source: gsdl/trunk/perllib/classify/List.pm@ 20825

Last change on this file since 20825 was 20825, checked in by kjdon, 13 years ago

check that the values of the partition_types_within_level option are valid

  • Property svn:keywords set to Author Date Id Revision
File size: 33.8 KB
Line 
1###########################################################################
2#
3# List.pm -- A general and flexible list classifier with most of
4# the abilities of AZCompactList, and better Unicode,
5# metadata and sorting capabilities.
6#
7# A component of the Greenstone digital library software
8# from the New Zealand Digital Library Project at the
9# University of Waikato, New Zealand.
10#
11# Author: Michael Dewsnip, NZDL Project, University of Waikato, NZ
12#
13# Copyright (C) 2005 New Zealand Digital Library Project
14#
15# This program is free software; you can redistribute it and/or modify
16# it under the terms of the GNU General Public License as published by
17# the Free Software Foundation; either version 2 of the License, or
18# (at your option) any later version.
19#
20# This program is distributed in the hope that it will be useful,
21# but WITHOUT ANY WARRANTY; without even the implied warranty of
22# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
23# GNU General Public License for more details.
24#
25# You should have received a copy of the GNU General Public License
26# along with this program; if not, write to the Free Software
27# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
28#
29# TO DO: - Remove punctuation from metadata values before sorting.
30# - Add an AZCompactList-style hlist option?
31#
32###########################################################################
33
34package List;
35
36
37use BaseClassifier;
38
39use strict;
40
41
42sub BEGIN {
43 @List::ISA = ('BaseClassifier');
44}
45
46my $partition_type_list =
47 [ { 'name' => "per_letter",
48 'desc' => "{List.level_partition.per_letter}" },
49 { 'name' => "constant_size",
50 'desc' => "{List.level_partition.constant_size}" },
51 { 'name' => "per_letter_fixed_size",
52 'desc' => "{List.level_partition.per_letter_fixed_size}"},
53 { 'name' => "none",
54 'desc' => "{List.level_partition.none}" } ];
55
56# following used to check types later on
57my $valid_partition_types = { 'per_letter' => 1,
58 'constant_size' => 1,
59 'per_letter_fixed_size' => 1,
60 'none' => 1};
61
62my $bookshelf_type_list =
63 [ { 'name' => "always",
64 'desc' => "{List.bookshelf_type.always}" },
65 { 'name' => "duplicate_only",
66 'desc' => "{List.bookshelf_type.duplicate_only}" },
67 { 'name' => "never",
68 'desc' => "{List.bookshelf_type.never}" } ];
69
70my $arguments =
71 [ { 'name' => "metadata",
72 'desc' => "{List.metadata}",
73 'type' => "metadata",
74 'reqd' => "yes" },
75
76 # The interesting options
77 { 'name' => "bookshelf_type",
78 'desc' => "{List.bookshelf_type}",
79 'type' => "enum",
80 'list' => $bookshelf_type_list,
81 'deft' => "never" },
82 { 'name' => "classify_sections",
83 'desc' => "{List.classify_sections}",
84 'type' => "flag" },
85 { 'name' => "partition_type_within_level",
86 'desc' => "{List.partition_type_within_level}",
87 'type' => "enumstring", # Must be enumstring because multiple values can be specified (separated by '/')
88 'list' => $partition_type_list,
89 'deft' => "per_letter" },
90 { 'name' => "partition_size_within_level",
91 'desc' => "{List.partition_size_within_level}",
92 'type' => "string" }, # Must be string because multiple values can be specified (separated by '/')
93 { 'name' => "partition_name_length",
94 'desc' => "{List.partition_name_length}",
95 'type' => "string" },
96 { 'name' => "sort_leaf_nodes_using",
97 'desc' => "{List.sort_leaf_nodes_using}",
98 'type' => "metadata",
99 'deft' => "Title" },
100 { 'name' => "sort_using_unicode_collation",
101 'desc' => "{List.sort_using_unicode_collation}",
102 'type' => "flag" },
103 { 'name' => "use_hlist_for",
104 'desc' => "{List.use_hlist_for}",
105 'type' => "string" },
106 { 'name' => "removeprefix",
107 'desc' => "{BasClas.removeprefix}",
108 'type' => "regexp" },
109 { 'name' => "removesuffix",
110 'desc' => "{BasClas.removesuffix}",
111 'type' => "regexp" } ];
112
113my $options = { 'name' => "List",
114 'desc' => "{List.desc}",
115 'abstract' => "no",
116 'inherits' => "yes",
117 'args' => $arguments };
118
119
120sub new
121{
122 my ($class) = shift(@_);
123 my ($classifierslist, $inputargs, $hashArgOptLists) = @_;
124 push(@$classifierslist, $class);
125
126 push(@{$hashArgOptLists->{"ArgList"}}, @{$arguments});
127 push(@{$hashArgOptLists->{"OptList"}}, $options);
128
129 my $self = new BaseClassifier($classifierslist, $inputargs, $hashArgOptLists);
130
131 if ($self->{'info_only'}) {
132 # don't worry about any options etc
133 return bless $self, $class;
134 }
135
136 # The metadata elements to use (required)
137 if (!$self->{'metadata'}) {
138 die "Error: No metadata fields specified for List.\n";
139 }
140 my @metadata_groups = split(/\//, $self->{'metadata'});
141 $self->{'metadata_groups'} = \@metadata_groups;
142
143 # The classifier button name (default: the first metadata element specified)
144 if (!$self->{'buttonname'}) {
145 my $first_metadata_group = $metadata_groups[0];
146 my $first_metadata_element = (split(/\;|,/, $first_metadata_group))[0];
147 $self->{'buttonname'} = $self->generate_title_from_metadata($first_metadata_element);
148 }
149
150 # Whether to group items into a bookshelf, (must be 'always' for all metadata fields except the last)
151 foreach my $metadata_group (@metadata_groups) {
152 $self->{$metadata_group . ".bookshelf_type"} = "always";
153 }
154 my $last_metadata_group = $metadata_groups[$#metadata_groups];
155 # Default: duplicate_only, ie. leave leaf nodes ungrouped (equivalent to AZCompactList -mingroup 2)
156 $self->{$last_metadata_group . ".bookshelf_type"} = $self->{'bookshelf_type'};
157
158 # Whether to use an hlist or a vlist for each level in the hierarchy (default: vlist)
159 foreach my $metadata_group (@metadata_groups) {
160 $self->{$metadata_group . ".list_type"} = "VList";
161 }
162 foreach my $metadata_group (split(/\,/, $self->{'use_hlist_for'})) {
163 $self->{$metadata_group . ".list_type"} = "HList";
164 }
165
166 # How the items are grouped into partitions (default: no partition)
167 # for each level (metadata group), separated by '/'
168 if (!$self->{"partition_type_within_level"}) {
169 foreach my $metadata_group (@metadata_groups) {
170 $self->{$metadata_group . ".partition_type_within_level"} = "none";
171 }
172 } else {
173 my @partition_type_within_levellist = split(/\//, $self->{'partition_type_within_level'});
174
175 my $first = 1;
176 foreach my $metadata_group (@metadata_groups) {
177 my $partition_type_within_levelelem = shift(@partition_type_within_levellist);
178 if (defined($partition_type_within_levelelem) && defined $valid_partition_types->{$partition_type_within_levelelem}) {
179 $self->{$metadata_group . ".partition_type_within_level"} = $partition_type_within_levelelem;
180 }
181 else {
182 if ($first) {
183 $self->{$metadata_group . ".partition_type_within_level"} = "none";
184 $first = 0;
185 } else {
186 $self->{$metadata_group . ".partition_type_within_level"} = $self->{$metadata_groups[0] . ".partition_type_within_level"};
187 }
188 if (defined($partition_type_within_levelelem)) {
189 # ie invalid entry
190 print STDERR "invalid partition type for level $metadata_group: $partition_type_within_levelelem, defaulting to ". $self->{$metadata_group . ".partition_type_within_level"} ."\n";
191 }
192 }
193 }
194 }
195
196 # The number of items in each partition
197 if (!$self->{'partition_size_within_level'}) {
198 # Default: 20
199 foreach my $metadata_group (@metadata_groups) {
200 $self->{$metadata_group . ".partition_size_within_level"} = 20;
201 }
202 }
203 else {
204 my @partition_size_within_levellist = split(/\//, $self->{'partition_size_within_level'});
205
206 # Assign values based on the partition_size_within_level parameter
207 foreach my $metadata_group (@metadata_groups) {
208 my $partition_size_within_levelelem = shift(@partition_size_within_levellist);
209 if (defined($partition_size_within_levelelem)) {
210 $self->{$metadata_group . ".partition_size_within_level"} = $partition_size_within_levelelem;
211 }
212 else {
213 $self->{$metadata_group . ".partition_size_within_level"} = $self->{$metadata_groups[0] . ".partition_size_within_level"};
214 }
215 }
216 }
217
218 # The removeprefix and removesuffix expressions
219 if ($self->{'removeprefix'}) {
220 # If there are more than one expressions, use '' to quote each experession and '/' to separate
221 my @removeprefix_exprs_within_levellist = split(/'\/'/, $self->{'removeprefix'});
222
223 foreach my $metadata_group (@metadata_groups) {
224 my $removeprefix_expr_within_levelelem = shift(@removeprefix_exprs_within_levellist);
225 if (defined($removeprefix_expr_within_levelelem) && $removeprefix_expr_within_levelelem ne "") {
226 # Remove the other ' at the beginning and the end if there is any
227 $removeprefix_expr_within_levelelem =~ s/^'//;
228 $removeprefix_expr_within_levelelem =~ s/'$//;
229 # Remove the extra ^ at the beginning
230 $removeprefix_expr_within_levelelem =~ s/^\^//;
231 $self->{$metadata_group . ".remove_prefix_expr"} = $removeprefix_expr_within_levelelem;
232 } else {
233 $self->{$metadata_group . ".remove_prefix_expr"} = $self->{$metadata_groups[0] . ".remove_prefix_expr"};
234 }
235 }
236 }
237 if ($self->{'removesuffix'}) {
238 my @removesuffix_exprs_within_levellist = split(/'\/'/, $self->{'removesuffix'});
239
240 foreach my $metadata_group (@metadata_groups) {
241 my $removesuffix_expr_within_levelelem = shift(@removesuffix_exprs_within_levellist);
242 if (defined($removesuffix_expr_within_levelelem) && $removesuffix_expr_within_levelelem ne "") {
243 $removesuffix_expr_within_levelelem =~ s/^'//;
244 $removesuffix_expr_within_levelelem =~ s/'$//;
245 # Remove the extra $ at the end
246 $removesuffix_expr_within_levelelem =~ s/\$$//;
247 $self->{$metadata_group . ".remove_suffix_expr"} = $removesuffix_expr_within_levelelem;
248 } else {
249 $self->{$metadata_group . ".remove_suffix_expr"} = $self->{$metadata_groups[0] . ".remove_suffix_expr"};
250 }
251 }
252 }
253
254 # The metadata elements to use to sort the leaf nodes (default: Title)
255 my @sort_leaf_nodes_using_metadata_groups = ( "Title" );
256 if ($self->{'sort_leaf_nodes_using'}) {
257 @sort_leaf_nodes_using_metadata_groups = split(/\|/, $self->{'sort_leaf_nodes_using'});
258 }
259 $self->{'sort_leaf_nodes_using_metadata_groups'} = \@sort_leaf_nodes_using_metadata_groups;
260
261 # Create an instance of the Unicode::Collate object if better Unicode sorting is desired
262 if ($self->{'sort_using_unicode_collation'}) {
263 # To use this you first need to download the allkeys.txt file from
264 # http://www.unicode.org/Public/UCA/latest/allkeys.txt and put it in the Perl
265 # Unicode/Collate directory.
266 require Unicode::Collate;
267 $self->{'unicode_collator'} = Unicode::Collate->new();
268 }
269
270 # An empty array for the document OIDs
271 $self->{'OIDs'} = [];
272
273 return bless $self, $class;
274}
275
276
277sub init
278{
279 # Nothing to do...
280}
281
282
283# Called for each document in the collection
284sub classify
285{
286 my $self = shift(@_);
287 my ($doc_obj,$edit_mode) = @_;
288
289 # If "-classify_sections" is set, classify every section of the document
290 if ($self->{'classify_sections'}) {
291 my $section = $doc_obj->get_next_section($doc_obj->get_top_section());
292 while (defined $section) {
293 $self->classify_section($doc_obj, $doc_obj->get_OID() . ".$section", $section, $edit_mode);
294 $section = $doc_obj->get_next_section($section);
295 }
296 }
297 # Otherwise just classify the top document section
298 else {
299 $self->classify_section($doc_obj, $doc_obj->get_OID(), $doc_obj->get_top_section(), $edit_mode);
300 }
301}
302
303
304sub classify_section
305{
306 my $self = shift(@_);
307 my ($doc_obj,$section_OID,$section,$edit_mode) = @_;
308
309 my @metadata_groups = @{$self->{'metadata_groups'}};
310
311 # Only classify the section if it has a value for one of the metadata elements in the first group
312 my $classify_section = 0;
313 my $first_metadata_group = $metadata_groups[0];
314 foreach my $first_metadata_group_element (split(/\;|,/, $first_metadata_group)) {
315 my $real_first_metadata_group_element = $self->strip_ex_from_metadata($first_metadata_group_element);
316 my $first_metadata_group_element_value = $doc_obj->get_metadata_element($section, $real_first_metadata_group_element);
317
318 # Remove prefix/suffix if requested
319 my $remove_prefix_expr = $self->{$first_metadata_group_element . ".remove_prefix_expr"};
320 if (defined $remove_prefix_expr && $remove_prefix_expr ne "") {
321 $first_metadata_group_element_value =~ s/^$remove_prefix_expr//;
322 }
323 my $remove_suffix_expr = $self->{$first_metadata_group_element . ".remove_suffix_expr"};
324 if (defined $remove_suffix_expr && $remove_suffix_expr ne "") {
325 $first_metadata_group_element_value =~ s/$remove_suffix_expr$//;
326 }
327
328 if (defined($first_metadata_group_element_value) && $first_metadata_group_element_value ne "") {
329 # This section must be included in the classifier
330 $classify_section = 1;
331 last;
332 }
333 }
334
335 # We're not classifying this section because it doesn't have the required metadata
336 return if (!$classify_section);
337
338 if (($edit_mode eq "delete") || ($edit_mode eq "update")) {
339 $self->oid_array_delete($section_OID,'OIDs');
340 if ($edit_mode eq "delete") {
341 return;
342 }
343 }
344
345 # Otherwise, include this section in the classifier
346 push(@{$self->{'OIDs'}}, $section_OID);
347
348 # Create a hash for the metadata values of each metadata element we're interested in
349 my %metadata_groups_done = ();
350 foreach my $metadata_group (@metadata_groups, @{$self->{'sort_leaf_nodes_using_metadata_groups'}}) {
351 # Take care not to do a metadata group more than once
352 unless ($metadata_groups_done{$metadata_group}) {
353 if ($edit_mode eq "update") {
354 # if we are updating, we delete all the old values before
355 # adding the new ones, otherwise, the section will end up in
356 # the classifier twice.
357 delete $self->{$metadata_group . ".list"}->{$section_OID};
358 }
359 foreach my $metadata_element (split(/\;|,/, $metadata_group)) {
360 my $real_metadata_element = $self->strip_ex_from_metadata($metadata_element);
361
362 my $remove_prefix_expr = $self->{$metadata_element . ".remove_prefix_expr"};
363 my $remove_suffix_expr = $self->{$metadata_element . ".remove_suffix_expr"};
364 my @metadata_values = @{$doc_obj->get_metadata($section, $real_metadata_element)};
365 foreach my $metadata_value (@metadata_values) {
366 # Strip leading and trailing whitespace
367 $metadata_value =~ s/^\s*//;
368 $metadata_value =~ s/\s*$//;
369
370 # Remove prefix/suffix if requested
371 if (defined $remove_prefix_expr && $remove_prefix_expr ne "") {
372 $metadata_value =~ s/^$remove_prefix_expr//;
373 }
374 if (defined $remove_suffix_expr && $remove_suffix_expr ne "") {
375 $metadata_value =~ s/$remove_suffix_expr$//;
376 }
377
378 # Convert the metadata value from a UTF-8 string to a Unicode string
379 # This means that length() and substr() work properly
380 # We need to be careful to convert classifier node title values back to UTF-8, however
381 my $metadata_value_unicode_string = $self->convert_utf8_string_to_unicode_string($metadata_value);
382
383 # Add the metadata value into the list for this combination of metadata group and section
384 push(@{$self->{$metadata_group . ".list"}->{$section_OID}}, $metadata_value_unicode_string);
385 }
386 last if (@metadata_values > 0);
387 }
388
389 $metadata_groups_done{$metadata_group} = 1;
390 }
391 }
392}
393
394
395sub get_classify_info
396{
397 my $self = shift(@_);
398
399 # The metadata groups to classify by
400 my @metadata_groups = @{$self->{'metadata_groups'}};
401 my $first_metadata_group = $metadata_groups[0];
402
403 # The OID values of the documents to include in the classifier
404 my @OIDs = @{$self->{'OIDs'}};
405
406 # Create the root node of the classification hierarchy
407 my %classifier_node = ( 'thistype' => "Invisible",
408 'childtype' => $self->{$first_metadata_group . ".list_type"},
409 'Title' => $self->{'buttonname'},
410 'contains' => [],
411 'mdtype' => $first_metadata_group );
412
413 # Recursively create the classification hierarchy, one level for each metadata group
414 $self->add_level(\@metadata_groups, \@OIDs, \%classifier_node);
415 return \%classifier_node;
416}
417
418
419sub add_level
420{
421 my $self = shift(@_);
422 my @metadata_groups = @{shift(@_)};
423 my @OIDs = @{shift(@_)};
424 my $classifier_node = shift(@_);
425 # print STDERR "\nAdding AZ list for " . $classifier_node->{'Title'} . "\n";
426
427 my $metadata_group = $metadata_groups[0];
428 # print STDERR "Processing metadata group: " . $metadata_group . "\n";
429 # print STDERR "Number of OID values: " . @OIDs . "\n";
430
431 if (!defined($self->{$metadata_group . ".list"})) {
432 print STDERR "Warning: No metadata values assigned to $metadata_group.\n";
433 return;
434 }
435
436 # Create a mapping from metadata value to OID
437 my $OID_to_metadata_values_hash_ref = $self->{$metadata_group . ".list"};
438 my %metadata_value_to_OIDs_hash = ();
439 foreach my $OID (@OIDs)
440 {
441 if ($OID_to_metadata_values_hash_ref->{$OID})
442 {
443 my @metadata_values = @{$OID_to_metadata_values_hash_ref->{$OID}};
444 foreach my $metadata_value (@metadata_values)
445 {
446 push(@{$metadata_value_to_OIDs_hash{$metadata_value}}, $OID);
447 }
448 }
449 }
450 # print STDERR "Number of distinct values: " . scalar(keys %metadata_value_to_OIDs_hash) . "\n";
451
452 # Partition the values (if necessary)
453 my $partition_type_within_level = $self->{$metadata_group . ".partition_type_within_level"};
454 if ($partition_type_within_level =~ /^per_letter$/i) {
455 # Generate one hlist for each letter
456 my @sortedmetadata_values = $self->sort_metadata_values_array(keys(%metadata_value_to_OIDs_hash));
457 my %metadata_value_to_OIDs_subhash = ();
458
459 my $lastpartition = substr($sortedmetadata_values[0], 0, 1);
460 foreach my $metadata_value (@sortedmetadata_values) {
461 my $metadata_valuepartition = substr($metadata_value, 0, 1);
462
463 # Is this the start of a new partition?
464 if ($metadata_valuepartition ne $lastpartition) {
465 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $lastpartition, \%metadata_value_to_OIDs_subhash);
466 %metadata_value_to_OIDs_subhash = ();
467 $lastpartition = $metadata_valuepartition;
468 }
469
470 $metadata_value_to_OIDs_subhash{$metadata_value} = $metadata_value_to_OIDs_hash{$metadata_value};
471 }
472
473 # Don't forget to add the last partition
474 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $lastpartition, \%metadata_value_to_OIDs_subhash);
475
476 # The partitions are stored in an HList
477 $classifier_node->{'childtype'} = "HList";
478 }
479 elsif ($partition_type_within_level =~ /^per_letter_fixed_size$/i) {
480 # Generate hlist based on the first letter of the metadata value (like per_letter) but with restriction on the partition size
481 # If a letter has fewer items than specified by the "partition_size_within_level", then group them together if possible
482 # If a letter has more items than specified, split into several hlists.
483 # Depends on the bookshelf_type, one item can be either a document (when bookshelf_type is "never") or a metadata value (otherwise)
484 my $partition_size_within_level = $self->{$metadata_group . ".partition_size_within_level"};
485 my @sortedmetadata_values = $self->sort_metadata_values_array(keys(%metadata_value_to_OIDs_hash));
486 my $bookshelf_type = $self->{$metadata_group . ".bookshelf_type"};
487
488 # Separate values by their first letter, each form a bucket, like the per_letter partition type
489 my $last_partition = substr($sortedmetadata_values[0], 0, 1);
490 my %partition_buckets = ();
491 my @metadata_values_in_bucket = ();
492 my $num_items_in_bucket = 0;
493 foreach my $metadata_value (@sortedmetadata_values) {
494 my $metadata_valuepartition = substr($metadata_value, 0, 1);
495 if ($metadata_valuepartition ne $last_partition) {
496 my @temp_array = @metadata_values_in_bucket;
497 # Cache the values that belong to this bucket, and the number of items in this bucket, not necessary to be the same number as the metadata values
498 my %partition_info = ();
499 $partition_info{'metadata_values'} = \@temp_array;
500 $partition_info{'size'} = $num_items_in_bucket;
501 $partition_buckets{$last_partition} = \%partition_info;
502
503 @metadata_values_in_bucket = ($metadata_value);
504 $num_items_in_bucket = $bookshelf_type eq "never" ? scalar(@{$metadata_value_to_OIDs_hash{$metadata_value}}) : scalar(@metadata_values_in_bucket);
505 $last_partition = $metadata_valuepartition;
506 } else {
507 $num_items_in_bucket += $bookshelf_type eq "never" ? scalar(@{$metadata_value_to_OIDs_hash{$metadata_value}}) : scalar(@metadata_values_in_bucket);
508 push (@metadata_values_in_bucket, $metadata_value);
509 }
510 }
511 # Last one
512 my %partition_info = ();
513 $partition_info{'metadata_values'} = \@metadata_values_in_bucket;
514 $partition_info{'size'} = $num_items_in_bucket;
515 $partition_buckets{$last_partition} = \%partition_info;
516
517 my @partition_keys = $self->sort_metadata_values_array(keys(%partition_buckets));
518 for (my $i = 0; $i < scalar(@partition_keys) - 1; $i++) {
519 my $partition = $partition_keys[$i];
520 my $items_in_partition = $partition_buckets{$partition}->{'size'};
521 # Merge small buckets together, but keep the numeric bucket apart
522 if ($items_in_partition < $partition_size_within_level) {
523 my $items_in_next_partition = $partition_buckets{$partition_keys[$i+1]}->{'size'};
524 if ($items_in_partition + $items_in_next_partition <= $partition_size_within_level
525 && !(($partition =~ /^[^0-9]/ && $partition_keys[$i+1] =~ /^[0-9]/)
526 || ($partition =~ /^[0-9]/ && $partition_keys[$i+1] =~ /^[^0-9]/))) {
527 foreach my $metadata_value_to_merge (@{$partition_buckets{$partition}->{'metadata_values'}}) {
528 push(@{$partition_buckets{$partition_keys[$i+1]}->{'metadata_values'}}, $metadata_value_to_merge);
529 }
530 $partition_buckets{$partition_keys[$i+1]}->{'size'} += $items_in_partition;
531 delete $partition_buckets{$partition};
532 }
533 }
534 }
535 @partition_keys = $self->sort_metadata_values_array(keys(%partition_buckets));
536
537 # Add partitions, and divide big bucket into several
538 my $last_partition_end = "";
539 my $partition_start = "";
540 foreach my $partition (@partition_keys) {
541 my @metadata_values = $self->sort_metadata_values_array(@{$partition_buckets{$partition}->{'metadata_values'}});
542 my $items_in_partition = $partition_buckets{$partition}->{'size'};
543 $partition_start = $self->generate_partition_start($metadata_values[0], $last_partition_end, $self->{"partition_name_length"});
544
545 if ($items_in_partition > $partition_size_within_level) {
546 my $items_done = 0;
547 my %metadata_values_to_OIDs_subhashes = ();
548 for (my $i = 0; $i < scalar(@metadata_values); $i++) {
549 my $metadata_value = $metadata_values[$i];
550 # If the bookshelf_type is "never", count the documents, otherwise count the distinct metadata values
551 my $items_for_this_md_value = $bookshelf_type eq "never" ? scalar(@{$metadata_value_to_OIDs_hash{$metadata_value}}) : 1;
552
553 my $partitionend = $self->generate_partition_end($metadata_value, $partition_start, $self->{"partition_name_length"});
554 my $partitionname = $partition_start;
555 if ($partitionend ne $partition_start) {
556 $partitionname = $partitionname . "-" . $partitionend;
557 }
558
559 # Start a new partition
560 if ($items_done + $items_for_this_md_value > $partition_size_within_level && $items_done != 0) {
561 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $partitionname, \%metadata_values_to_OIDs_subhashes);
562 $last_partition_end = $partitionend;
563 $partition_start = $self->generate_partition_start($metadata_value, $last_partition_end, $self->{"partition_name_length"});
564 $items_done = 0;
565 %metadata_values_to_OIDs_subhashes = ();
566 }
567
568 # If bookshelf_type is "never" and the current metadata value holds too many items, need to split into several partitions
569 if ($bookshelf_type eq "never" && $items_for_this_md_value > $partition_size_within_level) {
570 my $partitionname_for_this_value = $self->generate_partition_start($metadata_value, $last_partition_end, $self->{"partition_name_length"});
571 # Get the number of partitions needed for this value
572 my $num_splits = int($items_for_this_md_value / $partition_size_within_level);
573 $num_splits++ if ($items_for_this_md_value / $partition_size_within_level > $num_splits);
574
575 my @OIDs_for_this_value = @{$metadata_value_to_OIDs_hash{$metadata_value}};
576 for (my $i = 0; $i < $num_splits; $i++) {
577 my %OIDs_subhashes_for_this_value = ();
578 my @OIDs_for_this_partition = ();
579 for (my $d = $i * $partition_size_within_level; $d < (($i+1) * $partition_size_within_level > $items_for_this_md_value ? $items_for_this_md_value : ($i+1) * $partition_size_within_level); $d++) {
580 push (@OIDs_for_this_partition, $OIDs_for_this_value[$d]);
581 }
582
583 # The last bucket might have only a few items and need to be merged with buckets for subsequent metadata values
584 if ($i == $num_splits - 1 && scalar(@OIDs_for_this_partition) < $partition_size_within_level) {
585 $metadata_values_to_OIDs_subhashes{$metadata_value} = \@OIDs_for_this_partition;
586 $items_done += scalar(@OIDs_for_this_partition);
587 next;
588 }
589
590 # Add an HList for this bucket
591 $OIDs_subhashes_for_this_value{$metadata_value} = \@OIDs_for_this_partition;
592 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $partitionname_for_this_value, \%OIDs_subhashes_for_this_value);
593 $last_partition_end = $partitionname_for_this_value;
594 }
595 next;
596 }
597
598 $metadata_values_to_OIDs_subhashes{$metadata_value} = $metadata_value_to_OIDs_hash{$metadata_value};
599 $items_done += $bookshelf_type eq "never" ? scalar(@{$metadata_values_to_OIDs_subhashes{$metadata_value}}) : 1;
600
601 # The last partition
602 if($i == scalar(@metadata_values) - 1) {
603 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $partitionname, \%metadata_values_to_OIDs_subhashes);
604 }
605 }
606 }
607 else {
608 # The easier case, just add a partition
609 my %metadata_values_to_OIDs_subhashes = ();
610 for (my $i = 0; $i < scalar(@metadata_values); $i++) {
611 my $metadata_value = $metadata_values[$i];
612 $metadata_values_to_OIDs_subhashes{$metadata_value} = $metadata_value_to_OIDs_hash{$metadata_value};
613 }
614 my $last_metadata_value = $metadata_values[scalar(@metadata_values)-1];
615 my $partitionend = $self->generate_partition_end($last_metadata_value, $partition_start, $self->{"partition_name_length"});
616 my $partitionname = $partition_start;
617 if ($partitionend ne $partition_start) {
618 $partitionname = $partitionname . "-" . $partitionend;
619 }
620 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $partitionname, \%metadata_values_to_OIDs_subhashes);
621 $last_partition_end = $partitionend;
622 }
623 }
624 }
625 else {
626 # Generate hlists of a certain size
627 my $partition_size_within_level = $self->{$metadata_group . ".partition_size_within_level"};
628 if ($partition_type_within_level =~ /^constant_size$/i && scalar(keys %metadata_value_to_OIDs_hash) > $partition_size_within_level) {
629 my @sortedmetadata_values = $self->sort_metadata_values_array(keys(%metadata_value_to_OIDs_hash));
630 my $itemsdone = 0;
631 my %metadata_value_to_OIDs_subhash = ();
632 my $lastpartitionend = "";
633 my $partitionstart;
634 foreach my $metadata_value (@sortedmetadata_values) {
635 $metadata_value_to_OIDs_subhash{$metadata_value} = $metadata_value_to_OIDs_hash{$metadata_value};
636 $itemsdone++;
637 my $itemsinpartition = scalar(keys %metadata_value_to_OIDs_subhash);
638
639 # Is this the start of a new partition?
640 if ($itemsinpartition == 1) {
641 $partitionstart = $self->generate_partition_start($metadata_value, $lastpartitionend, $self->{"partition_name_length"});
642 }
643
644 # Is this the end of the partition?
645 if ($itemsinpartition == $partition_size_within_level || $itemsdone == @sortedmetadata_values) {
646 my $partitionend = $self->generate_partition_end($metadata_value, $partitionstart, $self->{"partition_name_length"});
647 my $partitionname = $partitionstart;
648 if ($partitionend ne $partitionstart) {
649 $partitionname = $partitionname . "-" . $partitionend;
650 }
651
652 $self->add_hlist_partition(\@metadata_groups, $classifier_node, $partitionname, \%metadata_value_to_OIDs_subhash);
653 %metadata_value_to_OIDs_subhash = ();
654 $lastpartitionend = $partitionend;
655 }
656 }
657
658 # The partitions are stored in an HList
659 $classifier_node->{'childtype'} = "HList";
660 }
661
662 # Otherwise just add all the values to a VList
663 else {
664 $self->add_vlist(\@metadata_groups, $classifier_node, \%metadata_value_to_OIDs_hash);
665 }
666 }
667}
668
669
670sub convert_utf8_string_to_unicode_string
671{
672 my $self = shift(@_);
673 my $utf8_string = shift(@_);
674
675 my $unicode_string = "";
676 foreach my $unicode_value (@{&unicode::utf82unicode($utf8_string)}) {
677 $unicode_string .= chr($unicode_value);
678 }
679 return $unicode_string;
680}
681
682
683sub convert_unicode_string_to_utf8_string
684{
685 my $self = shift(@_);
686 my $unicode_string = shift(@_);
687
688 my @unicode_array;
689 for (my $i = 0; $i < length($unicode_string); $i++) {
690 push(@unicode_array, ord(substr($unicode_string, $i, 1)));
691 }
692 return &unicode::unicode2utf8(\@unicode_array);
693}
694
695
696sub generate_partition_start
697{
698 my $self = shift(@_);
699 my $metadata_value = shift(@_);
700 my $lastpartitionend = shift(@_);
701 my $partition_name_length = shift(@_);
702
703 if ($partition_name_length) {
704 return substr($metadata_value, 0, $partition_name_length);
705 }
706
707 my $partitionstart = substr($metadata_value, 0, 1);
708 if ($partitionstart le $lastpartitionend) {
709 $partitionstart = substr($metadata_value, 0, 2);
710 # Give up after three characters
711 if ($partitionstart le $lastpartitionend) {
712 $partitionstart = substr($metadata_value, 0, 3);
713 }
714 }
715
716 return $partitionstart;
717}
718
719
720sub generate_partition_end
721{
722 my $self = shift(@_);
723 my $metadata_value = shift(@_);
724 my $partitionstart = shift(@_);
725 my $partition_name_length = shift(@_);
726
727 if ($partition_name_length) {
728 return substr($metadata_value, 0, $partition_name_length);
729 }
730
731 my $partitionend = substr($metadata_value, 0, length($partitionstart));
732 if ($partitionend gt $partitionstart) {
733 $partitionend = substr($metadata_value, 0, 1);
734 if ($partitionend le $partitionstart) {
735 $partitionend = substr($metadata_value, 0, 2);
736 # Give up after three characters
737 if ($partitionend le $partitionstart) {
738 $partitionend = substr($metadata_value, 0, 3);
739 }
740 }
741 }
742
743 return $partitionend;
744}
745
746
747sub add_hlist_partition
748{
749 my $self = shift(@_);
750 my @metadata_groups = @{shift(@_)};
751 my $classifier_node = shift(@_);
752 my $partitionname = shift(@_);
753 my $metadata_value_to_OIDs_hash_ref = shift(@_);
754
755 # Create an hlist partition
756 my %child_classifier_node = ( 'Title' => $self->convert_unicode_string_to_utf8_string($partitionname),
757 'childtype' => "VList",
758 'contains' => [] );
759
760 # Add the children to the hlist partition
761 $self->add_vlist(\@metadata_groups, \%child_classifier_node, $metadata_value_to_OIDs_hash_ref);
762 push(@{$classifier_node->{'contains'}}, \%child_classifier_node);
763}
764
765
766sub add_vlist
767{
768 my $self = shift(@_);
769 my @metadata_groups = @{shift(@_)};
770 my $classifier_node = shift(@_);
771 my $metadata_value_to_OIDs_hash_ref = shift(@_);
772
773 my $metadata_group = shift(@metadata_groups);
774 $classifier_node->{'mdtype'} = $metadata_group;
775
776 # Create an entry in the vlist for each value
777 foreach my $metadata_value ($self->sort_metadata_values_array(keys(%{$metadata_value_to_OIDs_hash_ref})))
778 {
779 my @OIDs = @{$metadata_value_to_OIDs_hash_ref->{$metadata_value}};
780
781 # If there is only one item and 'bookshelf_type' is not always (ie. never or duplicate_only), add the item to the list
782 if (@OIDs == 1 && $self->{$metadata_group . ".bookshelf_type"} ne "always") {
783 my $OID = $OIDs[0];
784
785 # Find the offset of this metadata value
786 my $offset = 0;
787 my $OID_to_metadata_values_hash_ref = $self->{$metadata_group . ".list"};
788 my @metadata_values = @{$OID_to_metadata_values_hash_ref->{$OID}};
789 for (my $i = 0; $i < scalar(@metadata_values); $i++) {
790 if ($metadata_value eq $metadata_values[$i]) {
791 $offset = $i;
792 last;
793 }
794 }
795 push(@{$classifier_node->{'contains'}}, { 'OID' => $OID, 'offset' => $offset });
796 }
797 # If 'bookshelf_type' is 'never', list all the items even if there are duplicated values
798 elsif ($self->{$metadata_group . ".bookshelf_type"} eq "never") {
799 $self->add_sorted_leaf_items(\@OIDs, $classifier_node);
800 }
801 # Otherwise create a sublist (bookshelf) for the metadata value
802 else {
803 my %child_classifier_node = ( 'Title' => $self->convert_unicode_string_to_utf8_string($metadata_value),
804 'childtype' => "VList",
805 'contains' => [] );
806
807 # If there are metadata elements remaining, recursively apply the process
808 if (@metadata_groups > 0) {
809 my $next_metadata_group = $metadata_groups[0];
810 $child_classifier_node{'childtype'} = $self->{$next_metadata_group . ".list_type"};
811 $self->add_level(\@metadata_groups, \@OIDs, \%child_classifier_node);
812 }
813 # Otherwise just add the documents as children of this list
814 else {
815 $self->add_sorted_leaf_items(\@OIDs, \%child_classifier_node);
816 }
817
818 # Add the sublist to the list
819 push(@{$classifier_node->{'contains'}}, \%child_classifier_node);
820 }
821 }
822}
823
824sub add_sorted_leaf_items
825{
826 my $self = shift(@_);
827 my @OIDs = @{shift(@_)};
828 my $classifier_node = shift(@_);
829
830 # Sort leaf nodes and add to list
831 my @sort_leaf_nodes_using_metadata_groups = @{$self->{'sort_leaf_nodes_using_metadata_groups'}};
832 foreach my $sort_leaf_nodes_usingmetaelem (reverse @sort_leaf_nodes_using_metadata_groups) {
833 my $OID_to_metadata_values_hash_ref = $self->{$sort_leaf_nodes_usingmetaelem . ".list"};
834 # Force a stable sort (Perl 5.6's sort isn't stable)
835 # !! The [0] bits aren't ideal (multiple metadata values) !!
836 @OIDs = @OIDs[ sort { $OID_to_metadata_values_hash_ref->{$OIDs[$a]}[0] cmp $OID_to_metadata_values_hash_ref->{$OIDs[$b]}[0] || $a <=> $b; } 0..$#OIDs ];
837 }
838
839 foreach my $OID (@OIDs) {
840 push(@{$classifier_node->{'contains'}}, { 'OID' => $OID });
841 }
842}
843
844
845sub sort_metadata_values_array
846{
847 my $self = shift(@_);
848 my @metadata_values = @_;
849
850 if ($self->{'unicode_collator'}) {
851 return $self->{'unicode_collator'}->sort(@metadata_values);
852 }
853 else {
854 return sort(@metadata_values);
855 }
856}
857
858
8591;
Note: See TracBrowser for help on using the repository browser.