
232 |
if ($tag->{'append'}) { |
if ($tag->{'append'}) { |
233 |
|
|
234 |
# I will delimit appended part with |
# I will delimit appended part with |
255 |
|
|
256 |
# TODO: name_sigular, name_plural |
# TODO: name_sigular, name_plural |
257 |
my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; |
my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; |
258 |
$row->{'name'} = $name ? $self->_x($name) : $field; |
my $row_name = $name ? $self->_x($name) : $field; |
259 |
|
|
260 |
# post-sort all values in field |
# post-sort all values in field |
261 |
if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { |
if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { |
262 |
$log->warn("sort at field tag not implemented"); |
$log->warn("sort at field tag not implemented"); |
263 |
} |
} |
264 |
|
|
265 |
push @ds, $row; |
$ds->{$row_name} = $row; |
266 |
|
|
267 |
$log->debug("row $field: ",sub { Dumper($row) }); |
$log->debug("row $field: ",sub { Dumper($row) }); |
268 |
} |
} |
269 |
|
|
270 |
} |
} |
271 |
|
|
272 |
if ($cache_file) { |
$self->{'db'}->save_ds( |
273 |
store { |
id => $id, |
274 |
ds => \@ds, |
ds => $ds, |
275 |
current_filename => $self->{'current_filename'}, |
prefix => $self->{prefix}, |
276 |
headline => $self->{'headline'}, |
) if ($self->{'db'}); |
|
}, $cache_file; |
|
|
$log->debug("created storable cache file $cache_file"); |
|
|
} |
|
|
|
|
|
return @ds; |
|
|
|
|
|
} |
|
|
|
|
|
=head2 apply_format |
|
|
|
|
|
Apply format specified in tag with C<format_name="name"> and |
|
|
C<format_delimiter=";;">. |
|
|
|
|
|
my $text = $webpac->apply_format($format_name,$format_delimiter,$data); |
|
|
|
|
|
Formats can contain C<lookup{...}> if you need them. |
|
|
|
|
|
=cut |
|
|
|
|
|
sub apply_format { |
|
|
my $self = shift; |
|
|
|
|
|
my ($name,$delimiter,$data) = @_; |
|
|
|
|
|
my $log = $self->_get_logger(); |
|
|
|
|
|
if (! $self->{'import_xml'}->{'format'}->{$name}) { |
|
|
$log->warn("<format name=\"$name\"> is not defined in ",$self->{'import_xml_file'}); |
|
|
return $data; |
|
|
} |
|
|
|
|
|
$log->warn("no delimiter for format $name") if (! $delimiter); |
|
277 |
|
|
278 |
my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); |
$log->debug("ds: ", sub { Dumper($ds) }); |
279 |
|
|
280 |
my @data = split(/\Q$delimiter\E/, $data); |
$log->logconfess("data structure returned is not array any more!") if wantarray; |
281 |
|
|
282 |
my $out = sprintf($format, @data); |
return $ds; |
|
$log->debug("using format $name [$format] on $data to produce: $out"); |
|
|
|
|
|
if ($self->{'lookup_regex'} && $out =~ $self->{'lookup_regex'}) { |
|
|
return $self->lookup($out); |
|
|
} else { |
|
|
return $out; |
|
|
} |
|
283 |
|
|
284 |
} |
} |
285 |
|
|
291 |
|
|
292 |
my $text = $webpac->parse($rec,'eval{"v901^a" eq "Deskriptor"}descriptor: v250^a', $i); |
my $text = $webpac->parse($rec,'eval{"v901^a" eq "Deskriptor"}descriptor: v250^a', $i); |
293 |
|
|
294 |
|
Filters are implemented here. While simple form of filters looks like this: |
295 |
|
|
296 |
|
filter{name_of_filter} |
297 |
|
|
298 |
|
but, filters can also have variable number of parametars like this: |
299 |
|
|
300 |
|
filter{name_of_filter(param,param,param)} |
301 |
|
|
302 |
=cut |
=cut |
303 |
|
|
304 |
|
my $warn_once; |
305 |
|
|
306 |
sub parse { |
sub parse { |
307 |
my $self = shift; |
my $self = shift; |
308 |
|
|
320 |
|
|
321 |
my @out; |
my @out; |
322 |
|
|
323 |
$log->debug("format: $format"); |
$log->debug("format: $format [$i]"); |
324 |
|
|
325 |
my $eval_code; |
my $eval_code; |
326 |
# remove eval{...} from beginning |
# remove eval{...} from beginning |
330 |
# remove filter{...} from beginning |
# remove filter{...} from beginning |
331 |
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
332 |
|
|
333 |
|
# did we found any (att all) field from format in row? |
334 |
|
my $found_any; |
335 |
|
# prefix before first field which we preserve it $found_any |
336 |
my $prefix; |
my $prefix; |
337 |
my $all_found=0; |
|
338 |
|
my $f_step = 1; |
339 |
|
|
340 |
while ($format =~ s/^(.*?)(v|s)(\d+)(?:\^(\w))?//s) { |
while ($format =~ s/^(.*?)(v|s)(\d+)(?:\^(\w))?//s) { |
341 |
|
|
342 |
my $del = $1 || ''; |
my $del = $1 || ''; |
343 |
$prefix ||= $del if ($all_found == 0); |
$prefix = $del if ($f_step == 1); |
344 |
|
|
345 |
|
my $fld_type = lc($2); |
346 |
|
|
347 |
# repeatable index |
# repeatable index |
348 |
my $r = $i; |
my $r = $i; |
349 |
$r = 0 if (lc("$2") eq 's'); |
if ($fld_type eq 's') { |
350 |
|
if ($found_any->{'v'}) { |
351 |
|
$r = 0; |
352 |
|
} else { |
353 |
|
return; |
354 |
|
} |
355 |
|
} |
356 |
|
|
357 |
my $found = 0; |
my $found = 0; |
358 |
my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found); |
my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found); |
359 |
|
|
360 |
if ($found) { |
if ($found) { |
361 |
push @out, $del; |
$found_any->{$fld_type} += $found; |
362 |
|
|
363 |
|
# we will skip delimiter before first occurence of field! |
364 |
|
push @out, $del unless($found_any->{$fld_type} == 1); |
365 |
push @out, $tmp; |
push @out, $tmp; |
|
$all_found += $found; |
|
366 |
} |
} |
367 |
|
$f_step++; |
368 |
} |
} |
369 |
|
|
370 |
return if (! $all_found); |
# test if any fields found? |
371 |
|
return if (! $found_any->{'v'} && ! $found_any->{'s'}); |
372 |
|
|
373 |
my $out = join('',@out); |
my $out = join('',@out); |
374 |
|
|
388 |
return if (! $self->_eval($eval)); |
return if (! $self->_eval($eval)); |
389 |
} |
} |
390 |
|
|
391 |
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
if ($filter_name) { |
392 |
$log->debug("about to filter{$filter_name} format: $out"); |
my @filter_args; |
393 |
$out = $self->{'filter'}->{$filter_name}->($out); |
if ($filter_name =~ s/(\w+)\((.*)\)/$1/) { |
394 |
return unless(defined($out)); |
@filter_args = split(/,/, $2); |
395 |
$log->debug("filter result: $out"); |
} |
396 |
|
if ($self->{'filter'}->{$filter_name}) { |
397 |
|
$log->debug("about to filter{$filter_name} format: $out with arguments: ", join(",", @filter_args)); |
398 |
|
unshift @filter_args, $out; |
399 |
|
$out = $self->{'filter'}->{$filter_name}->(@filter_args); |
400 |
|
return unless(defined($out)); |
401 |
|
$log->debug("filter result: $out"); |
402 |
|
} elsif (! $warn_once->{$filter_name}) { |
403 |
|
$log->warn("trying to use undefined filter $filter_name"); |
404 |
|
$warn_once->{$filter_name}++; |
405 |
|
} |
406 |
} |
} |
407 |
|
|
408 |
return $out; |
return $out; |
438 |
return @arr; |
return @arr; |
439 |
} |
} |
440 |
|
|
441 |
|
|
442 |
|
=head2 fill_in |
443 |
|
|
444 |
|
Workhourse of all: takes record from in-memory structure of database and |
445 |
|
strings with placeholders and returns string or array of with substituted |
446 |
|
values from record. |
447 |
|
|
448 |
|
my $text = $webpac->fill_in($rec,'v250^a'); |
449 |
|
|
450 |
|
Optional argument is ordinal number for repeatable fields. By default, |
451 |
|
it's assume to be first repeatable field (fields are perl array, so first |
452 |
|
element is 0). |
453 |
|
Following example will read second value from repeatable field. |
454 |
|
|
455 |
|
my $text = $webpac->fill_in($rec,'Title: v250^a',1); |
456 |
|
|
457 |
|
This function B<does not> perform parsing of format to inteligenty skip |
458 |
|
delimiters before fields which aren't used. |
459 |
|
|
460 |
|
This method will automatically decode UTF-8 string to local code page |
461 |
|
if needed. |
462 |
|
|
463 |
|
=cut |
464 |
|
|
465 |
|
sub fill_in { |
466 |
|
my $self = shift; |
467 |
|
|
468 |
|
my $log = $self->_get_logger(); |
469 |
|
|
470 |
|
my $rec = shift || $log->logconfess("need data record"); |
471 |
|
my $format = shift || $log->logconfess("need format to parse"); |
472 |
|
# iteration (for repeatable fields) |
473 |
|
my $i = shift || 0; |
474 |
|
|
475 |
|
$log->logdie("infitite loop in format $format") if ($i > ($self->{'max_mfn'} || 9999)); |
476 |
|
|
477 |
|
# FIXME remove for speedup? |
478 |
|
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
479 |
|
|
480 |
|
if (utf8::is_utf8($format)) { |
481 |
|
$format = $self->_x($format); |
482 |
|
} |
483 |
|
|
484 |
|
my $found = 0; |
485 |
|
my $just_single = 1; |
486 |
|
|
487 |
|
my $eval_code; |
488 |
|
# remove eval{...} from beginning |
489 |
|
$eval_code = $1 if ($format =~ s/^eval{([^}]+)}//s); |
490 |
|
|
491 |
|
my $filter_name; |
492 |
|
# remove filter{...} from beginning |
493 |
|
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
494 |
|
|
495 |
|
# do actual replacement of placeholders |
496 |
|
# repeatable fields |
497 |
|
if ($format =~ s/v(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,$i,\$found)/ges) { |
498 |
|
$just_single = 0; |
499 |
|
} |
500 |
|
|
501 |
|
# non-repeatable fields |
502 |
|
if ($format =~ s/s(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,0,\$found)/ges) { |
503 |
|
return if ($i > 0 && $just_single); |
504 |
|
} |
505 |
|
|
506 |
|
if ($found) { |
507 |
|
$log->debug("format: $format"); |
508 |
|
if ($eval_code) { |
509 |
|
my $eval = $self->fill_in($rec,$eval_code,$i); |
510 |
|
return if (! $self->_eval($eval)); |
511 |
|
} |
512 |
|
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
513 |
|
$log->debug("filter '$filter_name' for $format"); |
514 |
|
$format = $self->{'filter'}->{$filter_name}->($format); |
515 |
|
return unless(defined($format)); |
516 |
|
$log->debug("filter result: $format"); |
517 |
|
} |
518 |
|
# do we have lookups? |
519 |
|
if ($self->{'lookup'}) { |
520 |
|
if ($self->{'lookup'}->can('lookup')) { |
521 |
|
my @lookup = $self->{lookup}->lookup($format); |
522 |
|
$log->debug("lookup $format", join(", ", @lookup)); |
523 |
|
return @lookup; |
524 |
|
} else { |
525 |
|
$log->warn("Have lookup object but can't invoke lookup method"); |
526 |
|
} |
527 |
|
} else { |
528 |
|
return $format; |
529 |
|
} |
530 |
|
} else { |
531 |
|
return; |
532 |
|
} |
533 |
|
} |
534 |
|
|
535 |
|
|
536 |
=head2 fill_in_to_arr |
=head2 fill_in_to_arr |
537 |
|
|
538 |
Similar to C<fill_in>, but returns array of all repeatable fields. Usable |
Similar to C<fill_in>, but returns array of all repeatable fields. Usable |
556 |
my $i = 0; |
my $i = 0; |
557 |
my @arr; |
my @arr; |
558 |
|
|
559 |
while (my @v = $self->fill_in($rec,$format_utf8,$i++)) { |
while (my $v = $self->fill_in($rec,$format_utf8,$i++)) { |
560 |
push @arr, @v; |
push @arr, $v; |
561 |
} |
} |
562 |
|
|
563 |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
565 |
return @arr; |
return @arr; |
566 |
} |
} |
567 |
|
|
568 |
|
|
569 |
|
=head2 get_data |
570 |
|
|
571 |
|
Returns value from record. |
572 |
|
|
573 |
|
my $text = $self->get_data(\$rec,$f,$sf,$i,\$found); |
574 |
|
|
575 |
|
Arguments are: |
576 |
|
record reference C<$rec>, |
577 |
|
field C<$f>, |
578 |
|
optional subfiled C<$sf>, |
579 |
|
index for repeatable values C<$i>. |
580 |
|
|
581 |
|
Optinal variable C<$found> will be incremeted if there |
582 |
|
is field. |
583 |
|
|
584 |
|
Returns value or empty string. |
585 |
|
|
586 |
|
=cut |
587 |
|
|
588 |
|
sub get_data { |
589 |
|
my $self = shift; |
590 |
|
|
591 |
|
my ($rec,$f,$sf,$i,$found) = @_; |
592 |
|
|
593 |
|
if ($$rec->{$f}) { |
594 |
|
return '' if (! $$rec->{$f}->[$i]); |
595 |
|
no strict 'refs'; |
596 |
|
if ($sf && $$rec->{$f}->[$i]->{$sf}) { |
597 |
|
$$found++ if (defined($$found)); |
598 |
|
return $$rec->{$f}->[$i]->{$sf}; |
599 |
|
} elsif (! $sf && $$rec->{$f}->[$i]) { |
600 |
|
$$found++ if (defined($$found)); |
601 |
|
# it still might have subfield, just |
602 |
|
# not specified, so we'll dump all |
603 |
|
if ($$rec->{$f}->[$i] =~ /HASH/o) { |
604 |
|
my $out; |
605 |
|
foreach my $k (keys %{$$rec->{$f}->[$i]}) { |
606 |
|
my $v = $$rec->{$f}->[$i]->{$k}; |
607 |
|
$out .= "$v " if ($v); |
608 |
|
} |
609 |
|
return $out; |
610 |
|
} else { |
611 |
|
return $$rec->{$f}->[$i]; |
612 |
|
} |
613 |
|
} else { |
614 |
|
return ''; |
615 |
|
} |
616 |
|
} else { |
617 |
|
return ''; |
618 |
|
} |
619 |
|
} |
620 |
|
|
621 |
|
|
622 |
|
=head2 apply_format |
623 |
|
|
624 |
|
Apply format specified in tag with C<format_name="name"> and |
625 |
|
C<format_delimiter=";;">. |
626 |
|
|
627 |
|
my $text = $webpac->apply_format($format_name,$format_delimiter,$data); |
628 |
|
|
629 |
|
Formats can contain C<lookup{...}> if you need them. |
630 |
|
|
631 |
|
=cut |
632 |
|
|
633 |
|
sub apply_format { |
634 |
|
my $self = shift; |
635 |
|
|
636 |
|
my ($name,$delimiter,$data) = @_; |
637 |
|
|
638 |
|
my $log = $self->_get_logger(); |
639 |
|
|
640 |
|
if (! $self->{'import_xml'}->{'format'}->{$name}) { |
641 |
|
$log->warn("<format name=\"$name\"> is not defined in ",$self->{'import_xml_file'}); |
642 |
|
return $data; |
643 |
|
} |
644 |
|
|
645 |
|
$log->warn("no delimiter for format $name") if (! $delimiter); |
646 |
|
|
647 |
|
my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); |
648 |
|
|
649 |
|
my @data = split(/\Q$delimiter\E/, $data); |
650 |
|
|
651 |
|
my $out = sprintf($format, @data); |
652 |
|
$log->debug("using format $name [$format] on $data to produce: $out"); |
653 |
|
|
654 |
|
if ($self->{'lookup_regex'} && $out =~ $self->{'lookup_regex'}) { |
655 |
|
return $self->{'lookup'}->lookup($out); |
656 |
|
} else { |
657 |
|
return $out; |
658 |
|
} |
659 |
|
|
660 |
|
} |
661 |
|
|
662 |
=head2 sort_arr |
=head2 sort_arr |
663 |
|
|
664 |
Sort array ignoring case and html in data |
Sort array ignoring case and html in data |
685 |
} |
} |
686 |
|
|
687 |
|
|
688 |
|
=head1 INTERNAL METHODS |
689 |
|
|
690 |
=head2 _sort_by_order |
=head2 _sort_by_order |
691 |
|
|
692 |
Sort xml tags data structure accoding to C<order=""> attribute. |
Sort xml tags data structure accoding to C<order=""> attribute. |
706 |
|
|
707 |
=head2 _x |
=head2 _x |
708 |
|
|
709 |
Convert strings from C<conf/normalize> encoding into application specific |
Convert strings from C<conf/normalize/*.xml> encoding into application |
710 |
(optinally specified using C<code_page> to C<new> constructor. |
specific encoding (optinally specified using C<code_page> to C<new> |
711 |
|
constructor). |
712 |
|
|
713 |
my $text = $n->_x('normalize text string'); |
my $text = $n->_x('normalize text string'); |
714 |
|
|
735 |
|
|
736 |
=cut |
=cut |
737 |
|
|
738 |
1; # End of WebPAC::DB |
1; # End of WebPAC::Normalize |