1 |
dpavlin |
1 |
package jsFind; |
2 |
|
|
|
3 |
dpavlin |
14 |
use 5.005; |
4 |
dpavlin |
1 |
use strict; |
5 |
|
|
use warnings; |
6 |
dpavlin |
11 |
use HTML::Entities; |
7 |
dpavlin |
1 |
|
8 |
dpavlin |
39 |
our $VERSION = '0.07_01'; |
9 |
dpavlin |
1 |
|
10 |
dpavlin |
39 |
use Exporter; |
11 |
dpavlin |
12 |
use Carp; |
12 |
|
|
|
13 |
|
|
our @ISA = qw(Exporter); |
14 |
|
|
|
15 |
|
|
BEGIN { |
16 |
dpavlin |
39 |
Exporter::import 'jsFind::Node'; |
17 |
dpavlin |
12 |
} |
18 |
|
|
|
19 |
dpavlin |
1 |
=head1 NAME |
20 |
|
|
|
21 |
dpavlin |
33 |
jsFind - generate index for full text search engine in JavaScript |
22 |
dpavlin |
1 |
|
23 |
|
|
=head1 SYNOPSIS |
24 |
|
|
|
25 |
|
|
use jsFind; |
26 |
dpavlin |
12 |
my $t = new jsFind(B => 4); |
27 |
|
|
my $f = 1; |
28 |
|
|
foreach my $k (qw{minima ut dolorem sapiente voluptatem}) { |
29 |
|
|
$t->B_search(Key => $k, |
30 |
|
|
Data => { |
31 |
|
|
"path" => { |
32 |
|
|
t => "word $k", |
33 |
|
|
f => $f }, |
34 |
|
|
}, |
35 |
|
|
Insert => 1, |
36 |
|
|
Append => 1, |
37 |
|
|
); |
38 |
|
|
} |
39 |
dpavlin |
1 |
|
40 |
|
|
=head1 DESCRIPTION |
41 |
|
|
|
42 |
|
|
This module can be used to create index files for jsFind, powerful tool for |
43 |
|
|
adding a search engine to a CDROM archive or catalog without requiring the |
44 |
|
|
user to install anything. |
45 |
|
|
|
46 |
|
|
Main difference between this module and scripts delivered with jsFind are: |
47 |
|
|
|
48 |
|
|
=over 5 |
49 |
|
|
|
50 |
|
|
=item * |
51 |
|
|
|
52 |
|
|
You don't need to use swish-e to create index |
53 |
|
|
|
54 |
|
|
=item * |
55 |
|
|
|
56 |
dpavlin |
33 |
you can programatically (and incrementaly) create index for jsFind |
57 |
dpavlin |
1 |
|
58 |
dpavlin |
33 |
=item * |
59 |
|
|
|
60 |
|
|
you can create more than one index and search them using same C<search.html> |
61 |
|
|
page |
62 |
|
|
|
63 |
dpavlin |
1 |
=back |
64 |
|
|
|
65 |
dpavlin |
14 |
You can also examine examples which come as tests with this module, |
66 |
dpavlin |
33 |
for example C<t/04words.t> or C<t/10homer.t>. |
67 |
dpavlin |
14 |
|
68 |
dpavlin |
33 |
=head2 jsFind |
69 |
|
|
|
70 |
|
|
jsFind search engine was written by Shawn Garbett from eLucid Software. |
71 |
|
|
The search engine itself is a small piece of JavaScript (1.2 with level 2 |
72 |
|
|
DOM). It is easily customizable to fit into a current set of HTML. This |
73 |
|
|
JavaScript searches an XML index dataset for the appropriate links, and can |
74 |
|
|
filter and sort the results. |
75 |
|
|
|
76 |
|
|
JavaScript code distributed with this module is based on version 0.0.3 which |
77 |
|
|
was current when this module development started. Various changes where done |
78 |
|
|
on JavaScript code to fix bugs, add features and remove warnings. For |
79 |
|
|
complete list see C<Changes> file which comes with distribution. |
80 |
|
|
|
81 |
|
|
This module has been tested using C<html/test.html> with following browsers: |
82 |
|
|
|
83 |
|
|
=over 5 |
84 |
|
|
|
85 |
|
|
=item Mozilla FireFox 0.8 to 1.0 |
86 |
|
|
|
87 |
|
|
using DOM 2 C<document.implementation.createDocument> |
88 |
|
|
|
89 |
|
|
=item Internet Explorer 5.5 and 6.0 |
90 |
|
|
|
91 |
|
|
using ActiveX C<Microsoft.XMLDOM> or C<MSXML2.DOMDocument> |
92 |
|
|
|
93 |
|
|
=item Konqueror 3.3 |
94 |
|
|
|
95 |
|
|
using DOM 2 C<document.implementation.createDocument> |
96 |
|
|
|
97 |
|
|
=item Opera 7.54 (without Java) |
98 |
|
|
|
99 |
|
|
using experimental iframe implementation which is much slower than other methods. |
100 |
|
|
|
101 |
|
|
=back |
102 |
|
|
|
103 |
|
|
If searching doesn't work for your combination of operating system and |
104 |
|
|
browser, please open C<html/test.html> file and wait a while. It will search sample |
105 |
|
|
file included with distribution and report results. Reports with included |
106 |
|
|
test debugging are welcomed. |
107 |
|
|
|
108 |
dpavlin |
12 |
=head1 jsFind methods |
109 |
dpavlin |
1 |
|
110 |
dpavlin |
12 |
C<jsFind> is mode implementing methods which you, the user, are going to |
111 |
|
|
use to create indexes. |
112 |
dpavlin |
1 |
|
113 |
dpavlin |
12 |
=head2 new |
114 |
dpavlin |
1 |
|
115 |
|
|
Create new tree. Arguments are C<B> which is maximum numbers of keys in |
116 |
|
|
each node and optional C<Root> node. Each root node may have child nodes. |
117 |
|
|
|
118 |
|
|
All nodes are objects from C<jsFind::Node>. |
119 |
|
|
|
120 |
|
|
my $t = new jsFind(B => 4); |
121 |
|
|
|
122 |
|
|
=cut |
123 |
|
|
|
124 |
|
|
my $DEBUG = 1; |
125 |
|
|
|
126 |
|
|
sub new { |
127 |
|
|
my $package = shift; |
128 |
|
|
my %ARGV = @_; |
129 |
|
|
croak "Usage: {$package}::new(B => number [, Root => root node ])" |
130 |
|
|
unless exists $ARGV{B}; |
131 |
|
|
if ($ARGV{B} % 2) { |
132 |
|
|
my $B = $ARGV{B} + 1; |
133 |
|
|
carp "B must be an even number. Using $B instead."; |
134 |
|
|
$ARGV{B} = $B; |
135 |
|
|
} |
136 |
|
|
|
137 |
|
|
my $B = $ARGV{B}; |
138 |
|
|
my $Root = exists($ARGV{Root}) ? $ARGV{Root} : jsFind::Node->emptynode; |
139 |
|
|
bless { B => $B, Root => $Root } => $package; |
140 |
|
|
} |
141 |
|
|
|
142 |
dpavlin |
12 |
=head2 B_search |
143 |
dpavlin |
1 |
|
144 |
|
|
Search, insert, append or replace data in B-Tree |
145 |
|
|
|
146 |
dpavlin |
5 |
$t->B_search( |
147 |
|
|
Key => 'key value', |
148 |
|
|
Data => { "path" => { |
149 |
|
|
"t" => "title of document", |
150 |
|
|
"f" => 99, |
151 |
dpavlin |
7 |
}, |
152 |
dpavlin |
5 |
}, |
153 |
|
|
Insert => 1, |
154 |
|
|
Append => 1, |
155 |
|
|
); |
156 |
dpavlin |
1 |
|
157 |
|
|
Semantics: |
158 |
|
|
|
159 |
|
|
If key not found, insert it iff C<Insert> argument is present. |
160 |
|
|
|
161 |
|
|
If key B<is> found, replace existing data iff C<Replace> argument |
162 |
|
|
is present or add new datum to existing iff C<Append> argument is present. |
163 |
|
|
|
164 |
|
|
=cut |
165 |
|
|
|
166 |
|
|
sub B_search { |
167 |
|
|
my $self = shift; |
168 |
|
|
my %args = @_; |
169 |
|
|
my $cur_node = $self->root; |
170 |
|
|
my $k = $args{Key}; |
171 |
|
|
my $d = $args{Data}; |
172 |
|
|
my @path; |
173 |
|
|
|
174 |
|
|
if ($cur_node->is_empty) { # Special case for empty root |
175 |
|
|
if ($args{Insert}) { |
176 |
|
|
$cur_node->kdp_insert($k => $d); |
177 |
|
|
return $d; |
178 |
|
|
} else { |
179 |
|
|
return undef; |
180 |
|
|
} |
181 |
|
|
} |
182 |
|
|
|
183 |
|
|
# Descend tree to leaf |
184 |
|
|
for (;;) { |
185 |
|
|
|
186 |
|
|
# Didn't hit bottom yet. |
187 |
|
|
|
188 |
|
|
my($there, $where) = $cur_node->locate_key($k); |
189 |
|
|
if ($there) { # Found it! |
190 |
|
|
if ($args{Replace}) { |
191 |
|
|
$cur_node->kdp_replace($where, $k => $d); |
192 |
|
|
} elsif ($args{Append}) { |
193 |
|
|
$cur_node->kdp_append($where, $k => $d); |
194 |
|
|
} |
195 |
|
|
return $cur_node->data($where); |
196 |
|
|
} |
197 |
|
|
|
198 |
|
|
# Not here---must be in a subtree. |
199 |
|
|
|
200 |
|
|
if ($cur_node->is_leaf) { # But there are no subtrees |
201 |
|
|
return undef unless $args{Insert}; # Search failed |
202 |
|
|
# Stuff it in |
203 |
|
|
$cur_node->kdp_insert($k => $d); |
204 |
|
|
if ($self->node_overfull($cur_node)) { # Oops--there was no room. |
205 |
|
|
$self->split_and_promote($cur_node, @path); |
206 |
|
|
} |
207 |
|
|
return $d; |
208 |
|
|
} |
209 |
|
|
|
210 |
|
|
# There are subtrees, and the key is in one of them. |
211 |
|
|
|
212 |
|
|
push @path, [$cur_node, $where]; # Record path from root. |
213 |
|
|
|
214 |
|
|
# Move down to search the subtree |
215 |
|
|
$cur_node = $cur_node->subnode($where); |
216 |
|
|
|
217 |
|
|
# and start over. |
218 |
|
|
} # for (;;) ... |
219 |
|
|
|
220 |
|
|
croak ("How did I get here?"); |
221 |
|
|
} |
222 |
|
|
|
223 |
|
|
|
224 |
|
|
|
225 |
|
|
sub split_and_promote_old { |
226 |
|
|
my $self = shift; |
227 |
|
|
my ($cur_node, @path) = @_; |
228 |
|
|
|
229 |
|
|
for (;;) { |
230 |
|
|
my ($newleft, $newright, $kdp) = $cur_node->halves($self->B / 2); |
231 |
|
|
my ($up, $where) = @{pop @path}; |
232 |
|
|
if ($up) { |
233 |
|
|
$up->kdp_insert(@$kdp); |
234 |
|
|
my ($tthere, $twhere) = $up->locate_key($kdp->[0]); |
235 |
|
|
croak "Couldn't find key `$kdp->[0]' in node after just inserting it!" |
236 |
|
|
unless $tthere; |
237 |
|
|
croak "`$kdp->[0]' went into node at `$twhere' instead of expected `$where'!" |
238 |
|
|
unless $twhere == $where; |
239 |
|
|
$up->subnode($where, $newleft); |
240 |
|
|
$up->subnode($where+1, $newright); |
241 |
|
|
return unless $self->node_overfull($up); |
242 |
|
|
$cur_node = $up; |
243 |
|
|
} else { # We're at the top; make a new root. |
244 |
|
|
my $newroot = new jsFind::Node ([$kdp->[0]], |
245 |
|
|
[$kdp->[1]], |
246 |
|
|
[$newleft, $newright]); |
247 |
|
|
$self->root($newroot); |
248 |
|
|
return; |
249 |
|
|
} |
250 |
|
|
} |
251 |
|
|
|
252 |
|
|
} |
253 |
|
|
|
254 |
|
|
sub split_and_promote { |
255 |
|
|
my $self = shift; |
256 |
|
|
my ($cur_node, @path) = @_; |
257 |
|
|
|
258 |
|
|
for (;;) { |
259 |
|
|
my ($newleft, $newright, $kdp) = $cur_node->halves($self->B / 2); |
260 |
|
|
my ($up, $where) = @{pop @path} if (@path); |
261 |
|
|
if ($up) { |
262 |
|
|
$up->kdp_insert(@$kdp); |
263 |
|
|
if ($DEBUG) { |
264 |
|
|
my ($tthere, $twhere) = $up->locate_key($kdp->[0]); |
265 |
|
|
croak "Couldn't find key `$kdp->[0]' in node after just inserting it!" |
266 |
|
|
unless $tthere; |
267 |
|
|
croak "`$kdp->[0]' went into node at `$twhere' instead of expected `$where'!" |
268 |
|
|
unless $twhere == $where; |
269 |
|
|
} |
270 |
|
|
$up->subnode($where, $newleft); |
271 |
|
|
$up->subnode($where+1, $newright); |
272 |
|
|
return unless $self->node_overfull($up); |
273 |
|
|
$cur_node = $up; |
274 |
|
|
} else { # We're at the top; make a new root. |
275 |
|
|
my $newroot = new jsFind::Node([$kdp->[0]], |
276 |
|
|
[$kdp->[1]], |
277 |
|
|
[$newleft, $newright]); |
278 |
|
|
$self->root($newroot); |
279 |
|
|
return; |
280 |
|
|
} |
281 |
|
|
} |
282 |
|
|
} |
283 |
|
|
|
284 |
dpavlin |
12 |
=head2 B |
285 |
dpavlin |
1 |
|
286 |
|
|
Return B (maximum number of keys) |
287 |
|
|
|
288 |
|
|
my $max_size = $t->B; |
289 |
|
|
|
290 |
|
|
=cut |
291 |
|
|
|
292 |
|
|
sub B { |
293 |
|
|
$_[0]{B}; |
294 |
|
|
} |
295 |
|
|
|
296 |
dpavlin |
12 |
=head2 root |
297 |
dpavlin |
1 |
|
298 |
|
|
Returns root node |
299 |
|
|
|
300 |
|
|
my $root = $t->root; |
301 |
|
|
|
302 |
|
|
=cut |
303 |
|
|
|
304 |
|
|
sub root { |
305 |
|
|
my ($self, $newroot) = @_; |
306 |
|
|
$self->{Root} = $newroot if defined $newroot; |
307 |
|
|
$self->{Root}; |
308 |
|
|
} |
309 |
|
|
|
310 |
dpavlin |
12 |
=head2 node_overfull |
311 |
dpavlin |
1 |
|
312 |
|
|
Returns if node is overfull |
313 |
|
|
|
314 |
|
|
if ($node->node_overfull) { something } |
315 |
|
|
|
316 |
|
|
=cut |
317 |
|
|
|
318 |
|
|
sub node_overfull { |
319 |
|
|
my $self = shift; |
320 |
|
|
my $node = shift; |
321 |
|
|
$node->size > $self->B; |
322 |
|
|
} |
323 |
|
|
|
324 |
dpavlin |
12 |
=head2 to_string |
325 |
dpavlin |
1 |
|
326 |
|
|
Returns your tree as formatted string. |
327 |
|
|
|
328 |
|
|
my $text = $root->to_string; |
329 |
|
|
|
330 |
|
|
Mostly usefull for debugging as output leaves much to be desired. |
331 |
|
|
|
332 |
|
|
=cut |
333 |
|
|
|
334 |
|
|
sub to_string { |
335 |
|
|
$_[0]->root->to_string; |
336 |
|
|
} |
337 |
|
|
|
338 |
dpavlin |
12 |
=head2 to_dot |
339 |
dpavlin |
1 |
|
340 |
|
|
Create Graphviz graph of your tree |
341 |
|
|
|
342 |
|
|
my $dot_graph = $root->to_dot; |
343 |
|
|
|
344 |
|
|
=cut |
345 |
|
|
|
346 |
|
|
sub to_dot { |
347 |
|
|
my $self = shift; |
348 |
|
|
|
349 |
|
|
my $dot = qq/digraph dns {\nrankdir=LR;\n/; |
350 |
|
|
$dot .= $self->root->to_dot; |
351 |
|
|
$dot .= qq/\n}\n/; |
352 |
|
|
|
353 |
|
|
return $dot; |
354 |
|
|
} |
355 |
|
|
|
356 |
dpavlin |
12 |
=head2 to_jsfind |
357 |
dpavlin |
1 |
|
358 |
|
|
Create xml index files for jsFind. This should be called after |
359 |
|
|
your B-Tree has been filled with data. |
360 |
|
|
|
361 |
dpavlin |
35 |
$root->to_jsfind( |
362 |
|
|
dir => '/full/path/to/index/dir/', |
363 |
|
|
data_codepage => 'ISO-8859-2', |
364 |
|
|
index_codepage => 'UTF-8', |
365 |
|
|
output_filter => sub { |
366 |
|
|
my $t = shift || return; |
367 |
|
|
$t =~ s/è/e/; |
368 |
|
|
} |
369 |
|
|
); |
370 |
dpavlin |
1 |
|
371 |
dpavlin |
35 |
All options except C<dir> are optional. |
372 |
|
|
|
373 |
dpavlin |
1 |
Returns number of nodes in created tree. |
374 |
|
|
|
375 |
dpavlin |
35 |
Options: |
376 |
dpavlin |
9 |
|
377 |
dpavlin |
35 |
=over 4 |
378 |
dpavlin |
9 |
|
379 |
dpavlin |
35 |
=item dir |
380 |
dpavlin |
11 |
|
381 |
dpavlin |
35 |
Full path to directory for index (which will be created if needed). |
382 |
dpavlin |
11 |
|
383 |
dpavlin |
35 |
=item data_codepage |
384 |
|
|
|
385 |
|
|
If your imput data isn't in C<ISO-8859-1> encoding, you will have to specify |
386 |
|
|
this option. |
387 |
|
|
|
388 |
|
|
=item index_codepage |
389 |
|
|
|
390 |
|
|
If your index encoding is not C<UTF-8> use this option. |
391 |
|
|
|
392 |
|
|
If you are not using supplied JavaScript search code, or your browser is |
393 |
|
|
terribly broken and thinks that index shouldn't be in UTF-8 encoding, use |
394 |
|
|
this option to specify encoding for created XML index. |
395 |
|
|
|
396 |
|
|
=item output_filter |
397 |
|
|
|
398 |
|
|
B<this is just draft of documentation for option which is not implemented!> |
399 |
|
|
|
400 |
|
|
Code ref to sub which can do modifications on resulting XML file for node. |
401 |
|
|
Encoding of this data will be in L<index_codepage> and you have to take care |
402 |
|
|
not to break XML structure. Calling L<xmllint> on your result index |
403 |
|
|
(like C<t/90xmllint.t> does in this distribution) is a good idea after using |
404 |
|
|
this option. |
405 |
|
|
|
406 |
|
|
This option is also right place to plug in unaccenting function using |
407 |
|
|
L<Text::Unaccent>. |
408 |
|
|
|
409 |
|
|
=back |
410 |
|
|
|
411 |
dpavlin |
1 |
=cut |
412 |
|
|
|
413 |
dpavlin |
9 |
my $iconv; |
414 |
dpavlin |
11 |
my $iconv_l1; |
415 |
dpavlin |
9 |
|
416 |
dpavlin |
1 |
sub to_jsfind { |
417 |
|
|
my $self = shift; |
418 |
|
|
|
419 |
dpavlin |
35 |
my %arg = @_; |
420 |
dpavlin |
1 |
|
421 |
dpavlin |
35 |
confess "to_jsfind need path to your index directory !" unless ($arg{'dir'}); |
422 |
dpavlin |
11 |
|
423 |
dpavlin |
35 |
my $data_codepage = $arg{'data_codepage'}; |
424 |
|
|
my $index_codepage = $arg{'index_codepage'} || 'UTF-8'; |
425 |
dpavlin |
11 |
|
426 |
dpavlin |
35 |
# create ISO-8859-1 iconv for HTML::Entities decode |
427 |
|
|
$iconv_l1 = Text::Iconv->new('ISO-8859-1',$index_codepage); |
428 |
|
|
|
429 |
|
|
# create another iconv for data |
430 |
|
|
if ($data_codepage && $index_codepage) { |
431 |
|
|
$iconv = Text::Iconv->new($data_codepage,$index_codepage); |
432 |
dpavlin |
9 |
} |
433 |
|
|
|
434 |
dpavlin |
35 |
return $self->root->to_jsfind($arg{'dir'},"0"); |
435 |
dpavlin |
1 |
} |
436 |
|
|
|
437 |
|
|
|
438 |
|
|
# private, default cmd function |
439 |
|
|
sub default_cmp { |
440 |
|
|
$_[0] cmp $_[1]; |
441 |
|
|
} |
442 |
|
|
|
443 |
dpavlin |
12 |
=head2 _recode |
444 |
dpavlin |
9 |
|
445 |
|
|
This is internal function to recode charset. |
446 |
|
|
|
447 |
dpavlin |
12 |
It will also try to decode entities in data using L<HTML::Entities>. |
448 |
dpavlin |
11 |
|
449 |
dpavlin |
9 |
=cut |
450 |
|
|
|
451 |
|
|
sub _recode { |
452 |
|
|
my $self = shift; |
453 |
|
|
my $text = shift || return; |
454 |
|
|
|
455 |
dpavlin |
11 |
sub _decode_html_entities { |
456 |
|
|
my $data = shift || return; |
457 |
|
|
$data = $iconv_l1->convert(decode_entities($data)) || croak "entity decode problem: $data"; |
458 |
|
|
} |
459 |
|
|
|
460 |
dpavlin |
9 |
if ($iconv) { |
461 |
dpavlin |
11 |
$text = $iconv->convert($text) || $text && carp "convert problem: $text"; |
462 |
|
|
$text =~ s/(\&\w+;)/_decode_html_entities($1)/ges; |
463 |
dpavlin |
9 |
} |
464 |
dpavlin |
11 |
|
465 |
|
|
return $text; |
466 |
dpavlin |
9 |
} |
467 |
|
|
|
468 |
dpavlin |
1 |
##################################################################### |
469 |
|
|
|
470 |
dpavlin |
12 |
=head1 jsFind::Node methods |
471 |
dpavlin |
1 |
|
472 |
|
|
Each node has C<k> key-data pairs, with C<B> <= C<k> <= C<2B>, and |
473 |
|
|
each has C<k+1> subnodes, which might be null. |
474 |
|
|
|
475 |
|
|
The node is a blessed reference to a list with three elements: |
476 |
|
|
|
477 |
|
|
($keylist, $datalist, $subnodelist) |
478 |
|
|
|
479 |
|
|
each is a reference to a list list. |
480 |
|
|
|
481 |
|
|
The null node is represented by a blessed reference to an empty list. |
482 |
|
|
|
483 |
|
|
=cut |
484 |
|
|
|
485 |
|
|
package jsFind::Node; |
486 |
|
|
|
487 |
|
|
use warnings; |
488 |
|
|
use strict; |
489 |
|
|
|
490 |
|
|
use Carp; |
491 |
|
|
use File::Path; |
492 |
dpavlin |
9 |
use Text::Iconv; |
493 |
dpavlin |
15 |
use POSIX; |
494 |
dpavlin |
1 |
|
495 |
dpavlin |
9 |
use base 'jsFind'; |
496 |
|
|
|
497 |
dpavlin |
1 |
my $KEYS = 0; |
498 |
|
|
my $DATA = 1; |
499 |
|
|
my $SUBNODES = 2; |
500 |
|
|
|
501 |
dpavlin |
12 |
=head2 new |
502 |
dpavlin |
1 |
|
503 |
|
|
Create New node |
504 |
|
|
|
505 |
|
|
my $node = new jsFind::Node ($keylist, $datalist, $subnodelist); |
506 |
|
|
|
507 |
|
|
You can also mit argument list to create empty node. |
508 |
|
|
|
509 |
|
|
my $empty_node = new jsFind::Node; |
510 |
|
|
|
511 |
|
|
=cut |
512 |
|
|
|
513 |
|
|
sub new { |
514 |
|
|
my $self = shift; |
515 |
|
|
my $package = ref $self || $self; |
516 |
|
|
croak "Internal error: jsFind::Node::new called with wrong number of arguments." |
517 |
|
|
unless @_ == 3 || @_ == 0; |
518 |
|
|
bless [@_] => $package; |
519 |
|
|
} |
520 |
|
|
|
521 |
dpavlin |
12 |
=head2 locate_key |
522 |
dpavlin |
1 |
|
523 |
|
|
Locate key in node using linear search. This should probably be replaced |
524 |
|
|
by binary search for better performance. |
525 |
|
|
|
526 |
|
|
my ($found, $index) = $node->locate_key($key, $cmp_coderef); |
527 |
|
|
|
528 |
|
|
Argument C<$cmp_coderef> is optional reference to custom comparison |
529 |
|
|
operator. |
530 |
|
|
|
531 |
|
|
Returns (1, $index) if $key[$index] eq $key. |
532 |
|
|
|
533 |
|
|
Returns (0, $index) if key could be found in $subnode[$index]. |
534 |
|
|
|
535 |
|
|
In scalar context, just returns 1 or 0. |
536 |
|
|
|
537 |
|
|
=cut |
538 |
|
|
|
539 |
|
|
sub locate_key { |
540 |
|
|
# Use linear search for testing, replace with binary search. |
541 |
|
|
my $self = shift; |
542 |
|
|
my $key = shift; |
543 |
|
|
my $cmp = shift || \&jsFind::default_cmp; |
544 |
|
|
my $i; |
545 |
|
|
my $cmp_result; |
546 |
|
|
my $N = $self->size; |
547 |
|
|
for ($i = 0; $i < $N; $i++) { |
548 |
|
|
$cmp_result = &$cmp($key, $self->key($i)); |
549 |
|
|
last if $cmp_result <= 0; |
550 |
|
|
} |
551 |
|
|
|
552 |
|
|
# $i is now the index of the first node-key greater than $key |
553 |
|
|
# or $N if there is no such. $cmp_result is 0 iff the key was found. |
554 |
|
|
(!$cmp_result, $i); |
555 |
|
|
} |
556 |
|
|
|
557 |
|
|
|
558 |
dpavlin |
12 |
=head2 emptynode |
559 |
dpavlin |
1 |
|
560 |
|
|
Creates new empty node |
561 |
|
|
|
562 |
|
|
$node = $root->emptynode; |
563 |
|
|
$new_node = $node->emptynode; |
564 |
|
|
|
565 |
|
|
=cut |
566 |
|
|
|
567 |
|
|
sub emptynode { |
568 |
|
|
new($_[0]); # Pass package name, but not anything else. |
569 |
|
|
} |
570 |
|
|
|
571 |
dpavlin |
12 |
=head2 is_empty |
572 |
dpavlin |
1 |
|
573 |
|
|
Test if node is empty |
574 |
|
|
|
575 |
|
|
if ($node->is_empty) { something } |
576 |
|
|
|
577 |
|
|
=cut |
578 |
|
|
|
579 |
|
|
# undef is empty; so is a blessed empty list. |
580 |
|
|
sub is_empty { |
581 |
|
|
my $self = shift; |
582 |
|
|
!defined($self) || $#$self < 0; |
583 |
|
|
} |
584 |
|
|
|
585 |
dpavlin |
12 |
=head2 key |
586 |
dpavlin |
1 |
|
587 |
|
|
Return C<$i>th key from node |
588 |
|
|
|
589 |
|
|
my $key = $node->key($i); |
590 |
|
|
|
591 |
|
|
=cut |
592 |
|
|
|
593 |
|
|
sub key { |
594 |
|
|
# my ($self, $n) = @_; |
595 |
|
|
# $self->[$KEYS][$n]; |
596 |
|
|
|
597 |
|
|
# speedup |
598 |
|
|
$_[0]->[$KEYS][$_[1]]; |
599 |
|
|
} |
600 |
|
|
|
601 |
dpavlin |
12 |
=head2 data |
602 |
dpavlin |
1 |
|
603 |
|
|
Return C<$i>th data from node |
604 |
|
|
|
605 |
|
|
my $data = $node->data($i); |
606 |
|
|
|
607 |
|
|
=cut |
608 |
|
|
|
609 |
|
|
sub data { |
610 |
|
|
my ($self, $n) = @_; |
611 |
|
|
$self->[$DATA][$n]; |
612 |
|
|
} |
613 |
|
|
|
614 |
dpavlin |
12 |
=head2 kdp_replace |
615 |
dpavlin |
1 |
|
616 |
|
|
Set key data pair for C<$i>th element in node |
617 |
|
|
|
618 |
|
|
$node->kdp_replace($i, "key value" => { |
619 |
|
|
"data key 1" => "data value 1", |
620 |
|
|
"data key 2" => "data value 2", |
621 |
|
|
}; |
622 |
|
|
|
623 |
|
|
=cut |
624 |
|
|
|
625 |
|
|
sub kdp_replace { |
626 |
|
|
my ($self, $n, $k => $d) = @_; |
627 |
|
|
if (defined $k) { |
628 |
|
|
$self->[$KEYS][$n] = $k; |
629 |
|
|
$self->[$DATA][$n] = $d; |
630 |
|
|
} |
631 |
|
|
[$self->[$KEYS][$n], |
632 |
|
|
$self->[$DATA][$n]]; |
633 |
|
|
} |
634 |
|
|
|
635 |
dpavlin |
12 |
=head2 kdp_insert |
636 |
dpavlin |
1 |
|
637 |
dpavlin |
12 |
Insert key/data pair in tree |
638 |
dpavlin |
1 |
|
639 |
dpavlin |
12 |
$node->kdp_insert("key value" => "data value"); |
640 |
|
|
|
641 |
|
|
No return value. |
642 |
|
|
|
643 |
dpavlin |
1 |
=cut |
644 |
|
|
|
645 |
|
|
sub kdp_insert { |
646 |
|
|
my $self = shift; |
647 |
|
|
my ($k => $d) = @_; |
648 |
|
|
my ($there, $where) = $self->locate_key($k) unless $self->is_empty; |
649 |
|
|
|
650 |
|
|
if ($there) { croak("Tried to insert `$k => $d' into node where `$k' was already present."); } |
651 |
|
|
|
652 |
|
|
# undef fix |
653 |
|
|
$where ||= 0; |
654 |
|
|
|
655 |
|
|
splice(@{$self->[$KEYS]}, $where, 0, $k); |
656 |
|
|
splice(@{$self->[$DATA]}, $where, 0, $d); |
657 |
|
|
splice(@{$self->[$SUBNODES]}, $where, 0, undef); |
658 |
|
|
} |
659 |
|
|
|
660 |
dpavlin |
12 |
=head2 kdp_append |
661 |
dpavlin |
1 |
|
662 |
|
|
Adds new data keys and values to C<$i>th element in node |
663 |
|
|
|
664 |
|
|
$node->kdp_append($i, "key value" => { |
665 |
|
|
"added data key" => "added data value", |
666 |
|
|
}; |
667 |
|
|
|
668 |
|
|
=cut |
669 |
|
|
|
670 |
|
|
sub kdp_append { |
671 |
|
|
my ($self, $n, $k => $d) = @_; |
672 |
|
|
if (defined $k) { |
673 |
|
|
$self->[$KEYS][$n] = $k; |
674 |
|
|
my ($kv,$dv) = %{$d}; |
675 |
|
|
$self->[$DATA][$n]->{$kv} = $dv; |
676 |
|
|
} |
677 |
|
|
[$self->[$KEYS][$n], |
678 |
|
|
$self->[$DATA][$n]]; |
679 |
|
|
} |
680 |
|
|
|
681 |
dpavlin |
12 |
=head2 subnode |
682 |
dpavlin |
1 |
|
683 |
|
|
Set new or return existing subnode |
684 |
|
|
|
685 |
|
|
# return 4th subnode |
686 |
|
|
my $my_node = $node->subnode(4); |
687 |
|
|
|
688 |
|
|
# create new subnode 5 from $my_node |
689 |
|
|
$node->subnode(5, $my_node); |
690 |
|
|
|
691 |
|
|
=cut |
692 |
|
|
|
693 |
|
|
sub subnode { |
694 |
|
|
my ($self, $n, $newnode) = @_; |
695 |
|
|
$self->[$SUBNODES][$n] = $newnode if defined $newnode; |
696 |
|
|
$self->[$SUBNODES][$n]; |
697 |
|
|
} |
698 |
|
|
|
699 |
dpavlin |
12 |
=head2 is_leaf |
700 |
dpavlin |
1 |
|
701 |
|
|
Test if node is leaf |
702 |
|
|
|
703 |
|
|
if ($node->is_leaf) { something } |
704 |
|
|
|
705 |
|
|
=cut |
706 |
|
|
|
707 |
|
|
sub is_leaf { |
708 |
|
|
my $self = shift; |
709 |
|
|
! defined $self->[$SUBNODES][0]; # undefined subnode means leaf node. |
710 |
|
|
} |
711 |
|
|
|
712 |
dpavlin |
12 |
=head2 size |
713 |
dpavlin |
1 |
|
714 |
|
|
Return number of keys in the node |
715 |
|
|
|
716 |
|
|
my $nr = $node->size; |
717 |
|
|
|
718 |
|
|
=cut |
719 |
|
|
|
720 |
|
|
sub size { |
721 |
|
|
my $self = shift; |
722 |
|
|
return scalar(@{$self->[$KEYS]}); |
723 |
|
|
} |
724 |
|
|
|
725 |
dpavlin |
12 |
=head2 halves |
726 |
dpavlin |
1 |
|
727 |
dpavlin |
12 |
Split node into two halves so that keys C<0 .. $n-1> are in one node |
728 |
|
|
and keys C<$n+1 ... $size> are in the other. |
729 |
dpavlin |
1 |
|
730 |
dpavlin |
12 |
my ($left_node, $right_node, $kdp) = $node->halves($n); |
731 |
|
|
|
732 |
dpavlin |
1 |
=cut |
733 |
|
|
|
734 |
|
|
sub halves { |
735 |
|
|
my $self = shift; |
736 |
|
|
my $n = shift; |
737 |
|
|
my $s = $self->size; |
738 |
|
|
my @right; |
739 |
|
|
my @left; |
740 |
|
|
|
741 |
|
|
$left[$KEYS] = [@{$self->[$KEYS]}[0 .. $n-1]]; |
742 |
|
|
$left[$DATA] = [@{$self->[$DATA]}[0 .. $n-1]]; |
743 |
|
|
$left[$SUBNODES] = [@{$self->[$SUBNODES]}[0 .. $n]]; |
744 |
|
|
|
745 |
|
|
$right[$KEYS] = [@{$self->[$KEYS]}[$n+1 .. $s-1]]; |
746 |
|
|
$right[$DATA] = [@{$self->[$DATA]}[$n+1 .. $s-1]]; |
747 |
|
|
$right[$SUBNODES] = [@{$self->[$SUBNODES]}[$n+1 .. $s]]; |
748 |
|
|
|
749 |
|
|
my @middle = ($self->[$KEYS][$n], $self->[$DATA][$n]); |
750 |
|
|
|
751 |
|
|
($self->new(@left), $self->new(@right), \@middle); |
752 |
|
|
} |
753 |
|
|
|
754 |
dpavlin |
12 |
=head2 to_string |
755 |
dpavlin |
1 |
|
756 |
|
|
Dumps tree as string |
757 |
|
|
|
758 |
|
|
my $str = $root->to_string; |
759 |
|
|
|
760 |
|
|
=cut |
761 |
|
|
|
762 |
|
|
sub to_string { |
763 |
|
|
my $self = shift; |
764 |
|
|
my $indent = shift || 0; |
765 |
|
|
my $I = ' ' x $indent; |
766 |
|
|
return '' if $self->is_empty; |
767 |
|
|
my ($k, $d, $s) = @$self; |
768 |
|
|
my $result = ''; |
769 |
|
|
$result .= defined($s->[0]) ? $s->[0]->to_string($indent+2) : ''; |
770 |
|
|
my $N = $self->size; |
771 |
|
|
my $i; |
772 |
|
|
for ($i = 0; $i < $N; $i++) { |
773 |
|
|
# $result .= $I . "$k->[$i] => $d->[$i]\n"; |
774 |
|
|
$result .= $I . "$k->[$i]\n"; |
775 |
|
|
$result .= defined($s->[$i+1]) ? $s->[$i+1]->to_string($indent+2) : ''; |
776 |
|
|
} |
777 |
|
|
$result; |
778 |
|
|
} |
779 |
|
|
|
780 |
|
|
=begin comment |
781 |
|
|
|
782 |
|
|
use Data::Dumper; |
783 |
|
|
|
784 |
|
|
sub to_string { |
785 |
|
|
my $self = shift; |
786 |
|
|
my $indent = shift || 0; |
787 |
|
|
my $path = shift || '0'; |
788 |
|
|
return '' if $self->is_empty; |
789 |
|
|
my ($k, $d, $s) = @$self; |
790 |
|
|
my $result = ''; |
791 |
|
|
$result .= defined($s->[0]) ? $s->[0]->to_string($indent+1,"$path/0") : ''; |
792 |
|
|
my $N = $self->size; |
793 |
|
|
for (my $i = 0; $i < $N; $i++) { |
794 |
|
|
my $dump = Dumper($d->[$i]); |
795 |
|
|
$dump =~ s/[\n\r\s]+/ /gs; |
796 |
|
|
$dump =~ s/\$VAR1\s*=\s*//; |
797 |
|
|
$result .= sprintf("%-5s [%2d] %2s: %s => %s\n", $path, $i, $indent, $k->[$i], $dump); |
798 |
|
|
$result .= defined($s->[$i+1]) ? $s->[$i+1]->to_string($indent+1,"$path/$i") : ''; |
799 |
|
|
} |
800 |
|
|
$result; |
801 |
|
|
} |
802 |
|
|
|
803 |
|
|
=end comment |
804 |
|
|
|
805 |
dpavlin |
12 |
=head2 to_dot |
806 |
dpavlin |
1 |
|
807 |
|
|
Recursivly walk nodes of tree |
808 |
|
|
|
809 |
|
|
=cut |
810 |
|
|
|
811 |
|
|
sub to_dot { |
812 |
|
|
my $self = shift; |
813 |
|
|
my $parent = shift; |
814 |
|
|
|
815 |
|
|
return '' if $self->is_empty; |
816 |
|
|
|
817 |
|
|
my $dot = ''; |
818 |
|
|
|
819 |
|
|
my ($k, $d, $s) = @$self; |
820 |
|
|
my $N = $self->size; |
821 |
|
|
|
822 |
|
|
my @dot_keys; |
823 |
|
|
|
824 |
|
|
my $node_name = $parent || '_'; |
825 |
|
|
$node_name =~ s/\W+//g; |
826 |
|
|
$node_name .= " [$N]"; |
827 |
|
|
|
828 |
|
|
for (my $i = 0; $i <= $N; $i++) { |
829 |
|
|
if (my $key = $k->[$i]) { |
830 |
|
|
push @dot_keys, qq{<$i>$key}; |
831 |
|
|
} |
832 |
|
|
$dot .= $s->[$i]->to_dot(qq{"$node_name":$i}) if ($s->[$i]); |
833 |
|
|
} |
834 |
|
|
push @dot_keys, qq{<$N>...} if (! $self->is_leaf); |
835 |
|
|
|
836 |
|
|
my $label = join("|",@dot_keys); |
837 |
|
|
$dot .= qq{"$node_name" [ shape=record, label="$label" ];\n}; |
838 |
|
|
|
839 |
|
|
$dot .= qq{$parent -> "$node_name";\n} if ($parent); |
840 |
|
|
|
841 |
|
|
$dot; |
842 |
|
|
} |
843 |
|
|
|
844 |
dpavlin |
12 |
=head2 to_xml |
845 |
dpavlin |
11 |
|
846 |
|
|
Escape <, >, & and ", and to produce valid XML |
847 |
|
|
|
848 |
|
|
=cut |
849 |
|
|
|
850 |
|
|
my %escape = ('<'=>'<', '>'=>'>', '&'=>'&', '"'=>'"'); |
851 |
|
|
my $escape_re = join '|' => keys %escape; |
852 |
|
|
|
853 |
|
|
sub to_xml { |
854 |
|
|
my $self = shift || confess "you should call to_xml as object!"; |
855 |
|
|
|
856 |
|
|
my $d = shift || return; |
857 |
|
|
$d = $self->SUPER::_recode($d); |
858 |
|
|
confess "escape_re undefined!" unless ($escape_re); |
859 |
|
|
$d =~ s/($escape_re)/$escape{$1}/g; |
860 |
|
|
return $d; |
861 |
|
|
} |
862 |
|
|
|
863 |
dpavlin |
36 |
=head2 base_x |
864 |
dpavlin |
15 |
|
865 |
dpavlin |
36 |
Convert number to base x (used for jsFind index filenames). |
866 |
dpavlin |
15 |
|
867 |
dpavlin |
36 |
my $n = $tree->base_x(50); |
868 |
dpavlin |
15 |
|
869 |
|
|
=cut |
870 |
|
|
|
871 |
dpavlin |
36 |
sub base_x { |
872 |
dpavlin |
15 |
my $self = shift; |
873 |
|
|
|
874 |
|
|
my $value = shift; |
875 |
|
|
|
876 |
|
|
confess("need non-negative number") if (! defined($value) || $value < 0); |
877 |
|
|
|
878 |
|
|
my @digits = qw( |
879 |
|
|
0 1 2 3 4 5 6 7 8 9 |
880 |
|
|
a b c d e f g h i j k l m n o p q r s t u v w x y z |
881 |
|
|
); |
882 |
|
|
|
883 |
|
|
my $base = scalar(@digits); |
884 |
|
|
my $out = ""; |
885 |
|
|
my $pow = 1; |
886 |
|
|
my $pos = 0; |
887 |
|
|
|
888 |
|
|
|
889 |
|
|
if($value == 0) { |
890 |
|
|
return "0"; |
891 |
|
|
} |
892 |
|
|
|
893 |
|
|
while($value > 0) { |
894 |
|
|
$pos = $value % $base; |
895 |
|
|
$out = $digits[$pos] . $out; |
896 |
|
|
$value = floor($value/$base); |
897 |
|
|
$pow *= $base; |
898 |
|
|
} |
899 |
|
|
|
900 |
|
|
return $out; |
901 |
|
|
} |
902 |
|
|
|
903 |
dpavlin |
12 |
=head2 to_jsfind |
904 |
dpavlin |
1 |
|
905 |
|
|
Create jsFind xml files |
906 |
|
|
|
907 |
dpavlin |
8 |
my $nr=$tree->to_jsfind('/path/to/index','0'); |
908 |
dpavlin |
1 |
|
909 |
|
|
Returns number of elements created |
910 |
|
|
|
911 |
|
|
=cut |
912 |
|
|
|
913 |
|
|
sub to_jsfind { |
914 |
|
|
my $self = shift; |
915 |
|
|
my ($path,$file) = @_; |
916 |
|
|
|
917 |
|
|
return 0 if $self->is_empty; |
918 |
|
|
|
919 |
dpavlin |
8 |
confess("path is undefined.") unless ($path); |
920 |
|
|
confess("file is undefined. Did you call \$t->root->to_jsfind(..) instead of \$t->to_jsfind(..) ?") unless (defined($file)); |
921 |
|
|
|
922 |
dpavlin |
36 |
$file = $self->base_x($file); |
923 |
dpavlin |
15 |
|
924 |
dpavlin |
1 |
my $nr_keys = 0; |
925 |
|
|
|
926 |
|
|
my ($k, $d, $s) = @$self; |
927 |
|
|
my $N = $self->size; |
928 |
|
|
|
929 |
|
|
my ($key_xml, $data_xml) = ("<n>","<d>"); |
930 |
|
|
|
931 |
|
|
for (my $i = 0; $i <= $N; $i++) { |
932 |
|
|
my $key = lc($k->[$i]); |
933 |
|
|
|
934 |
|
|
if ($key) { |
935 |
dpavlin |
11 |
$key_xml .= '<k>'.$self->to_xml($key).'</k>'; |
936 |
|
|
$data_xml .= '<e>'; |
937 |
dpavlin |
1 |
#use Data::Dumper; |
938 |
|
|
#print Dumper($d->[$i]); |
939 |
|
|
foreach my $path (keys %{$d->[$i]}) { |
940 |
dpavlin |
11 |
$data_xml .= '<l f="'.($d->[$i]->{$path}->{'f'} || 1).'" t="'.$self->to_xml($d->[$i]->{$path}->{'t'} || 'no title').'">'.$self->to_xml($path).'</l>'; |
941 |
dpavlin |
1 |
$nr_keys++; |
942 |
|
|
} |
943 |
dpavlin |
11 |
$data_xml .= '</e>'; |
944 |
dpavlin |
1 |
} |
945 |
|
|
|
946 |
|
|
$nr_keys += $s->[$i]->to_jsfind("$path/$file","$i") if ($s->[$i]); |
947 |
|
|
} |
948 |
|
|
|
949 |
dpavlin |
11 |
$key_xml .= '</n>'; |
950 |
|
|
$data_xml .= '</d>'; |
951 |
dpavlin |
1 |
|
952 |
|
|
if (! -e $path) { |
953 |
|
|
mkpath($path) || croak "can't create dir '$path': $!"; |
954 |
|
|
} |
955 |
|
|
|
956 |
|
|
open(K, "> ${path}/${file}.xml") || croak "can't open '$path/$file.xml': $!"; |
957 |
|
|
open(D, "> ${path}/_${file}.xml") || croak "can't open '$path/_$file.xml': $!"; |
958 |
|
|
|
959 |
dpavlin |
11 |
print K $key_xml; |
960 |
|
|
print D $data_xml; |
961 |
dpavlin |
1 |
|
962 |
|
|
close(K); |
963 |
|
|
close(D); |
964 |
|
|
|
965 |
|
|
return $nr_keys; |
966 |
|
|
} |
967 |
|
|
|
968 |
|
|
1; |
969 |
|
|
__END__ |
970 |
|
|
|
971 |
|
|
=head1 SEE ALSO |
972 |
|
|
|
973 |
|
|
jsFind web site L<http://www.elucidsoft.net/projects/jsfind/> |
974 |
|
|
|
975 |
|
|
B-Trees in perl web site L<http://perl.plover.com/BTree/> |
976 |
|
|
|
977 |
dpavlin |
14 |
This module web site L<http://www.rot13.org/~dpavlin/jsFind.html> |
978 |
|
|
|
979 |
dpavlin |
1 |
=head1 AUTHORS |
980 |
|
|
|
981 |
|
|
Mark-Jonson Dominus E<lt>mjd@pobox.comE<gt> wrote C<BTree.pm> which was |
982 |
|
|
base for this module |
983 |
|
|
|
984 |
|
|
Shawn P. Garbett E<lt>shawn@elucidsoft.netE<gt> wrote jsFind |
985 |
|
|
|
986 |
|
|
Dobrica Pavlinusic E<lt>dpavlin@rot13.orgE<gt> wrote this module |
987 |
|
|
|
988 |
|
|
=head1 COPYRIGHT AND LICENSE |
989 |
|
|
|
990 |
|
|
Copyright (C) 2004 by Dobrica Pavlinusic |
991 |
|
|
|
992 |
|
|
This program is free software; you can redistribute it and/or modify it |
993 |
|
|
under the terms of the GNU General Public License as published by the Free |
994 |
|
|
Software Foundation; either version 2 of the License, or (at your option) |
995 |
|
|
any later version. This program is distributed in the hope that it will be |
996 |
|
|
useful, but WITHOUT ANY WARRANTY; without even the implied warranty of |
997 |
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General |
998 |
|
|
Public License for more details. |
999 |
|
|
|
1000 |
|
|
=cut |