1 |
package WebPAC::Output::Estraier; |
2 |
|
3 |
use warnings; |
4 |
use strict; |
5 |
|
6 |
use base qw/WebPAC::Common/; |
7 |
|
8 |
use HyperEstraier; |
9 |
use Text::Iconv; |
10 |
use Data::Dumper; |
11 |
use LWP; |
12 |
use URI::Escape; |
13 |
|
14 |
=head1 NAME |
15 |
|
16 |
WebPAC::Output::Estraier - Create Hyper Estraier full text index |
17 |
|
18 |
=head1 VERSION |
19 |
|
20 |
Version 0.08 |
21 |
|
22 |
=cut |
23 |
|
24 |
our $VERSION = '0.08'; |
25 |
|
26 |
=head1 SYNOPSIS |
27 |
|
28 |
Create full text index using Hyper Estraier index from data with |
29 |
type C<search>. |
30 |
|
31 |
=head1 FUNCTIONS |
32 |
|
33 |
=head2 new |
34 |
|
35 |
Connect to Hyper Estraier index using HTTP |
36 |
|
37 |
my $est = new WebPAC::Output::Estraier( |
38 |
masterurl => 'http://localhost:1978/', |
39 |
user => 'admin', |
40 |
passwd => 'admin', |
41 |
database => 'demo', |
42 |
encoding => 'iso-8859-2', |
43 |
clean => 1, |
44 |
); |
45 |
|
46 |
Options are: |
47 |
|
48 |
=over 4 |
49 |
|
50 |
=item masterurl |
51 |
|
52 |
URI to C<estmaster> node |
53 |
|
54 |
=item user |
55 |
|
56 |
C<estmaster> user with sufficient rights |
57 |
|
58 |
=item passwd |
59 |
|
60 |
password for user |
61 |
|
62 |
=item database |
63 |
|
64 |
name of database from which data comes |
65 |
|
66 |
=item encoding |
67 |
|
68 |
character encoding of C<data_structure> if it's differenet than C<ISO-8859-2> |
69 |
(and it probably is). This encoding will be converted to C<UTF-8> for |
70 |
Hyper Estraier. |
71 |
|
72 |
=back |
73 |
|
74 |
Name of database will be used to form URI of documents in index. |
75 |
|
76 |
=cut |
77 |
|
78 |
sub new { |
79 |
my $class = shift; |
80 |
my $self = {@_}; |
81 |
bless($self, $class); |
82 |
|
83 |
my $log = $self->_get_logger; |
84 |
|
85 |
#$log->debug("self: ", sub { Dumper($self) }); |
86 |
|
87 |
foreach my $p (qw/masterurl user passwd database/) { |
88 |
$log->logdie("need $p") unless ($self->{$p}); |
89 |
} |
90 |
|
91 |
my $url = $self->{masterurl} . '/node/' . $self->{database}; |
92 |
$self->{url} = $url; |
93 |
|
94 |
if ($self->{clean}) { |
95 |
$log->debug("nodedel $self->{database}"); |
96 |
$self->master( action => 'nodedel', name => $self->{database} ); |
97 |
} else { |
98 |
$log->debug("opening index $self->{url}"); |
99 |
} |
100 |
|
101 |
my $nodes = $self->master( action => 'nodelist' ); |
102 |
|
103 |
$log->debug("nodes found: $nodes"); |
104 |
|
105 |
if ($nodes !~ m/^$self->{database}\t/sm) { |
106 |
$log->warn("creating index $url"); |
107 |
$self->master( |
108 |
action => 'nodeadd', |
109 |
name => $self->{database}, |
110 |
label => "WebPAC $self->{database}", |
111 |
) || $log->logdie("can't create Hyper Estraier node $self->{database}"); |
112 |
} |
113 |
|
114 |
$self->{'db'} = HyperEstraier::Node->new($self->{url}); |
115 |
$self->{'db'}->set_auth($self->{'user'}, $self->{passwd}); |
116 |
|
117 |
my $encoding = $self->{'encoding'} || 'ISO-8859-2'; |
118 |
$log->info("using index $self->{url} with encoding $encoding"); |
119 |
|
120 |
$self->{'iconv'} = new Text::Iconv($encoding, 'UTF-8') or |
121 |
$log->logdie("can't create conversion from $encoding to UTF-8"); |
122 |
|
123 |
$self ? return $self : return undef; |
124 |
} |
125 |
|
126 |
|
127 |
=head2 add |
128 |
|
129 |
Adds one entry to database. |
130 |
|
131 |
$est->add( |
132 |
id => 42, |
133 |
ds => $ds, |
134 |
type => 'display', |
135 |
text => 'optional text from which snippet is created', |
136 |
); |
137 |
|
138 |
This function will create entries in index using following URI format: |
139 |
|
140 |
C<file:///type/database%20name/000> |
141 |
|
142 |
Each tag in C<data_structure> with specified C<type> will create one |
143 |
attribute and corresponding hidden text (used for search). |
144 |
|
145 |
=cut |
146 |
|
147 |
sub add { |
148 |
my $self = shift; |
149 |
|
150 |
my $args = {@_}; |
151 |
|
152 |
my $log = $self->_get_logger; |
153 |
|
154 |
my $database = $self->{'database'} || $log->logconfess('no database in $self'); |
155 |
$log->logconfess('need db in object') unless ($self->{'db'}); |
156 |
|
157 |
foreach my $p (qw/id ds type/) { |
158 |
$log->logdie("need $p") unless ($args->{$p}); |
159 |
} |
160 |
|
161 |
my $type = $args->{'type'}; |
162 |
my $id = $args->{'id'}; |
163 |
|
164 |
my $uri = "file:///$type/$database/$id"; |
165 |
$log->debug("creating $uri"); |
166 |
|
167 |
my $doc = HyperEstraier::Document->new; |
168 |
$doc->add_attr('@uri', $self->{'iconv'}->convert($uri) ); |
169 |
|
170 |
$log->debug("ds = ", sub { Dumper($args->{'ds'}) } ); |
171 |
|
172 |
# filter all tags which have type defined |
173 |
my @tags = grep { |
174 |
ref($args->{'ds'}->{$_}) eq 'HASH' && defined( $args->{'ds'}->{$_}->{$type} ) |
175 |
} keys %{ $args->{'ds'} }; |
176 |
|
177 |
$log->debug("tags = ", join(",", @tags)); |
178 |
|
179 |
return unless (@tags); |
180 |
|
181 |
foreach my $tag (@tags) { |
182 |
|
183 |
my $vals = join(" ", @{ $args->{'ds'}->{$tag}->{$type} }); |
184 |
|
185 |
$log->logconfess("no values for $tag/$type") unless ($vals); |
186 |
|
187 |
$vals = $self->{'iconv'}->convert( $vals ) or |
188 |
$log->logdie("can't convert '$vals' to UTF-8"); |
189 |
|
190 |
$doc->add_attr( $tag, $vals ); |
191 |
$doc->add_hidden_text( $vals ); |
192 |
} |
193 |
|
194 |
my $text = $args->{'text'}; |
195 |
if ( $text ) { |
196 |
$text = $self->{'iconv'}->convert( $text ) or |
197 |
$log->logdie("can't convert '$text' to UTF-8"); |
198 |
$doc->add_text( $text ); |
199 |
} |
200 |
|
201 |
$log->debug("adding ", sub { $doc->dump_draft } ); |
202 |
$self->{'db'}->put_doc($doc) || $log->warn("can't add document $uri with draft " . $doc->dump_draft . " to node " . $self->{url} . " status: " . $self->{db}->status()); |
203 |
|
204 |
return 1; |
205 |
} |
206 |
|
207 |
# |
208 |
# REST parametars validation data |
209 |
# |
210 |
|
211 |
my $estraier_rest = { |
212 |
master => { |
213 |
userdel => [ qw/name/ ], |
214 |
nodelist => [], |
215 |
nodeadd => [ qw/name label/ ], |
216 |
nodedel => [ qw/name/ ], |
217 |
}, |
218 |
node => { |
219 |
_set_link => [ qw/url label credit/ ], |
220 |
}, |
221 |
}; |
222 |
|
223 |
=head2 master |
224 |
|
225 |
Issue administrative commands to C<estmaster> process and receive response |
226 |
as array of lines |
227 |
|
228 |
my $nodelist = $est->master( action => 'nodelist' ); |
229 |
|
230 |
=cut |
231 |
|
232 |
sub master { |
233 |
my $self = shift; |
234 |
|
235 |
my $args = {@_}; |
236 |
my $log = $self->_get_logger; |
237 |
|
238 |
my $action = $args->{action} || $log->logconfess("no action specified"); |
239 |
|
240 |
$log->logdie("action '$action' isn't supported") unless ($estraier_rest->{master}->{$action}); |
241 |
|
242 |
$log->debug("master action: $action"); |
243 |
|
244 |
return $self->estcall( |
245 |
validate => 'master', |
246 |
rest_url => $self->{masterurl} . '/master?action=' . $action , |
247 |
action => $action, |
248 |
%{ $args }, |
249 |
); |
250 |
} |
251 |
|
252 |
=head2 add_link |
253 |
|
254 |
$est->add_link( |
255 |
from => 'ps', |
256 |
to => 'webpac2', |
257 |
credit => 10000, |
258 |
); |
259 |
|
260 |
=cut |
261 |
|
262 |
sub add_link { |
263 |
my $self = shift; |
264 |
|
265 |
my $args = {@_}; |
266 |
my $log = $self->_get_logger; |
267 |
|
268 |
my @labels = $self->master( action => 'nodelist' ); |
269 |
|
270 |
$log->debug("got labels: ", join("|", @labels)); |
271 |
|
272 |
@labels = grep(/^$args->{to}\t/, @labels); |
273 |
my $label = shift @labels; |
274 |
(undef,$label) = split(/\t/, $label) if ($label); |
275 |
|
276 |
if (! $label) { |
277 |
$log->warn("can't find label for $args->{to}, skipping link creaton"); |
278 |
return; |
279 |
} |
280 |
|
281 |
$log->debug("using label $label for $args->{to}"); |
282 |
|
283 |
return $self->estcall( |
284 |
validate => 'node', |
285 |
action => '_set_link', |
286 |
rest_url => $self->{masterurl} . '/node/' . $args->{from} . '/_set_link' , |
287 |
url => $self->{masterurl} . '/node/' . $args->{to}, |
288 |
label => $label, |
289 |
credit => $args->{credit}, |
290 |
); |
291 |
} |
292 |
|
293 |
=head2 estcall |
294 |
|
295 |
Workhourse which does actual calls to Hyper Estraier |
296 |
|
297 |
$self->estcall( |
298 |
rest_url => '/master?action=' . $action, |
299 |
validate => 'master', |
300 |
# ... |
301 |
); |
302 |
|
303 |
C<rest_url> is relative URL to C<estmaster> and C<validate> is entry into |
304 |
internal hash which will check if all parametars are available before |
305 |
calling function. |
306 |
|
307 |
=cut |
308 |
|
309 |
sub estcall { |
310 |
my $self = shift; |
311 |
my $args = {@_}; |
312 |
my $log = $self->_get_logger; |
313 |
|
314 |
$log->debug("estcall: ",Dumper($args)); |
315 |
|
316 |
foreach my $p (qw/rest_url validate action/) { |
317 |
$log->die("ectcall needs $p parametar") unless ($args->{$p}); |
318 |
} |
319 |
|
320 |
my $url = $args->{rest_url}; |
321 |
my $del = '?'; |
322 |
$del = '&' if ($url =~ m#\?#); |
323 |
|
324 |
my $url_args; |
325 |
|
326 |
foreach my $arg (@{ $estraier_rest->{ $args->{validate} }->{ $args->{action} } }) { |
327 |
$log->logdie("missing parametar $arg for action $args->{action}") unless ($args->{$arg}); |
328 |
$url_args .= $del . $arg . '=' . uri_escape( $args->{$arg} ); |
329 |
$del = '&'; |
330 |
} |
331 |
|
332 |
$url .= $url_args if ($url_args); |
333 |
|
334 |
$log->debug("calling $url"); |
335 |
|
336 |
my $res = $self->est_ua()->get($url); |
337 |
|
338 |
if ($res->is_success) { |
339 |
#$log->debug( $res->content ); |
340 |
return split(/\n/, $res->content) if wantarray; |
341 |
return $res->content || 0E0; |
342 |
} else { |
343 |
$log->warn("unable to call $url: " . $res->status_line); |
344 |
return; |
345 |
} |
346 |
|
347 |
} |
348 |
|
349 |
=head2 est_ua |
350 |
|
351 |
This is helper function to create C<LWP::UserAgent> object with Super User |
352 |
priviledges |
353 |
|
354 |
my $ua = $self->est_ua( user => 'admin', passwd => 'admin' ); |
355 |
|
356 |
=cut |
357 |
|
358 |
|
359 |
|
360 |
sub est_ua { |
361 |
my $self = shift; |
362 |
|
363 |
return $self->{_master_ua} if ($self->{_master_ua}); |
364 |
|
365 |
{ |
366 |
package AdminUserAgent; |
367 |
use base qw/LWP::UserAgent/; |
368 |
sub new { |
369 |
my $self = LWP::UserAgent::new(@_); |
370 |
$self->agent("webpac/$VERSION"); |
371 |
$self; |
372 |
} |
373 |
sub get_basic_credentials { |
374 |
my($self, $realm, $uri) = @_; |
375 |
return ($self->{user}, $self->{passwd}); |
376 |
} |
377 |
sub set_basic_credentials { |
378 |
my ($self, $user, $passwd) = @_; |
379 |
$self->{user} = $user; |
380 |
$self->{passwd} = $passwd; |
381 |
} |
382 |
}; |
383 |
|
384 |
$self->{_master_ua} = AdminUserAgent->new( ) || sub { |
385 |
my $log = $self->_get_logger; |
386 |
$log->logdie("can't create LWP::UserAgent: $!"); |
387 |
}; |
388 |
|
389 |
$self->{_master_ua}->set_basic_credentials($self->{user}, $self->{passwd}); |
390 |
|
391 |
return $self->{_master_ua}; |
392 |
} |
393 |
|
394 |
=head1 AUTHOR |
395 |
|
396 |
Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >> |
397 |
|
398 |
=head1 COPYRIGHT & LICENSE |
399 |
|
400 |
Copyright 2005 Dobrica Pavlinusic, All Rights Reserved. |
401 |
|
402 |
This program is free software; you can redistribute it and/or modify it |
403 |
under the same terms as Perl itself. |
404 |
|
405 |
=cut |
406 |
|
407 |
1; # End of WebPAC::Output::Estraier |