2 |
|
|
3 |
use warnings; |
use warnings; |
4 |
use strict; |
use strict; |
5 |
|
use blib; |
6 |
|
use WebPAC::Common; |
7 |
|
use base 'WebPAC::Common'; |
8 |
use Data::Dumper; |
use Data::Dumper; |
9 |
|
|
10 |
=head1 NAME |
=head1 NAME |
11 |
|
|
12 |
WebPAC::Normalize - normalisation of source file |
WebPAC::Normalize - data mungling for normalisation |
13 |
|
|
14 |
=head1 VERSION |
=head1 VERSION |
15 |
|
|
16 |
Version 0.01 |
Version 0.08 |
17 |
|
|
18 |
=cut |
=cut |
19 |
|
|
20 |
our $VERSION = '0.01'; |
our $VERSION = '0.08'; |
21 |
|
|
22 |
=head1 SYNOPSIS |
=head1 SYNOPSIS |
23 |
|
|
24 |
This package contains code that could be helpful in implementing different |
This package contains code that mungle data to produce normalized format. |
25 |
normalisation front-ends. |
|
26 |
|
It contains several assumptions: |
27 |
|
|
28 |
|
=over |
29 |
|
|
30 |
|
=item * |
31 |
|
|
32 |
|
format of fields is defined using C<v123^a> notation for repeatable fields |
33 |
|
or C<s123^a> for single (or first) value, where C<123> is field number and |
34 |
|
C<a> is subfield. |
35 |
|
|
36 |
|
=item * |
37 |
|
|
38 |
|
source data records (C<$rec>) have unique identifiers in field C<000> |
39 |
|
|
40 |
|
=item * |
41 |
|
|
42 |
|
optional C<eval{length('v123^a') == 3}> tag at B<beginning of format> will be |
43 |
|
perl code that is evaluated before producing output (value of field will be |
44 |
|
interpolated before that) |
45 |
|
|
46 |
|
=item * |
47 |
|
|
48 |
|
optional C<filter{filter_name}> at B<begining of format> will apply perl |
49 |
|
code defined as code ref on format after field substitution to producing |
50 |
|
output |
51 |
|
|
52 |
|
There is one built-in filter called C<regex> which can be use like this: |
53 |
|
|
54 |
|
filter{regex(s/foo/bar/)} |
55 |
|
|
56 |
|
=item * |
57 |
|
|
58 |
|
optional C<lookup{...}> will be then performed. See C<WebPAC::Lookups>. |
59 |
|
|
60 |
|
=item * |
61 |
|
|
62 |
|
at end, optional C<format>s rules are resolved. Format rules are similar to |
63 |
|
C<sprintf> and can also contain C<lookup{...}> which is performed after |
64 |
|
values are inserted in format. |
65 |
|
|
66 |
|
=back |
67 |
|
|
68 |
|
This also describes order in which transformations are applied (eval, |
69 |
|
filter, lookup, format) which is important to undestand when deciding how to |
70 |
|
solve your data mungling and normalisation process. |
71 |
|
|
72 |
|
|
73 |
|
|
74 |
|
|
75 |
=head1 FUNCTIONS |
=head1 FUNCTIONS |
76 |
|
|
79 |
Create new normalisation object |
Create new normalisation object |
80 |
|
|
81 |
my $n = new WebPAC::Normalize::Something( |
my $n = new WebPAC::Normalize::Something( |
82 |
cache_data_structure => './cache/ds/', |
filter => { |
83 |
|
'filter_name_1' => sub { |
84 |
|
# filter code |
85 |
|
return length($_); |
86 |
|
}, ... |
87 |
|
}, |
88 |
|
db => $db_obj, |
89 |
lookup_regex => $lookup->regex, |
lookup_regex => $lookup->regex, |
90 |
|
lookup => $lookup_obj, |
91 |
|
prefix => 'foobar', |
92 |
); |
); |
93 |
|
|
94 |
Optional parameter C<cache_data_structure> defines path to directory |
Parametar C<filter> defines user supplied snippets of perl code which can |
95 |
in which cache file for C<data_structure> call will be created. |
be use with C<filter{...}> notation. |
96 |
|
|
97 |
|
C<prefix> is used to form filename for database record (to support multiple |
98 |
|
source files which are joined in one database). |
99 |
|
|
100 |
Recommended parametar C<lookup_regex> is used to enable parsing of lookups |
Recommended parametar C<lookup_regex> is used to enable parsing of lookups |
101 |
in structures. |
in structures. If you pass this parametar, you must also pass C<lookup> |
102 |
|
which is C<WebPAC::Lookup> object. |
103 |
|
|
104 |
=cut |
=cut |
105 |
|
|
108 |
my $self = {@_}; |
my $self = {@_}; |
109 |
bless($self, $class); |
bless($self, $class); |
110 |
|
|
111 |
$self->setup_cache_dir( $self->{'cache_data_structure'} ); |
my $r = $self->{'lookup_regex'} ? 1 : 0; |
112 |
|
my $l = $self->{'lookup'} ? 1 : 0; |
|
$self ? return $self : return undef; |
|
|
} |
|
|
|
|
|
=head2 setup_cache_dir |
|
|
|
|
|
Check if specified cache directory exist, and if not, disable caching. |
|
|
|
|
|
$setup_cache_dir('./cache/ds/'); |
|
113 |
|
|
114 |
If you pass false or zero value to this function, it will disable |
my $log = $self->_get_logger(); |
|
cacheing. |
|
|
|
|
|
=cut |
|
115 |
|
|
116 |
sub setup_cache_dir { |
# those two must be in pair |
117 |
my $self = shift; |
if ( ($r & $l) != ($r || $l) ) { |
118 |
|
my $log = $self->_get_logger(); |
119 |
|
$log->logdie("lookup_regex and lookup must be in pair"); |
120 |
|
} |
121 |
|
|
122 |
my $dir = shift; |
$log->logdie("lookup must be WebPAC::Lookup object") if ($self->{'lookup'} && ! $self->{'lookup'}->isa('WebPAC::Lookup')); |
123 |
|
|
124 |
my $log = $self->_get_logger(); |
$log->warn("no prefix defined. please check that!") unless ($self->{'prefix'}); |
125 |
|
|
126 |
if ($dir) { |
$log->debug("using lookup regex: ", $self->{lookup_regex}) if ($r && $l); |
|
my $msg; |
|
|
if (! -e $dir) { |
|
|
$msg = "doesn't exist"; |
|
|
} elsif (! -d $dir) { |
|
|
$msg = "is not directory"; |
|
|
} elsif (! -w $dir) { |
|
|
$msg = "not writable"; |
|
|
} |
|
127 |
|
|
128 |
if ($msg) { |
if (! $self->{filter} || ! $self->{filter}->{regex}) { |
129 |
undef $self->{'cache_data_structure'}; |
$log->debug("adding built-in filter regex"); |
130 |
$log->warn("cache_data_structure $dir $msg, disabling..."); |
$self->{filter}->{regex} = sub { |
131 |
} else { |
my ($val, $regex) = @_; |
132 |
$log->debug("using cache dir $dir"); |
eval "\$val =~ $regex"; |
133 |
} |
return $val; |
134 |
} else { |
}; |
|
$log->debug("disabling cache"); |
|
|
undef $self->{'cache_data_structure'}; |
|
135 |
} |
} |
136 |
|
|
137 |
|
$self ? return $self : return undef; |
138 |
} |
} |
139 |
|
|
140 |
|
|
145 |
|
|
146 |
This structures are used to produce output. |
This structures are used to produce output. |
147 |
|
|
148 |
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. |
|
149 |
|
|
150 |
=cut |
=cut |
151 |
|
|
157 |
my $rec = shift; |
my $rec = shift; |
158 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
159 |
|
|
160 |
|
$log->debug("data_structure rec = ", sub { Dumper($rec) }); |
161 |
|
|
162 |
|
$log->logdie("need unique ID (mfn) in field 000 of record " . Dumper($rec) ) unless (defined($rec->{'000'})); |
163 |
|
|
164 |
|
my $id = $rec->{'000'}->[0] || $log->logdie("field 000 isn't array!"); |
165 |
|
|
166 |
my $cache_file; |
my $cache_file; |
167 |
|
|
168 |
if (my $cache_path = $self->{'cache_data_structure'}) { |
if ($self->{'db'}) { |
169 |
my $id = $rec->{'000'}; |
my $ds = $self->{'db'}->load_ds( id => $id, prefix => $self->{prefix} ); |
170 |
$id = $rec->{'000'}->[0] if ($id =~ m/^ARRAY/o); |
$log->debug("load_ds( rec = ", sub { Dumper($rec) }, ") = ", sub { Dumper($ds) }); |
171 |
unless (defined($id)) { |
return $ds if ($ds); |
172 |
$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'}; |
|
|
} |
|
|
} |
|
|
} |
|
|
} |
|
173 |
} |
} |
174 |
|
|
|
undef $self->{'currnet_filename'}; |
|
|
undef $self->{'headline'}; |
|
|
|
|
175 |
my @sorted_tags; |
my @sorted_tags; |
176 |
if ($self->{tags_by_order}) { |
if ($self->{tags_by_order}) { |
177 |
@sorted_tags = @{$self->{tags_by_order}}; |
@sorted_tags = @{$self->{tags_by_order}}; |
180 |
$self->{tags_by_order} = \@sorted_tags; |
$self->{tags_by_order} = \@sorted_tags; |
181 |
} |
} |
182 |
|
|
183 |
my @ds; |
my $ds; |
184 |
|
|
185 |
$log->debug("tags: ",sub { join(", ",@sorted_tags) }); |
$log->debug("tags: ",sub { join(", ",@sorted_tags) }); |
186 |
|
|
191 |
#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'}}); |
192 |
|
|
193 |
foreach my $tag (@{$self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}}) { |
foreach my $tag (@{$self->{'import_xml'}->{'indexer'}->{$field}->{$self->{'tag'}}}) { |
194 |
my $format = $tag->{'value'} || $tag->{'content'}; |
my $format; |
195 |
|
|
196 |
$log->debug("format: $format"); |
$log->logdie("expected tag HASH and got $tag") unless (ref($tag) eq 'HASH'); |
197 |
|
$format = $tag->{'value'} || $tag->{'content'}; |
198 |
|
|
199 |
my @v; |
my @v; |
200 |
if ($self->{'lookup_regex'} && $format =~ $self->{'lookup_regex'}) { |
if ($self->{'lookup_regex'} && $format =~ $self->{'lookup_regex'}) { |
201 |
@v = $self->fill_in_to_arr($rec,$format); |
@v = $self->_rec_to_arr($rec,$format,'fill_in'); |
202 |
|
} else { |
203 |
|
@v = $self->_rec_to_arr($rec,$format,'parse'); |
204 |
|
} |
205 |
|
if (! @v) { |
206 |
|
$log->debug("$field <",$self->{tag},"> format: $format no values"); |
207 |
|
next; |
208 |
} else { |
} else { |
209 |
@v = $self->parse_to_arr($rec,$format); |
$log->debug("$field <",$self->{tag},"> format: $format values: ", join(",", @v)); |
210 |
} |
} |
|
next if (! @v); |
|
211 |
|
|
212 |
if ($tag->{'sort'}) { |
if ($tag->{'sort'}) { |
213 |
@v = $self->sort_arr(@v); |
@v = $self->sort_arr(@v); |
218 |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
@v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; |
219 |
} |
} |
220 |
|
|
|
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! |
|
|
} |
|
|
|
|
221 |
# delimiter will join repeatable fields |
# delimiter will join repeatable fields |
222 |
if ($tag->{'delimiter'}) { |
if ($tag->{'delimiter'}) { |
223 |
@v = ( join($tag->{'delimiter'}, @v) ); |
@v = ( join($tag->{'delimiter'}, @v) ); |
224 |
} |
} |
225 |
|
|
226 |
# default types |
# default types |
227 |
my @types = qw(display swish); |
my @types = qw(display search); |
228 |
# override by type attribute |
# override by type attribute |
229 |
@types = ( $tag->{'type'} ) if ($tag->{'type'}); |
@types = ( $tag->{'type'} ) if ($tag->{'type'}); |
230 |
|
|
231 |
foreach my $type (@types) { |
foreach my $type (@types) { |
232 |
# append to previous line? |
# append to previous line? |
233 |
$log->debug("type: $type ",sub { join(" ",@v) }, $row->{'append'} || 'no append'); |
$log->debug("tag $field / $type [",sub { join(",",@v) }, "] ", $row->{'append'} || 'no append'); |
234 |
if ($tag->{'append'}) { |
if ($tag->{'append'}) { |
235 |
|
|
236 |
# I will delimit appended part with |
# I will delimit appended part with |
257 |
|
|
258 |
# TODO: name_sigular, name_plural |
# TODO: name_sigular, name_plural |
259 |
my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; |
my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; |
260 |
$row->{'name'} = $name ? $self->_x($name) : $field; |
my $row_name = $name ? $self->_x($name) : $field; |
261 |
|
|
262 |
# post-sort all values in field |
# post-sort all values in field |
263 |
if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { |
if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { |
264 |
$log->warn("sort at field tag not implemented"); |
$log->warn("sort at field tag not implemented"); |
265 |
} |
} |
266 |
|
|
267 |
push @ds, $row; |
$ds->{$row_name} = $row; |
268 |
|
|
269 |
$log->debug("row $field: ",sub { Dumper($row) }); |
$log->debug("row $field: ",sub { Dumper($row) }); |
270 |
} |
} |
271 |
|
|
272 |
} |
} |
273 |
|
|
274 |
if ($cache_file) { |
$self->{'db'}->save_ds( |
275 |
store { |
id => $id, |
276 |
ds => \@ds, |
ds => $ds, |
277 |
current_filename => $self->{'current_filename'}, |
prefix => $self->{prefix}, |
278 |
headline => $self->{'headline'}, |
) if ($self->{'db'}); |
|
}, $cache_file; |
|
|
$log->debug("created storable cache file $cache_file"); |
|
|
} |
|
279 |
|
|
280 |
return @ds; |
$log->debug("ds: ", sub { Dumper($ds) }); |
281 |
|
|
282 |
} |
$log->logconfess("data structure returned is not array any more!") if wantarray; |
283 |
|
|
284 |
=head2 apply_format |
return $ds; |
|
|
|
|
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); |
|
|
|
|
|
my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); |
|
|
|
|
|
my @data = split(/\Q$delimiter\E/, $data); |
|
|
|
|
|
my $out = sprintf($format, @data); |
|
|
$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; |
|
|
} |
|
285 |
|
|
286 |
} |
} |
287 |
|
|
293 |
|
|
294 |
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); |
295 |
|
|
296 |
|
Filters are implemented here. While simple form of filters looks like this: |
297 |
|
|
298 |
|
filter{name_of_filter} |
299 |
|
|
300 |
|
but, filters can also have variable number of parametars like this: |
301 |
|
|
302 |
|
filter{name_of_filter(param,param,param)} |
303 |
|
|
304 |
=cut |
=cut |
305 |
|
|
306 |
|
my $warn_once; |
307 |
|
|
308 |
sub parse { |
sub parse { |
309 |
my $self = shift; |
my $self = shift; |
310 |
|
|
311 |
my ($rec, $format_utf8, $i) = @_; |
my ($rec, $format_utf8, $i, $rec_size) = @_; |
312 |
|
|
313 |
return if (! $format_utf8); |
return if (! $format_utf8); |
314 |
|
|
322 |
|
|
323 |
my @out; |
my @out; |
324 |
|
|
325 |
$log->debug("format: $format"); |
$log->debug("format: $format [$i]"); |
326 |
|
|
327 |
my $eval_code; |
my $eval_code; |
328 |
# remove eval{...} from beginning |
# remove eval{...} from beginning |
332 |
# remove filter{...} from beginning |
# remove filter{...} from beginning |
333 |
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
334 |
|
|
335 |
|
# did we found any (att all) field from format in row? |
336 |
|
my $found_any; |
337 |
|
# prefix before first field which we preserve it $found_any |
338 |
my $prefix; |
my $prefix; |
339 |
my $all_found=0; |
|
340 |
|
my $f_step = 1; |
341 |
|
|
342 |
while ($format =~ s/^(.*?)(v|s)(\d+)(?:\^(\w))?//s) { |
while ($format =~ s/^(.*?)(v|s)(\d+)(?:\^(\w))?//s) { |
343 |
|
|
344 |
my $del = $1 || ''; |
my $del = $1 || ''; |
345 |
$prefix ||= $del if ($all_found == 0); |
$prefix = $del if ($f_step == 1); |
346 |
|
|
347 |
|
my $fld_type = lc($2); |
348 |
|
|
349 |
# repeatable index |
# repeatable index |
350 |
my $r = $i; |
my $r = $i; |
351 |
$r = 0 if (lc("$2") eq 's'); |
if ($fld_type eq 's') { |
352 |
|
if ($found_any->{'v'}) { |
353 |
|
$r = 0; |
354 |
|
} else { |
355 |
|
return; |
356 |
|
} |
357 |
|
} |
358 |
|
|
359 |
my $found = 0; |
my $found = 0; |
360 |
my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found); |
my $tmp = $self->get_data(\$rec,$3,$4,$r,\$found,$rec_size); |
361 |
|
|
362 |
if ($found) { |
if ($found) { |
363 |
push @out, $del; |
$found_any->{$fld_type} += $found; |
364 |
|
|
365 |
|
# we will skip delimiter before first occurence of field! |
366 |
|
push @out, $del unless($found_any->{$fld_type} == 1); |
367 |
push @out, $tmp; |
push @out, $tmp; |
|
$all_found += $found; |
|
368 |
} |
} |
369 |
|
$f_step++; |
370 |
} |
} |
371 |
|
|
372 |
return if (! $all_found); |
# test if any fields found? |
373 |
|
return if (! $found_any->{'v'} && ! $found_any->{'s'}); |
374 |
|
|
375 |
my $out = join('',@out); |
my $out = join('',@out); |
376 |
|
|
390 |
return if (! $self->_eval($eval)); |
return if (! $self->_eval($eval)); |
391 |
} |
} |
392 |
|
|
393 |
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
if ($filter_name) { |
394 |
$log->debug("about to filter{$filter_name} format: $out"); |
my @filter_args; |
395 |
$out = $self->{'filter'}->{$filter_name}->($out); |
if ($filter_name =~ s/(\w+)\((.*)\)/$1/) { |
396 |
return unless(defined($out)); |
@filter_args = split(/,/, $2); |
397 |
$log->debug("filter result: $out"); |
} |
398 |
|
if ($self->{'filter'}->{$filter_name}) { |
399 |
|
$log->debug("about to filter{$filter_name} format: $out with arguments: ", join(",", @filter_args)); |
400 |
|
unshift @filter_args, $out; |
401 |
|
$out = $self->{'filter'}->{$filter_name}->(@filter_args); |
402 |
|
return unless(defined($out)); |
403 |
|
$log->debug("filter result: $out"); |
404 |
|
} elsif (! $warn_once->{$filter_name}) { |
405 |
|
$log->warn("trying to use undefined filter $filter_name"); |
406 |
|
$warn_once->{$filter_name}++; |
407 |
|
} |
408 |
} |
} |
409 |
|
|
410 |
return $out; |
return $out; |
411 |
} |
} |
412 |
|
|
413 |
=head2 parse_to_arr |
=head2 fill_in |
414 |
|
|
415 |
|
Workhourse of all: takes record from in-memory structure of database and |
416 |
|
strings with placeholders and returns string or array of with substituted |
417 |
|
values from record. |
418 |
|
|
419 |
|
my $text = $webpac->fill_in($rec,'v250^a'); |
420 |
|
|
421 |
|
Optional argument is ordinal number for repeatable fields. By default, |
422 |
|
it's assume to be first repeatable field (fields are perl array, so first |
423 |
|
element is 0). |
424 |
|
Following example will read second value from repeatable field. |
425 |
|
|
426 |
|
my $text = $webpac->fill_in($rec,'Title: v250^a',1); |
427 |
|
|
428 |
|
This function B<does not> perform parsing of format to inteligenty skip |
429 |
|
delimiters before fields which aren't used. |
430 |
|
|
431 |
|
This method will automatically decode UTF-8 string to local code page |
432 |
|
if needed. |
433 |
|
|
434 |
Similar to C<parse>, but returns array of all repeatable fields |
There is optional parametar C<$record_size> which can be used to get sizes of |
435 |
|
all C<field^subfield> combinations in this format. |
436 |
|
|
437 |
my @arr = $webpac->parse_to_arr($rec,'v250^a'); |
my $text = $webpac->fill_in($rec,'got: v900^a v900^x',0,\$rec_size); |
438 |
|
|
439 |
=cut |
=cut |
440 |
|
|
441 |
sub parse_to_arr { |
sub fill_in { |
442 |
my $self = shift; |
my $self = shift; |
443 |
|
|
|
my ($rec, $format_utf8) = @_; |
|
|
|
|
444 |
my $log = $self->_get_logger(); |
my $log = $self->_get_logger(); |
445 |
|
|
446 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
my ($rec,$format,$i,$rec_size) = @_; |
|
return if (! $format_utf8); |
|
447 |
|
|
448 |
my $i = 0; |
$log->logconfess("need data record") unless ($rec); |
449 |
my @arr; |
$log->logconfess("need format to parse") unless($format); |
450 |
|
|
451 |
|
# iteration (for repeatable fields) |
452 |
|
$i ||= 0; |
453 |
|
|
454 |
|
$log->logdie("infitite loop in format $format") if ($i > ($self->{'max_mfn'} || 9999)); |
455 |
|
|
456 |
while (my $v = $self->parse($rec,$format_utf8,$i++)) { |
# FIXME remove for speedup? |
457 |
push @arr, $v; |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
458 |
|
|
459 |
|
if (utf8::is_utf8($format)) { |
460 |
|
$format = $self->_x($format); |
461 |
} |
} |
462 |
|
|
463 |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
my $found = 0; |
464 |
|
my $just_single = 1; |
465 |
|
|
466 |
return @arr; |
my $eval_code; |
467 |
|
# remove eval{...} from beginning |
468 |
|
$eval_code = $1 if ($format =~ s/^eval{([^}]+)}//s); |
469 |
|
|
470 |
|
my $filter_name; |
471 |
|
# remove filter{...} from beginning |
472 |
|
$filter_name = $1 if ($format =~ s/^filter{([^}]+)}//s); |
473 |
|
|
474 |
|
# do actual replacement of placeholders |
475 |
|
# repeatable fields |
476 |
|
if ($format =~ s/v(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,$i,\$found,$rec_size)/ges) { |
477 |
|
$just_single = 0; |
478 |
|
} |
479 |
|
|
480 |
|
# non-repeatable fields |
481 |
|
if ($format =~ s/s(\d+)(?:\^(\w))?/$self->get_data(\$rec,$1,$2,0,\$found,$rec_size)/ges) { |
482 |
|
return if ($i > 0 && $just_single); |
483 |
|
} |
484 |
|
|
485 |
|
if ($found) { |
486 |
|
$log->debug("format: $format"); |
487 |
|
if ($eval_code) { |
488 |
|
my $eval = $self->fill_in($rec,$eval_code,$i); |
489 |
|
return if (! $self->_eval($eval)); |
490 |
|
} |
491 |
|
if ($filter_name && $self->{'filter'}->{$filter_name}) { |
492 |
|
$log->debug("filter '$filter_name' for $format"); |
493 |
|
$format = $self->{'filter'}->{$filter_name}->($format); |
494 |
|
return unless(defined($format)); |
495 |
|
$log->debug("filter result: $format"); |
496 |
|
} |
497 |
|
# do we have lookups? |
498 |
|
if ($self->{'lookup'}) { |
499 |
|
if ($self->{'lookup'}->can('lookup')) { |
500 |
|
my @lookup = $self->{lookup}->lookup($format); |
501 |
|
$log->debug("lookup $format", join(", ", @lookup)); |
502 |
|
return @lookup; |
503 |
|
} else { |
504 |
|
$log->warn("Have lookup object but can't invoke lookup method"); |
505 |
|
} |
506 |
|
} else { |
507 |
|
return $format; |
508 |
|
} |
509 |
|
} else { |
510 |
|
return; |
511 |
|
} |
512 |
} |
} |
513 |
|
|
|
=head2 fill_in_to_arr |
|
514 |
|
|
515 |
Similar to C<fill_in>, but returns array of all repeatable fields. Usable |
=head2 _rec_to_arr |
516 |
|
|
517 |
|
Similar to C<parse> and C<fill_in>, but returns array of all repeatable fields. Usable |
518 |
for fields which have lookups, so they shouldn't be parsed but rather |
for fields which have lookups, so they shouldn't be parsed but rather |
519 |
C<fill_id>ed. |
C<paste>d or C<fill_id>ed. Last argument is name of operation: C<paste> or C<fill_in>. |
520 |
|
|
521 |
my @arr = $webpac->fill_in_to_arr($rec,'[v900];;[v250^a]'); |
my @arr = $webpac->fill_in_to_arr($rec,'[v900];;[v250^a]','paste'); |
522 |
|
|
523 |
=cut |
=cut |
524 |
|
|
525 |
sub fill_in_to_arr { |
sub _rec_to_arr { |
526 |
my $self = shift; |
my $self = shift; |
527 |
|
|
528 |
my ($rec, $format_utf8) = @_; |
my ($rec, $format_utf8, $code) = @_; |
529 |
|
|
530 |
my $log = $self->_get_logger(); |
my $log = $self->_get_logger(); |
531 |
|
|
532 |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
$log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); |
533 |
return if (! $format_utf8); |
return if (! $format_utf8); |
534 |
|
|
535 |
|
$log->debug("using $code on $format_utf8"); |
536 |
|
|
537 |
my $i = 0; |
my $i = 0; |
538 |
|
my $max = 0; |
539 |
my @arr; |
my @arr; |
540 |
|
my $rec_size = {}; |
541 |
|
|
542 |
while (my @v = $self->fill_in($rec,$format_utf8,$i++)) { |
while ($i <= $max) { |
543 |
push @arr, @v; |
my @v = $self->$code($rec,$format_utf8,$i++,\$rec_size); |
544 |
|
if ($rec_size) { |
545 |
|
foreach my $f (keys %{ $rec_size }) { |
546 |
|
$max = $rec_size->{$f} if ($rec_size->{$f} > $max); |
547 |
|
} |
548 |
|
$log->debug("max set to $max"); |
549 |
|
undef $rec_size; |
550 |
|
} |
551 |
|
if (@v) { |
552 |
|
push @arr, @v; |
553 |
|
} else { |
554 |
|
push @arr, ''; |
555 |
|
} |
556 |
} |
} |
557 |
|
|
558 |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
$log->debug("format '$format_utf8' returned ",--$i," elements: ", sub { join(" | ",@arr) }) if (@arr); |
560 |
return @arr; |
return @arr; |
561 |
} |
} |
562 |
|
|
563 |
|
|
564 |
|
=head2 get_data |
565 |
|
|
566 |
|
Returns value from record. |
567 |
|
|
568 |
|
my $text = $self->get_data(\$rec,$f,$sf,$i,\$found,\$rec_size); |
569 |
|
|
570 |
|
Required arguments are: |
571 |
|
|
572 |
|
=over 8 |
573 |
|
|
574 |
|
=item C<$rec> |
575 |
|
|
576 |
|
record reference |
577 |
|
|
578 |
|
=item C<$f> |
579 |
|
|
580 |
|
field |
581 |
|
|
582 |
|
=item C<$sf> |
583 |
|
|
584 |
|
optional subfield |
585 |
|
|
586 |
|
=item C<$i> |
587 |
|
|
588 |
|
index offset for repeatable values ( 0 ... $rec_size->{'400^a'} ) |
589 |
|
|
590 |
|
=item C<$found> |
591 |
|
|
592 |
|
optional variable that will be incremeted if preset |
593 |
|
|
594 |
|
=item C<$rec_size> |
595 |
|
|
596 |
|
hash to hold maximum occurances of C<field^subfield> combinations |
597 |
|
(which can be accessed using keys in same format) |
598 |
|
|
599 |
|
=back |
600 |
|
|
601 |
|
Returns value or empty string, updates C<$found> and C<rec_size> |
602 |
|
if present. |
603 |
|
|
604 |
|
=cut |
605 |
|
|
606 |
|
sub get_data { |
607 |
|
my $self = shift; |
608 |
|
|
609 |
|
my ($rec,$f,$sf,$i,$found,$cache) = @_; |
610 |
|
|
611 |
|
return '' unless ($$rec->{$f} && ref($$rec->{$f}) eq 'ARRAY'); |
612 |
|
|
613 |
|
if (defined($$cache)) { |
614 |
|
$$cache->{ $f . ( $sf ? '^' . $sf : '' ) } ||= scalar @{ $$rec->{$f} }; |
615 |
|
} |
616 |
|
|
617 |
|
return '' unless ($$rec->{$f}->[$i]); |
618 |
|
|
619 |
|
{ |
620 |
|
no strict 'refs'; |
621 |
|
if (defined($sf)) { |
622 |
|
$$found++ if (defined($$found) && $$rec->{$f}->[$i]->{$sf}); |
623 |
|
return $$rec->{$f}->[$i]->{$sf}; |
624 |
|
} else { |
625 |
|
$$found++ if (defined($$found)); |
626 |
|
# it still might have subfields, just |
627 |
|
# not specified, so we'll dump some debug info |
628 |
|
if ($$rec->{$f}->[$i] =~ /HASH/o) { |
629 |
|
my $out; |
630 |
|
foreach my $k (keys %{$$rec->{$f}->[$i]}) { |
631 |
|
$out .= '$' . $k .':' . $$rec->{$f}->[$i]->{$k}." "; |
632 |
|
} |
633 |
|
return $out; |
634 |
|
} else { |
635 |
|
return $$rec->{$f}->[$i]; |
636 |
|
} |
637 |
|
} |
638 |
|
} |
639 |
|
} |
640 |
|
|
641 |
|
|
642 |
|
=head2 apply_format |
643 |
|
|
644 |
|
Apply format specified in tag with C<format_name="name"> and |
645 |
|
C<format_delimiter=";;">. |
646 |
|
|
647 |
|
my $text = $webpac->apply_format($format_name,$format_delimiter,$data); |
648 |
|
|
649 |
|
Formats can contain C<lookup{...}> if you need them. |
650 |
|
|
651 |
|
=cut |
652 |
|
|
653 |
|
sub apply_format { |
654 |
|
my $self = shift; |
655 |
|
|
656 |
|
my ($name,$delimiter,$data) = @_; |
657 |
|
|
658 |
|
my $log = $self->_get_logger(); |
659 |
|
|
660 |
|
if (! $self->{'import_xml'}->{'format'}->{$name}) { |
661 |
|
$log->warn("<format name=\"$name\"> is not defined in ",$self->{'import_xml_file'}); |
662 |
|
return $data; |
663 |
|
} |
664 |
|
|
665 |
|
$log->warn("no delimiter for format $name") if (! $delimiter); |
666 |
|
|
667 |
|
my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); |
668 |
|
|
669 |
|
my @data = split(/\Q$delimiter\E/, $data); |
670 |
|
|
671 |
|
my $out = sprintf($format, @data); |
672 |
|
$log->debug("using format $name [$format] on $data to produce: $out"); |
673 |
|
|
674 |
|
if ($self->{'lookup_regex'} && $out =~ $self->{'lookup_regex'}) { |
675 |
|
return $self->{'lookup'}->lookup($out); |
676 |
|
} else { |
677 |
|
return $out; |
678 |
|
} |
679 |
|
|
680 |
|
} |
681 |
|
|
682 |
=head2 sort_arr |
=head2 sort_arr |
683 |
|
|
684 |
Sort array ignoring case and html in data |
Sort array ignoring case and html in data |
705 |
} |
} |
706 |
|
|
707 |
|
|
708 |
|
=head1 INTERNAL METHODS |
709 |
|
|
710 |
=head2 _sort_by_order |
=head2 _sort_by_order |
711 |
|
|
712 |
Sort xml tags data structure accoding to C<order=""> attribute. |
Sort xml tags data structure accoding to C<order=""> attribute. |
726 |
|
|
727 |
=head2 _x |
=head2 _x |
728 |
|
|
729 |
Convert strings from C<conf/normalize> encoding into application specific |
Convert strings from C<conf/normalize/*.xml> encoding into application |
730 |
(optinally specified using C<code_page> to C<new> constructor. |
specific encoding (optinally specified using C<code_page> to C<new> |
731 |
|
constructor). |
732 |
|
|
733 |
my $text = $n->_x('normalize text string'); |
my $text = $n->_x('normalize text string'); |
734 |
|
|
755 |
|
|
756 |
=cut |
=cut |
757 |
|
|
758 |
1; # End of WebPAC::DB |
1; # End of WebPAC::Normalize |